Haematologica, Volume 103, issue 6

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Ancient Greek

The origin of a name that reflects Europe’s cultural roots.

Scientific Latin

aÂma [haima] = blood a·matow [haimatos] = of blood lÒgow [logos]= reasoning

Scientific Latin

haematologicus (adjective) = related to blood

Modern English

haematologica (adjective, plural and neuter, used as a noun) = hematological subjects The oldest hematology journal, publishing the newest research results.

2016 JCR impact factor = 7.702

Haematologica, as the journal of the European Hematology Association (EHA), aims not only to serve the scientific community, but also to promote European cultural identify.

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation Editor-in-Chief Luca Malcovati (Pavia)

Managing Director Antonio Majocchi (Pavia)

Associate Editors Omar I. Abdel-Wahab (New York), Hélène Cavé (Paris), Simon Mendez-Ferrer (Cambridge), Pavan Reddy (Ann Arbor), Andreas Rosenwald (Wuerzburg), Monika Engelhardt (Freiburg), Davide Rossi (Bellinzona), Jacob Rowe (Haifa, Jerusalem), Wyndham Wilson (Bethesda), Paul Kyrle (Vienna), Swee Lay Thein (Bethesda), Pieter Sonneveld (Rotterdam)

Assistant Editors Anne Freckleton (English Editor), Cristiana Pascutto (Statistical Consultant), Rachel Stenner (English Editor), Kate O’Donohoe (English Editor), Ziggy Kennell (English Editor)

Editorial Board Jeremy Abramson (Boston); Paolo Arosio (Brescia); Raphael Bejar (San Diego); Erik Berntorp (Malmö); Dominique Bonnet (London); Jean-Pierre Bourquin (Zurich); Suzanne Cannegieter (Leiden); Francisco Cervantes (Barcelona); Nicholas Chiorazzi (Manhasset); Oliver Cornely (Köln); Michel Delforge (Leuven); Ruud Delwel (Rotterdam); Meletios A. Dimopoulos (Athens); Inderjeet Dokal (London); Hervé Dombret (Paris); Peter Dreger (Hamburg); Martin Dreyling (München); Kieron Dunleavy (Bethesda); Dimitar Efremov (Rome); Sabine Eichinger (Vienna); Jean Feuillard (Limoges); Carlo Gambacorti-Passerini (Monza); Guillermo Garcia Manero (Houston); Christian Geisler (Copenhagen); Piero Giordano (Leiden); Christian Gisselbrecht (Paris); Andreas Greinacher (Greifswals); Hildegard Greinix (Vienna); Paolo Gresele (Perugia); Thomas M. Habermann (Rochester); Claudia Haferlach (München); Oliver Hantschel (Lausanne); Christine Harrison (Southampton); Brian Huntly (Cambridge); Ulrich Jaeger (Vienna); Elaine Jaffe (Bethesda); Arnon Kater (Amsterdam); Gregory Kato (Pittsburg); Christoph Klein (Munich); Steven Knapper (Cardiff); Seiji Kojima (Nagoya); John Koreth (Boston); Robert Kralovics (Vienna); Ralf Küppers (Essen); Ola Landgren (New York); Peter Lenting (Le Kremlin-Bicetre); Per Ljungman (Stockholm); Francesco Lo Coco (Rome); Henk M. Lokhorst (Utrecht); John Mascarenhas (New York); Maria-Victoria Mateos (Salamanca); Giampaolo Merlini (Pavia); Anna Rita Migliaccio (New York); Mohamad Mohty (Nantes); Martina Muckenthaler (Heidelberg); Ann Mullally (Boston); Stephen Mulligan (Sydney); German Ott (Stuttgart); Jakob Passweg (Basel); Melanie Percy (Ireland); Rob Pieters (Utrecht); Stefano Pileri (Milan); Miguel Piris (Madrid); Andreas Reiter (Mannheim); Jose-Maria Ribera (Barcelona); Stefano Rivella (New York); Francesco Rodeghiero (Vicenza); Richard Rosenquist (Uppsala); Simon Rule (Plymouth); Claudia Scholl (Heidelberg); Martin Schrappe (Kiel); Radek C. Skoda (Basel); Gérard Socié (Paris); Kostas Stamatopoulos (Thessaloniki); David P. Steensma (Rochester); Martin H. Steinberg (Boston); Ali Taher (Beirut); Evangelos Terpos (Athens); Takanori Teshima (Sapporo); Pieter Van Vlierberghe (Gent); Alessandro M. Vannucchi (Firenze); George Vassiliou (Cambridge); Edo Vellenga (Groningen); Umberto Vitolo (Torino); Guenter Weiss (Innsbruck).

Editorial Office Simona Giri (Production & Marketing Manager), Lorella Ripari (Peer Review Manager), Paola Cariati (Senior Graphic Designer), Igor Ebuli Poletti (Senior Graphic Designer), Marta Fossati (Peer Review), Diana Serena Ravera (Peer Review)

Affiliated Scientific Societies SIE (Italian Society of Hematology, www.siematologia.it) SIES (Italian Society of Experimental Hematology, www.siesonline.it)

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Information for readers, authors and subscribers Haematologica (print edition, pISSN 0390-6078, eISSN 1592-8721) publishes peer-reviewed papers on all areas of experimental and clinical hematology. The journal is owned by a non-profit organization, the Ferrata Storti Foundation, and serves the scientific community following the recommendations of the World Association of Medical Editors (www.wame.org) and the International Committee of Medical Journal Editors (www.icmje.org).

Haematologica publishes editorials, research articles, review articles, guideline articles and letters. Manuscripts should be prepared according to our guidelines (www.haematologica.org/information-for-authors), and the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, prepared by the International Committee of Medical Journal Editors (www.icmje.org).

Manuscripts should be submitted online at http://www.haematologica.org/.

Conflict of interests. According to the International Committee of Medical Journal Editors (http://www.icmje.org/#conflicts), “Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing, peer review, and editorial decision making”. The ad hoc journal’s policy is reported in detail online (www.haematologica.org/content/policies).

Transfer of Copyright and Permission to Reproduce Parts of Published Papers. Authors will grant copyright of their articles to the Ferrata Storti Foundation. No formal permission will be required to reproduce parts (tables or illustrations) of published papers, provided the source is quoted appropriately and reproduction has no commercial intent. Reproductions with commercial intent will require written permission and payment of royalties.

Detailed information about subscriptions is available online at www.haematologica.org. Haematologica is an open access journal. Access to the online journal is free. Use of the Haematologica App (available on the App Store and on Google Play) is free.

For subscriptions to the printed issue of the journal, please contact: Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, E-mail: info@haematologica.org).

Rates of the International edition for the year 2018 are as following: Print edition

Institutional Euro 600

Personal Euro 150

Advertisements. Contact the Advertising Manager, Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, e-mail: marketing@haematologica.org).

Disclaimer. Whilst every effort is made by the publishers and the editorial board to see that no inaccurate or misleading data, opinion or statement appears in this journal, they wish to make it clear that the data and opinions appearing in the articles or advertisements herein are the responsibility of the contributor or advisor concerned. Accordingly, the publisher, the editorial board and their respective employees, officers and agents accept no liability whatsoever for the consequences of any inaccurate or misleading data, opinion or statement. Whilst all due care is taken to ensure that drug doses and other quantities are presented accurately, readers are advised that new methods and techniques involving drug usage, and described within this journal, should only be followed in conjunction with the drug manufacturer’s own published literature.

Direttore responsabile: Prof. Edoardo Ascari; Autorizzazione del Tribunale di Pavia n. 63 del 5 marzo 1955.

Printing: Press Up, zona Via Cassia Km 36, 300 Zona Ind.le Settevene - 01036 Nepi (VT)

haematologica calendar of events

Journal of the European Hematology Association Published by the Ferrata Storti Foundation 35th International Congress of the ISBT The International Society of Blood Transfusion (ISBT) Chairs: K Pavenski, E van der Schoot, E Wood June 2-6, 2018 Toronto, Canada 11th PDWP, IDWP, IEWP & Paediatric Nurses Group Meeting European Society for Blood and Marrow Transplantations (EBMT) Chairs: P Bader, S Corbacioglu, M Ansari, A Balduzzi June 7-9, 2018 Verona, Italy 20 Congress of the European Society for Haemapheresis European Society for Haemapheresis June 13-14, 2018 Valencia, Spain th

23rd Congress of EHA June 14-17, 2018 Stockholm, Sweden

EHA-SAH Hematology Tutorial on lymphoid Malignancies and Plasma Cell Dyscrasias September 14-15, 2018 Buenos Aires, Argentina 14th Educational Course of the Lymphoma Working Party European Society for Blood and Marrow Transplantations (EBMT) Lymphoma Working Party Chairs: S Montoto, A Sureda, L Bento September 27-28, 2018 Palma, Spain EHA-SWG Scientific Meeting on Aging and Hematology Chair: D Bron October 12-14, 2018 Warsaw, Poland

Calendar of Events updated on May 7, 2018

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Table of Contents Volume 103, Issue 6: June 2018 Cover Figure

Peripheral blood smear from a patient with Sézary syndrome showing a monocyte and four Sézary cells containing rings of coarse Periodic acid-Schiff positive granules. Courtesy of Prof. Rosangela Invernizzi.

Editorials 919

Hematopoietic stem cells made BETter by inhibition Ludovica Marando and Brian J. P. Huntly

921

NSG-S mice for acute myeloid leukemia, yes. For myelodysplastic syndrome, no.

Emmanuel Griessinger and Michael Andreeff

924

Defining the elusive boundaries of chronic active Epstein-Barr virus infection Sebastian Fernandez-Pol et al.

Perspective 928

Present results and future perspectives in optimizing chronic myeloid leukemia therapy Angelo M. Carella et al.

Review Article 931

Morphological, immunophenotypic, and genetic features of chronic lymphocytic leukemia with trisomy 12: a comprehensive review Francesco Autore et al.

Articles Hematopoiesis

939

BET-inhibition by JQ1 promotes proliferation and self-renewal capacity of hematopoietic stem cells Mark Wroblewski et al.

Red Cell Biology & its Disorders

949

Recurring mutations in RPL15 are linked to hydrops fetalis and treatment independence in Diamond-Blackfan anemia Marcin W. Wlodarski et al.

Myelodysplastic Syndromes

959

Cytokines increase engraftment of human acute myeloid leukemia cells in immunocompromised mice but not engraftment of human myelodysplastic syndrome cells Maria Krevvata et al.

Myeloproliferative Neoplasms

972

Impact of hydroxycarbamide and interferon-α on red cell adhesion and membrane protein expression in polycythemia vera Mégane Brusson et al.

Acute Myeloid Leukemia

982

A novel regimen for relapsed/refractory adult acute myeloid leukemia using a KMT2A partial tandem duplication targeted therapy: results of phase 1 study NCI 8485 Alice S. Mims et al.

988

Dexamethasone in hyperleukocytic acute myeloid leukemia Sarah Bertoli et al.

Haematologica 2018; vol. 103 no. 6 - June 2018 http://www.haematologica.org/

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation Acute Lymphoblastic Leukemia

999

Fit αβ T-cell receptor suppresses leukemogenesis of Pten-deficient thymocytes Stéphanie Gon et al.

1008

Leukemia reconstitution in vivo is driven by cells in early cell cycle and low metabolic state Luca Trentin et al.

Non-Hodgkin Lymphoma

1018

A distinct subtype of Epstein-Barr virus-positive T/NK-cell lymphoproliferative disorder: adult patients with chronic active Epstein-Barr virus infection-like features Keisuke Kawamoto et al.

1029

Highly sensitive MYD88L265P mutation detection by droplet digital polymerase chain reaction in Waldenström macroglobulinemia Daniela Drandi et al.

Chronic Lymphocytic Leukemia

1038

DNA polymerase ν gene expression influences fludarabine resistance in chronic lymphocytic leukemia independently of p53 status Srdana Grgurevic et al.

Plasma Cell Disorders

1047

Treatment to suppression of focal lesions on positron emission tomography-computed tomography is a therapeutic goal in newly diagnosed multiple myeloma Faith E. Davies et al.

Stem Cell Transplantation

1054

A second-generation 15-PGDH inhibitor promotes bone marrow transplant recovery independently of age, transplant dose and granulocyte colony-stimulating factor support Amar Desai et al.

Cell Therapy & Immunotherapy

1065

Targeting the anion exchanger 2 with specific peptides as a new therapeutic approach in B lymphoid neoplasms Jon Celay et al.

Coagulation & its Disorders

1073

Structural and cellular mechanisms of peptidyl-prolyl isomerase Pin1-mediated enhancement of Tissue Factor gene expression, protein half-life, and pro-coagulant activity Kondababu Kurakula et al.

1083

Mass spectrometry-assisted identification of ADAMTS13-derived peptides presented on HLA-DR and HLA-DQ Johana Hrdinová et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/103/6.toc

e230

Memory CD8+ T cells support the maintenance of hematopoietic stem cells in the bone marrow Sulima Geerman et al. http://www.haematologica.org/content/103/6/e230

e234

Shortened telomeres in essential thrombocythemia: clinicopathological and treatment correlations Samah Alimam et al. http://www.haematologica.org/content/103/6/e234

e237

Outcomes of a large cohort of individuals with clinically ascertained high-count monoclonal B-cell lymphocytosis Sameer A. Parikh et al. http://www.haematologica.org/content/103/6/e237

Haematologica 2018; vol. 103 no. 6 - June 2018 http://www.haematologica.org/

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation e241

Liquid biopsy for the identification of intravascular large B-cell lymphoma Yasuhito Suehara et al. http://www.haematologica.org/content/103/6/e241

e245

Circulating tumor DNA as a liquid biopsy in plasma cell dyscrasias Bernhard Gerber et al. http://www.haematologica.org/content/103/6/e245

e249

Survival adjusting for crossover: phase 3 study of ibrutinib vs. chlorambucil in older patients with untreated chronic lymphocytic leukemia/small lymphocytic lymphoma Steven Coutre et al. http://www.haematologica.org/content/103/6/e249

e252

Non-myeloablative allogeneic hematopoietic cell transplantation for relapsed or refractory Waldenström macroglobulinemia: evidence for a graft-versus-lymphoma effect Enrico Maffini et al. http://www.haematologica.org/content/103/6/e252

e256

The contribution of platelet glycoprotein receptors to inflammatory bleeding prevention is stimulus and organ dependent Julie Rayes et al. http://www.haematologica.org/content/103/6/e256

e259

A novel variant Glanzmann thrombasthenia due to co-inheritance of a loss- and a gain-of-function mutation of ITGB3: evidence of a dominant effect of gain-of-function mutations Loredana Bury et al. http://www.haematologica.org/content/103/6/e259

e264

Long-term management of leukocyte adhesion deficiency type III without hematopoietic stem cell transplantation Paul Saultier et al. http://www.haematologica.org/content/103/6/e264

Comments Comments are available online only at www.haematologica.org/content/103/6.toc

e268

Comment to the article by Arai Y, Jo T, Matsui H, Kondo T and Takaori-Kondo A: “Comparison of up-front treatments for newly diagnosed immune thrombocytopenia – a systematic review and network meta-analysis”. Haematologica 2018;103(1):163-171. Need to direct immune thrombocytopenia therapy towards shared goals Lorenzo Cirasino et al. http://www.haematologica.org/content/103/6/e268

e269

Response to the Comment by Cirasino L and Semeraro S: “Need to direct immune thrombocytopenia therapy towards shared goals” Direct and indirect comparisons to determine the first choice for newly diagnosed primary immune thrombocytopenia in adults Yasuyuki Arai et al http://www.haematologica.org/content/103/6/e269

Case Reports Case Reports are available online only at www.haematologica.org/content/103/6.toc

e270

Lenalidomide results in a durable complete remission in acute myeloid leukemia accompanied by persistence of somatic mutations and a T-cell infiltrate in the bone marrow Dhruv Bansal et al. http://www.haematologica.org/content/103/6/e270

e274

Down syndrome-like acute megakaryoblastic leukemia in a patient with Cornelia de Lange syndrome Yoann Vial et al. http://www.haematologica.org/content/103/6/e274

e277

Safety and efficacy of daratumumab in dialysis-dependent renal failure secondary to multiple myeloma Serena Rocchi et al. http://www.haematologica.org/content/103/6/e277

Haematologica 2018; vol. 103 no. 6 - June 2018 http://www.haematologica.org/

EDITORIALS Hematopoietic stem cells made BETter by inhibition Ludovica Marando1,2,3 and Brian J. P. Huntly1,2,3 1

Wellcome Trust‐MRC Cambridge Stem Cell Institute; 2Department of Haematology, University of Cambridge and 3Cambridge Institute for Medical Research, Cambridge Biomedical Campus, UK E-mail: bjph2@cam.ac.uk doi:10.3324/haematol.2018.193706

B

romodomain and extra terminal (BET) proteins comprise the ubiquitously expressed BRD2, BRD3, BRD4 and the testes specific BRDT.1,2 These multipurpose proteins contain tandem N-terminal bromodomains that bind acetylated lysine residues of histone (and non-histone) proteins and other protein modules, such as the extra terminal domain, and in some (BRD4, BRDT), a C-terminal domain.

They also mediate a number of effects including transcriptional activation via recruitment of other partner proteins.3 Inhibitors of these proteins are emerging as exciting new therapies for the treatment of hematologic and solid malignancies, offering the possibility of specifically targeting epigenetic readers. We and others have already demonstrated the preclinical efficacy of BET inhibitors in acute myeloid leukemia (AML),4-6 while several other papers have documented similar efficacy in myeloma,7 non-Hodgkin lymphoma,8 and acute lymphoblastic leukemia.9 These observations have led to several clinical trials that are currently underway to confirm the efficacy of these drugs in AML and other malignancies. Even though the most mature trials have recently reported limited objective responses of monotherapy in heavily pre-treated AML, lymphoma and myeloma patients,10-12 early data suggest that combination therapies with other small molecules or more conventional cytotoxic agents might be particularly promising.13-15 BET proteins have multiple functions. Amongst these, and considered critical for the maintenance of malignant transcription, they are implicated in the regulation of large or “super” enhancers that control a number of critical genes, including oncogenes pivotal for the maintenance of leukemia, such as BCL-2, IRF8, and c-MYC. Their downregulation upon treatment with small molecule BET inhibitors at least partially explains the effects of BET inhibition observed in hematologic and solid malignancies.4,13,15,16 However, despite several ongoing clinical trials, we know surprisingly little about the consequences of disrupting BET protein function in normal tissues. In this issue of the Journal, Wroblewski et al.,17 therefore, address a highly relevant topic and describe the effects of the prototypic BET inhibitor JQ1 on normal hematopoiesis.

Surprisingly, upon JQ1 treatment, Wroblewski et al. identify an increase in phenotypic HSC proliferation and mobilization in mice. In addition, these effects seem sustained, and functionally, in the setting of competitive transplantation, JQ1 treated HSC appear to contribute more to hematopoiesis in primary and, importantly, in secondary recipients with no evidence of exhaustion, albeit follow up was only for 12 months following transplantation. This increased proliferation does not enhance radiosensitivity. On the contrary, JQ1 treated mice show faster count recovery following sublethal irradiation compared to untreated controls. Although the authors have not studied potential underlying mechanisms, they postulate that the effects might be haematologica | 2018; 103(6)

mediated by JQ1 dependant suppression of Myb,18 given the phenotypic similarity between JQ1 treatment and a mouse model of reduced Myb activity.19 Due to its pharmacokinetic properties, JQ1 does not lead to sustained target inhibition, and due to its limited efficacy in pre-clinical models, JQ1 has never been tested in clinical trials.20 Therefore, whether JQ1 induced expansion of the normal HSC pool described here is a “class effect” shared by more potent BET inhibitors needs to be carefully addressed by further studies. This is further called into question by the toxicity reports from the clinical trials that have consistently reported hematologic toxicity, in particular, a dose-dependent, non-cumulative, reversible thrombocytopenia,10,11 and by opposing reports in the literature regarding the effects of BET inhibition on normal HSC. In a mouse model of controlled BRD4 inhibition, using an inducible transgenic shRNA, Brd4 silencing caused a significant reduction in Lineage- Sca-1+ c-kit+ hematopoietic stem cells 12 weeks after hematopoietic reconstitution.21 However, as RNAi would not only lead to almost complete loss of Brd4, but would also target the non-bromodomain dependent functions of the protein, differences between these models, where inhibitors would only intermittently target the bromodomains, might be expected.

Certainly the possibility that BET inhibitors, in addition to exerting antitumor effects, could enhance recovery of normal hematopoiesis, especially after combination chemotherapy, is intriguing and could open up new avenues for the use of BET inhibitors in clinical practice (Figure 1). Other areas of utility for BET inhibition, as suggested by Wroblewski et al.,17 could include mobilization of peripheral blood (PB) HSC from donors who fail more standard approaches.

However, validation of this study with orally available BET inhibitors and specific testing of these hypotheses will be necessary.

BET inhibitors might also have a role against acute graftversus-host disease (GvHD) and, in fact, were initially designed as immunosuppressives. Wroblewski et al.17 and others have shown that BET inhibitors impair T-cell function. Specifically, Wroblewski et al.17 find that JQ1 treatment causes increased apoptosis in T cells, and this is associated with reduced expression of the antiapoptotic BCL-2 as measured by RT-qPCR. Others have demonstrated that treatment with I-BET151 results in a reduction in the secretion of IL-6, TNF-α, and IL-12 by stimulated dendritic cells (DCs).22 Both T cells and DCs are important mediators of GvHD in recipients of allogeneic stem cell transplant, suggesting that BET inhibitors may serve as a prophylactic therapy against acute GvHD. Again, the results suggested by the pre-clinical models need to be validated in clinical trials and the benefits must be weighed against the risks of hematologic toxicity, particularly thrombocytopenia and neutropenia.

In conclusion, BRD4/BET inhibition is becoming a novel 919

Editorials

Figure 1. Effects of BET inhibition on normal and diseased hematopoiesis. Hematopoiesis is a finely regulated process precisely balancing hematopoietic stem cell (HSC) self-renewal and differentiation. In a preclinical model, BET inhibition seems to affect this tightly regulated mechanism, resulting (as shown) in increased HSC self-renewal and mobilization but also altered lymphocyte differentiation. Advantageous effects may occur on mobilization of peripheral blood HSC donors or could enhance restoration of normal hematopoiesis after chemotherapy through effects on normal HSC and by altering the balance between normal and leukemic stem cells (LSCs). However, caution is required due to the demonstrated emergence of resistance within the leukemia stem cell compartment under sustained BET inhibition in recent preclinical models.

and exciting treatment option in many hematologic and solid malignancies, and may have novel promising effects in bone marrow transplantation. However, little is known about the effects of such treatment in normal tissues and an ongoing concern has been about the size of the therapeutic window. However, some reassurances are provided by the studies performed by Wroblewski et al. that have addressed this issue and suggest potentially beneficial effects on normal hematopoiesis; further studies are warranted to validate and extend these findings. However, a note of potential caution comes from models of BET inhibitor resistance that suggest that under the continuous and sustained selective pressures of BET inhibition, a small proportion of leukemic stem cells (LSC) survive and that over time these cells become the dominant clone.

Mechanistically, this appears to occur through the activation of the ancillary WNT/β-Catenin pathway that allows escape from BET inhibition. Importantly, resistance to BET inhibition in this model is only partially reversible on suspending I-BET.23 The evidence from the Fong et al. study,23 coupled with the evidence emerging from clinical trials and 920

early study of combination therapy, strengthens the argument for developing rational combination strategies of BET inhibitors and other agents for patients with hematologic and solid malignancies. The Wroblewski et al. study17 promises an unexpected but welcome positive effect on normal hematopoiesis.

References 1 Wu SY, Chiang CM. The double bromodomain-containing chromatin adaptor Brd4 and transcriptional regulation. J Biol Chem. 2007;282 (18):13141-13145.

2. Pivot-Pajot C, Caron C, Govin J, Vion A, Rousseaux S, Khochbin S. Acetylation-dependent chromatin reorganization by BRDT: a testisspecific bromodomain-containing protein. Mol Cell Biol. 2003;23(15): 5354-5365.

3. Taverna SD, Li H, Ruthenburg AJ, Allis CD, Patel DJ. How chromatinbinding modules interpret histone modifica- tions: lessons from professional pocket pickers. Nat Struct Mol Biol. 2007;14(11):1025-1040.

4. Dawson MA, Gudgin EJ, Horton SJ, et al. Recurrent mutations: including NPM1c, activate a BRD4- dependent core transcriptional program in acute myeloid leukemia. Leukemia. 2014;28(2):311-320.

5 Dawson MA, Prinjha RK, Dittmann A, et al. Inhibition of BET recruitment to chromatine as an effective treatment for MLL fusion

haematologica | 2018; 103(6)

Editorials

leukaemia. Nature. 2011;478(7370):529-533.

6. Zuber J, Shi J, Wang E, et al. RNA1 screen identifies Brd4 as a therapeutic target in Acute Myeloid Leukaemia. Nature. 2011;478(7370):524528.

7. Delmore JE, Issa GC, Lemieux ME, et al. BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell. 2011;146(6):904-917.

8. Mertz JA, Conery AR, Bryant BM, et al. Targeting MYC dependence in cancer by inhibiting BET bromodomains. Proc Natl Acad Sci USA. 2011;108(40):16669-16674.

9. Ott CJ, Kopp N, Bird L, et al. BET bromodomain inhibition targets both c-Myc and IL7R in high-risk acute lymphoblastic leukemia. Blood.

2012;120(14):2843-2852.

10. Berthon C, Raffoux E, Thomas X, et al. Bromodomain inhibitor OTX015 in patients with acute leukaemia: a dose-escalation, phase 1 study. Lancet Haematol. 2016;3(4):e186-195 11. Amorim S, Stathis A, Gleeson M, et al. Bromodomain inhibitor OTX015 in patients with lymphoma or multiple myeloma: a doseescalation, open-label, pharmacokinetic, phase 1 study. Lancet Haematol. 2016;3(4):e196-204.

12. Dawson M, Stein EM, Huntly BJP, et al. A Phase I Study of GSK525762, a Selective Bromodomain (BRD) and Extra Terminal Protein (BET) Inhibitor: Results from Part 1 of Phase I/II Open Label Single Agent Study in Patients with Acute Myeloid Leukemia (AML). Blood.

2017;130(Suppl 1):1377.

13. Fiskus W, Sharma S, Qi J, et al. Bet protein antagonist JQ1 is synergistically lethal with FLT3 tyrosine kinase inhibitor (TKI) and overcomes resistance to FLT3-TKI in AML cells expressing FLT-ITD. Mol Cancer Ther. 2014;13(10):2315-2327.

14. Fiskus W, Sharma S, Qi J, et al. Highly active combination of BRD4 antagonist and histone deacetylase inhibitor against human acute myelogenous leukemia cells. Mol Cancer Ther. 2014;13(5):1142-1154.

15. Herrmann H, Blatt K, Shi J, et al. Small-molecule inhibition of BRD4 as a new potent approach to eliminate leukemic stem- and progenitor cells in acute myeloid leukemia AML. Oncotarget. 2012;3(12):15881599.

16. Lovén J, Hoke HA, Lin CY, et al. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell. 2013;153(2):320-334.

17. Wroblewski M, Scheller-Wendorff M, Udonta F, et al. BET-inhibition

by JQ1 promotes proliferation and self-renewal capacity of hematopoietic stem cells. Haematologica 2018;103(6):939-948.

18. Roe JS, Mercan F, Rivera K, Pappin DJ, Vakoc CR. BET Bromodomain Inhibition Suppresses the Function of Hematopoietic Transcription Factors in Acute Myeloid Leukemia. Mol Cell. 2015;58(6):1028-1039.

19. Sandberg ML, et al. c-Myb and p300 regulate hematopoietic stem cell proliferation and differentiation. Dev Cell. 2005;8(2):153-166.

20. Trabucco SE, Gerstein RM, Evens AM, et al. Inhibition of bromodomain proteins for the treatment of human diffuse large B-cell lymphoma. Clin Cancer Res. 2015;21(1):113-122.

21. Bolden JE, Tasdemir N, Dow LE, et al. Inducible in vivo silencing of Brd4 identifies potential toxicities of sustained BET protein inhibition.

Cell Rep. 2014;8(6):1919-1929.

22. Sun Y, Wang Y, Toubai T, et al. BET bromodomain inhibition suppresses graft-versus-host disease after allogeneic bone marrow transplantation in mice. Blood. 2015;125(17):2724-2748.

23. Fong CY, Gilan O, Lam EY, et al. BET inhibitor resistance emerges from leukaemia stem cells. Nature. 2015;525(7570):538-542.

NSG-S mice for acute myeloid leukemia, yes. For myelodysplastic syndrome, no.

Emmanuel Griessinger,1,2 Michael Andreeff3,4 1

INSERM U1065, Mediterranean Centre for Molecular Medicine (C3M), Team 4 Leukemia: Molecular Addictions, Resistances & Leukemic Stem Cells, Nice, France; 2Faculté de Médecine, Université de Nice Sophia Antipolis, France; 3Section of Molecular Hematology and Therapy, Department of Leukemia, University of Texas MD Anderson Cancer Center, Houston, TX, USA and 4 Department of Stem Cell Transplantation and Cellular Therapy, University of Texas MD Anderson Cancer Center, Houston, TX, USA E-mail: emmanuel.griessinger@gmail.com or mandreef@mdanderson.org doi:10.3324/haematol.2018.193847

R

esearch on primary patient cells is a compelling challenge for scientists. Although initially limited to short experiments over hours or days, engrafting these primary human cells in immunodeficient mice today allows even more informative investigation to be carried out over weeks and months. This experiment is fascinating, probably first because it gives rise to personal and moral questions about the patient’s avatar. Also, in basic research, the xenograft is the model to be used to reveal the stemness properties of a certain population of cancer cells.1 Although today there are some ex vivo alternatives, the xenograft remains the gold standard technique to study cancer stem cells which are responsible for cancer initiation propagation, maintenance and evolution. Uncovering the presence of primary human leukemic cells in a sample of mouse tissue 10-16 weeks after injection, demonstrating the initial engraftment of leukemia initiating cells (LICs) causes an exhilarating sensation known to only a few lucky scientists. Absence of graft triggers the opposite sensation of complete disappointment, which has led several teams to focus their attention on this particular problem with the xenograft approach. In this issue of Haematologica, Krevvata et al. put forward fundamental new insights to haematologica | 2018; 103(6)

help improve xenograft of acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS).2 Myelodysplastic syndrome and acute myeloid leukemia are myeloid neoplasms that disrupt normal hematopoiesis. This group of myeloid leukemias could be considered as a continuum consisting of a multitude of different leukemias, including all possible myeloid abnormalities. This results in a wide range of severity and patient overall survival (OS). MDS patients have globally better OS than AML patients, and some MDS evolve inevitably towards AML. Interestingly, the first attempts at AML/MDS xenograft quickly revealed, through the repartition of samples engrafting and non-engrafting the mice, that the engraftment potential was perfectly linked with the aggressiveness of the leukemia, since AML samples are usually more easy to engraft than MDSs.3,4 Many independent studies have offered different reasons for engraftment failure, but none can satisfactorily explain it.

Possible explanations are either related to the host immune environment or to the defect of the grafted cells or to the graft and host compatibility.

The innate and adaptive immune response of the host environment is an obvious and very clear obstacle for the 921

Editorials

graft. This was the goal and the reports accross the different generation on immunodeficient recipient. Indeed previous comparisons of different models with increased immunodeficiency, NOD/SCID mice (non-obese diabetic/severe combined immunodeficiency mice), NOD/SCID beta2 (β2-microglobulin-deficient NOD-scid mice), NSG (NOD-Scid-IL-2Rgcnull) mice, showed that the degree of immunodeficiency clearly matters but is not the key to explain xenograft failure of some samples.

Fewer AML cells are necessary to initiate the graft and a better engraftment is reached using more immunodeficient recipients as compared with first generation NOD/SCID mice.5-7 However, changing the permissiveness of the recipient maintains the sample stratification in terms of engraftment potential. This means that samples with the potential to engraft better remain better engrafters and the poorer remain poorer. (In the context of this Editorial, non-engrafters remained mostly nonengrafters using different strains of immunodeficient mice.7) Thus, further increasing the immunodeficiency of the recipient would actually jeopardize the recipient viability without improving the overall engraftment rate.

Independently of the recipient used, it was shown that the xenograft potential of AML samples was linked to intrinsic properties of the cells injected. Engraftment failure is related to good prognosis AML and, inversely, xenograft potential is a poor prognosis marker.8,9 Paczulla et al.10 and our own team recently showed that increasing the incubation period from 10 to 30 weeks allows some successful leukemic xenografts of good prognosis-related samples incapable of engrafting NSG mice during a conventional 10-12 week period. Actually, these samples have a lower frequency of stem-progenitor cells associated with a lower expansion capacity ex vivo compared to poor prognosis-related samples’ cells efficiently engrafting NSG mice.11 These data suggest that the non-engrafter samples might just have a slower progression, and that the recipient residual immunity is not an insurmountable obstacle for these samples.

Eventually, beyond the recipient immunity or the grafted cells defect, the last explanation for xenograft failure is the lack of a human specific microenvironment support for some categories of leukemia samples. In the last decade, one strategy approaching the issue from two different directions was adopted to try to improve the compatibility of the animal for human cells. The approach had been to humanize the murine recipient either genetically (by forcing the expression of human cytokines) or by injecting the mice with cellular components of the human bone marrow microenvironment (BMME).

Different immunodeficient mouse strains expressing various human cytokines have been generated over the last decade and are reviewed by Theocharides et al.12 Among them, the NSG-S mice used by Krevvata et al.2 is an engineered strain, with knock-in for human stem cell factor (SCF), granulocyte-macrophage colony-stimulating factor (GM-CSF), and IL-3 in the background of the NSG recipient.13 Alternatively, humanizing recipient BMME is achieved by injection of stromal cells of the human BMME, such as mesenchymal stem cells (MSC), or endothelial cells, or osteoblast progenitors.14-16 For different reasons, intravenous (IV) or intra-bone marrow bolus 922

of stromal cells is less and less used for the benefit of humanized ectopic ossicle approaches, subcutaneously implanted by surgery with either matrigel scaffold, sponges, or ceramic seeded with human stromal cells.

These methods have certainly improved the situation since they recently resulted in successful engraftment of samples previously defined as non-engrafters, including good pronosis AML.15,16 However two disadvantages were reported for these models. First, these ectopic leukemic grafts of good prognosis-related samples were reported to not invade recipient bone marrows, thus actually limiting the size of the human leukemic population in this particular case.15,16 Secondly, some protocols are quite demanding to handle in routine lab practice, such as the pre-treatment of the ectopic niche with parathyroid hormone to favor the osteoblastic differentiation of the MSC prior to the introduction of the leukemic cells.15 Thus, direct IV injection will probably remain the most common in vivo protocol to explore the LIC compartment, and increasing the chances of successful engraftment in this setting is of particular interest.

Krevvata et al. show on a large cohort (n=77) of AML patients that 82% of AML samples engraft NSG-S recipient versus 50% in the NSG strain.2 Sixty-seven percent of non-engrafter AML in the NSG strain become engrafters in the NSG-S strain during a conventional incubation period. This was also true for good prognosis inv16 AML, which are core binding factor (CBF) mutated AML known to repeatedly fail xenograft procedure. NSG-S also presents the advantages of faster engraftment and a leukemic burden present in the peripheral blood similar to that of patients, allowing simple blood sampling for longitudinal monitoring. However, the downside of the NSG-S model is the management of the leukemia progression that reduces viability of the cohorts. Xenografted with the same sample in NSG-S mice die faster than in NSG. Although the swiftness and quantity of engraftment is clearly shown in this model, further comparative tests should investigate the quality of the graft, for example, to exclude LIC alteration and exhaustion. The Authors eventually found that 18% of the samples remain nonengrafters in NSG-S, providing opportunities for further investigation into graft failures. Engrafting good prognosis AML samples, including CBF-AML, is a big step forward, opening up opportunities for previously impossible investigation, such as identifying the phenotype of their LICs, or analyzing their in situ behavior in the endostoeum by intravital microscopy, or the possibility of comparison of clonal architecture and clonal evolution in vivo with poor prognosis AML samples. This model could also allow the in vivo comparison of drug resistance mechanisms of these two groups of patients on the condition of first determining whether NSG-S mice can support an induction regimen, as Farge et al. have recently shown for NSG mice.17 The second part of the study of Krevvata et al. reports negative results but is nonetheless equally important. The Authors have performed a deep analysis of low- and high-risk MDS sample engraftment in NSG-S mice with or without MSCs co-injection. MSCs from different origins were tested: healthy donor-derived MSC (normal), allogeneic patient-derived MSC (allo), or patient-derived haematologica | 2018; 103(6)

Editorials

autologous MSC (auto). These results are very interesting because previous publications from two other groups arrived at opposite conclusions.18-20 Krevvata et al. confirm the observation of Rouault-Pierre et al. and demonstrate that MDS only offer transient benefit from the cytokine stimulation in the NSG-S model, and actually tend to exhaust their engraftment level over time.2 It is also the first study presenting a comprehensive paired analysis of engraftment that clearly establishes that MDS engraftment is not enhanced by co-injection of MSC, in contrast to previous reports. Overall, this work suggests that improving the MDS xenograft model remains a key challenge. Further testing should be performed using other newly developed immunodeficient mouse models, such as the four genes encoding human cytokines MISTRG (M-CSFh/h IL-3/GM-CSFh/h hSIRPAtg TPOh/h Rag2-/Il2rg-/-) strain or NBSGW mice (mouse stem cell factor receptor mutated in the background of NSG), to eventually develop better MDS xenografts. In the perspective of this study, future investigations could explore and try to understand how and why an IL3, SCF and GM-SCF cytokine cocktail can be beneficial for supporting LICs of good prognosis AML but not of MDS, and how their distinct epigenetic regulators and DNA methylation patterns might be involved in this differential response.

Acknowledgments EG is supported by a grant from the Fondation de France.

References 1 Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997;3(7):730-737.

2. Krevvata M, Shan X, Zhou C, et al. Cytokines increase engraftment of human acute myeloid leukemia cells in immunocompromised mice but not engraftment of human myelodysplastic syndrome cells.

Haematologica. 2018;103(6):959-971 3. Martin MG, Welch JS, Uy GL, et al. Limited engraftment of low-risk myelodysplastic syndrome cells in NOD/SCID gamma-C chain knockout mice. Leukemia. 2010;24(9):1662-1664.

4. Muguruma Y, Matsushita H, Yahata T, et al. Establishment of a xenograft model of human myelodysplastic syndromes.

Haematologica. 2011;96(4):543-551.

haematologica | 2018; 103(6)

5. Ishikawa F, Livingston AG, Wingard JR, Nishikawa S, Ogawa M. An assay for long-term engrafting human hematopoietic cells based on newborn NOD/SCID/beta2-microglobulin(null) mice. Exp Hematol.

2002;30(5):488-494.

6. Feuring-Buske M, Gerhard B, Cashman J, Humphries RK, Eaves CJ, Hogge DE. Improved engraftment of human acute myeloid leukemia progenitor cells in beta 2-microglobulin-deficient NOD/SCID mice and in NOD/SCID mice transgenic for human growth factors. Leukemia.

2003;17(4):760-763.

7. Vargaftig J, Taussig DC, Griessinger E, et al. Frequency of leukemic initiating cells does not depend on the xenotransplantation model used.

Leukemia. 2012;26(4):858-860.

8. Pearce DJ, Taussig D, Zibara K, et al. AML engraftment in the NOD/SCID assay reflects the outcome of AML: implications for our understanding of the heterogeneity of AML. Blood. 2006;107(3):11661173.

9. Monaco G, Konopleva M, Munsell M, et al. Engraftment of acute myeloid leukemia in NOD/SCID mice is independent of CXCR4 and predicts poor patient survival. Stem Cells. 2004;22(2):188-201.

10. Paczulla AM, Dirnhofer S, Konantz M, et al. Long-term observation reveals high-frequency engraftment of human acute myeloid leukemia in immunodeficient mice. Haematologica. 2017;102(5):854-864.

11. Griessinger E, Anjos-Afonso F, Vargaftig J, et al. Frequency and Dynamics of Leukemia-Initiating Cells during Short-term Ex Vivo Culture Informs Outcomes in Acute Myeloid Leukemia Patients.

Cancer Res. 2016;76(8):2082-2086.

12. Theocharides AP, Rongvaux A, Fritsch K, Flavell RA, Manz MG. Humanized hemato-lymphoid system mice. Haematologica.

2016;101(1):5-19.

13. Wunderlich M, Mizukawa B, Chou FS, et al. AML cells are differentially sensitive to chemotherapy treatment in a human xenograft model.

Blood. 2013;121(12):e90-97.

14. Chen Y, Jacamo R, Shi Y, et al. Human extramedullary bone marrow in mice: a novel in vivo model of genetically controlled hematopoietic microenvironment. Blood. 2012;119(21):4971-4980.

15. Reinisch A, Thomas D, Corces MR, et al. A humanized bone marrow ossicle xenotransplantation model enables improved engraftment of healthy and leukemic human hematopoietic cells. Nat Med. 2016;22(7):812-821.

16. Abarrategi A, Foster K, Hamilton A, et al. Versatile humanized niche model enables study of normal and malignant human hematopoiesis. J Clin Invest. 2017;127(2):543-548.

17. Farge T, Saland E, de Toni F, et al. Chemotherapy-Resistant Human Acute Myeloid Leukemia Cells Are Not Enriched for Leukemic Stem Cells but Require Oxidative Metabolism. Cancer Discov.

2017;7(7):716-735.

18. Medyouf H, Mossner M, Jann JC, et al. Myelodysplastic cells in patients reprogram mesenchymal stromal cells to establish a transplantable stem cell niche disease unit. Cell Stem Cell. 2014;14(6):824837.

19. Rouault-Pierre K, Mian SA, Goulard M, et al. Preclinical modeling of myelodysplastic syndromes. Leukemia. 2017;31(12):2702-2708.

20. Rouault-Pierre K, Smith AE, Mian SA, et al. Myelodysplastic syndrome can propagate from the multipotent progenitor compartment.

Haematologica. 2017;102(1):e7-10.

923

EDITORIALS & PERSPECTIVES Defining the elusive boundaries of chronic active Epstein-Barr virus infection Sebastian Fernandez-Pol, Oscar Silva and Yasodha Natkunam Department of Pathology, Stanford University School of Medicine, CA, USA E-mail: yaso@stanford.edu doi:10.3324/haematol.2018.193714

T

he Epstein-Barr virus (EBV) infects more than 90% of people by early adulthood. While childhood EBV infection tends to be asymptomatic, adolescents and young adults often develop infectious mononucleosis (IM). With rare exceptions, IM usually resolves spontaneously. EBV is also implicated in the development of several hematolymphoid malignancies derived from B-, T- and natural killer (NK)-cells, including Burkitt lymphoma, classic Hodgkin lymphoma, extranodal NK-/T-cell lymphoma, nasal type (ENKTL), and immunodeficiency-associated lymphoproliferative disorders (LPD).1-5 Primarily because of their rarity, systemic EBV+ T-cell and NK-cell LPDs are not completely understood. Criteria to separate clinical risk groups at diagnosis are confounded by several important factors: 1) non-specific histopathological and immunophenotypic features that overlap with infectious and inflammatory conditions as well as with other T- and NK-LPDs; 2) lack of markers to reliably predict clinical behavior that ranges from self-limiting proliferations to those that are rapidly fatal; 3) molecular clonality, although often detected, is not necessarily indicative of malignancy; and 4) inconsistent association with hemophagocytic lymphohistiocytosis (HLH), whose presence is often life-threatening.1-3,6-11 Large, clinically well-annotated cohorts of patients offer a tremendous opportunity to investigate rare subtypes of EBV+ T- and NK-cell proliferations, define diagnostic and prognostic criteria, and explore disease boundaries. To that end, the paper by Kawamoto et al. in this issue of Haematologica addresses a rare type of EBV+ T- and NK-cell proliferation in adult patients with chronic active EBV-like features (adult-onset CAEBV).12 Chronic active EBV is characterized by EBV+ T- or NK-cell proliferations and includes systemic as well as cutaneous forms. Systemic CAEBV presents with fever, lymphadenopathy and splenomegaly, and typically develops in immunocompetent patients following primary EBV infection. The initial phase resembles an IM-like illness. These lymphoid proliferations may be polyclonal, oligoclonal or monoclonal, and have a propensity to evolve into a systemic EBV+ T- or NK-cell lymphoma. CAEBV most often occurs in children and adolescents without a history of immunodeficiency or autoimmunity. The clinical course and prognosis of CAEBV is highly variable: while some patients exhibit a protracted disease course, others experience a fulminant form of the disease accompanied by HLH.1-3,8-10 Chronic active EBV of T- and NK-cell types have a strong ethnic predisposition, and are most frequent in East Asia and in indigenous populations of Central and South America.1,6-11 It is rare in Western and African populations.1,13 The etiology of CAEBV is unknown, although susceptibility is thought to result from defective cytotoxic T-cell or NK-cell activity against EBV-infected cells. In addition to a defective host 924

immune response, EBV viral load is a key factor that impacts the severity of the disease.1,2 Rare CAEBV cases with EBV+ Bcell proliferations have also been reported in Asia and in the United States.2,13 Although CAEBV typically affects children and adolescents, rare cases of adult-onset CAEBV have been reported; these are associated with a worse prognosis.1,2 Kawamoto et al. describe a cohort of 54 patients with adult-onset CAEBV (defined as onset >15 years of age).12 For comparison, they utilized control groups of patients including pediatric onset CAEBV (n=75), and ENKTL of nasal (n=37) and non-nasal (n=45) types. The diagnostic criteria for CAEBV as defined in the recently revised World Health Organization classification include persistent IM-like symptoms for more than three months, increased EBV DNA (>102.5 copies/mg) in peripheral blood, histological evidence of organ disease, and EBV RNA or viral protein in affected tissues.1 All patients in the Kawamoto et al. study met these criteria. Patients with prior immunodeficiency, including HIV infection, were rigorously excluded. A median age of onset at 39 years (range 16-86 years) and the period from estimated onset to diagnosis of over one year typified this cohort. In addition, there was a bimodal age distribution for adult-onset CAEBV, a finding that had not been previously appreciated.

The EBV+ lymphoid infiltrates of CAEBV most often lack atypical features and may mimic non-specific inflammatory or infectious conditions. As such, a strong index of suspicion is necessary to perform appropriate testing for EBV DNA in peripheral blood and EBV RNA (in situ hybridization for EBER) in affected tissue to establish a diagnosis. T-cell CAEBV patients typically have high titers of IgG antibodies against EBV viral capsid antigen and early antigen, and demonstrate a worse prognosis when compared to NK-cell type CAEBV.1-3 In the Kawamoto et al. study,12 EBV DNA viral load was measured in peripheral blood as well as in tissue by in situ hybridization for EBER RNA. Lymphadenopathy was more frequent in patients with T-cell CAEBV, while skin involvement was more frequent in NKcell CAEBV. There were no significant differences in outcome between the T-cell and NK-cell cases among adultonset CAEBV patients. When patients aged over 50 years were analyzed separately, no significant differences were found in comparison to all adult-onset CAEBV cases.

Cutaneous CAEBV includes hydroa vacciniforme-like lymphoproliferative disorder (HV-LPD), which typically occurs in children. It is most often a cytotoxic CD8+ T-cell proliferation, although CD4, CD56 and CD30 are positive in subsets of cases. The clinical course is variable, and the severity may depend on photosensitivity.1,2,9,14-16 Pathological diagnosis may be confounded by the lack of distinctive histological features. Clinical correlation and testing for EBV is paramount to making the diagnosis. Severe allergy to moshaematologica | 2018; 103(6)

Editorials

quito bites is an NK-cell proliferation that results from local and systemic symptoms following a mosquito bite.

Immunophenotypic features, particularly CD30 expression in a subset may overlap with features of lymphomatoid papulosis.1,2,9,17 Recognition and timely diagnosis is important because patients with severe allergy to mosquito bites may progress to systemic NK-cell type CAEBV, aggressive NK-cell leukemia, or ENKTL.17 In comparison to pediatric onset CAEBV, the adult cohort analyzed by Kawamoto et al. had significantly lower frequency of fever and greater frequency of skin lesions, including HV-LPD and severe allergy to mosquito bites.12 Distinguishing CAEBV from more aggressive EBV+ Tand NK-cell proliferations can be challenging, particularly given the propensity of CAEBV to progress to a more

aggressive LPD. It is, however, important to recognize and accurately diagnose LPDs in this continuum because early intervention may offer the only solution for patients progressing toward a fulminant disease. Therefore, defining the boundary between CAEBV and more aggressive EBV+ T- and NK-LPDs, and the development of guidelines for the management of such patients, are urgently needed. In the Kawamoto et al. study, although no specific markers of disease severity or progression were identified, thrombocytopenia (platelets <100x109/L), high EBNA titer (â&#x2030;Ľ40), and HLH at initial diagnosis were associated with a worse overall clinical outcome. HLH was diagnosed using the HLH 2004 guidelines and was more frequent (46%) in the bone marrows of adult-onset CAEBV. Most importantly, allogeneic stem cell transplan-

Figure1. Spectrum of Epstein-Barr virus-positive (EBV+) T-cell and natural killer (NK)-cell lymphoproliferative disorders (LPD). Characteristic features of EBV+ T-cell and NK-cell LPD are shown. Hemophagocytic lymphohistiocytosis (HLH) in the bone marrow of a 3-year old Asian child with EBV infection. Core biopsy highlights HLH within bone marrow sinuses (A and B). Chronic active EBV infection causing an oral cavity lesion in an otherwise asymptomatic 14-year old Caucasian adolescent showing a polymorphous lymphoid infiltrate with EBV+ cells in an angiocentric distribution (C and D). Systemic EBV T-cell lymphoma in an otherwise asymptomatic 8-year old child presenting with persistent fever and chills. A CD56 stain highlights the atypical infiltrate and underscores the importance of testing for EBV in a patient with unexplained fevers (E and F). Aggressive NK leukemia replacing the bone marrow in a 42-year old man who presented with pancytopenia and massive hepatosplenomegaly. EBER in situ hybridization shows an exuberant EBV+ interstitial infiltrate in the bone marrow (G and H). Extranodal NK/T-cell lymphoma, nasal type causing an ulcerated palatal lesion in a 51-year old man from Guatemala. Prominent necrosis and angiodestruction characterize the atypical lymphoid infiltrate (I and J).

haematologica | 2018; 103(6)

925

Editorials

tation was found to be the most effective treatment choice for prolonged overall survival.12 Of particular interest in the study by Kawamoto et al.12 is their comparison of overall survival in patients with pediatric and adult-onset CAEBV with nasal and nonnasal ENKTL. Adult-onset CAEBV showed a worse overall survival in comparison to pediatric onset CAEBV and ENKTL. In fact, adult-onset CAEBV had an overall survival comparable to that of non-nasal ENKTL, which is generally more aggressive than its nasal counterpart.

In the pediatric age group, systemic EBV+ T-cell lymphoma of childhood, a de novo CD8+ T-cell lymphoma with an aggressive clinical course, must be separated from CAEBV.1-3,8 One key feature that helps distinguish CAEBV from systemic EBV+ T-cell lymphoma of child-

hood is that the T cells in CAEBV are predominantly CD4+ and less frequently of cytotoxic CD8 origin. In addition, CAEBV can show clinical and histological overlap with aggressive NK-cell leukemia and ENKTL. Age of onset, together with immunophenotypic differences help in the diagnosis: sCD3+, CD56-, CD8+ and/or CD4+ in systemic EBV+ T-cell lymphoma of childhood; sCD3-, cCD3+, CD16+, CD56dim, CD8- in aggressive NK-cell leukemia; and CD2+, CD5-, CD56+/-, cCD3+, EBER in ENKTL. The presence of T-cell clonality in systemic EBV+ T-cell lymphoma can further help distinguish that entity from aggressive NK-cell leukemia, which lacks T-cell clones.

Markers associated with NK-cell differentiation, such as CD16, can also be useful, as CD16 is positive in 75% of aggressive NK leukemia but usually absent in ENKTL.

Table 1. Characteristic features of Epstein-Barr virus (EBV)-positive (EBV+) T-cell and natural killer (NK)-cell lymphoproliferative disorders.

EBV-associated hemophagocytic lymphohistiocytosis (HLH) • Non-neoplastic proliferation typically seen in the pediatric age group • Acute presentation without prior history of immunodeficiency or EBV infection • Most common in patients of Asian descent • T-cell monoclonality is often detected but is not an indication of malignancy • Hyperbilirubinemia (>1.8 mg/dL) and hyperferritinemia (>20,300 ng/mL) at diagnosis are poor prognostic factors • HLH 2004 protocol is used for management; lack of response may necessitate bone marrow transplantation • May be self-limiting in some cases Chronic active EBV (CAEBV) infection • Cutaneous and systemic forms exist, although symptoms can overlap • Cutaneous CAEBV includes severe mosquito bite allergy and hydroa vacciniforme-like lymphoproliferative disorders • Systemic CAEBV presents with fever, lymphadenopathy and splenomegaly, and typically follows primary EBV infection • Most common in children and young adults • Most common in patients of Asian or Hispanic descent • Follows a chronic, relapsing clinical course with a risk of progression to systemic EBV+ T- or NK-cell lymphoma • CAEBV of T-cell origin has high IgG titers against EBV viral capsid and early antigens and a worse prognosis • Adult-onset CAEBV is associated with a worse prognosis Systemic EBV+ T-cell lymphoma • Acute, de novo T-cell lymphoma of EBV+ cytotoxic T cells • Most common in children and young adults • Most common in Asians and in indigenous populations of Central and South America • Most patients are immunocompetent • Most patients develop HLH • Follows a fulminant clinical course with poor response to chemotherapy and death within days to weeks of diagnosis Aggressive NK leukemia • Systemic neoplastic proliferation of NK cells frequently associated with EBV • Most commonly involves peripheral blood, bone marrow, liver and spleen • Most common in young to middle-aged adults with a median age of 40 years • Most common in patients of Asian descent • May evolve from CAEBV, particularly in younger patients • Follows a fulminant clinical course with multi-organ failure and poor response to chemotherapy, with a median survival of <2 months Extranodal NK/T cell lymphoma, nasal type • Extranodal lymphoma of EBV+ cytotoxic T or NK cells associated with angiodestruction and prominent necrosis • Affects immunocompetent as well as immunocompromised patients • Most commonly involves the upper aerodigestive tract, skin, soft tissue, gastrointestinal tract and testes • Most common in adults with a median age of 44-54 years • Most common in Asians and in indigenous populations of Central and South America • Clinical course is variable: some respond well to chemotherapy and up-front radiation while others succumb to disseminated disease • Advanced stage, invasion of bone or skin, EBV+ cells in the bone marrow and a high proliferation index are unfavorable prognostic factors • Extranasal disease is highly aggressive and is refractory to treatment • Some cutaneous cases may have a protracted clinical course Nodal peripheral T-cell lymphoma, EBV+ • Other EBV+ peripheral T-cell lymphomas (not further discussed in the report)

926

haematologica | 2018; 103(6)

Editorials

Recently, recurring somatic mutations, including DDX3X, STAT3, STAT5B, JAK3, and TP53, have been identified in ENKTL.18-20 Mutational profiles in other subtypes of EBV+ T- and NK-cell LPDs will be necessary to translate these discoveries for diagnostic purposes. It is hoped that clinically well-annotated cohorts, such as that described by Kawamoto et al., will be used to define molecular profiles in future studies.

In summary, Kawamoto et al. present a comprehensive characterization of a rare EBV+ T- and NK-cell LPD that occurs primarily in adults and has features of CAEBV.12 As is typical of rare diseases, precise diagnostic criteria and classification can be challenging unless sufficient numbers of patients become available for study. By providing a relatively large cohort of patients with detailed clinical annotation, this paper serves to extend knowledge in the field, and raises the possibility that adult-onset CAEBV may be among the most aggressive of EBV+ T- and NKcell LPDs. It remains to be seen if future molecular characterization of adult-onset CAEBV is likely to offer insights that will allow accurate and early diagnosis and lead to improved outcomes for these patients.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

References 1 Swerdlow SH, Campo E, Harris NL, et al. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. (4th ed.) IARC, Lyon, 2017.

2. Kimura H, Ito Y, Kawabe S, et al. EBV-associated T/NK-cell lymphoproliferative diseases in nonimmunocompromised hosts: prospective analysis of 108 cases. Blood. 2012;119(3):673-686.

3. Swerdlow SH, Jaffe ES, Brousset P, et al. Cytotoxic T-cell and NK-cell lymphomas: current questions and controversies. Am J Surg Pathol.

2014;38(10):60-71.

4. Natkunam Y, Gratzinger D, de Jong D, et al. Immunodeficiency and Dysregulation: Report of the 2015 Workshop of the Society for Hematopathology/European Association for Haematopathology. Am J Clin Pathol. 2017;147(2):124-128.

5. Gratzinger D, de Jong D, Jaffe ES, et al. T- and NK-Cell Lymphomas and Systemic Lymphoproliferative Disorders and the Immuno-

haematologica | 2018; 103(6)

16.

17.

18.

19.

20.

deficiency Setting: 2015 SH/EAHP Workshop Report - Part 4. Am J Clin Pathol. 2017;147(2):188-203.

Kikuta H, Sakiyama, Y, Matsumoto S, et al. Fatal Epstein-Barr virus associated hemophagocytic syndrome. Blood. 1993;82(11):3259-3264.

Su IJ, Chen RL, Lin DT, et al. Epaterin-Barr virus (EBV) infects T lymphocytes in childhood EBV-associated hemophagocytic syndrome in Taiwan. Am J Pathol. 1994;144(6):1219-1225.

Quintanilla-Martinez L, Kumar S, Fend F, et al. Fulminant EBV(+) T-cell lymphoproliferative disorder following acute/chronic EBV infection: a distinct clinicopathologic syndrome. Blood. 2000;96(2):443-451.

Kimura H, Hoshino Y, Kanegane H, et al. Clinical and virologic characteristics of chronic active Epstein-Barr virus infection. Blood.

2001;98(2):280-286.

Suzuki K, Ohshima K, Karube K, et al. Clinicopathological states of Epsterin-Barr virus associated T/NK-cell lymphoproliferative disorders (severe chronic active EBV infection) of children and young adults. Int J Oncol. 2004;24(5):1165-1174.

Suzuki R, Suzumiya J, Nakamura S, et al. Aggressive natural killer-cell leukemia revisited: large granular lymphocyte leukemia of cytotoxic NK cells. Leukemia. 2004;18(4):763-770.

Kawamoto K Miyoshi H, SuzukiT, et al. A distinct subtype of Epstein Barr virus positive T/NK cell lymphoproliferative disorder: Adult patients with Epstein Barr virus infection-like features. Haematologica.

2018;103(6):1018-1028.

Cohen JI, Jaffe ES, Dale JK, et al. Characterization and treatment of chronic active Epstein-Barr virus disease: a 8-year expereince in the United States. Blood. 2011;117(22):5835-5849.

Iwatsuki K, Xu Z, Takata M, et al. The association of latent EpsteinBarr virus infection with hydroa vacciniforme. Br J Dermatol.

1999;140(4):715-721.

Gupta G, Man I, Kemmett D. Hydroa vacciniforme: A clinical and follow-up study of 17 cases. J Am Acad Dermatol. 2000;42(2 Pt 1):208213.

Quintanilla-Martinez L, Ridaura C, Nagl F, et al. Hydroa vacciniformelike lymphoma: a chronic EBV+ lymphoproliferative disorder with risk to develop a systemic lymphoma. Blood. 2013;122(18):3101-3110.

Kempf W, Kazakov D V, Schärer L, et al. Angioinvasive lymphomatoid papulosis: a new variant simulating aggressive lymphomas. Am J Surg Pathol. 2013;37(1):1-13.

Koo GC, Tan SY, Tang T, et al. Janus kinase 3-activating mutations identified in natural killer/T-cell lymphoma. Cancer Discov.

2012;2(7):591–597.

Küçük C, Jiang B, Hu X, et al. Activating mutations of STAT5B and STAT3 in lymphomas derived from gδ-T or NK cells. Nat Commun.

2015;6:6025.

Jiang L, Gu Z-H, Yan Z-X, et al. Exome sequencing identifies somatic mutations of DDX3X in natural killer/T-cell lymphoma. Nat Genet.

2015;47(9):1061-1066.

927

PERSPECTIVES Present results and future perspectives in optimizing chronic myeloid leukemia therapy Angelo M. Carella,1 Giuseppe Saglio,2 Xavier F. Mahon3 and Michael J. Mauro4 1

Hematology and BMT Unit, Ospedale Policlinico San Martino, Genova,Italy; 2Dipartimento di Scienze Cliniche e Biologiche Università di Torino and Ospedale Mauriziano di Torino, Italy; 3Bergonié Cancer Institute, University of Bordeaux, France and 4Myeloproliferative Neoplasms Program, Memorial Sloan Kettering Cancer Center, New York, NY, USA E-mail: angelomcarella@gmail.com doi:10.3324/haematol.2017.182022

T

he presence of the Philadelphia chromosome and BCR/ABL1 fusion, combined with an elevated leukocyte count and other less specific clinical and hematological features, defines chronic myeloid leukemia (CML). In the last two decades, tyrosine kinase inhibitors (TKIs) have revolutionized CML treatment.1 The challenge now is to eradicate the disease. Three main crucial questions remain: first, which TKI should be used as first-line therapy, given the ‘embarrassment of riches’ with regard to the choices; second, who should discontinue TKI, and when; and last, what about the future?

Which TKI should be chosen as first–line therapy?

Being the first TKI to receive approval for the treatment of CML, imatinib has definitely changed the natural history of this disease. The recently published update of the IRIS trial shows that patients assigned to receive imatinib 400 mg per day have an estimated overall survival (OS) rate of 83.3% at 10 years, almost matching that of a control population without CML.2 However only approximately half of the originally enrolled patients (48.3%) completed study treatment with imatinib with such a long follow-up; discontinuation occurred in 15.9% of patients because of insufficient therapeutic effect and in 6.9% of cases because of adverse events; in addition, during study treatment 6.9% of the patients progressed to accelerated or blast crisis. Most of the patients who discontinued imatinib and who were not transplanted were subsequently moved to alternate treatment.

As it is well known that the final outcome in terms of OS and progression free survival (PFS) of CML patients correlates to the depth and rapidity of the cytogenetic and molecular responses achieved, the European LeukemiaNet (ELN) recommendations have established molecular and cytogenetic parameters to be achieved by patients at specific timepoints after the start of TKI therapy. If these parameters are not achieved, a switch of TKI, if possible, is recommended.3 The second-generation TKI dasatinib, initially approved in 2006 as second-line treatment for patients resistant or intolerant to imatinib,4 was finally approved as a first-line therapy in 2010 following the results of the phase 3 DASISION study.5 In this study the cumulative one-year major molecular response (MMR; 0.1% BCR-ABL) rate was 46% for dasatinib and 28% for imatinib. At five years, the cumulative MMR rate remained higher with dasatinib (76%) than with imatinib (64%). Whereas only 16% of the patients treated with dasatinib failed to achieve, at three months, early molecular response (EMR; a threshold of BCR-ABL is seen to correlate with OS and PFS) this percentage was 36% with imatinib.6 Finally, very deep molecular responses (DMR; 928

MR4, MR4.5), which were of greater interest due to the understandable desire of many patients to achieve treatment free remission (TFR), were definitely higher with dasatinib than with imatinib (MR4.5 42% vs. 32%). However, the five-year OS and PFS rates for dasatinib were 91% and 85%, respectively, and did not differ from that observed with imatinib; in addition, dasatinib therapy was associated with risk of developing pleural effusion.6 Nilotinib is a second-generation TKI which was accepted in 2007 for the treatment of CML resistant or intolerant to imatinib.7 It was approved for first-line treatment of chronic phase CML in 2010, following the positive results from the phase 3 ENESTnd study.8 In the aforementioned trial, after a minimum follow-up of five years, the rates of MMR and MR4.5 continue to be significantly higher in both nilotinib arms versus the imatinib arm (MMR: 77 and 77.2% versus 60%; MR4.5: 53.5 and 52.3% versus 31.4%), with more than half of the nilotinib-treated patients achieving MR4.5 by five years. Comparing nilotinib 300 mg twice daily (BID) with imatinib 400 mg daily, several differences emerge at the three month landmark on therapy: 91% of nilotinib-treated versus 67% of imatinib-treated patients achieved BCR-ABL transcript levels ≤10% EMR; 56% of nilotinib versus only 16% of imatinib patients achieved BCR-ABL transcript levels ≤1%. Although rates of freedom from progression to accelerated phase and blastic phase (AP/BC) remain statistically higher in the nilotinib-treated patients (96.3% and 97.8% for nilotinib versus 92.1% for imatinib), the estimated rate of OS is statistically superior only for nilotinib 400 mg BID arm patients compared to imatinib. Importantly, the occurrence of metabolic changes such as worsening glycemic control and lipid increase as well as cardiovascular events (CVEs) increasing over time with follow-up has been more frequently observed in both nilotinib arms. Although mainly observed in patients with an increased Framingham risk score, predictive of CVEs,9 increased attention to cardiovascular risk assessment and comorbidities for all CML patients is warranted.

Frontline use of bosutinib was initially investigated in the phase 3 BELA trial, but this study failed to meet its primary endpoint of complete cytogenetic response (CCyR) at 12 months.10 Bosutinib has been subsequently re-investigated versus imatinib for chronic phase CML patients at a dosage of 400mg per day in the phase 3 BFORE trial. In this study, the proportion of patients who achieved MMR at 12 months (primary endpoint) was greater with regard to statistical significance in the bosutinib group, compared with the imatinib group: 47.2 versus 36.9 percent, respectively.10 Deeper molecular responses were also higher in the bosutinib group haematologica | 2018; 103(6)

Perspective

compared to the imatinib group: MR4, 20.7% vs. 12.0% (P=0.01) and MR4.5, 8.1% vs. 3.3% (P=0.02). Such findings confirm the efficacy of second-generation TKIs and their ability to induce faster and deeper molecular responses relative to that observed with imatinib.11 In conclusion, imatinib has indeed changed the landscape of CML. Subsequently developed TKIs, dasatinib, nilotinib, and bosutinib, are potential alternatives to imatinib as first-line therapy, mainly due to the deeper and faster molecular responses they induce. In our opinion, initial treatment with second-generation TKIs should be offered initially to patients with higher risk of progression. Alternatively, imatinib could be the initial therapy for all patients, with the incorporation of early switch to second-generation TKIs if optimal response is not achieved at three months.

Cessation of treatment Until now the recommendation for TKI therapy in CML was to continue treatment indefinitely. However, there are numerous justifiable reasons for stopping TKI therapy. Off-target effects of TKIs and severe adverse drug reactions have been increasingly reported. These side effects may not only impair the quality of life, but some of them, such as pulmonary arterial hypertension, pleural effusion, or vascular occlusive events may potentially modify life expectancy. In addition, it is forbidden to administer TKIs to pregnant women, and experience in pediatric CML cases reveals growth disruption resulting from TKI therapy. The patientsâ&#x20AC;&#x2122; requests are also important; the question of whether TKI therapy is necessary lifelong is frequently asked.

Clinical trials have demonstrated the feasibility of stopping TKIs in patients with durable and deep MR beyond MMR.12-15 The convincing results of all of these studies have validated the concept of TFR, which has increasingly become the main focus of clinical trials in CML.16 The sine qua non condition for proposing TKI cessation is the achievement of a sustained DMR. A certified laboratory is necessary to perform and validate the robust molecular monitoring needed for safety during TFR studies. Great reassurance is to be found in the reproducibility of TFR studies over time, and with the adaption of more pragmatic and applicable criteria for patient consideration.17 Half of the patients who are eligible for TKI discontinuation remain treatment-free, while the other 50% recover optimal response upon therapy re-introduction. Most of the molecular recurrence occurs in the first six months following TKI cessation. One cannot overstate the importance of the safety observed to date in TFR trials, evidenced by preserved TKI sensitivity and prompt re-induction of molecular response in patients rechallenged after molecular recurrence. Of interest is a peculiar transitory TKI withdrawal syndrome reported in a minority of patients; the mechanism is unknown.18 A key question emerging from the experience with TFR trials relates to the evidence of persisting leukemic cells (e.g., persistent BCR-ABL detection) without exhibiting true relapse, and how this status should be defined (functional cure?).

The number of patients who are stopping TKI treatment is increasing over time and it may be possible to reattempt TKI cessation more than once in the same haematologica | 2018; 103(6)

patient. Encouraging results have emerged from a multicenter study entitled RE-STI, where eligible patients, i.e., those with a second sustained DMR, remained in TFR in one third of cases.19 Although extensive long-term experience is limited, substantial knowledge accumulated during the last years justifies moving TFR strategies from research to clinical practice.

The future of CML The future of CML therapy must answer those questions pending from the torrid pace of advance over the last 20 years. While treatment choices are many, selecting the safest and optimal path to cure still needs perfection.

One point oft forgotten is the longitudinal cost of TKI therapy in varying health care systems; with movement from indefinite to, ideally, defined duration therapy, this debate will evolve. Continued development of more precise prognostic information derived at diagnosis, such as BCR-ABL fusion type,20 the role of cytogenetic abnormalities aside from the Ph chromosome,21 and the formulation of an idealized risk score best able to predict outcome, including that of survival,22 is ongoing. It is possible that a more meticulous scrutiny of early response and a more individualized assessment of response will better aid decision making regarding any need for change in treatment. Particular attention must be paid to comorbid conditions, especially cardiovascular disease, both at diagnosis and with therapy change, given the available data on the increase in risk and impact of comorbidities on outcome.23 During therapy, the past had us focus on the quantitation of BCR-ABL â&#x20AC;&#x2DC;burdenâ&#x20AC;&#x2122; by conventional means (cytogenetic) whereas the present is sharply focused on the molecular assessment of disease burden. Such assessment increasingly utilizes next-generation sequencing, allowing for increased sensitivity, clarity regarding ABL kinase domain mutations, and clonal hierarchy. This technique may open the door to a broader consideration of molecular changes during the course of CML, including the impact of non-BCR-ABL clonal markers and clonal hematopoiesis on CML and other comorbid condition risks.

Therapy should continue to evolve, with advances including non-ATP site allosteric inhibitors with ABL001 (asciminib) showing great promise in single-agent phase I data,24 bolstered by its ability to be safely combined with available TKIs for potential synergy. In addition, later generation ABL kinase inhibitors have also moved into early clinical development, including two agents with expected activity against the T315I mutation, namely PF114 (Fusion Pharma) and K0706 (Sun Pharma Advanced Research Company [SPARC]). Unmet needs continue to consist of the treatment of advanced phase disease, with novel options emerging for lymphoid transformation, and more needed for myeloid blast phase. In addition, particular focus is being concentrated on novel agents able to either deepen suboptimal molecular response to TKIs or facilitate a second TFR. While the future is very bright for those diagnosed with CML, risks remain, and success requires informed choice, careful navigation of adverse events and response mile929

Perspective

stones, prompt recognition of progressive disease, continued utilization of allografting in the proper settings, and prudent and timely selection of candidates for treatment cessation.

References 1 Carella AM, Branford S, Deininger M, et al. What challenges remain in Chronic Myeloid Leukemia research? Haematologica. 2013: 98(8): 1168–1172.

2. Hochhaus A, Larson RA, Guilhot F, et al. IRIS Investigators. Long-term outcomes of imatinib treatment for chronic myeloid leukemia. N Engl J Med. 2017;376(10):917-927.

3. Baccarani M, Deininger MW, Rosti G, et al. European LeukemiaNet recommendations for the management of chronic myeloid leukemia.

Blood, 2013;122(6):872-874 4. Shah NP, Rousselot P, Schiffer C, et al. Dasatinib in imatinib-resistant or -intolerant chronic-phase, chronic myeloid leukemia patients: 7year follow-up of study CA180-034. Am J Hematol. 2016;91(9):869874.

5. Kantarjian H, Shah NP, Hochhaus A, et al. Dasatinib versus imatinib in newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med. 2010;362(24):2260-2270.

6. Cortes JE, Saglio G, Kantarjian HM, et al. Final 5-year study results of DASISION: the dasatinib versus imatinib study in treatment-naïve chronic myeloid leukemia patients trial. J Clin Oncol.

2016;34(20):2333-2340.

7. Kantarjian HM, Giles F, Gattermann N, et al. Nilotinib (formerly AMN107), a highly selective BCR-ABL tyrosine kinase inhibitor, is effective in patients with Philadelphia chromosome-positive chronic myelogenous leukemia in chronic phase following imatinib resistance and intolerance. Blood. 2007;110(10):3540-3546.

8. Hochhaus A, Saglio G, Hughes TP, et al. ENESTnd Investigators.

Nilotinib versus imatinib for newly diagnosed chronic myeloid leukemia. N Engl J Med. 2010;362(24):2251-2259.

9. Hochhaus A, Saglio G, Hughes TP, et al. Long-term benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial.

Leukemia. 2016;30(5):1044-1054.

10. Cortes JE, Kim DW, Kantarjian HM, et al. Bosutinib versus imatinib in newly diagnosed chronic-phase chronic myeloid leukemia: results from the BELA trial. J Clin Oncol. 2012;30(28):3486-3492.

11. Cortes JE, Gambacorti-Passerini C, Deininger MW, et al. Bosutinib (BOS) versus imatinib (IM) for newly diagnosed chronic myeloid leukemia (CML): initial results from the BFORE trial. Abstract #7002.

Presented at the 2017 American Society of Clinical Oncology Annual Meeting, 2017; Chicago, Illinois.

930

12. Rousselot P, Huguet F, Rea D, et al. Imatinib mesylate discontinuation in patients with chronic myelogenous leukemia in complete molecular remission for more than 2 years. Blood. 2007;109(1):58-60.

13. Mahon FX, Réa D, Guilhot J, et al. Intergroupe Français des Leucémies Myéloïdes Chroniques. Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre Stop Imatinib (STIM) trial. Lancet Oncol. 2010;11(11):1029-1035.

14. Ross DM, Branford S, Seymour, et al. Patients with chronic myeloid leukemia who maintain a complete molecular response after stopping imatinib treatment have evidence of persistent leukemia by DNA PCR. Leukemia. 2010;24(10):1719-1724.

15. Saußele S, Richter J, Hochhaus A, Mahon FX. The concept of treatment-free remission in chronic myeloid leukemia. Leukemia.

2016;30(8):1638-1647.

16. Etienne G, Guilhot J, Rea D, et al. Long-term follow-up of the French Stop Imatinib (STIM1) study in patients with chronic myeloid leukemia. J Clin Oncol. 2017;35(3):298-305.

17. Imagawa J, Tanaka H, Okada M, et al. DADI Trial Group.

Discontinuation of dasatinib in patients with chronic myeloid leukaemia who have maintained deep molecular response for longer than 1 year (DADI trial): a multicentre phase 2 trial. Lancet Haematol.

2015;2(12):e528-535.

18. Richter J, Söderlund S, Lübking A, et al. Musculoskeletal pain in patients with chronic myeloid leukemia after discontinuation of imatinib: a tyrosine kinase inhibitor withdrawal syndrome? J Clin Oncol.

2014;32(25):2821-2823.

19. Legros L, Nicolini FE, Etienne G, et al. French Intergroup for Chronic Myeloid Leukemias. Second tyrosine kinase inhibitor discontinuation attempt in patients with chronic myeloid leukemia. Cancer.

2017;123(22):4403-4410.

20. Jain P, Kantarjian H, Patel KP, et al. Impact of BCR-ABL transcript type on outcome in patients with chronic-phase CML treated with tyrosine kinase inhibitors. Blood. 2016;127(10):1269-1275.

21. Alhuraiji A, Kantarjian H, Boddu P, et al. Prognostic significance of additional chromosomal abnormalities at the time of diagnosis in patients with chronic myeloid leukemia treated with frontline tyrosine kinase inhibitors. Am J Hematol. 2018;93(1):84-90.

22. Hoffmann VS, Baccarani M, Hasford J, et al. Treatment and outcome of 2904 CML patients from the EUTOS population-based registry.

Leukemia. 2017;31(3):593-601.

23. Saussele S, Krauss MP, Hehlmann R, et al. Schweizerische Arbeitsgemeinschaft für Klinische Krebsforschung and the German CML Study Group. Impact of comorbidities on overall survival in patients with chronic myeloid leukemia: results of the randomized CML study IV. Blood. 2015;126(1):42-49.

24. Hughes TP, Goh Y-T, Ottmann OG, et al. Expanded phase 1 study of ABL001, a potent, allosteric inhibitor of BCR-ABL, reveals significant and durable responses in patients with CML-chronic phase with failure of prior TKI therapy. Blood. 2016;128(22):625.

haematologica | 2018; 103(6)

REVIEW ARTICLE

Morphological, immunophenotypic, and genetic features of chronic lymphocytic leukemia with trisomy 12: a comprehensive review

Ferrata Storti Foundation

Francesco Autore,1 Paolo Strati,2 Luca Laurenti1 and Alessandra Ferrajoli2

Hematology Institute, Catholic University of the Sacred Heart, Fondazione Policlinico A. Gemelli, Rome, Italy and 2Department of Leukemia, The University of Texas, MD Anderson Cancer Center, Houston, TX, USA 1

ABSTRACT

C

hronic lymphocytic leukemia is an extremely heterogeneous disease and prognostic factors such as chromosomal abnormalities are important predictors of time to first treatment and survival.

Trisomy 12 is the second most frequent aberration detected by fluorescence in situ hybridization at the time of diagnosis (10-25%), and it confers an intermediate prognostic risk, with a median time to first treatment of 33 months and a median overall survival of 114 months. Here, we review the unique morphological, immunophenotypic, and genetic characteristics of patients with chronic lymphocytic leukemia and trisomy 12.

These patients carry a significantly higher expression of CD19, CD22, CD20, CD79b, CD24, CD27, CD38, CD49d, sIgM, sIgk, and sIgÎť and lower expression of CD43 compared with patients with normal karyotype. Circulating cells show increased expression of the integrins CD11b, CD18, CD29, and ITGB7, and of the adhesion molecule CD323.

Patients with chronic lymphocytic leukemia and trisomy 12 frequently have unmutated IGHV, ZAP-70 positivity, and closely homologous stereotyped B-cell receptors. They rarely show TP53 mutations but frequently have NOTCH1 mutations, which can be identified in up to 40% of those with a rapidly progressive clinical course.

Introduction Chronic lymphocytic leukemia (CLL), the most prevalent adult leukemia in the Western world, is characterized by the progressive accumulation of mature B cells, and its clinical course is very heterogeneous.1-5 Therefore, the identification of prognostic and predictive factors for CLL is critical and this is a field of active investigation.6-12 Genetic abnormalities in particular have the potential to serve as prognostic factors in CLL. Until recently, the low mitotic index of most CLL cells made the use of metaphase cytogenetics difficult, even when metaphases could be generated, because the cellularity was often poor, and abnormalities were detected in only 4050% of cases. Recently, analyzable karyotypes have been obtained in patients with CLL using mitogenic stimulation of CLL specimens with CpG oligonucleotides.13-15 Furthermore, the analysis of aberrant chromosomal regions using specific DNA probes through fluorescence in situ hybridization (FISH) has made it possible to detect clonal aberrations in more than 80% of CLL patients.11,16,17 The most common chromosomal abnormalities identified by FISH in CLL are 13q14 deletion (del13q), trisomy 12 (+12), 11q23.3-q23.1 deletion (del11q), and 17p13 deletion (del17p).11,16,18-22 A recently published prospective cohort study of 1494 patients across 199 US centers has shown that del13q is present in approximately 46% of CLL cases, and +12 is present in 21% of CLL cases. Del17p and del11q are observed in fewer cases: 12% and 18% of CLL cases, respectively.23 The frequency of detection of each chromosomal abnormality is influenced by several factors, such as the methods and probes used, the frequency of neoplastic B cells in the specimen analyzed, and the cohort of patients investigated.19,20 haematologica | 2018; 103(6)

Haematologica 2018 Volume 103(6):931-938

Correspondence: aferrajo@mdanderson.org

Received: January 17, 2018.

Accepted: April 16, 2018.

Pre-published: May 10, 2018. doi:10.3324/haematol.2017.186684 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6931 Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

931

F. Autore et al.

Here, we review the unique morphological, immunophenotypic, and genetic characteristics of patients with CLL carrying +12.

Cytogenetic abnormalities concomitant with +12 +12 is the second most frequent cytogenetic abnormality identified by FISH in patients with CLL, occurring in 16% of these patients at the time of initial evaluation.18 When identified by FISH, +12 is the sole aberration in approximately 70% of +12 CLL cases; +12 has been associated with del13q, del11q, or del17p in 18%, 8%, and 4% of +12 CLL cases, respectively.18,19,22,24 When identified by chromosome banding analysis, +12 is the sole abnormality in 30% of +12 CLL cases, but it can be associated with trisomy 18 (5% of cases) or del14q (3% of cases); associations with t(14;19)(q32;q13), trisomy 19, del17p, del13q, and del11q have also been reported.11,25-27 +12 is even more frequent in patients with small lymphocytic leukemia than in those with CLL (28-36% vs.

13-16%).28,29 In fact, increased +12 frequency and reduced del13q frequency may represent specific features of small lymphocytic lymphoma with a different prognosis and may distinguish it from classical CLL.28 +12 as a single aberration in CLL confers an intermediate prognostic risk, with a median time to first treatment (TTFT) of 33 months and a median overall survival time (OS) of 114 months.18 +12 that is associated with additional chromosomal abnormalities, including t(14;18)(q32;q21) and trisomy 18, portends a poor prognosis independent of NOTCH1 mutation, which is more common in cases with isolated +12.30 Patients with CLL who develop Richter syndrome have a particularly poor prognosis, and +12 is frequent in patients with this syndrome, with an incidence of up to 50%.31 Yi et al. have retrospectively analyzed 330 non-selected CLL patients, using a panel DNA probe to detect cytogenetic abnormalities by FISH, 70 patients were positive for +12. In patients who carried a favorable or neutral cytogenetic abnormality, including +12, the co-occurrence of several chromosome abnormalities associated with shorter TTFT and OS. However, in patients with adverse cytogenetic abnormalities, such as del17p or del11q, the co-occurrence of a favorable or neutral clone, including again +12, portended an improved prognosis. This study challenged the classical hierarchical FISH classification used in CLL, highlighting the prognostic relevance of the intra-tumoral cytogenetic heterogeneity.32 +12 is not related to a specific abnormality. However, +12 as a 'driver' mutation can facilitate the appearance of NOTCH1 and FBXW7 mutations: disruptions of FBXW7 function may lead to constitutional activation of NOTCH1 and then to cell proliferation and evasion from apoptosis. The presence of abnormalities of FBXW7 may be implicated in the pathogenesis of the CLL and determine the selection of treatment-resistant clones.33,34

Cytopenia in patients with +12 CLL Cytopenias are common in patients with +12 CLL. It has been reported that up to 24% of patients with +12 CLL will develop cytopenias during the course of their disease.35 Cytopenias in CLL can result from either bone marrow failure or autoimmune disease.36,37 Two studies showed that CLL patients with +12 have a higher incidence of autoimmune cytopenias than infiltrative cytopenias.36,37 932

Morphological features of patients with +12 CLL Atypical morphology, defined as the presence of cleaved nuclei and/or lymphoplasmacytoid features in more than 15% of cells, can be observed in up to 20% of patients with CLL.38,39 Several studies have revealed that +12 is the most common cytogenetic abnormality in these cases.38 Athanasiadou et al. evaluated the presence of atypical and typical CLL morphology in a population of 100 CLL patients.38 In the +12 subgroup, 17 of the 23 cases analyzed (74%) were considered morphologically atypical.

Atypical morphology was also frequent in the del11q/del17p subgroup (5 of 7 cases, 71%). All del13q cases and 45 of 51 cases (88%) with normal karyotypes had typical morphology.38 Ruptured B cells, or smudge cells, are commonly seen in blood smears of patients with CLL.40 Levels of smudge cells have been correlated with low CD45 expression on CLL cells by flow cytometry.41 Smudge cells are believed to reflect intrinsic CLL cell fragility and cytoskeletal abnormalities, linked to reduced expression of vimentin, and some authors have proposed using percentages of smudge cells to stage CLL.42 In patients with +12 CLL, fewer smudge cells are observed than in patients with CLL overall, along with normal levels of CD45 expression.43

Immunophenotype of patients with +12 CLL Chronic lymphocytic leukemia patients show an immunophenotypic panel that is typically CD5+, CD23+, FMC7-, surface immunoglobulin (sIg)dim, CD22dim-, and CD79bdim-. Quijano et al.19 established the presence of an atypical immunophenotypic pattern in some subgroups of 180 CLL patients by using combinations of monoclonal antibodies. The atypical immunophenotypic pattern was characterized by a distinct and unusual pattern of expression regarding CD22, CD20, CD79d, FMC7, sIgM, sIgj, sIgk, and CD38. Cases with +12 (23% of the series), compared with cases with normal karyotypes (38%), showed significantly higher expression of CD19, CD22, CD20, CD79b, CD24, CD27, CD38, sIgM, sIgk, and sIgÎť and lower expression of CD43. Among various cytogenetic subgroups, the atypical immunophenotypic pattern identified in overall CLL patients was more frequent in the +12 and del17p subgroups and less frequent in the del13q subgroup; the del11q subgroup presented an equal distribution of the typical and atypical patterns.19 An atypical immunophenotype can be defined using the Matutes score,39 which is calculated from the following cell surface markers: positive expression of CD5 and CD23, negative expression of FMC7, and weak expression of CD22 and sIg. A modification of the scoring system, called the modified Matutes score, replaces CD22 with CD79b, and this score is considered atypical if less than 4.39 Interestingly, with the modified Matutes score, +12 defines a subgroup of CLL with more frequent atypical morphology than with the original scoring system.39,44

CD38 CD38 expression45 is significantly more frequent in the +12 and del11q/del17p subgroups of CLL patients (10 of 26 cases and 5 of 10 cases, respectively) than in patients with normal karyotype or in the del13q subgroup of CLL patients (10 of 72 cases and 1 of 5 cases, respectively).38 High CD38 expression on CLL cells has been used as a haematologica | 2018; 103(6)

Chronic lymphocytic leukemia with trisomy 12

surrogate marker of unmutated IGHV genes.46 When the threshold for CD38 positivity was set at the standard 30%, higher expression of CD38 was not associated with a significantly impaired TTFT, but a threshold of 40% for CD38 expression retained its prognostic value for TTFT (P=0.008); so the threshold of CD38 positivity may be raised to 40% in +12 CLL cases to preserve its prognostic value.46 CD38 inhibitors may have particular efficacy in this cytogenetic subtype.47 CD3848-50 also showed higher expression among CLL cases with both +12 and del11q,19 which could explain the adverse prognosis of this subgroup compared to CLL patients with +12.50-52

ZAP-70 The negative prognostic impact of ZAP-70 expression and the association of ZAP-70 with unmutated IGHV are maintained in patients with +12 CLL.46,53,54

CD20 CLL patients with +12 have higher expression of CD20 than CLL patients with del17p, del11q, del13q, or negative FISH results; this expression of CD20 may predict a better response to rituximab-based regimens.55

FMC7 FMC7, an epitope on the CD20 molecule itself, is significantly (P=0.035) more frequent in the +12 subgroup of CLL (32.0%) than in the del13q CLL subgroup, normal patients, and the del11q/del17p CLL subgroup (in which 20.0%, 11.1%, and 10.0% were FMC7 positive, respectively).38

CD49 CD49d is expressed in a vast majority of +12 CLL cases and emerged as a negative prognostic factor.56-58 Using a 30% cut off, an analysis was performed on 1200 CLL patients:59 735 cases (61.2%) were CD49d-, whereas 465 cases (38.8%) were CD49d+. A significantly (P<0.001) higher percentage of cases were CD49d+ (89.4%) in the +12 subgroup than in the other cytogenetic subgroups.

Among the CD49d+ CLL patients, those with +12 expressed CD49d at higher mean fluorescence intensity (MFI) levels [median MFI 2200; 95% confidence interval (CI) 1810-2546] compared with patients without +12 (median MFI 1386; 95%CI: 1050-1673; P<0.001). CD49d overexpression in +12 CLL is associated with ITGA4 hypomethylation, as shown by quantitative real-time polymerase chain reaction analysis in 74 CLL patients (31 patients who were CD49d- and lacked +12, and 43 patients who were CD49d+ and had +12). Patients with +12 who were CD49d+ showed a shorter TTFT than that of patients with +12 who were CD49d- (P<0.001).56 A specific tropism of +12 CLL cells toward lymph nodes has been confirmed by the higher proportion of +12 CLL cells in lymph nodes than in peripheral blood or bone marrow.60 CD49d overexpression may provide a molecular basis for the peculiar biological behavior of +12 CLL and may predict the development of additional cytogenetic lesions.61

CD23 CD23, a low-affinity receptor for IgE, is expressed on B cells and on monocytes and eosinophils.62 CD23 expression varies in CLL:63 low CD23 expression, which was correlated with the presence of a prolymphocyte infiltrate haematologica | 2018; 103(6)

in the bone marrow, a higher white blood cell count, and an advanced stage of disease, was reported to be a negative prognostic factor for CLL.64 There are two isotypes of CD23: CD23a, whose expression is higher than that of CD23b, has a role in survival, while CD23b enhances proliferation.65 CLL patients can be divided into two groups: one group expressing both isotypes of CD23 at a high level and CD20, CD22, and CD38 at low levels, and one group expressing both CD23 isotypes at a low level, most of them carrying +12.66 Therefore, a high CD23a/CD23b ratio and low CD23 expression combined with high CD20 and CD38 expression may be a useful tool in predicting +12 in CLL and could serve as a marker for a high likelihood of +12.

Integrins Circulating +12 CLL cells show increased expression of the integrins CD11b, CD18, CD29, and ITGB7, and of the adhesion molecule CD323 compared with other CLL cytogenetic groups. These changes are modulated by NOTCH1 mutation status; NOTCH1-mutated +12 cases have lower expression of CD11a, CD11b, and CD18 than wild-type cases. +12 CLL cells also have upregulation of signaling pathways that increase ligand binding and enhance VLA-4–directed adhesion and motility.46 Also, increased expression of the α-integrins CD11a and CD11b was associated with a shortened TTFT (P=0.002 for CD11a; P=0.027 for CD11b). The presence of NOTCH1 mutations in the context of +12 was associated with decreased CLL cell expression of CD11a, CD11b, and CD18 to levels similar to those of CLL cells without +12. Notably, the presence of a NOTCH1 mutation had no impact on CD29, CD49d, or ITGB7 expression on CLL cells. The observed heterogeneity of expression of β2integrins in +12 CLL cases can be explained largely by the presence of NOTCH1 mutations.46 The expression of the β2-integrins CD11a/CD18 (LFA-1) and CD11b/CD18 (Mac-1) is down-regulated by the co-existence of NOTCH1 mutations, indicating a novel interaction that may be of importance in aggressive, high-risk CLL.46

Adhesion molecules Other cell surface markers with a possible prognostic role in +12 CLL are CD25 (also called activation-associated interleukin-2 receptor), which has a role in B-cell proliferation; CD54 (i.e. intercellular adhesion molecule-1), which has a role in cell adhesion and in the homing process; and CD95 (i.e. FasR), which has a role in regulation of apoptosis in lymphoid tissues. Patients with levels of 30% or higher of CD25+ B-CLL cells had a shorter median TTFT (P=0.01).67 Low CD54 expression was associated with a prolonged median TTFT (P=0.004), low leukocyte count (P<0.05), and low serum lactate dehydrogenase level (P=0.03). CD54 could be a prognostic marker for CLL68,69 and might be a determiner of whether patients present with localized or widespread lymphadenopathies.70 High CD95 expression has been correlated with elevated serum lactate dehydrogenase level (P=0.02) and lymphadenopathy (P=0.02). Hjalmar et al. correlated these markers on CD19+ cells with the occurrence of +12 and atypical lymphocyte morphology.71 They found that the mean percentage of CD25+ B cells in CLL with atypical morphology and in patients with +12 was lower than that in CLL with typical morphology (P<0.02) and in patients with disomic tumor cells 933

F. Autore et al.

(P<0.03). Published data regarding the association between CD25 and +12 report conflicting findings.72-74

Stereotyped B-cell receptor prevalence in patients with +12 CLL Recurrent 'stereotyped' patterns,75,76 known as subsets, have been reported in +12 CLL patients, including a higher prevalence of the IGHV 4-39 gene, particularly in patients who developed Richter Syndrome.77-79 Falisi et al. highlighted the fact that 44% of +12 CLL patients carried stereotyped B-cell receptor33 and found a significantly higher prevalence of stereotyped IGHV configurations in +12 CLL patients than in controls (27%; P=0.01). The most prevalent stereotyped subset was subset #1, followed by subsets #8, #10, #28 and #59. The most frequently observed IGHV gene was V4-39*01, followed by V1-3*01. IGHV 4-39*01 was the most common gene in +12 CLL patients in the literature, confirming the strong association between +12, IGHV 4-39*01 stereotype, and Richter syndrome.77-80

Immunoglobulin heavy chain mutational status in patients with +12 CLL Chronic lymphocytic leukemia subgroups with del11q/del17p and +12 showed greater percentages of unmutated IGHV cases (80.0% and 53.8%, respectively) than those with del13q or with normal karyotype (37.5% and 31.5%, respectively).38 Cytogenetic subgroups also showed differences in the IGHV repertoire: IGHV1 genes were observed more frequently in the +12 subgroup (26.9%) than in other subgroups (20.0% in del11q/del17p and in del13q subgroups and 12.3% in the normal karyotype). IGHV4 genes were noted at similar frequencies in the +12 subgroup and in the normal group (34.6% and 30.1%, respectively), and they were not identified in any cases in the del11q/del17p subgroup.38

Molecular abnormalities in patients with +12 CLL Chronic lymphocytic leukemia patients with +12 or normal karyotype at diagnosis have an approximately 2fold higher probability of developing poor-risk genetic lesions and entering the intermediate- or high-risk subgroups associated with shorter survival, compared with low-risk CLL patients harboring del13q only.12 Chronic lymphocytic leukemia patients with +12 rarely show TP53 mutations or acquire them over time.12,81,82 On the contrary, NOTCH1 mutations can be identified in 3040% of CLL patients carrying +12.83-86 NOTCH1 is a human gene encoding a single-pass transmembrane protein, whose signaling network regulates interactions between physically adjacent cells. Following the pivotal study that identified NOTCH1 mutations in CLL and provided initial evidence on the unfavorable clinical outcome associated with NOTCH1 alterations,87 two additional independent studies of the CLL coding genome have identified activating mutations of NOTCH1 gene in approximately 10% of CLL at diagnosis.88,89 The biological contribution of NOTCH1 mutations in determining the aggressive behavior of CLL could be due to the over-representation of genes involved in the cell cycle and proliferation. In at least four series, NOTCH1 mutations had an adverse impact on outcome independently of other clinico-biological features.87-90 In an analysis by Del Giudice et al.,91 NOTCH1 mutations clustered within 104 untreated +12 CLL patients accounted for 24% of those cases, and 934

were frequently detected in cases with unfavorable biological markers, associated with a particular gene expression profile predicting a poor outcome. NOTCH1 mutations are not associated with TP53 disruption or del11q; this mutual exclusivity could be useful in a genetic hierarchical prognostic model.91 NOTCH1 mutations were represented in all cases included by frameshift deletions, in particular c.7544_7545delCT in 88% of cases, and were preferentially associated with unmutated IGHV (84% of cases; P=0.003).91 NOTCH1 mutations were not associated with sex or ZAP-70 positivity (mutated NOTCH1 cases were 62.5% ZAP-70+, and wild-type NOTCH1 cases were 44.2% ZAP-70+; P=0.116). NOTCH1 mutations have been reported at CLL diagnosis in approximately 20% of chemoresistant cases.33 NOTCH1 mutations conferred a higher cumulative probability of Richter syndrome.88,92,93 The impact of NOTCH1 mutations on prognosis is less relevant in the presence of concurrent chromosomal aberrations, as worse outcomes are observed among patients with +12 associated with additional chromosomal abnormalities irrespective of NOTCH1 mutation status.30 Landau et al. have published the results of a large retrospective cohort study using whole-exome sequencing to analyze the frequency of somatic mutation and copy number abnormalities in 160 patients with CLL. They found that favorable alterations, such as MYD88 mutation, del13q and +12, represented early clonal drivers of disease; instead, more unfavorable alterations, such as ATM, TP53 and SF3B1 mutations, were subclonal during the early stages of disease, and became clonal at time of first treatment or treatment failure. The same group has more recently published prospective genetic data collected from 278 patients with CLL; while confirming previously published retrospective data, this study found also an association between +12 and mutations in BIRC3 and BCOR. Interestingly, neither BIRC3 nor BCOR mutation was associated with worse survival, in contrast to what had previously been published in smaller studies.34,35 Maura et al. studied the association of CLL with IGF1R expression.94,95 IGF1R-IGF1/2 interaction is involved in intracellular cell signaling pathways such as RAS/RAF/MEK/ERK kinases and PI3K-Akt. IGF1R, which is implicated in the activation of the PI3K/Akt and MAPK pathways, was generally over-expressed in CLL cells compared with healthy B cells.96 IGF1R expression was strongly associated with unfavorable/intermediate-risk cytogenetic features, in particular +12, independently from unmutated IGHV, NOTCH1 mutation, or clinical monoclonal B-cell lymphocytosis status. Thus, IGF1R represents a potential novel candidate for specific targeted therapy in +12 CLL patients. In a prospective series, a group of +12 patients, showing low IGF1R expression, mutated IGHV, and wild-type NOTCH1 status had an indolent clinical course, whereas the others, characterized by unmutated IGHV and/or high CD38, mutated NOTCH1, and high IGF1R, had a more aggressive evolution and shorter TTFT. Thus, IGF1R may have a role in discriminating aggressive CLL from intermediate-risk or favorable CLL in the presence of +12.95 The overexpression of ATF5 in CLL was significantly (P<0.001) associated with chromosomal abnormalities including del11q and +12.97 ATF5 is known for its role in the regulation of cell cycle progression and of differentiation and apoptosis. ATF5 overexpression was most comhaematologica | 2018; 103(6)

Chronic lymphocytic leukemia with trisomy 12

Table 1. Clinical and laboratory characteristics at diagnosis of patients with +12 chronic lymphocytic leukemia.

Characteristic

González-Gascón y Marín (n=289)82

Strati (n=250)25

Del Giudice (n=104)91

Falisi (n=44)33

68 (22-88) 61.6% 71.1% 23.6% 5.4% 19 (4.5-294.2) 16.8% 15.5% 5.7% 28.9%a 53.8% 37.4% 55.8% 3.9% 6.1% 17.0%

60 (32-87) 58.8% 71% 13% 16% 18 (2-681) 26.0% 16.0% 6.0% 18.1%b 55.1% 55.0% 45.1% 0% 0% 17.6%

65 (56-72) 54 (52%)

62.8 (45-86) 32 (73%)

Age (range), years Male sex Binet stage A Binet stage B Binet stage C Median white blood cell count, x109/L Absolute lymphocyte count, >30*109/L Splenomegaly Hepatomegaly High β2 microglobulin Unmutated IGHV CD38 positivity ZAP-70 positivity +12 and del11q +12 and del17p +12 and del13q

High β2 microglobulin level defined as above normal range. bHigh β2 microglobulin level defined as ≥4 mg/L.

36% Binet B to C 36% Binet B to C

58.0%

70.6%

8.0% 5.0% 18.0%

4.5% 2.3% 9.1%

a

mon in the del11q subgroup and second most common in the +12 subgroup, which had poor outcomes with the second shortest TTFT. ATF5 may be a key gene responsible for inducing cell proliferation in CLL patients, and it may explain the mechanisms of disease progression due to cytogenetic abnormalities in these patients.97

TTFT and OS in patients with +12 CLL Few reports detail the clinical outcomes of CLL cases with +12. Characteristics of the patients in these main series are summarized in Table 1.25,33,82,91 Here, we review data from the two largest series of patients with +12 CLL published so far. Outcomes from these studies are summarized in Table 2.

A multicenter analysis by González-Gascón y Marín et al. included 2561 patients diagnosed with CLL from 25 Spanish institutions, of whom 289 (11.3%) patients showed +12 at diagnosis; the analysis excluded patients with monoclonal B-cell lymphocytosis, patients who acquired +12 as clonal evolution, and patients with inadequate follow up.82 +12 patients had a median TTFT of 42 months (95%CI: 34-49 months) and a median OS of 129 months (95%CI: 100-158 months). The findings in this series are similar to those of previous studies that included a lower frequency of +12 cases, approximately 1520%, with a median TTFT of 32 months and a median OS of 111 months.18 These patients were a median 68 years old, were predominantly male, and had a high frequency of low lymphocyte counts (83%), low β2 microglobulin levels (71%), early Binet stage (71%), low levels of CD38 (63%), and an absence of significant organomegaly.

González-Gascón y Marín et al.82 correlated the proportion of +12 CLL cells with prognosis and identified 60% as the cut-off value associated with survival. One hundred and seventy-four patients (60.2%) carried +12 in less than 60% of cells, whereas 115 patients (39.8%) carried haematologica | 2018; 103(6)

+12 in 60% or more of cells. The patients with +12 in 60% or more of cells had more marked leukocytosis (P=0.001), higher lymphocyte counts (P=0.006), higher levels of serum lactate dehydrogenase (P=0.03), more pronounced splenomegaly (P=0.001), and more advanced Binet stage (P=0.04). In the cohort of 289 +12 patients, 175 (60.6%) required treatment initiation. The patients with +12 in less than 60% of cells had a lower rate of treatment initiation and longer TTFT (51.2% with a median TTFT of 49 months) than did patients with +12 in 60% or more of cells (75.7% with a median TTFT of 30 months; P<0.001) in both univariate and multivariate analyses. At the time of analysis, 69 patients (23.9%) had died, and there was a significant difference in frequency of deaths (P=0.009) between patients with less than 60% +12 cells (16.7% with a median OS of 159 months) and patients with 60% or more +12 cells (29.6% with a median OS of 96 months). On multivariate analysis, only Binet stage (P=0.04), del11q in addition to +12 (P=0.01), and high β2 microglobulin levels (P=0.03) were significantly associated with shorter OS. TTFT and OS were shorter in +12 and del11q patients compared with those with +12 as a unique aberration (23 months vs. 44 months, P=0.02 for TTFT; 44 months vs. 159 months, P=0.02 for OS), but no difference could be found in TTFT or OS when +12 was accompanied by del17p or del13q.

Strati et al. analyzed a single-center experience of 250 untreated CLL patients with +12 followed for nine years and compared their outcomes with those of 516 treatment-naïve CLL patients who lacked common recurrent abnormalities on FISH analysis (negative for del11q, del13q, del17p, and +12), evaluated at the same institution during the same period.25 On multivariate analysis, factors significantly associated with +12 were a platelet count of less than 100x109/L [odds ratio (OR) 2.4; P=0.03], positivity for CD38 (OR 2.4; P=0.001), and a Matutes score of less than 4 (OR 2.4; P<0.001). At a median follow 935

F. Autore et al.

Table 2. Clinical outcomes of patients with +12 chronic lymphocytic leukemia.

Outcome

González-Gascón y Marín (n=289)82

Median follow up (range), months Treatments Median TTFT (95% CI), months Deaths Median OS (95% CI), months

41 (1-197) 61.0% 42 (34-49) 21.8% 129 (100 to 158)

Strati (n=250)25 51 (1-105) 57.0% 38 (27-48) 7.0% Not reached

CI: Confidence Interval; TTFT: time to first treatment; OS: overall survival.

up of 51 months (range 1-105 months), 142 patients (56.8%) required initiation of treatment. The median TTFT was 38 months (95%CI: 27-48 months), significantly shorter than that of FISH-negative patients (82 months; P<0.001). The multivariable model for TTFT showed that a Rai stage of III-IV [Hazard Ratio (HR) 3.3; P=0.02], palpable splenomegaly (HR 2.3; P=0.007), and chromosome banding analysis positivity for del14q (HR 3.5; P=0.004) were independently associated with TTFT. The estimated median OS was not reached, as only 18 patients died, and did not differ significantly between patients with or without +12 by FISH (P=0.22).

Followed over time, 6 of these patients (2.4%) developed histologically confirmed Richter syndrome after a median ten months (95%CI: 5-15 months) from diagnosis. This incidence was significantly higher in treated than in untreated patients (4% vs. 0%; P=0.04) and in patients with +10 compared with the control cohort of patients with normal karyotype (2.0% vs. 0.4%; P=0.02). Twentytwo patients (8.8%) developed a malignancy other than Richter syndrome after a median of 19 months (95%CI: 6-32 months) from diagnosis. Among second neoplasms,

References 1. Ghia P, Ferreri AM, Caligaris-Cappio F. Chronic lymphocytic leukemia. Crit Rev Oncol Hematol. 2007;64(3):234-246.

2. Thurmes P, Call T, Slager S, et al. Comorbid conditions and survival in unselected, newly diagnosed patients with chronic lymphocytic leukemia.

Leuk Lymphoma.

2008;49(1):49-56.

3. Hallek M, Cheson BD, Catovsky D, et al. A report from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute-Working Group 1996 guidelines. Blood. 2008;111(12): 5446-5456.

4. Zenz T, Mertens D, Kuppers R, Döhner H, Stilgenbauer S. From pathogenesis to treatment of chronic lymphocytic leukaemia.

Nat Rev Cancer. 2010;10(1):37-50.

5. Eichhorst B, Dreyling M, Robak T, Montserrat E, Hallek M; ESMO Guidelines Working Group. Chronic lymphocytic leukemia: ESMO Clinical Practice

936

6.

7.

8.

9.

10.

5 were hematologic malignancies, and 17 were nonhematologic malignancies. In the multivariable model, an absolute lymphocyte count of more than 30x109/L (HR 14.5; P=0.04) and development of a second malignant neoplasm (HR 23.8; P=0.002) remained independently associated with OS. A correlation between CLL and second malignant neoplasms has been reported, but the mechanism linking +12 to the onset of second neoplasms remains unknown.98 Eighteen patients (7.2%) died during the observation time, 15 of whom had received treatment for CLL. Seven patients died of CLL-related (n=4) or Richter syndrome-related (n=3) causes; 6 patients died of second neoplasms, and 5 died of unrelated causes. There was no significant difference in distribution of the causes of death between treated and untreated patients (P=0.24). The mortality of patients with +12 was only partly related to complications of CLL; the leading causes of death were Richter syndrome and second malignant neoplasms. However, given the evidence of a worse outcome for second malignant neoplasms in CLL, increased surveillance of patients in this specific group should be considered.99

Conclusions The unique morphology, immunophenotype, and cytogenetics of +12 CLL easily distinguish it from other CLL cytogenetic subtypes. +12 in CLL confers an intermediate prognostic risk, characterized by a median TTFT of 33 months and a median OS of 114 months. Genomic alterations, especially NOTCH1 mutations, can portend a worse prognosis and an increased risk of Richter syndrome.

Acknowledgments The authors would like to thank Sarah Bronson, ELS Department of Scientific Publications The University of Texas MD Anderson Cancer Center for editorial assistance.

Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2011;22 Suppl 6:vi5054.

Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN, Pasternack BS. Clinical staging of chronic lymphocytic leukemia. Blood.

1975;46(2):219-234.

Binet JL, Auquier A, Dighiero G, et al. A new prognostic classification of chronic lymphocytic leukemia derived from a multivariate survival analysis.

Cancer.

1981;48(1):198-206.

Zenz T, Fröhling S, Mertens D, Döhner H, Stilgenbauer S. Moving from prognostic to predictive factors in chronic lymphocytic leukaemia (CLL). Best Pract Res Clin Haematol. 2010;23(1):71-84.

Furman RR. Prognostic markers and stratification of chronic lymphocytic leukemia.

Hematology Am Soc Hematol Educ Program. 2010; 2010:77-81.

Pospisilova S, Gonzalez D, Malcikova J, et al. ERIC recommendations on TP53 mutation analysis in chronic lymphocytic leukemia. Leukemia. 2012;26(7):1458-1461.

11. Haferlach C, Dicker F, Schnittger S, Kern W, Haferlach T. Comprehensive genetic characterization of CLL: a study on 506 cases analysed with chromosome banding analysis, interphase FISH, IgV(H) status and immunophenotyping. Leukemia. 2007;21 (12):2442-2451.

12. Rossi D, Rasi S, Spina V, et al. Integrated mutational and cytogenetic analysis identifies new prognostic subgroups in chronic lymphocytic leukemia. Blood. 2013;121 (8):1403-1412.

13. Dubuc AM, Davids MS, Pulluqi M, et al.

FISHing in the dark: How the combination of FISH and conventional karyotyping improves the diagnostic yield in CpG-stimulated chronic lymphocytic leukemia. Am J Hematol. 2016;91(10):978-983.

14. Thompson PA, O’Brien SM, Wierda WG, et al. Complex karyotype is a stronger predictor than del(17p) for an inferior outcome in relapsed or refractory chronic lymphocytic leukemia patients treated with ibrutinibbased regimens. Cancer. 2015;121(20):36123621.

haematologica | 2018; 103(6)

Chronic lymphocytic leukemia with trisomy 12 15. Jaglowski SM, Ruppert AS, Heerema NA, et al. Complex karyotype predicts for inferior out- comes following reduced-intensity conditioning allogeneic transplant for chronic lymphocytic leukaemia. Br J Haematol.

2012;159(1):82-87.

16. Buhmann R, Kurzeder C, Rehklau J, et al.

CD40L stimulation enhances the ability of conventional metaphase cytogenetics to detect chromosome aberrations in B-cell chronic lymphocytic leukaemia cells. Br J Haematol. 2002;118(4):968-975.

17. Dicker F, Schnittger S, Haferlach T, Kern W, Schoch C. Immunostimulatory oligonucleotide-induced metaphase cytogenetics detect chromosomal aberrations in 80% of CLL patients: a study of 132 CLL cases with correlation to FISH, IgVH status, and CD38 expression. Blood. 2006;108(9):3152-3160.

18. Döhner H, Stilgenbauer S, Benner A, et al.

Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):1910-1916.

19. Quijano S, Lopez A, Rasillo A, et al. Impact of trisomy 12, del(13q), del(17p), and del(11q) on the immunophenotype, DNA ploidy status, and proliferative rate of leukemic B-cells in chronic lymphocytic leukemia. Cytometry B Clin Cytom.

2008;74(3):139-149.

20. Stilgenbauer S, Dohner K, Bentz M, Lichter P, Dohner H. Molecular cytogenetic analysis of B-cell chronic lymphocytic leukemia. Ann Hematol. 1998;76(3-4):101-110.

21. Cuneo A, Bigoni R, Rigolin GM, et al. Late appearance of the 11q22.3-23.1 deletion involving the ATM locus in B-cell chronic lymphocytic leukemia and related disorders. Clinico-biological significance.

Haematologica. 2002;87(1):44-51.

22. Dohner H, Stilgenbauer S, Fischer K, Bentz M, Lichter P. Cytogenetic and molecular cytogenetic analysis of B cell chronic lymphocytic leukemia: Specific chromosome aberrations identify prognostic subgroups of patients and point to loci of candidate genes.

Leukemia. 1997;11 Suppl 2:S19-S24.

23. Mato A, Nabhan C, Kay NE, et al. Realworld clinical experience in the Connect chronic lymphocytic leukaemia registry: a prospective cohort study of 1494 patients across 199 US centres. Br J Haematol.

2016;175(5):892-903.

24. Sindelarova L, Michalova K, Zemanova Z, et al. Incidence of chromosomal anomalies detected with FISH and their clinical correlations in B-chronic lymphocytic leukemia.

Cancer Genet Cytogenet. 2005;160(1):2734.

25. Strati P, Abruzzo LV, Wierda WG, O'Brien S, Ferrajoli A, Keating MJ. Second cancers and Richter transformation are the leading causes of death in patients with trisomy 12 chronic lymphocytic leukemia. Clin Lymphoma Myeloma Leuk. 2015;15(7):420427.

26. Huh YO, Schweighofer CD, Ketterling RP, et al. Chronic lymphocytic leukemia with t(14; 19)(q32; q13) is characterized by atypical morphologic and immunophenotypic features and distinctive genetic features. Am J Clin Pathol. 2011;135(5):686-696.

27. Sellmann L, Gesk S, Walter C, et al. Trisomy 19 is associated with trisomy 12 and mutated IGHV genes in B-chronic lymphocytic leukaemia. Br J Haematol. 2007;138(2):217220.

28. Daudignon A, Poulain S, Morel P, et al.

Increased trisomy 12 frequency and a biased IgVH 3-21 gene usage characterize small lymphocytic lymphoma. Leuk Res.

haematologica | 2018; 103(6)

2010;34(5):580-584.

29. Santos FP, O’Brien S. Small lymphocytic lymphoma and chronic lymphocytic leukemia: are they the same disease? Cancer J. 2012;18(5):396-403.

30. Lopez C, Delgado J, Costa D, et al. Different distribution of NOTCH1 mutations in chronic lymphocytic leukemia with isolated trisomy 12 or associated with other chromosomal alterations. Genes Chromosomes Cancer. 2012;51(9):881-889.

31. Rossi D, Cerri M, Capello D, et al. Biological and clinical risk factors of chronic lymphocytic leukaemia transformation to Richter syndrome. Br J Haematol. 2008;142(2):202215.

32. Yi S, Li Z, Zou D, et al. Intratumoral genetic heterogeneity and number of cytogenetic aberrations provide additional prognostic significance in chronic lymphocytic leukemia. Genet Med. 2017;19(2):182-191.

33. Falisi E, Novella E, Visco C, et al. B-cell receptor configuration and mutational analysis of patients with chronic lymphocytic leukaemia and trisomy 12 reveal recurrent molecular abnormalities. Hematol Oncol. 2014;32(1):22-30.

34. Landau DA, Carter SL, Stojanov P, et al.

Evolution and impact of subclonal mutations in chronic lymphocytic leukemia. Cell.

2013;152(4):714-726.

35. Landau DA, Tausch E, Taylor-Weiner AN, et al. Mutations driving CLL and their evolution in progression and relapse. Nature.

2015;526(7574):525-530.

36. Zent CS, Ding W, Schwager SM, et al. The prognostic significance of cytopenia in chronic lymphocytic leukaemia/small lymphocytic lymphoma. Br J Haematol.

2008;141(5):615-621.

37. Strati P, Caligaris-Cappio F. A matter of debate in chronic lymphocytic leukemia: is the occurrence of autoimmune disorders an indicator of chronic lymphocytic leukemia therapy? Curr Opin Oncol. 2011;23(5):455460.

38. Athanasiadou A, Stamatopoulos K, Tsompanakou A, et al. Clinical, immunophenotypic, and molecular profiling of trisomy 12 in chronic lymphocytic leukemia and comparison with other karyotypic subgroups defined by cytogenetic analysis. Cancer Genet Cytogenet. 2006;168 (2):109-119.

39. Matutes E, Oscier D, Garcia-Marco J, et al.

Trisomy 12 defines a group of CLL with atypical morphology: Correlation between cytogenetic, clinical and laboratory features in 544 patients. Br J Haematol. 1996;92(2): 382-388.

40. Nowakowski GS, Hoyer JD, Shanafelt TD, et al. Percentage of smudge cells on routine blood smear predicts survival in chronic lymphocytic leukemia. J Clin Oncol.

2009;27(11):1844-1849.

41. Saunders AE, Johnson P. Modulation of immune cell signalling by the leukocyte common tyrosine phosphatase, CD45. Cell Signal. 2010;22(3):339-348.

42. Johansson P, Eisele L, Klein-Hitpass L, et al.

Percentage of smudge cells determined on routine blood smears is a novel prognostic factor in chronic lymphocytic leukemia.

Leuk Res. 2010;34(7):892-898.

43. Rizzo D, Lotay A, Gachard N, et al. Very low levels of surface CD45 reflect CLL cell fragility, are inversely correlated with trisomy 12 and are associated with increased treatment-free survival. Am J Hematol.

2013;88(9):747-753.

44. Cro L, Ferrario A, Lionetti M, et al. The clin-

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

ical and biological features of a series of immunophenotypic variant of B-CLL. Eur J Haematol. 2010;85(2):120-129.

Deaglio S, Morra M, Mallone R, et al.

Human CD38 (ADP-ribosyl cyclase) is a counter-receptor of CD31, an Ig superfamily member. J Immunol. 1998;160(1):395-402.

Riches JC, O’Donovan CJ, Kingdon SJ, et al.

Trisomy 12 chronic lymphocytic leukemia cells exhibit upregulation of integrin signaling that is modulated by NOTCH1 mutations. Blood. 2014;123(26):4101-4110.

Moreau C, Liu Q, Graeff R, et al. CD38 structure-based inhibitor design using the 1cyclic inosine 50-diphosphate ribose template. PLoS One. 2013;8(6):e66247.

Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):18401847.

Hamblin TJ, Orchard JA, Ibbotson RE, et al.

CD38 expression and immunoglobulin variable region mutations are independent prognostic variables in chronic lymphocytic leukemia, but CD38 expression may vary during the course of the disease. Blood.

2002;99(3):1023-1029.

Jelinek DF, Tschumper RC, Geyer SM, et al.

Analysis of clonal B-cell CD38 and immunoglobulin variable region sequence status in relation to clinical outcome for Bchronic lymphocytic leukaemia. Br J Haematol. 2001;115(4):854-861.

D’Arena G, Musto P, Cascavilla N, et al.

CD38 expression correlates with adverse biological features and predicts poor clinical outcome in B-cell chronic lymphocytic leukemia. Leuk Lymphoma. 2001;42(12):109-114.

Brachtl G, Pinon Hofbauer J, Greil R, Hartmann TN. The pathogenic relevance of the prognostic markers CD38 and CD49d in chronic lymphocytic leukemia. Ann Hematol. 2014;93(3):361-374.

Wiestner A, Rosenwald A, Barry TS, et al.

ZAP-70 expression identifies a chronic lymphocytic leukemia subtype with unmutated immunoglobulin genes, inferior clinical outcome, and distinct gene expression profile.

Blood. 2003;101(12):4944-4951.

Crespo M, Bosch F, Villamor N, et al. ZAP70 expression as a surrogate for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia. N Engl J Med. 2003;348(18):1764-1775.

Tam CS, Otero-Palacios J, Abruzzo LV, et al.

Chronic lymphocytic leukaemia CD20 expression is dependent on the genetic subtype: a study of quantitative flow cytometry and fluorescent in-situ hybridization in 510 patients. Br J Haematol. 2008;141(1):36-40.

Gattei V, Bulian P, Del Principe MI, et al.

Relevance of CD49d protein expression as overall survival and progressive disease prognosticator in chronic lymphocytic leukemia. Blood. 2008;111(2):865-873.

Shanafelt TD, Geyer SM, Bone ND, et al.

CD49d expression is an independent predictor of overall survival in patients with chronic lymphocytic leukaemia: a prognostic parameter with therapeutic potential. Br J Haematol. 2008;140(5):537-546.

Dal Bo M, Bomben R, Zucchetto A, et al.

Microenvironmental interactions in chronic lymphocytic leukemia: hints for pathogenesis and identification of targets for rational therapy. Curr Pharm Des. 2012;18(23):33233334.

Zucchetto A, Caldana C, Benedetti D, et al.

CD49d is overexpressed by trisomy 12

937

F. Autore et al.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

938

chronic lymphocytic leukemia cells: evidence for a methylation-dependent regulation mechanism. Blood. 2013;122(19):33173321.

Strati P, Parikh SA, Chaffee KG, et al. CD49d associates with nodal presentation and subsequent development of lymphadenopathy in patients with chronic lymphocytic leukaemia. Br J Haematol. 2017;178(1):99105.

Shanafelt TD, Hanson C, Dewald GW, et al.

Karyotype evolution on fluorescent in situ hybridization analysis is associated with short survival in patients with chronic lymphocytic leukemia and is related to CD49d expression. J Clin Oncol. 2008;26(14):e5-e6.

Conrad DH, Ford JW, Sturgill JL, Gibb DR. CD23: an overlooked regulator of allergic disease. Curr Allergy Asthma Rep. 2007;7(5):331-337.

Barna G, Reiniger L, Tatrai P, Kopper L, Matolcsy A. The cutoff levels of CD23 expression in the differential diagnosis of MCL and CLL. Hematol Oncol. 2008;26 (3):167-170.

Jurisic V, Colovic N, Kraguljac N, Atkinson HD, Colovic M. Analysis of CD23 antigen expression in B-chronic lymphocytic leukaemia and its correlation with clinical parameters. Med Oncol. 2008;25(3):315-322.

Fournier S, Yang LP, Delespesse G, Rubio M, Biron G, Sarfati M. The two CD23 isoforms display differential regulation in chronic lymphocytic leukaemia. Br J Haematol.

1995;89(2):373-379.

Kriston C, Bödör C, Szenthe K, et al. Low CD23 expression correlates with high CD38 expression and the presence of trisomy 12 in CLL. Hematol Oncol. 2017;35(1):58-63.

Burton J, Kay NE. Does IL-2 receptor expression and secretion in chronic B-cell leukemia have a role in downregulation of the immune system? Leukemia. 1994;8(1):92-96.

Molica S, Levato D, Dell'Olio M, et al.

Clinico-prognostic implications of increased levels of soluble CD54 in the serum of B-cell chronic lymphocytic leukemia patients.

Results of a multivariate survival analysis.

Haematologica. 1997;82(2):148-151.

Lúcio PJ, Faria MT, Pinto AM, et al.

Expression of adhesion molecules in chronic B-cell lymphoproliferative disorders.

Haematologica. 1998;83(2):104-111.

Angelopoulou MK, Kontopidou FN, Pangalis GA. Adhesion molecules in Bchronic lymphoproliferative disorders.

Semin Hematol. 1999;36(2):178-197.

Hjalmar V, Hast R, Kimby E. Cell surface expression of CD25, CD54, and CD95 on Band T-cells in chronic lymphocytic leukaemia in relation to trisomy 12, atypical morphology and clinical course. Eur J Haematol. 2002;68(3):127-134.

Tefferi A, Bartholmai B, Witzig TE, et al.

Clinical correlations of immunophenotypic variations and the presence of trisomy 12 in B-cell chronic lymphocytic leukemia. Cancer Genet Cytogenet. 1997;95(2):173-177.

Knauf WU, Knuutila S, Zeigmeister B, Thiel

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

E. Trisomy 12 in B-cell chronic lymphocytic leukemia: Correlation with advanced disease, atypical morphology, high levels of soluble CD25, and with refractoriness to treatment. Leuk Lymphoma. 1995;19(3-4):289294.

Knauf WU, Langenmayer I, Ehlers B, et al.

Serum levels of soluble CD23, but not soluble CD25, predict disease progression in early stage B-cell chronic lymphocytic leukemia. Leuk Lymphoma. 1997;89(11): 4241-4242.

Darzentas N, Hadzidimitriou A, Murray F, et al. A different ontogenesis for chronic lymphocytic leukemia cases carrying stereotyped antigen receptors: molecular and computational evidence. Leukemia. 2010;24(1): 125-132.

Rosén A, Murray F, Evaldsson C, Rosenquist R. Antigens in chronic lymphocytic leukemia-implications for cell origin and leukemogenesis. Semin Cancer Biol.

2010;20(6):400-409.

Maura F, Cutrona G, Fabris S, et al.

Relevance of stereotyped B-cell receptors in the context of the molecular, cytogenetic and clinical features of chronic lymphocytic leukemia. PLoS One. 2011;6(8):e24313.

Rossi D, Spina V, Cerri M, et al. Stereotyped B-cell receptor is an independent risk factor of chronic lymphocytic leukemia transformation to Richter syndrome. Clin Cancer Res. 2009;15(13):4415-4422.

Stamatopoulos K, Belessi C, Moreno C, et al. Over 20% of patients with chronic lymphocytic leukemia carry stereotyped receptors: pathogenetic implications and clinical correlations. Blood. 2007;109(1):259-270.

Athanasiadou A, Stamatopoulos K, Gaitatzi M, Stavroyianni N, Fassas A, Anagnostopoulos A. Recurrent cytogenetic findings in subsets of patients with chronic lymphocytic leukemia expressing IgGswitched stereotyped immunoglobulins.

Haematologica. 2008;93(3):473-474.

Zenz T, Vollmer D, Trbusek M, et al. TP53 mutation profile in chronic lymphocytic leukemia: evidence for a disease specific profile from a comprehensive analysis of 268 mutations. Leukemia. 2010;24(12):20722079.

González-Gascón y Marín I, HernándezSánchez M, Rodríguez-Vicente AE, et al. A high proportion of cells carrying trisomy 12 is associated with a worse outcome in patients with chronic lymphocytic leukemia. Hematol Oncol. 2016;34(2):84-92.

Villamor N, Conde L, Martínez-Trillos A, et al. NOTCH1 mutations identify a genetic subgroup of chronic lymphocytic leukemia patients with high risk of transformation and poor outcome.

Leukemia.

2013;27(5):1100-1106.

Balatti V, Bottoni A, Palamarchuk A, et al.

NOTCH1 mutations in CLL associated with trisomy 12. Blood. 2012;119(2):329-331.

Balatti V, Lerner S, Rizzotto L, et al. Trisomy 12 CLLs progress through NOTCH1 mutations. Leukemia. 2013;27(3):740-743.

86. Weissmann S, Roller A, Jeromin S, et al.

Prognostic impact and landscape of NOTCH1 mutations in chronic lymphocytic leukemia (CLL): a study on 852 patients.

Leukemia. 2013;27(12):2393-2396.

87. Sportoletti P, Baldoni S, Cavalli L, et al.

NOTCH1 PEST domain mutation is an adverse prognostic factor in B-CLL. Br J Haematol. 2010;151(4):404-406.

88. Fabbri G, Rasi S, Rossi D, et al. Analysis of the chronic lymphocytic leukemia coding genome: role of NOTCH1 mutational activation. J Exp Med. 2011;208(7):1389-1401.

89. Puente XS, Pinyol M, Quesada V, et al.

Whole genome sequencing identifies recurrent mutations in chronic lymphocytic leukaemia. Nature. 2011;475(7354):101-105.

90. Oscier DG, Rose-Zerilli MJ, Winkelmann N, et al. The clinical significance of NOTCH1 and SF3B1 mutations in the UK LRF CLL4 trial. Blood. 2013;121(3):468-475.

91. Del Giudice I, Rossi D, Chiaretti S, et al.

NOTCH1 mutations in +12 chronic lymphocytic leukemia (CLL) confer an unfavorable prognosis, induce a distinctive transcriptional profiling and refine the intermediate prognosis of +12 CLL. Haematologica.

2012;97(3):437-441.

92. Rossi D, Rasi S, Fabbri G, et al. Mutations of NOTCH1 are an independent predictor of survival in chronic lymphocytic leukemia.

Blood. 2012;119(2):521-529.

93. Rossi D, Spina V, Deambrogi C, et al. The genetics of Richter syndrome reveals disease heterogeneity and predicts survival after transformation. Blood. 2011;117(12):33913401.

94. Yaktapour N, Übelhart R, Schüler J, et al.

Insulin-like growth factor-1 receptor (IGF1R) as a novel target in chronic lymphocytic leukemia. Blood. 2013;122(9):1621-1633.

95. Maura F, Mosca L, Fabris S, et al. Insulin growth factor 1 receptor expression is associated with NOTCH1 mutation, trisomy 12 and aggressive clinical course in chronic lymphocytic leukaemia. PLoS One. 2015;10 (3):e0118801.

96. Medyouf H, Gusscott S, Wang H, et al.

High-level IGF1R expression is required for leukemia-initiating cell activity in T-ALL and is supported by Notch signaling. J Exp Med. 2011;208(9):1809-1822.

97. Mittal AK, Hegde GV, Aoun P, et al.

Molecular basis of aggressive disease in chronic lymphocytic leukemia patients with 11q deletion and trisomy 12 chromosomal abnormalities. Int J Mol Med. 2007;20(4): 461-469.

98. Tsimberidou AM, Wen S, McLaughlin P, et al. Other malignancies in chronic lymphocytic leukemia/small lymphocytic lymphoma. J Clin Oncol. 2009;27(6):904-910.

99. Solomon BM, Rabe KG, Slager SL, Brewer JD, Cerhan JR, Shanafelt TD. Overall and cancer-specific survival of patients with breast, colon, kidney, and lung cancers with and without chronic lymphocytic leukemia: a SEER population-based study. J Clin Oncol. 2013;31(7):930-937.

haematologica | 2018; 103(6)

ARTICLE

Hematopoiesis

BET-inhibition by JQ1 promotes proliferation and self-renewal capacity of hematopoietic stem cells Mark Wroblewski,1,2* Marina Scheller-Wendorff,1,2,3* Florian Udonta,1,2 Raimund Bauer,1,2 Jara Schlichting,1,2 Lin Zhao,1,2,4 Isabel Ben Batalla,1,2 Victoria Gensch,1,2 Sarina Päsler,1,2 Lei Wu,5,6 Marek Wanior,7 Hanna Taipaleenmäki,8 Simona Bolamperti,8 Zeynab Najafova,9 Klaus Pantel,2 Carsten Bokemeyer,1 Jun Qi,5,6 Eric Hesse,8 Stefan Knapp,7,10,11 Steven Johnsen2,9 and Sonja Loges1,2

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):939-948

Department of Hematology and Oncology with Sections BMT and Pneumology, Hubertus Wald Tumorzentrum, University Comprehensive Cancer Center Hamburg, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 2Institute of Tumor Biology, Center of Experimental Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 3Department of Medicine V, Hematology, Oncology and Rheumatology, University of Heidelberg, Germany; 4Department of Oncology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China; 5Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA, USA; 6Department of Medicine, Harvard Medical School, Boston, MA, USA; 7Institute for Pharmaceutical Chemistry, Johann Wolfgang Goethe-University and Buchmann Institute for Molecular Life Sciences, Frankfurt am Main, Germany; 8Heisenberg-Group for Molecular Skeletal Biology, Department of Trauma, Hand & Reconstructive Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 9Department of General, Visceral and Pediatric Surgery, University Medical Center Göttingen, Germany; 10Nuffield Department of Clinical Medicine, Structural Genomics Consortium and Target Discovery Institute, University of Oxford, Old Road Campus Research Building, UK and 11German Cancer Consortium (DKTK) Frankfurt am Main, Germany 1

*MWr and MS-W contributed equally to this work.

ABSTRACT

A

lthough inhibitors of bromodomain and extra terminal domain (BET) proteins show promising clinical activity in different hematologic malignancies, a systematic analysis of the consequences of pharmacological BET inhibition on healthy hematopoietic (stem) cells is urgently needed. We found that JQ1 treatment decreases the numbers of pre-, immature and mature B cells while numbers of early pro-B cells remain constant. In addition, JQ1 treatment increases apoptosis in T cells, all together leading to reduced cellularity in thymus, bone marrow and spleen. Furthermore, JQ1 induces proliferation of long-term hematopoietic stem cells, thereby increasing stem cell numbers. Due to increased numbers, JQ1-treated hematopoietic stem cells engrafted better after stem cell transplantation and repopulated the hematopoietic system significantly faster after sublethal myeloablation. As quantity and functionality of hematopoietic stem cells determine the duration of life-threatening myelosuppression, BET inhibition might benefit patients in myelosuppressive conditions.

Introduction The molecular mechanisms that govern hematopoietic stem cell (HSC) activity and lineage specification are increasingly well known and it has been demonstrated that abnormalities in pathways controlling these functions are a major cause of malignant transformation.1 Moreover, investigation of HSC biology has changed the view of cancer, and it is now believed that tumors are sustained by cells with a cancer stem cell phenotype, a highly malignant subpopulation which maintains the uncontrolled production of less malignant progeny.2 It is, therefore, essential to identify pathways that control key stem cell functions in order to better understand genes and mechanisms which are involved in transformation, tumor progression haematologica | 2018; 103(6)

Correspondence: s.loges@uke.de

Received: October 12, 2017.

Accepted: March 15, 2018.

Pre-published: March 22, 2018. doi:10.3324/haematol.2017.181354 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/939 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

939

M. Wroblewski et al.

and relapse.

In recent years, increasing evidence has linked epigenetic dysregulation with aberrant gene expression ultimately leading to the development of cancer. It was shown that 50% of human cancers harbor mutations in enzymes that are important for proper chromatin organization.3 Members of the family of bromodomain and extra terminal domain (BET) proteins include BRD2, BRD3, BRD4 and BRDT which function as epigenetic readers and have been shown to facilitate transcription through interaction with acetylated histones.4 Chromatin binding of BET proteins is known to drive MYC expression5 which is an oncogenic driver in several hematologic malignancies such as acute myeloid leukemia (AML),6 lymphoma7 or multiple myeloma (MM).8 Consequently, epigenetic regulators have become attractive therapeutic targets and, despite their toxic and non-specific side effects, drugs interfering with epigenetic pathways have entered clinical practice.9 The BET inhibitor JQ1 binds with highest affinity to bromodomain 1 of BRD4 and prevents it binding to acetylated histones at promoters and linage specific enhancers thereby decreasing transcription of some lineage specific genes.10 At a therapeutic level, it exerts potent anti-cancer effects in a broad range of human AML subtypes,5 acute lymphoblastic leukemia (ALL),11 and MM,12 and the structurally similar BET inhibitor (OTX015) recently showed promising clinical results in these patients.13,14 Although it has recently been shown that JQ1 decreases pluripotency of embryonic stem cells by suppression of Nanog and Lefty1,15 the consequences of BET inhibition for adult HSC have not yet been analyzed in detail.

The fact that BET inhibitors are already used in clinical studies today underlines the need for a systematic approach to understand the impact of pharmacological BET inhibition on healthy hematopoiesis.

We, therefore, addressed this issue and show for the first time that BET inhibition by JQ1 induces expansion of HSC thereby promoting recovery of the hematopoietic system after myelosuppression or stem cell transplantation.

Methods More detailed information can be found in the Online Supplementary Appendix.

Animals All experiments were carried out in accordance with the institutional guidelines for the welfare of animals and were approved by the local licensing authority (Behörde für Soziales, Gesundheit, Familie, Verbraucherschutz; Amt für Gesundheit und Verbraucherschutz, Hamburg, Germany, project n. G128/16).

Treatments and mixed bone marrow chimeras Ly5.1 animals received daily intraperitoneal (i.p.) injections with 50 mg/kg JQ1 or control for 21 days. To generate bone marrow (BM) chimeras, BM of treated Ly5.1 mice was mixed with 2x105 untreated Ly5.2 supporter BM cells and injected intravenously (i.v.) into lethally irradiated Ly5.2 mice (9 Gy). Transplantations were performed with total numbers of BM cells per condition and not corrected for equal numbers of HSC. Detailed information about the normalization in secondary recipients and the calculation of repopulating units can be found in the Online Supplementary Appendix.

940

Hematopoietic recovery after sublethal myeloablation Mice received daily i.p. injections with 50 mg/kg JQ1 or control for 21 days. Myeloablation was then induced via sublethal irradiation (5 Gy). Hematopoietic recovery was monitored in peripheral blood (PB) and BM.

Colony formation assays Colony formation assays for quantification of hematopoietic stem and progenitor cells (GF M3434, H4034 Optimum) or megakaryocytes (MegaCult-C) were performed according to the manufacturer’s instructions.

ELISA and RT-qPCR Transcript levels were analyzed using Power SYBR Green RTqPCR and protein concentrations were determined using ELISA according to the manufacturer’s instruction.

Flow cytometry Different cell populations in thymus, PB, BM and spleen were stained with antibodies and analyzed on a FACS Canto II flow cytometer.

Statistical analysis Data represent mean±Standard Error of Mean (SEM) of representative experiments, which were carried out at least in duplicates, unless otherwise stated. Statistical significance was calculated by Student's t-test or ANOVA. Statistical significance in KaplanMeier plots was calculated by log-rank test.

Results JQ1 treatment alters normal hematopoiesis in a cell context-dependent manner As a first step, we analyzed the mRNA expression patterns of Brd2-4 in all major hematopoietic subpopulations, because BET family members represent the main molecular targets of JQ1.5 Expression of Brd4 mRNA was most abundant in megakaryocytes (MK), followed by B cells, T cells and HSC (Figure 1A). Brd2 expression showed a similar distribution whereas Brd3 expression was highest in MK and T cells (Online Supplementary Figure S1A and B).

To characterize the effect of BET inhibition by JQ1 on hematopoietic development in vivo, we analyzed 7-week old mice which were treated with 50 mg/kg JQ1 for three weeks, a dose level previously used in literature.5,16 As reported previously, JQ1 treatment resulted in delayed growth of mice (Online Supplementary Figure S2A).

Moreover, JQ1-treated mice were leukopenic (Figure 1B) and showed reduced numbers of mononucleated cells (MNC) in the BM (Figure 1C). In contrast, the numbers of erythrocytes and platelets were not affected by JQ1 (Online Supplementary Figure S2B and C). Spleen and thymus weight were reduced upon treatment, indicating a JQ1-induced reduction of immune cells in secondary lymphoid organs in agreement with the literature16 (Figure 1D and E). Indeed, flow-cytometric analysis of lymphoid and myeloid cell populations showed that absolute cell numbers of both B220+ B cells and CD3+ T cells were lower after treatment (Figure 1F and G). In contrast, absolute cell numbers of myeloid cells remained constant (Online Supplementary Figure S2D) indicating a more lymphoidspecific and cell context-dependent role of BET inhibition

by JQ1.

It has been shown that JQ1-mediated effects occur due haematologica | 2018; 103(6)

JQ1 induces proliferation of HSC

to downregulation of c-Myb, which is known to be a potent negative modulator of megakaryopoiesis.17 As Brd4 is also highly expressed in MK (Figure 1A), we examined their prevalence in the BM and found numerically increased absolute numbers of CD41+CD61+ MK with enhanced colony forming unit-megakaryocyte (CFU-MK) upon JQ1 treatment (Figure 1H and I). Wrightâ&#x20AC;&#x2122;s-Giemsa staining of BM and splenocytes confirmed the presence of more MK upon JQ1 treatment (Figure 1J). Thus, JQ1 treat-

A

ment induces quantitative and phenotypic changes within cell populations showing the highest level of Brd4 expression.

JQ1 blocks B-cell maturation and induces T-cell apoptosis Since BET inhibition by JQ1 reduced the numbers of mature B cells, we next analyzed the different stages of Bcell development in the BM and spleen. The earliest pro-

B

C

D

E

F

H

I

J

G

Figure 1. JQ1 treatment affects normal hematopoiesis in cell context-dependent manner. (A) Sorted megakaryocytes (MK), hematopoietic stem cells (HSC), granulocyte-monocyte progenitors (GMP), Gr1+ cells (GR1+) and monocytes (Mo) from bone marrow (BM) of mice were analyzed for baseline RNA expression of Brd4 via qRT-PCR (n=2-3; *P<0.05). (B-J) Animals received daily intraperitoneal (i.p.) injections of 50 mg/kg JQ1 for 21 days after which the different parameters were analyzed. (B) Mononucleated cells in peripheral blood (n=8; *P<0.05). (C) Mononucleated cells in BM (n=8; *P<0.05). (D and E) Spleen (D) and thymus (E) weight (n=78; *P<0.05). (F and G) B-cell (F) and T-cell (G) counts in BM (n=8; *P<0.05). (H) Flowcytometric quantification of MK in BM (n=4; *P<0.05). (I) Colony formation assay from BM to quantify megakaryocytic progenitors (n=5-7; *P<0.05). (J) Representative images showing increased numbers of MK (arrows) upon JQ1 treatment in BM. Scale bars represent 100 mm.

haematologica | 2018; 103(6)

941

M. Wroblewski et al.

B-cell progenitors (Lin-B220+CD43+IgM–) were present in normal numbers after treatment (Figure 2A). However, numbers of late pre-B (Lin–B220+CD43–IgM–) and immature B cells (Lin–B220+IgM+) were reduced by 2- and 5-fold, respectively (Figure 2B and C).

Moreover, JQ1-mediated BET inhibition caused depletion of absolute numbers of T cells (Figure 1G). Analysis of T-cell development in the thymus showed that, despite the reduction of T-cell numbers in the BM (Figure 1G), there were no developmental differences in double negative (DN) or double positive (DP) thymocyte subpopulations (Figure 2D and E). Within the developmental stages of DN thymocytes, JQ1 did not influence the proportions of DN1, DN2, DN3 and DN4 progenitor subpopulations (Figure 2F). Notably, all differentiation stages of T-cell development showed increased apoptosis rates, whereas apoptosis of B-cell subsets was unchanged (Figure 2G and H).

On a transcriptional level, JQ1 treatment decreased the expression of anti-apoptotic mediator Bcl2 in the majority of sorted T-cell subsets (Figure 2I). In contrast, treatment did not affect the expression of pro-apoptotic Bak1 in these cell types (Online Supplementary Figure S1E). Further analyses revealed no changes in differential expression of other BH3 genes important for apoptosis during particular steps of the development of DN (Mcl1, Bcl2a1a, Bcl2l11), DP (Mcl1, Bclxl, Bcl2l11) or single positive CD4+ or CD8+ (Bclxl, Mcl1, Bcl2l11) T cells18-21 (Online Supplementary Figure S1F-I). This suggests that JQ1 induces apoptosis in T cells by reducing Bcl2 expression.

To further characterize their proliferation and ability to preserve stemness, we serially replated the CFU multiple times (Figure 3H). Intriguingly, and in contrast to control, JQ1-treated BM gave rise to constant CFU numbers after each replating, whereas CFU counts decreased rapidly in the control group. This indicates that, in addition to increasing proliferation, JQ1 treatment helps maintain the regenerative potential of hematopoietic progenitors, indicating a preservation of stemness.

To investigate potential species-to-species variations, we treated purified human CD34+ cells from BM with JQ1 in vitro and assed the colony-forming capacity via colony formation assays. Interestingly, JQ1 induced an increase in CFU numbers when compared to placebo treatment (Online Supplementary Figure S2J), indicating that JQ1 has a similar effect on both murine and human HSC. To determine whether the accumulation of HSC results from increased proliferation or reduced differentiation, we analyzed BrdU incorporation into HSC in vivo. We found that LT-HSC from JQ1-treated mice incorporated more BrdU than those from control-treated littermates (Figure 3I), suggesting a JQ1-induced increase in HSC-cycling.

Interestingly, more differentiated ST-HSC or the pool of LSK cells did not show increased cell cycle activity, indicating specificity of the JQ1-induced effect in LT-HSC (Online Supplementary Figure S2G and H). In addition, HSC proliferation was accompanied by a marginal increase in apoptosis of HSC (Figure 3J). These data strongly suggest that JQ1 promotes HSC proliferation, ultimately leading to an increase in the size of the HSC pool.

JQ1 treatment results in expansion and mobilization of HSC

JQ1 treatment increases hematopoietic reconstitution after HSC transplantation

Prompted by the high expression of BET family members in HSCs, we further dissected their expression pattern in different stages during HSC differentiation, such as LT-HSC, ST-HSC, GMP and MEP (Figure 3A).

Interestingly, only Brd4 expression was highest in LT-HSC and this decreased with further differentiation via GMP to MEP, whereas no differential expression of Brd2 and 3 was detectable (Online Supplementary Figure S1C and D).

Therefore, we examined the impact of JQ1 treatment on hematopoietic stem and progenitor cells (HSPC) in vivo.

Despite reduced cellularity in BM, the absolute numbers of phenotypic lineage-ckit+sca1+ cells (LSK) were increased 3-fold upon JQ1 treatment (Figure 3B). To rule out the possibility that JQ1 treatment only affects developing hematopoiesis in young mice, we validated the JQ1induced increase in LSK cells in adult mice (Online Supplementary Figure S2I). Interestingly, an increased cell count was only present in LT-HSC (CD34–CD135–LSK), ST-HSC (CD34+CD135–LSK) and MPP (CD34+CD135+ LSK) (Figure 3C and D), but not in more differentiated lineages such as GMP (lin–sca1–ckit+CD34+FcgRII+) or MEP (lin–sca1–ckit+ CD34–FcgRII–) (Online Supplementary Figure S2E and F).

To assess the functionality of hematopoietic progenitors, we evaluated the colony forming capacity of JQ1treated MNC from BM and spleen in vitro. We found a 2to 3-fold increase in total numbers of colony forming units (CFU) in BM and spleen in mice that had received JQ1 treatment (Figure 3E and F). Moreover, JQ1 treatment mobilized early hematopoietic progenitors into the PB, where we could detect a 2.5-fold increase in CFU per mL of blood (Figure 3G).

To measure HSC numbers and multipotency at a functional level, we carried out limiting dilution competitive repopulation transplantations. Therefore, we treated CD45.1 mice for three weeks with JQ1 or control before sacrifice. Subsequently, a mix of supporter BM from CD45.2 mice and JQ1- or control-treated BM from CD45.1 mice was transplanted into lethally irradiated CD45.2 recipients. Engraftment of HSC was evaluated by analyzing recipient PB contribution over four months.

Consistent with our in vitro studies, mice transplanted with JQ1-treated BM had higher levels of donor-derived chimerism during the whole observation period, indicating an increased fraction of HSPC in the transplants (Figure 4A and B). Relative calculation of repopulating units (RU) revealed 2.5-fold more HSC in transplants when donors received JQ1 treatment [RUD (DMSO) = 2.15; RUD (JQ1) = 4.9]. In addition, extreme limiting dilution analysis (ELDA) showed a 3-fold increase of HSC frequency in BM upon JQ1 treatment (DMSO: 1/19480 vs.

JQ1: 1/6284; n=8-12; *P<0.05) (Figure 4C). Importantly, JQ1 and DMSO-treated HSC reconstituted all major hematopoietic lineages with similar frequencies (Online Supplementary Figure S3A-C). We, therefore, conclude that JQ1 induces HSC pool expansion in donors without induction of lineage priming after transplantation.

After six months, primary recipients were sacrificed and their BM was transplanted into secondary recipients. As it is well known that HSC engraftment is less effective in secondary recipients, we normalized the measured frequencies of CD45.1+ cells in secondary recipients (Online Supplementary Figure S4) to the known frequencies of their respective donor animals; this allowed us to analyze

942

haematologica | 2018; 103(6)

JQ1 induces proliferation of HSC

potential changes of HSC contribution to hematopoiesis independent of their numbers (Figure 4D). In line with our observations following primary transplantation, JQ1treated HSC also contributed more to hematopoiesis in secondary recipients, indicating a sustained long-term effect of JQ1 on HSC. The contribution of DMSO-treated HSC to hematopoiesis decreased in secondary recipients over the course of 12 months in agreement with literature.22 Interestingly, the decline was slower upon transplantation of JQ1-treated HSC, and after 12 months they still significantly contributed to hematopoiesis while any contribution of DMSO-treated HSC to the host hematopoiesis was no longer detectable (Figure 4D).

Therefore, the relative contribution of JQ1 over DMSOtreated HSC to hematopoiesis increased over time, indicating a sustained long-term effect of JQ1 on HSC (1 month: 1.9-fold, 3 months: 3.4-fold, 6 months: 5.2-fold, 12 months: 11.8-fold). Furthermore, the JQ1-induced increase in HSC proliferation did not translate into enhanced exhaustion of HSC, as there was no significant difference

A

B

E

F

H

haematologica | 2018; 103(6)

C

in long-term survival of secondary recipients from both groups (Figure 4E).

JQ1 treatment accelerates blood recovery after sublethal irradiation As JQ1 treatment increases HSC numbers by inducing proliferation, we asked whether the pro-stem-cell phenotype elicited by JQ1 might translate into increased hematopoietic recovery after sublethal irradiation, or if this beneficial effect is compromised by increased radiosensitivity of HSC due to increased cell cycle activity.

To verify this, mice were treated with JQ1 for three weeks before they received sublethal irradiation and hematopoietic recovery was monitored. As we found that JQ1 also mediates HSC expansion in 11-week old mice that have been treated with JQ1 for three weeks (Online Supplementary Figure S2I), we used adult mice for these experiments.

Analysis of the BM of irradiated mice revealed a strong decrease in hematopoietic stem and progenitor cells within two weeks after irradiation (Figure 5A and B).

D

G

I

Figure 2. JQ1 induces T-cell apoptosis and inhibits B-cell maturation.

Animals received daily intraperitoneal (i.p.) injections of 50 mg/kg JQ1 for 21 days after which the different parameters were analyzed.

(A-C) Flowcytometric quantification of Pro-B (A), Pre-B (B) and immature B cells to analyze B-cell development in bone marrow (BM) (n=4; *P<0.05). (D-F) Flowcytometric quantification of CD4–CD8– double negative (D), CD4+CD8+ double positive (E) and all differentiation states of CD4–CD8– double negative T cells (DN1-4) in thymus (n=4; *P<0.05).

(G) Flowcytometric quantification of apoptosis in CD4–CD8– double negative, CD4+CD8+ double positive and CD4/CD8 single positive T cells using Annexin V (n=3; *P<0.05). (H) Flowcytometric quantification of apoptosis in pro-B, pre-B, immatureB (Imm) and mature B cells using Annexin V (n=4; *P<0.05). (I) FACSsorted T-cell subpopulations from thymus of placebo or JQ1-treated mice were analyzed for mRNA expression via qRT-PCR (n=5; *P<0.05).

943

M. Wroblewski et al.

Before irradiation, BM of JQ1-treated mice contained 1.6fold higher numbers of HSC when compared to placebotreated mice. Interestingly, irradiation led to a similar reduction in HSC in both groups two weeks after treatment, indicated by an unchanged HSC ratio of 1.6-fold (Figure 5A).

Similar observations were made for MPP (Figure 5B). This indicates that, despite increasing cell cycle activity, JQ1 does not sensitize HSC to irradiation at a dose level of 5 Gy. In response to irradiation, cell counts of all major hematopoietic lineages decreased by 75-99% in both groups, with B and T cells being most affected one week after irradiation (Online Supplementary Figure S5A).

Interestingly, and in line with the increased numbers of HSC and MPP, JQ1-treated mice showed a significantly faster recovery of all lineages when compared to animals before irradiation.

Within the whole post-irradiation period, the relative recovery of leukocytes was significantly faster in JQ1-treat-

A

D

H

944

E

I

ed mice than in the control cohort (Figure 5C). Six weeks after irradiation, total leukocyte counts of JQ1-treated mice fully recovered to levels before irradiation (94Âą7%; n=5; *P<0.05), whereas recovery was slower in the control group (47Âą5%; n=5; *P<0.05). Strikingly, recovery of most hematopoietic lineages such as B cells, T cells, monocytes, thrombocytes or granulocytes was accelerated when animals received JQ1 treatment before irradiation (Figure 5DH). In contrast, the recovery of erythrocytes was delayed upon JQ1 treatment; however, levels never dropped to critical values and fully recovered after irradiation (Online Supplementary Figure S5B).

Although JQ1 induces cycling of HSC, this mild increase does not seem to be sufficient to render HSC more susceptible to irradiation with 5 Gy. Still harboring increased numbers of HSC two weeks after irradiation, JQ1-pre-treated animals recover faster from sublethal myelosuppression than their placebo-treated littermates.

B

C

F

G

J

Figure 3. JQ1 treatment results in expansion and mobilization of hematopoietic stem cells (HSC).

Animals received daily intraperitoneal (i.p.) injections of 50 mg/kg JQ1 for 21 days after which the different parameters were analyzed.

(A) Different FACS-sorted hematopoietic stem and progenitor cells from bone marrow (BM) of mice were analyzed for baseline mRNA expression of Brd4 via qRTPCR (n=2-3; *P<0.05). (B-D) Flowcytometric quantification of phenotypic lineage-ckit+sca1+ cells (LSK) (B), ST-HSC and LT-HSC (C), as well as MPP (D) in BM (n=5-9; *P<0.05). (E-G) Colony formation assays to quantify hematopoietic progenitors in BM (E), spleen (F), and peripheral blood (G) (n=3-8; *P<0.05). (H) Replating of CFU from BM (n=3; *P<0.05). (I) Flowcytometric quantification of BrdU incorporation in LT-HSC in BM (n=5; *P<0.05).

(J) Flowcytometric quantification of apoptosis in LT-HSC in BM using Annexin V (n=5; *P<0.05).

haematologica | 2018; 103(6)

JQ1 induces proliferation of HSC

Discussion

bopoiesis exerted by JQ1 in mice and by other BET inhibitors in humans.

In agreement with previous literature, JQ1 treatment led to reduced cellularity in peripheral blood and BM due to reduced numbers of B and T cells.16 We could refine these analyzes showing that JQ1 treatment only led to a reduction of pre-, immature and mature B-cell numbers; however, no decrease in pro-B cells was observed. This might indicate a differentiation block occurring at the pro-B to pre-B transition, which would be in line with the high expression of BET proteins starting at the pro-B-cell level.

Despite this hypothesis, we cannot rule out the possibility that JQ1 specifically targets all B-cell subsets except for pro-B cells, thereby reducing their numbers.

Furthermore, JQ1 treatment increased apoptosis in all analyzed T- but not B-cell subsets. We have been able to show that JQ1 decreases mRNA expression levels of the anti-apoptotic BH3 protein BCL2 in the majority of T-cell subsets, whereas no changes in the expression of proapoptotic genes could be detected. However, future studies are needed to understand whether this differential regulation of BCL2 mRNA is directly mediated by the effect of JQ1

Despite the emerging clinical importance of BET inhibitors, a systematic study examining their effects in the adult hematopoietic compartment is urgently needed.

We have been able to demonstrate high expression of Brd4 in MK, B cells, T cells and HSC, while it was lower in GMP, monocytes and granulocytes. The expression of Brd2 showed a similar distribution, whereas Brd3 expression was highest in MK and T cells.

In contrast to data with BET inhibitors analyzed in clinical studies, JQ1 treatment of mice did not lead to thrombocytopenia, which might be indicative of species-tospecies variations. Interestingly, this comparison is impaired by the fact that JQ1 has never been used in clinical studies due to its short half-life in plasma.23 On the other hand, the different effects on thrombopoiesis seen with JQ1 and other BET inhibitors could also be due to minor structural differences between JQ1 and clinical compounds or the drug administration methods.

Further experiments are warranted in order to determine the reasons for the different effects on throm-

A

D

B

C

E

Figure 4. JQ1 treatment increases hematopoietic repopulation after hematopoietic stem cell (HSC) transplantation. Limiting dilution competitive repopulation transplantations were performed by transplanting bone marrow (BM) from CD45.1+ JQ1- or control-treated mice together with supporter BM from CD45.2+ mice into lethally irradiated CD45.2+ recipients. (A and B) Analysis of chimerism after competitive repopulation (n=12; *P<0.05). (C) Extreme limiting dilution analysis (ELDA) for determining the frequency of HSC in transplants (n=8-12; *P<0.05). (D and E) Monitoring of chimerism (D) and survival (E) after retransplantation of the BM of primary recipients into the second generation of mice and over the course of 12 months.

haematologica | 2018; 103(6)

945

M. Wroblewski et al.

on known targets including MYC and MYB or by other mechanisms. The finding of reduced numbers of B cells at the pre-B-cell level while pro-B cells are similar upon JQ1 treatment suggests there is a differentiation block, but this needs further clarification.

Although JQ1 induced numerical expansion of megakaryocytes in the BM, other myeloid cells, platelets and RBC remained constant upon treatment. Thus, JQ1 treatment induces profound cell type-specific changes in hematopoietic cell populations displaying high mRNA expression of BET proteins. As JQ1 targets multiple proteins of the BET family, the individual contribution of specific molecular targets to the observed phenotype has to be identified in future studies.

JQ1 induces LT-HSC proliferation, thereby leading to an expansion of LT-HSC, ST-HSC and MPP in BM and spleen. Our data further indicate that JQ1 induces HSC mobilization. At the functional level, JQ1 treatment increased the quantity of HSC leading to better engraftment and faster recovery of the hematopoietic system without inducing HSC exhaustion. It is worthy of note that we did not challenge the secondary recipients of JQ1or control-treated BM with myelosuppressive stimuli, but only measured their lifespan as a read-out for HSC

exhaustion. It might, therefore, be conceivable that transient myeloablation of secondary recipients would reveal potential differences in hematopoietic recovery between both groups. Such experiments are especially important as we found marginal induction of apoptosis in HSC upon JQ1 treatment. Therefore, this aspect should be analyzed in future studies.

Interestingly, recent reports have shown that MYB plays an important role in hematopoiesis by recruiting BRD4, which is functionally suppressed by BET inhibitors such as JQ1.24 In line with this, the MYB polymorphism M303V disrupts the interaction of MYB with its transcriptional coactivator p300, thereby inducing megakaryocytosis, lymphopenia, as well as a pronounced increase in HSC.25 The striking phenotypic similarities of reduced MYB transactivation and JQ1 treatment suggest that the results reported in this study could result from JQ1-mediated suppression of MYB. Furthermore, and in line with our observations, impaired function of MYB was reported to increase cycling of LT-HSC.25 In any case, additional studies are warranted to verify whether JQ1 induces its effects on hematopoiesis via MYB. With respect to BRD4 inhibition, a previous study demonstrated that short-term reconstitution of LSK cells

A

B

C

D

E

F

G

H Figure 5. JQ1 treatment accelerates blood recovery after myeloablation. Animals were injected daily with 50 mg/kg JQ1 intraperitoneal (i.p.) for 21 days before receiving sublethal whole body irradiation with 5 Gy. (A and B) Flowcytometric quantification of hematopoietic stem cells (HSC) (A) and MPP (B) over the course of six weeks after myelosuppression (n=3-5; *P<0.05). (C-H) Analysis of the recovery of leukocytes (C), B cells (D), T cells (E), monocytes (F), thrombocytes (G) and granulocytes (H) after myelosuppression over the course of six weeks (n=3-5; *P<0.05).

946

haematologica | 2018; 103(6)

JQ1 induces proliferation of HSC

after lethal irradiation is impaired when the transplanted hematopoietic stem and progenitor cells harbor an shRNA-mediated knockdown of Brd4 (Brd4 KD).26 The reduced repopulative capacity of Brd4 KD HSCP at the LSK level initially seems to contradict our findings. One obvious reason for this discrepancy might be that the shRNA strategy only focuses on the inhibition of Brd4, whereas JQ1 targets the entire BET family. In addition, BRD4 contains other protein domains such as an extra-terminal domain, the function of which is not influenced by JQ1.27 Although the effect of JQ1 described in our study most likely results from the increased number of HSC, other data link it to increased HSC fitness. In colony formation assays, the increased CFU numbers arise from the expanded pool of stem and progenitor cells. However, their increased regenerative potential seen in replating assays may also result from increased HSC fitness. Similarly, the increased contribution of JQ1-treated HSC to hematopoiesis in secondary recipients results from the increased HSC numbers in the graft. However, the decline of the relative contribution to hematopoiesis was slower in the animals that received JQ1-treated HSC, which might indicate enhanced fitness. As in our study it is difficult to discriminate between the effects of JQ1 on HSC quantity and quality, future experiments transplanting equal numbers of JQ1- or control-treated HSC are warranted.

It is worthy of note that the supporter BM cells from the first transplantation outcompeted the DMSO-treated HSC in secondary recipients starting at 12 months after transplantation. Although it is well known that the fitness of HSC dramatically decreases in secondary recipients,22 we also observed an unusually reduced median survival, particularly in the DMSO arm, although this effect was not statistically significant; this has to be considered when interpreting the data. Therefore, future studies are needed to identify the underlying factors for this effect.

Although allogeneic or autologous HSCT is an important option for the treatment of leukemia, severe autoimmunity or myelosuppression,28 due to different limitations, not all patients benefit. It has been shown that the quantity and quality of HSCs are critical factors determining patient survival and time required for engraftment after HSCT.29 Between 2-5% of stem cell donors do not adequately respond to G-CSF treatment and the number of CD34+

References 1. Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, and cancer stem cells.

Nature. 2001;414(6859):105-111.

2. Nguyen LV, Vanner R, Dirks P, Eaves CJ. Cancer stem cells: an evolving concept. Nat Rev Cancer. 2012;12(2):133-143.

3. Jones PA, Issa JP, Baylin S. Targeting the cancer epigenome for therapy. Nat Rev Genet. 2016;17(10):630-641.

4. Wu SY, Chiang CM. The double bromodomain-containing chromatin adaptor Brd4 and transcriptional regulation. J Biol Chem.

2007;282(18):13141-13145.

haematologica | 2018; 103(6)

HSC mobilized is not sufficient; these donors are called ‘poor mobilizers’.30-33 As a donor’s response to G-CSF varies,34 and HSCT programs have to, therefore, deal with these obvious uncertainties, alternative approaches are needed to obtain sufficient numbers of HSC for transplantation. We were able to show that, besides HSC proliferation, BET inhibition by JQ1 also increases the mobilization of HSC into the PB. Our data, therefore, indicate that BET inhibition might represent an alternative approach to yield increased numbers of HSC. However, future studies are necessary to verify whether this approach is safe or applicable for the G-CSF-independent mobilization of stem cells in humans.

In this context, it should be considered that JQ1 treatment has opposing effects on some mature hematopoietic cell populations and HSC. This leads to the question as to whether the reduction of B and T cells might compromise the beneficial effects on HSC during repopulation of the hematopoietic system. In cases in which the detrimental effect on B and T cells persists even after discontinuation of BET inhibitor treatment in a transplantation setting, this would make HSCT recipients more prone to infections, thereby counteracting the beneficial effects of increased HSC numbers in the grafts. Therefore, further experiments are warranted to clarify the consequences of (transient) BET inhibition on B and T cells and the outcome of HSCT. Acknowledgments The authors would like to thank the FACS Core Facility (UKE, Hamburg, Germany) for helping with flow cytometry.

Funding SL was supported by the German Research Council (DFG LO1863/3-1), by the Margarethe Clemens Stiftung, by a Starting Grant of the European Research Council and is the recipient of a Heisenberg professorhip (DFG, LO1863/4-1); MWr was supported by the Medical Faculty of the University of Hamburg (FFM program); RB is an Erwin-Schrödinger fellow of the Austrian Science Fund (FWF); SK is grateful for support by the Structural Genomics Consortium (SGC); MWa is supported by the Research Training Group Translational Research Innovation - Pharma (TRIP), supported by the Else KrönerFresenius Foundation (EKFS). SAJ is supported by the German Research Council (DFG, JO 815/3-2) and the German Cancer Aid-funded PiPAC consortium (70112505). JQ and LW are supported by the NIH/NCI P01 CA066996.

5. Zuber J, Shi J, Wang E, et al. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature.

2011;478(7370):524-528.

6. Salvatori B, Iosue I, Djodji Damas N, et al.

Critical Role of c-Myc in Acute Myeloid Leukemia Involving Direct Regulation of miR-26a and Histone Methyltransferase EZH2. Genes Cancer. 2011;2(5):585-592.

7. Ott G, Rosenwald A, Campo E. Understanding MYC-driven aggressive Bcell lymphomas: pathogenesis and classification. Hematology Am Soc Hematol Educ Program. 2013;2013:575-583.

8. Holien T, Vatsveen TK, Hella H, Waage A, Sundan A. Addiction to c-MYC in multiple

myeloma. Blood. 2012;120(12):2450-2453.

9. Wagner JM, Hackanson B, Lubbert M, Jung M. Histone deacetylase (HDAC) inhibitors in recent clinical trials for cancer therapy.

Clin Epigenetics. 2010;1(3-4):117-136.

10. Filippakopoulos P, Qi J, Picaud S, et al.

Selective inhibition of BET bromodomains.

Nature. 2010;468(7327):1067-1073.

11. Da Costa D, Agathanggelou A, Perry T, et al. BET inhibition as a single or combined therapeutic approach in primary paediatric B-precursor acute lymphoblastic leukaemia. Blood Cancer J. 2013;3:e126.

12. Delmore JE, Issa GC, Lemieux ME, et al.

BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell.

947

M. Wroblewski et al.

2011;146(6):904-917.

13. Amorim S, Stathis A, Gleeson M, et al.

Bromodomain inhibitor OTX015 in patients with lymphoma or multiple myeloma: a dose-escalation, open-label, pharmacokinetic, phase 1 study. Lancet Haematol. 2016;3(4):e196-204.

14. Berthon C, Raffoux E, Thomas X, et al.

Bromodomain inhibitor OTX015 in patients with acute leukaemia: a dose-escalation, phase 1 study. Lancet Haematol.

2016;3(4):e186-195.

15. Horne GA, Stewart HJ, Dickson J, et al.

Nanog requires BRD4 to maintain murine embryonic stem cell pluripotency and is suppressed by bromodomain inhibitor JQ1 together with Lefty1. Stem Cells Dev. 2015;24(7):879-891.

16. Lee DU, Katavolos P, Palanisamy G, et al.

Nonselective inhibition of the epigenetic transcriptional regulator BET induces marked lymphoid and hematopoietic toxicity in mice. Toxicol Appl Pharmacol.

2016;300:47-54.

17. Bianchi E, Bulgarelli J, Ruberti S, et al. MYB controls erythroid versus megakaryocyte lineage fate decision through the miR-4863p-mediated downregulation of MAF. Cell Death Differ. 2015;22(12):1906-1921.

18. Opferman JT, Letai A, Beard C, et al.

Development and maintenance of B and T lymphocytes requires antiapoptotic MCL1. Nature. 2003;426(6967):671-676.

19. Tripathi P, Koss B, Opferman JT, Hildeman DA. Mcl-1 antagonizes Bax/Bak to promote effector CD4(+) and CD8(+) T-cell responses. Cell Death Differ. 2013; 20(8):998-1007.

948

20. Ma A, Pena JC, Chang B, et al. Bclx regulates the survival of double-positive thymocytes. Proc Natl Acad Sci USA. 1995; 92(11):4763-4767.

21. Hu Q, Sader A, Parkman JC, Baldwin TA. Bim-mediated apoptosis is not necessary for thymic negative selection to ubiquitous self-antigens.

J Immunol.

2009; 183(12):7761-7767.

22. Carnevalli LS, Scognamiglio R, CabezasWallscheid N, et al. Improved HSC reconstitution and protection from inflammatory stress and chemotherapy in mice lacking granzyme B. J Exp Med. 2014;211(5):769779.

23. Trabucco SE, Gerstein RM, Evens AM, et al. Inhibition of bromodomain proteins for the treatment of human diffuse large B-cell lymphoma. Clin Cancer Res. 2015; 21(1):113-122.

24. Roe JS, Mercan F, Rivera K, Pappin DJ, Vakoc CR. BET Bromodomain Inhibition Suppresses the Function of Hematopoietic Transcription Factors in Acute Myeloid Leukemia. Mol Cell. 2015;58(6):1028-1039.

25. Sandberg ML, Sutton SE, Pletcher MT, et al. c-Myb and p300 regulate hematopoietic stem cell proliferation and differentiation.

Dev Cell. 2005;8(2):153-166.

26. Bolden JE, Tasdemir N, Dow LE, et al.

Inducible in vivo silencing of Brd4 identifies potential toxicities of sustained BET protein inhibition. Cell Rep. 2014;8(6):19191929.

27. Rahman S, Sowa ME, Ottinger M, et al.

The Brd4 extraterminal domain confers transcription activation independent of pTEFb by recruiting multiple proteins,

28.

29.

30.

31.

32.

33.

34.

including NSD3. Mol Cell Biol.

2011;31(13):2641-2652.

Hamadani M, Craig M, Awan FT, Devine SM. How we approach patient evaluation for hematopoietic stem cell transplantation.

Bone Marrow Transplant. 2010;45(8):12591268.

Ogonek J, Kralj Juric M, Ghimire S, et al.

Immune Reconstitution after Allogeneic Hematopoietic Stem Cell Transplantation.

Front Immunol. 2016;7:507.

Ings SJ, Balsa C, Leverett D, et al. Peripheral blood stem cell yield in 400 normal donors mobilised with granulocyte colony-stimulating factor (G-CSF): impact of age, sex, donor weight and type of G-CSF used. Br J Haematol. 2006;134(5):517-525.

Anderlini P, Rizzo JD, Nugent ML, et al.

Peripheral blood stem cell donation: an analysis from the International Bone Marrow Transplant Registry (IBMTR) and European Group for Blood and Marrow Transplant (EBMT) databases. Bone Marrow Transplant. 2001;27(7):689-692.

Anderlini P, Champlin RE. Biologic and molecular effects of granulocyte colonystimulating factor in healthy individuals: recent findings and current challenges.

Blood. 2008;111(4):1767-1772.

Anderlini P, Donato M, Chan KW, et al.

Allogeneic blood progenitor cell collection in normal donors after mobilization with filgrastim: the M.D. Anderson Cancer Center experience. Transfusion. 1999; 39(6):555-560.

Holig K. G-CSF in Healthy Allogeneic Stem Cell Donors. Transfus Med Hemother.

2013;40(4):225-235.

haematologica | 2018; 103(6)

ARTICLE

Red Cell Biology & its Disorders

Recurring mutations in RPL15 are linked to hydrops fetalis and treatment independence in Diamond-Blackfan anemia

Marcin W. Wlodarski,1,2 Lydie Da Costa,3,4,5,6 Marie-Françoise O'Donohue,7 Marc Gastou,3,4,8 Narjesse Karboul,3,5 Nathalie Montel-Lehry,7 Ina Hainmann,1 Dominika Danda,1,9 Amina Szvetnik,1 Victor Pastor,1,10 Nahuel Paolini,11 Franca M. di Summa,11 Hannah Tamary,12,13 Abed Abu Quider,14 Anna Aspesi,15 Riekelt H. Houtkooper,16 Thierry Leblanc,17 Charlotte M. Niemeyer,1,2 Pierre-Emmanuel Gleizes7 and Alyson W. MacInnes16

Department of Pediatrics and Adolescent Medicine, Division of Pediatric Hematology and Oncology, Medical Center, Faculty of Medicine, University of Freiburg, Germany; 2 German Cancer Consortium (DKTK), Freiburg, Germany and German Cancer Research Center (DKFZ), Heidelberg, Germany; 3University Paris VII Denis Diderot, Faculté de Médecine Xavier Bichat, Paris, France; 4Laboratory of Excellence for Red Cell, LABEX GR-Ex, Paris, France; 5Inserm Unit 1149, CRI, Paris, France; 6Hematology Laboratory, Robert Debré Hospital, Paris, France; 7LBME, Centre de Biologie Intégrative, Université de Toulouse, CNRS, UPS, France; 8UMR1170, Gustave Roussy, Villejuif, France; 9 Department of Tumor Pathology, Centre of Oncology, Maria Sklodowska-Curie Memorial Institute, Poland; 10Faculty of Biology, University of Freiburg, Germany; 11Department of Hematopoiesis, Sanquin and Landsteiner Laboratory, AMC/UvA, CX Amsterdam, the Netherlands; 12Hematology Unit, Schneider Children's Medical Center of Israel, Petach Tikva, Israel; 13Sackler School of Medicine, Tel Aviv University, Israel; 14Pediatric Hematology/Oncology Department, Soroka Medical Center, Faculty of Medicine, BenGurion University, Beer Sheva, Israel; 15Dipartimento di Scienze della Salute, Università del Piemonte Orientale, Novara, Italy; 16Laboratory Genetic Metabolic Diseases, Academic Medical Center, Amsterdam, the Netherlands and 17Pediatric Hematology Service, Robert-Debré Hospital and EA-3518, Université Paris Diderot - Institut Universitaire d'Hématologie, Paris, France 1

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):949-958

ABSTRACT

D

iamond-Blackfan anemia (DBA) is a rare inherited bone marrow failure disorder linked predominantly to ribosomal protein gene mutations. Here the European DBA consortium reports novel mutations identified in the RPL15 gene in 6 unrelated individuals diagnosed with DBA. Although point mutations have not been previously reported for RPL15, we identified 4 individuals with truncating mutations p.Tyr81* (in 3 of 4) and p.Gln29*, and 2 with missense variants p.Leu10Pro and p.Lys153Thr. Notably, 75% (3 of 4) of truncating mutation carriers manifested with severe hydrops fetalis and required intrauterine transfusions. Even more remarkable is the observation that the 3 carriers of p.Tyr81* mutation became treatment-independent between four and 16 months of life and maintained normal blood counts until their last follow up. Genetic reversion at the DNA level as a potential mechanism of remission was not observed in our patients. In vitro studies revealed that cells carrying RPL15 mutations have pre-rRNA processing defects, reduced 60S ribosomal subunit formation, and severe proliferation defects. Red cell culture assays of RPL15-mutated primary erythroblast cells also showed a severe reduction in cell proliferation, delayed erythroid differentiation, elevated TP53 activity, and increased apoptosis. This study identifies a novel subgroup of DBA with mutations in the RPL15 gene with an unexpected high rate of hydrops fetalis and spontaneous, long-lasting remission.

Introduction Diamond-Blackfan anemia (DBA) (OMIM# 105650) is an inherited bone marrow failure disorder that typically manifests in children under the age of one year.

While the central phenotype is pure red cell aplasia, developmental delay and a haematologica | 2018; 103(6)

Correspondence: marcin.wlodarski@uniklinik-freiburg.de

Received: August 4, 2017.

Accepted: March 6, 2018.

Pre-published: March 29, 2018. doi:10.3324/haematol.2017.177980 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/949 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

949

M.W. Wlodarski et al.

number of physical malformations are also linked to DBA.1 These include (but are not limited to) craniofacial malformations, growth retardation, abnormalities in the extremities, heart defects, as well as urogenital defects.2,3 While most patients (>90%) are diagnosed within the first year of life, a minority present with anemia at birth.

Hydrops fetalis due to severe intrauterine anemia is considered a very rare manifestation of DBA.4 To date, a total of 10 cases of DBA-associated hydrops have been reported in single cases.5-13 The clinical outcome of these patients was poor as compared to typical DBA (3 patients died perinatally, 4 patients were steroid unresponsive, 2 patients required steroid therapy, and 1 had unknown outcome).

Almost all of the mutations linked to DBA have been found in genes coding for ribosomal proteins (RPs).14 These RPs include: eS7 (RPS7), uS8 (RPS15A), eS10 (RPS10), eS17 (RPS17), eS19 (RPS19), eS24 (RPS24), eS26 (RPS26), eS27 (RPS27), eS28 (RPS28), uS14 (RPS29), uL18 (RPL5), uL5 (RPL11), eL15 (RPL15), eL18 (RPL18), uL24 (RPL26), eL27 (RPL27), eL31 (RPL31), uL29 (RPL35) and eL33 (RPL35A).15-28 The RPL15 gene has so far been reported in one patient who carried a large monoallelic microdeletion involving this gene.23 Non-RP genes linked to DBA, albeit very rarely involved, are TSR2 and GATA1.26,29 RP gene aberrations lead to haploinsufficiency, and in many cases result in reduction in the respective RP which impairs ribosome biogenesis, affecting both the processing pathways of pre-rRNA maturation and the assembly of the large or small ribosomal subunit.30,31 TP53 is a tumor suppressor protein and transcription factor that stabilizes in response to cell stress, such as DNA damage or nucleolar stress induced by ribosome biogenesis defects.32,33 Depending on the level of stress, stabilized TP53 will induce cell arrest, DNA repair, senescence and/or apoptosis. This pathway has been prominently implicated in the pathogenesis of DBA, with a number of studies suggesting that TP53 stabilization lies at the heart of the loss of erythroid progenitor cells in DBA bone marrow.34-36 In addition to activating apoptosis, DBA-linked RP gene mutations can impair cellular differentiation, alter the landscape of mRNAs on ribosomes, and induce autophagy.36-38 Therapy in DBA depends on the severity of anemia and on the response to oral glucocorticosteroids (GCS). GCSnon-responders receive chronic blood transfusions or can undergo hematopoietic stem cell transplantation (HSCT) which is often delayed because patients can achieve treatment independence.39 For unknown reasons, 20% of all DBA cases can become independent of steroid treatment and/or blood transfusions by the age of 25 years.39,40 To date, however, there are no predictive genetic markers that could improve decision making for timely HSCT. Given the significant risks that are associated with HSCT,41 especially if a matched sibling donor is not available, such predictive markers would be very valuable.

Here we describe 6 DBA cases associated with mutations in ribosomal protein gene RPL15. These patients display an unexpectedly high rate of hydrops fetalis and treatment independence, pointing to a new genotypephenotype correlation in DBA which could be important for risk stratification.

950

Methods Patients Diagnosis of DBA was made based on typical features including aregenerative anemia with erythroid hypoplasia.1 Written informed consent was obtained from patients and/or parents prior to inclusion in this study, which was performed in accordance with the ethical standards of the Declaration of Helsinki.

The study was approved by the institutional ethics committee (University of Freiburg, IRB n. 205/99 and n. 351/17).

Cell culture Lymphoblastoid cell lines (LCLs) were derived from EpsteinBarr virus (EBV)-immortalization of peripheral mononuclear cells isolated from whole blood using Ficoll (GE Life Sciences) and grown in RPMI (Gibco) containing 10% fetal calf serum (FCS), 1% L-glutamine, and 1% penicillin/streptomycin, as previously described.42

DNA sequencing and bioinformatics Targeted Sanger sequencing was performed as previously reported.43 Primer sequences are available upon request.

Pathogenicity of mutations, evolutionary conservation across species, and the physicochemical difference between amino acids were evaluated using standard prediction tools as outlined in Online Supplementary Table S1.

Details of pre-rRNA processing analysis, Northern blots, polysome profiling, measurement of growth rate and de novo protein synthesis are available in the Online Supplementary Methods.

Erythroid red cell culture assays Erythroid red cell culture assays were performed as previously described.36,44 Antibodies and stains for FACS analysis were PC7 or PE conjugated CD34 (Beckman coulter, Brea, CA, USA), APC conjugated CD36 (BD Biosciences San Jose, CA, USA), PE conjugated α4 integrin (Miltenyi, Paris, France), APC conjugated Band 3 (kindly provided by Mohandas Narla’s lab, NYBC, New York, USA), PE/Cy7 conjugated IL-3R (Miltennyi, Paris, France), PE/Cy7 (PE)-coupled GPA (Life Technologies Carlsbad, CA, USA), and DAPI (Sigma). FACS analysis was conducted on a BD Biosciences Influx flow cytometer (BD Biosciences San Jose, CA, USA). Data were analyzed using Kaluza software (Beckman Coulter, Brea, CA, USA). Antibodies for western blotting were TP53 (Sigma #5816), phospho-TP53 (Ser15, Cell Signaling), p21 (Cell Signaling #2947), actin (Sigma #Ac-15), and eL15 (Abcam #ab130992). Primers for real-time PCR used are as follows: TaqMhupRPL15F: CAGCCATCAGGTAAGCCAAGA; TaqMhuRPL15R: CAGCGGACCCTCAGAAGAAA; TaqMhupp21F: TTGCTGCCGCATGGG, TaqMhup21R: CCTTGTGGAGCCGGAGCT; TaqMhupactinF: CTGGAACGGTGAAGGTGACA;TaqMhupactinR: AAGGGACTTCCTGTAACAACGCA; TGTAGTGGATGGTGGTACAGTCAGA; TaqMB2MhF: GCGGCATCTTCAAACCTCC; TaqMB2MhR: TGACTTTGTCACAGCCCAAGATA. Real-time PCR was performed using the ABI 7900 Real Time PCR system and Taqman PCR mastermix (Life Technologies Carlsbad, CA, USA).

Quantification of gene amplification was performed in duplicate using DDCt method. The expression level of each gene was normalized using the housekeeping genes: actin, and β2 microglobulin.

haematologica | 2018; 103(6)

RPL15 mutations in DBA

Results Identification of novel RPL15 mutations in 6 patients within EuroDBA registries Approximately 30% of all registered DBA patients who have been tested for mutations in the most common DBA-linked genes (RPS19, RPL5, RPL11, RPS10, RPS26, RPS7, RPS17, RPS24, and RPL35a) still do not have an established genotype. Because a whole gene deletion of RPL15 has been reported before in one DBA patient but not in our cohorts,23 we used targeted Sanger sequencing of RPL15 to determine if mutations in this gene could be driving disease in patients without an established genotype. The national patient registries from EuroDBA partners in Germany, France, Italy and Israel were included in this study. As of November 2017, these cohorts represent a total of 985 patients. A complete description of the history and composition of the EuroDBA consortium has been recently published.45 Study outline and screening strategy are illustrated in Online Supplementary Figure S1.

Four unrelated DBA patients from the German, and one each from the French and Israeli registries were identified carrying mutations in the RPL15 gene (NM_001253379.1), which encodes ribosomal protein eL15/RPL15 (Figure 1A and B). Out of the 6 RPL15-mutated patients, 3 unrelated patients (P1-3) carried the same novel indel mutation c.242dupA altering the tyrosine at position 81 to a stop codon: p.Tyr81* (Table 1 and Online Supplementary Table S1). Patient 4 (P4) carried another stop-gain mutation in RPL15 resulting in an even earlier protein truncation: c.85C>T; p.Gln29*. The Exome Aggregation Consortium (ExAC) reports that RPL15 is very intolerant to loss-offunction mutations with no reported cases present in over 60,000 individuals (pLI score =0.96), strongly suggesting the novel mutations p.Gln29* and p.Tyr81* found in our patients are highly deleterious (Online Supplementary Table S1). In addition, the analysis of family trios revealed that hotspot mutations arose de novo in two families (Table 1 and Figure 1A), while one patient (P1/DE071) inherited the mutation from her father who had been categorized as

A

B

Figure 1. Mutations in RPL15 are identified in patients with Diamond-Blackfan anemia (DBA). (A) Six unrelated pedigrees of individuals affected by DBA associated with RPL15 mutations. All families have one DBA-affected individual who is also a mutation carrier, as indicated with filled squares (male) or circles (female).

Unaffected individuals are indicated by unfilled symbols. Unaffected mutation carriers are denoted by a dot symbol (ď&#x201A;&#x;). NA: unaffected family members who were not investigated for the presence of mutations. Families 1-4 harbor heterozygous stopgain mutations in RPL15; families 5-6 carry heterozygous missense RPL15 mutations. (B) Schematic representation of human RPL15 depicting localization of the mutations identified in families 1-6.

haematologica | 2018; 103(6)

951

M.W. Wlodarski et al.

a DBA silent carrier due to very high erythrocyte adenosine deaminase (eADA) levels. The remaining two point mutations c.29T>C; p.Leu10Pro (P5) and c.458A>C; p.Lys153Thr (P6) affect highly conserved residues and, based on results from in silico prediction, are probably deleterious (Online Supplementary Table S1). Biomuta, DMDM, the Exac/GnomAD databases, and NCBI do not report any variants in the codon for Leu10. The GnomAD population database reports one variant in Lys153 (Lys153Arg; rs370700905) identified in 33 out of 232840 total alleles.

Genotype-phenotype association for truncating RPL15 mutations: severe hematologic phenotype and rapid acquisition of treatment independence All of the individuals with mutations in RPL15 presented with typical bone marrow erythroid hypoplasia, ele-

vated eADA, and most of them presented with increased fetal hemoglobin (HbF) levels (Table 1). Notably, hydrops fetalis (considered the most severe hematologic phenotype of DBA) was associated only with truncating RPL15 mutations. The affected fetuses P2-4 required between four and nine intrauterine transfusions (Table 1). One individual (P2) with the p.Tyr81* hotspot mutation presented with various physical malformations, while the dysmorphic features in other patients were less severe (Table 1).

Unexpectedly, all 3 patients carrying p.Tyr81* substitution (P1-3) attained a rapid treatment independence both with and without steroid treatment (Figure 2A), while the fourth patient with the RPL15 mutation c.85C>T; p.Gln29* responded to steroids; however, the therapy was discontinued due to overt toxicity.

Based on published observations of genetic revertant mosaicism as a “repair mechanism” in other bone marrow

Table 1. Clinical characteristics of patients with RPL15 mutations.

Pat/ ID (sex)

RPL15 gene; Hematology and therapies mutation

1/DE071 (F)

c.242dupA; p.Tyr81*

2/DE189 (M)

c.242dupA; p.Tyr81*

3/DE115 (F)

c.242dupA; p.Tyr81*

4/DE202 (M)

c.85C>T; p.Gln29*

5/IL (F)

c.29T>C; p.Leu10Pro

6/FR (M)

c.458A>C; p.Lys153Thr

Gestational age; malformations; other

Age and status at last follow up

Onset: 3 months old; Hb 4.6g/dL 6.5 years, Lab: MCV↑, eADA↑ (925U/Iec), HbF normal normal Hb, Evolution: spontaneous recovery after 1 transfusion 37 weeks, IUGR no therapies at 6 months old. Relapse at 5.7 years (Hb 3.6g/dL), achieved remission after short course of steroids Onset: prenatal (4 intrauterine transfusions) 34+1 weeks, hydrops fetalis, ptosis, 5 years, Lab: MCV↑, eADA↑ (1526U/Iec), HbF (6.6%) flat nose, deep set ears, normal Hb, Evolution: 4 transfusions (birth-16 months), intermittent AV-block, duplex no therapies achieved remission after short course left kidney, hypogonadism, of steroids intersexual genitalia (46XY), microcephaly, left cerebellar hypoplasia, developmental disorder with mental retardation and cerebral palsy Onset: prenatal (6 intrauterine transfusions), 35+6 weeks, hydrops fetalis, 16 years, Lab: MCV↑, eADA↑ (1284U/Iec), HbF↑ (10.6%) IUGR, PFO normal Hb, Evolution: 2 transfusions after birth, achieved no therapies spontaneous remission at age of 4 months Onset: prenatal (9 intrauterine transfusions) 32+3 weeks, 18 years, Lab: MCV↑, eADA ↑ (2628U/Iec), HbF↑ (2%) hydrops fetalis, regular Evolution: after birth irregular transfusions, hypogammaglobulinemia, transfusions steroid-responsive (4-9 years), discontinued plantar warts due to toxicity, transfusion-dependent from age during steroid therapy of 9 years Onset: 4 months old, Hb 5g/dL 27 weeks 2 years, Lab: MCV↑, eADA and HbF unknown (placenta previa), steroids Evolution: transfusion dependent, recently started none steroids with good response Onset: 6 months old, Hb 7g/dL Lab: MCV normal, eADA unknown, HbF↑ (15%) Evolution: initially steroid-responsive, transfusion dependent from age of 5 years

41 weeks, low-set hair line; growth retardation and mental retardation

22 years, regular transfusions

Family history Father mutation carrier, eADA↑ (1396U/Iec) but clinically silent Parents and sister wild type

Parents wild type

Parents and sibling healthy (carrier status unknown), eADA/Hbf unknown Mother mutation carrier: Hb, MCV and HbF normal, eADA unknown Parents and sibling healthy (carrier status unknown)

All patients presented with DBA-typical erythroblastopenia in bone marrow. Pat: patient number; ID: patient identifier in respective national registry; RP: ribosomal protein; F: female, M: male; Hb: hemoglobin; Lab: supportive laboratory parameters; MCV: mean corpuscular volume; eADA: erythrocyte adenosine deaminase; Evolution: evolution of disease and therapies; ↑ elevated for age; HbF: fetal hemoglobin; Remission: treatment independence; IUGR: intrauterine growth restriction; PFO: persistent foramen ovale.

952

haematologica | 2018; 103(6)

RPL15 mutations in DBA

failure syndromes,46 we speculated that the hematologic remission in our patients was attained via genetic reversion (e.g. uniparental disomy). Sequencing of DNA in P1-3 either at multiple time points or after the acquisition of treatment independence did not reveal any changes in mutation load (Figure 2B).

Mutations in RPL15 recapitulate specific pre-rRNA processing defects found in RP depleted cells Mutations in DBA-linked RP genes result in haploinsufficiency of the encoded RP. Since the majority of RPs incorporate into pre-ribosomal particles in a progressive

manner concurrently with pre-rRNA maturation, a lack of one RP impairs pre-rRNA processing in an explicit way that is reproducible in most cell types.47 Online Supplementary Figure S2 illustrates the normal pre-rRNA processing pathway that begins with a single transcribed strand of 47S pre-rRNA that then undergoes a complex series of cleavages and trimmings to ultimately result in the mature 18S, 28S, and 5.8S strands of rRNA found in ribosomes. To investigate the functional consequences of the newly identified RP gene mutations reported here, EBV-immortalized lymphoblast cell lines (LCLs) were generated and the pre-rRNA processing of these LCLs was

A

B

Figure 2. Longitudinal mutations in RPL15 are identified in individuals diagnosed with Diamond-Blackfan anemia (DBA). (A) Clinical evolution of P1-P3 carrying truncating hotspot mutation RPL15 p.Tyr81*. Patient 1 manifested with DBA after birth and after one transfusion achieved spontaneous remission at the age of six months. A relapse occurred five years later and after a short course of steroids the patient attained treatment independence. Patients 2 and 3 had a similar clinical course with hydrops fetalis and prenatal intrauterine transfusions, and achieved treatment independence either spontaneously or after one course of steroids. (B) Sanger sequencing results of the recurring mutation in RPL15 from initial diagnoses as well as post remission. Arrows indicate the inserted A nucleotide. FUP: follow up; BM: bone marrow; PB: peripheral blood; LCL: lymphoblastoid cell lines.

haematologica | 2018; 103(6)

953

M.W. Wlodarski et al.

compared to HeLa cells depleted of eL15.

As previously reported,23 depletion of eL15 in HeLa cells resulted in a decrease in 32S and 12S pre-rRNAs (Figure 3A and C). This was accompanied by the accumulation of the 36S and 36S-C precursors, which are inconspicuous in normal cells, and a concomitant drop in 32.5S pre-rRNA (Figure 3B). In addition, these cells displayed lower levels of 30S pre-rRNA and higher levels of 41S and 18S-E prerRNAs (Figure 3A and C). This phenotype indicates that depletion of eL15 affects cleavage of the ITS1 at site 2, which promotes direct cleavage of early precursors at site E and formation of 18S-E and 36S pre-rRNAs. The 36S precursor is then trimmed by the 5’-3’ exoRNase XRN2 to produce the 36S-C and 32.5S pre-rRNAs.48 Sucrose gradient analyses confirmed the efficiency of the eL15 siRNAs in reducing the free 60S subunit peak and inducing the formation of half-mers, as expected from the deficiency of an RPL protein (Online Supplementary Figure S3). Consistent with eL15 partial loss-of-function, an increase of 36S and

A

36S-C precursors was observed in the patient LCLs carrying the eL15 p.Tyr81* variant (Figure 3B). In addition, both cell lines displayed a marked increase of the amount of 18S-E precursors translating into a higher 18S-E/21S ratio, similar to eL15-depleted cells (Figure 3A and C). This phenotype is similar to that previously observed in LCLs harboring a large heterozygous deletion in RPL15.23

RPL15 mutations impair 60S ribosomal subunit formation, cell proliferation, and de novo protein synthesis Defective processing of pre-rRNA in cells often results in cells that are unable to fully form ribosomal subunits.

As such, pathogenic variants of RPs linked to DBA very often impair biogenesis of ribosomal subunits in LCLs.37 Polysome profiling was performed to determine if the observed pre-rRNA biogenesis defects resulted in impaired biogenesis of large ribosomal subunits. Figure 4A shows polysome profiles of LCL extracts derived from a

B

C

Figure 3. Mutations in RPL15 recapitulate specific pre-rRNA processing defects found in eL15 depleted cells. (A) Northern blot analysis of siRNA-treated HeLa cells or lymphoblastoid cell lines (LCLs) derived from individuals with Diamond-Blackfan anemia (DBA). Radiolabeled probes against ITS2 (top panel), ITS1 (middle panel), 18S or 28S (lower panel) rRNA sequences were used to blot 3µg total RNA isolated from cells. (B) Longer exposures of upper molecular weight pre-rRNA species observed in the northern blots from (A). Intensity profiles of the lanes is shown in the right-hand boxes. (C) Quantification of rRNA precursors in siRNA-treated HeLa cells (top) or LCLs (bottom) derived from individuals with DBA. The results of single experiments performed for each sample are displayed as multiple bars. Pre-rRNA ratios are normalized by dividing by the mean of the control samples.

954

haematologica | 2018; 103(6)

RPL15 mutations in DBA

healthy individual or patients carrying the p.Tyr81* variant in eL15. As expected, the LCLs from the healthy individual reveal an equivalent ratio of 40S to 60S peaks, while the LCLs derived from patients reveal a substantial reduction of 60S peaks compared to 40S peaks. These results suggest the pre-rRNA defects depicted in Figure 3 go on to impair biogenesis of large 60S ribosomal subunits.

The impairment of ribosome biogenesis in cells can slow the rate of protein synthesis and cell proliferation. RP gene mutations linked to DBA are reported to slow the rate of cell proliferation and increase apoptosis.36 In agreement, we found that LCLs carrying the eL15 p.Tyr81* vari-

A

ant proliferated far more slowly than healthy control LCLs (Figure 4B). Moreover, measurement of de novo protein synthesis by Click-iTÂŽ labeling analysis revealed a severe reduction in the synthesis rate of LCLs carrying the eL15 variant (Figure 4C and Online Supplementary Figure S4).

Erythroid cell culture assays of primary RPL15 c.242dupA cells reveal severe proliferation defects, differentiation delays, and TP53-related apoptosis Erythroid cell culture assays show that hematopoietic progenitor cells can reveal reduced proliferation rates, delayed differentiation, and increased TP53-induced apoptosis in a manner that is largely dependent on which

B

C

Figure 4. RPL15 mutations impair cell proliferation, de novo protein synthesis, and 60S ribosomal subunit formation. (A) Representative (n=3) polysome profiles of lymphoblastoid cell line (LCL) extracts derived from healthy individuals or individuals with Diamond-Blackfan anemia (DBA) carrying the RPL15 c.242dupA mutation. The 40S small subunit, 60S subunit, 80S monosome, and polysomes are labeled. Arrows point to the reduced 60S peaks in cells with RPL15 mutations. (B) Growth curve of LCLs derived from individuals with DBA or healthy controls over six days. Standard Deviations for healthy control-1 cells on days 1-5 are 2.6e4, 1.5e4, 3.8e4, 6.9e4, 1.0e5; healthy control-2 cells are 8.8e3, 3.0e4, 2.9e4, 7.5e4, 8.5e4; RPL15 c.242dupA-1 are 2.1e4, 1.9e4, 1.6e4, 2.3e4, 2.2e4; and RPL15 c.242dupA-2 are 9.5e3, 1.7e4, 1.7e4, 2.2e4, 2.5e4. (C) Measurement of the amount of de novo protein synthesis in 30 minutes in LCLs derived from healthy individuals or a DBA patient carrying the eL15 Tyr81* variant using Click-iTÂŽ analysis.

haematologica | 2018; 103(6)

955

M.W. Wlodarski et al.

RP gene is haploinsufficient.36 We performed these assays in CD34+ cells isolated from bone marrow mononuclear cells (BM-MNC) of 2 patients with truncating RPL15 mutations p.Tyr81*. Compared to healthy control,

A

B

C

D

E

F

H

I

patients BM-MNC showed a higher rate of cell death and apoptosis (Figures 5B and Online Supplementary Figure S5).

FACS analysis revealed CD36 downregulation and CD34 upregulation on both day 7 and day 9 in the Tyr81*

G

Figure 5. Erythroid cell culture assays of primary RPL15 c.242dupA erythroid progenitor and precursor cells reveal severe erythroid proliferation defects, differentiation delays, and TP53-related apoptosis. (A) Proliferation curve of erythroid cells isolated from CD34+ cells from peripheral blood of 2 individuals with DiamondBlackfan anemia (DBA) and an RPL15 mutation or a healthy control over nine days in liquid culture medium. (B) FACS analysis results of the percent of dead cells or apoptotic cells staining positive for Annexin V on days 7, 10, and 13 after plating in red cell culture medium. (C) FACS analysis results on days 7 and 9 of the percent of cells staining positive for CD36 and CD34. (D) FACS analysis results on days 7 and 9 of cells staining positive for IL3R and GPA. (E) FACS analysis results on days 7 and 9 of cells staining positive for Alpha-4 and Band-3. (F) Real-time PCR results of the ratio between RPL15 mRNA and actin mRNA in cells. (G) Real-time PCR results of the ratio between p21 mRNA and actin mRNA in cells. (H) Western blot analysis of cells from a healthy control or an individual with DBA and a mutation in RPL15 using antibodies against phosphorylated TP53, p21, eL15, or actin on day 7. (I) Western blot analysis of cells from a healthy control or an RPL15 with DBA and a mutation in RPL15 using antibodies against TP53 or actin on days 7 and 9.

956

haematologica | 2018; 103(6)

RPL15 mutations in DBA

mutant cells as compared to healthy controls (Figures 5C and Online Supplementary Figure S5). In addition, Tyr81* mutant cells retain a higher expression of the IL3 receptor that is normally down-regulated in the erythroid lineage, and express less erythroid-specific glycophorin A (GPA), Band-3, or alpha-4 integrin compared to healthy control cells (Figure 5D and E and Online Supplementary Figure S5).

Additionally, expression of RPL15 mRNA was lower (Figure 5F) and of p21 mRNA higher in Tyr81* mutant cells (Figure 5G). The observed p21 overexpression is likely TP53-mediated, as supported by western blot analysis confirming an increase in TP53 phosphorylation and increased p21 protein expression (Figure 5H). Finally, TP53 protein stabilization was observed in Tyr81* mutant patient cells (Figure 5I).

Discussion At present there are 19 RP genes associated with DBA, representing almost one-quarter of the 80 cytoplasmic RPs in human cells. Here, we describe a novel genetic subgroup of DBA due to mutations in the RPL15 gene and identify the truncating mutation c.242dupA; p.Tyr81* as a recurrent genetic cause in 3 unrelated patients. The experimental results show that LCLs carrying the p.Tyr81* variant reveal an array of molecular defects typically associated with a ribosomopathy phenotype. These include impaired pre-rRNA processing, reduced 60S ribosomal subunit formation, a reduction in de novo protein synthesis, and impaired cell proliferation.

Further, mutant LCLs phenocopied biological changes observed in HeLa cells depleted of eL15. These results, similar to previous reports that DBA-linked RP mutations impair pre-rRNA processing, indicate that the RPL15 mutations reported here are likely pathogenic and result in haploinsufficiency.

The altered ribosome biogenesis may explain the observed reduction in global protein synthesis, substantiating previous reports discussing altered translation of specific mRNAs in DBA-mutant cells as a key component of disease pathogenesis.49,50 In support of this pathogenesis, we also found that hematopoietic stem cells with the eL15 Tyr81* variant that were induced to differentiate into erythrocytes revealed a substantial decrease in the number of erythroid colonies and delays in differentiation. These findings, in addition to increased apoptosis, TP53 stabilization, and p21 overexpression in RPL15-mutant hematopoiesis, suggest that eL15 plays a critical role in ribosome biogenesis and that the reduction of the protein by genetic haploinsufficiency drives severe stress in the nucleolus.

Hydrops fetalis arising from severe intrauterine anemia is an uncommon manifestation of DBA with 10 cases reported so far. The clinical outcome of these patients was poor and, unexpectedly, no spontaneous remission was observed. In contrast, 50% (3 of 6) of all patients with RPL15 mutations reported in our study showed prenatal manifestation with the necessity of intrauterine transfusions. Remarkably, the hydrops cases were observed only in the subgroup of patients carrying protein truncating mutations p.Tyr81* and p.Gln29*, with prevalence of hydrops reaching 75% (3 of 4).

haematologica | 2018; 103(6)

A substantial percentage of individuals with DBA spontaneously achieve treatment independence at some point; however, the reasons for this remain elusive and no predictive biomarkers exist.39 In line with this, another unexpected finding was the rapid and sustained treatment independence achieved within four to 16 months after birth in all 3 unrelated DBA patients with the recurrent p.Tyr81* mutation. Although Patient 1 relapsed five years after achieving spontaneous remission, a short steroid course was very effective and resulted in treatment independence. These observations suggest that p.Tyr81* mutation carriers can achieve treatment independence with or without a previous course of steroids, and that a genotype itself may be an important biomarker for the expected clinical course. Our findings might help prospective clinical stratification to determine which individuals are more likely to become treatment independent without the need for HSCT. The reasons why the specific mutation p.Tyr81* drives treatment independence remain unknown and cannot be explained by a potential genetic reversion, which was not observed in our patients. One tempting speculation is that the remaining wild-type allele compensates for the haploinsufficiency (e.g. by epigenetic mechanisms) in the adult hematopoiesis, but the compensation might not be present or sufficient in the fetus.

These findings warrant further studies on the future availability of more patient material.

In conclusion, our study establishes germline point mutations in the RPL15 gene as novel genetic etiology of DBA. Half of the individuals carrying these mutations manifest with hydrops fetalis, a phenotype that is very rarely observed in DBA patients. Finally, we establish that a recurrent DBA genotype, eL15 p.Tyr81*, is linked to treatment independence.

Acknowledgments Very special thanks go to all the individuals who consented to participate in this study and their families. We thank Alexandra Fischer (Freiburg) for data management, Sandra Zolles, Sophia Hollander, Dirk Lebrecht, and Gunda Ruzaike (Freiburg) for technical assistance, Pritam Kumar Panda (Freiburg) for bioinformatic analysis, and Dr. Marije Bartels (UMC Utrecht) for critical reading of the manuscript.

Funding AM, ML, MW, LDC, NK, HT, PEG, and MFOD are supported under the framework of E-Rare-2, ERA-Net for Research on Rare Diseases (ZonMW #113301205 and #40-44000-981008 in the Netherlands; #BMBF 01GM1301 and 01GM1609 in Germany; Chief Scientist Office, Israeli Ministry of Health # 3-12844 in Israel; #ANR-15-RAR3-0007-04 and #ANR-12-RARE-0007-02 in France). MW and CN are supported by DKTK German Cancer Consortium, fot molecular diagnostics of pediatric malignancies. LDC was supported by the Laboratory of Excellence for Red Cells [(LABEX GR-Ex)-ANR Avenir-11-LABX-0005-02], the French National PHRC OFABD (DBA registry and molecular biology), and the “Fondation ARC pour la recherche contre le cancer”. LDC, MFOD and PEG are funded by the French National Research Agency [ANR-DBA-Multigenes-ANR-2015-AAP génériqueCE12]. AA was supported by DBA Telethon grant GGP13177 and Fondazione Europea per la DBA and Gruppo di Sostegno DBA Italia.

957

M.W. Wlodarski et al.

References 18.

1. Vlachos A, Ball S, Dahl N, et al. Diagnosing and treating Diamond Blackfan anaemia: results of an international clinical consensus conference. Br J Haematol. 2008; 142(6):859-876.

2. Ball SE, McGuckin CP, Jenkins G, GordonSmith EC. Diamond-Blackfan anaemia in the U.K.: analysis of 80 cases from a 20year birth cohort. Br J Haematol.

1996;94(4):645-653.

3. Lipton JM, Atsidaftos E, Zyskind I, Vlachos A. Improving clinical care and elucidating the pathophysiology of Diamond Blackfan anemia: an update from the Diamond Blackfan Anemia Registry. Pediatr Blood Cancer. 2006;46(5):558-564.

4. Vlachos A, Klein GW, Lipton JM. The Diamond Blackfan Anemia Registry: tool for investigating the epidemiology and biology of Diamond-Blackfan anemia. J Pediatr Hematol Oncol. 2001;23(6):377382.

5. Semmekrot BA, Haraldsson A, Weemaes CM, Smeets DF, Geven WB, Brunner HG. Absent thumb, immune disorder, and congenital anemia presenting with hydrops fetalis. Am J Med Genet. 1992;42(5):736740.

6. Van Hook JW, Gill P, Cyr D, Kapur RP. Diamond-Blackfan anemia as an unusual cause of nonimmune hydrops fetalis: a case report. J Reprod Med. 1995;40(12):850-854.

7. McLennan AC, Chitty LS, Rissik J, Maxwell DJ. Prenatal diagnosis of BlackfanDiamond syndrome: case report and review of the literature. Prenat Diagn.

1996;16(4):349-353.

8. Rogers BB, Bloom SL, Buchanan GR. Autosomal dominantly inherited Diamond-Blackfan anemia resulting in nonimmune hydrops. Obstet Gynecol.

1997;89(5 Pt 2):805-807.

9. Scimeca PG, Weinblatt ME, Slepowitz G, Harper RG, Kochen JA. Diamond-Blackfan syndrome: an unusual cause of hydrops fetalis. Am J Pediatr Hematol Oncol. 1988 Fall;10(3):241-243.

10. Dunbar AE, 3rd, Moore SL, Hinson RM. Fetal Diamond-Blackfan anemia associated with hydrops fetalis. Am J Perinatol.

2003;20(7):391-394.

11. Saladi SM, Chattopadhyay T, Adiotomre PN. Nomimmune hydrops fetalis due to Diamond-Blackfan anemia. Indian Pediatr.

2004;41(2):187-188.

12. Valsky DV, Daum H, Yagel S. Reversal of mirror syndrome after prenatal treatment of Diamond-Blackfan anemia. Prenat Diagn. 2007;27(12):1161-1164.

13. Da Costa L, Chanoz-Poulard G, Simansour M, et al. First de novo mutation in RPS19 gene as the cause of hydrops fetalis in Diamond-Blackfan anemia. Am J Hematol.

2013;88(4):340-341.

14. Boria I, Garelli E, Gazda HT, et al. The ribosomal basis of Diamond-Blackfan Anemia: mutation and database update. Hum Mutat. 2010;31(12):1269-1279.

15. Ban N, Beckmann R, Cate JH, et al. A new system for naming ribosomal proteins.

Curr Opin Struct Biol. 2014;24:165-169.

16. Cmejla R, Cmejlova J, Handrkova H, Petrak J, Pospisilova D. Ribosomal protein S17 gene (RPS17) is mutated in DiamondBlackfan anemia. Hum Mutat. 2007; 28(12):1178-1182.

17. Doherty L, Sheen MR, Vlachos A, et al.

Ribosomal protein genes RPS10 and RPS26 are commonly mutated in Diamond-

958

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Blackfan anemia. Am J Hum Genet. 2010; 86(2):222-228.

Draptchinskaia N, Gustavsson P, Andersson B, et al. The gene encoding ribosomal protein S19 is mutated in DiamondBlackfan anaemia.

Nat Genet.

1999;21(2):169-175.

Farrar JE, Nater M, Caywood E, et al.

Abnormalities of the large ribosomal subunit protein, Rpl35a, in Diamond-Blackfan anemia. Blood. 2008;112(5):1582-1592.

Gazda HT, Grabowska A, Merida-Long LB, et al. Ribosomal protein S24 gene is mutated in Diamond-Blackfan anemia. Am J Hum Genet. 2006;79(6):1110-1118.

Gazda HT, Preti M, Sheen MR, et al.

Frameshift mutation in p53 regulator RPL26 is associated with multiple physical abnormalities and a specific pre-ribosomal RNA processing defect in diamond-blackfan anemia. Hum Mutat. 2012;33(7):10371044.

Gazda HT, Sheen MR, Vlachos A, et al.

Ribosomal protein L5 and L11 mutations are associated with cleft palate and abnormal thumbs in Diamond-Blackfan anemia patients. Am J Hum Genet. 2008;83(6):769780.

Landowski M, O'Donohue MF, Buros C, et al. Novel deletion of RPL15 identified by array-comparative genomic hybridization in Diamond-Blackfan anemia. Hum Genet.

2013;132(11):1265-1274.

Wang R, Yoshida K, Toki T, et al. Loss of function mutations in RPL27 and RPS27 identified by whole-exome sequencing in Diamond-Blackfan anaemia. Br J Haematol.

2015;168(6):854-864.

Mirabello L, Macari ER, Jessop L, et al.

Whole-exome sequencing and functional studies identify RPS29 as a novel gene mutated in multicase Diamond-Blackfan anemia families. Blood. 2014;124(1):24-32.

Gripp KW, Curry C, Olney AH, et al.

Diamond-Blackfan anemia with mandibulofacial dystostosis is heterogeneous, including the novel DBA genes TSR2 and RPS28. Am J Med Genet A. 2014;164A(9):2240-2249.

Ikeda F, Yoshida K, Toki T, et al. Exome sequencing identified RPS15A as a novel causative gene for Diamond-Blackfan anemia. Haematologica. 2017;102(3):e93-e96.

Mirabello L, Khincha PP, Ellis SR, et al.

Novel and known ribosomal causes of Diamond-Blackfan anaemia identified through comprehensive genomic characterisation. J Med Genet. 2017;54(6):417-425.

Sankaran VG, Ghazvinian R, Do R, et al.

Exome sequencing identifies GATA1 mutations resulting in Diamond-Blackfan anemia. J Clin Invest. 2012;122(7):2439-2443.

Choesmel V, Bacqueville D, Rouquette J, et al. Impaired ribosome biogenesis in Diamond-Blackfan anemia.

Blood.

2007;109(3):1275-1283.

Ferreira-Cerca S, Poll G, Gleizes PE, Tschochner H, Milkereit P. Roles of eukaryotic ribosomal proteins in maturation and transport of pre-18S rRNA and ribosome function. Mol Cell. 2005;20(2):263-275.

Golomb L, Volarevic S, Oren M. p53 and ribosome biogenesis stress: the essentials.

FEBS Lett. 2014;588(16):2571-2579.

Kastan MB, Onyekwere O, Sidransky D, Vogelstein B, Craig RW. Participation of p53 protein in the cellular response to DNA damage. Cancer Res. 1991;51(23 Pt 1):63046311.

Dutt S, Narla A, Lin K, et al.

Haploinsufficiency for ribosomal protein genes causes selective activation of p53 in

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

human erythroid progenitor cells. Blood.

2011;117(9):2567-2576.

Jaako P, Flygare J, Olsson K, et al. Mice with ribosomal protein S19 deficiency develop bone marrow failure and symptoms like patients with Diamond-Blackfan anemia.

Blood. 2011;118(23):6087-6096.

Moniz H, Gastou M, Leblanc T, et al.

Primary hematopoietic cells from DBA patients with mutations in RPL11 and RPS19 genes exhibit distinct erythroid phenotype in vitro. Cell Death Dis. 2012; 3:e356.

Pereboom TC, Bondt A, Pallaki P, et al.

Translation of branched-chain aminotransferase-1 transcripts is impaired in cells haploinsufficient for ribosomal protein genes.

Exp Hematol. 2014;42(5):394-403.e4.

Heijnen HF, van Wijk R, Pereboom TC, et al. Ribosomal protein mutations induce autophagy through S6 kinase inhibition of the insulin pathway. PLoS Genet. 2014; 10(5):e1004371.

Vlachos A, Muir E. How I treat DiamondBlackfan anemia. Blood. 2010;116(19): 3715-3723.

Narla A, Vlachos A, Nathan DG. Diamond Blackfan anemia treatment: past, present, and future. Semin Hematol. 2011; 48(2):117-123.

Peffault de Latour R, Peters C, Gibson B, et al. Recommendations on hematopoietic stem cell transplantation for inherited bone marrow failure syndromes. Bone Marrow Transplant. 2015;50(9):1168-1172.

Hui-Yuen J, McAllister S, Koganti S, Hill E, Bhaduri-McIntosh S. Establishment of Epstein-Barr virus growth-transformed lymphoblastoid cell lines. J Vis Exp. 2011;(57).

Hirabayashi S, Flotho C, Moetter J, et al.

Spliceosomal gene aberrations are rare, coexist with oncogenic mutations, and are unlikely to exert a driver effect in childhood MDS and JMML. Blood. 2012; 119(11):e96-99.

Hu J, Liu J, Xue F, et al. Isolation and functional characterization of human erythroblasts at distinct stages: implications for understanding of normal and disordered erythropoiesis in vivo. Blood. 2013; 121(16):3246-3253.

Da Costa L, O'Donohue MF, van Dooijeweert B, et al. Molecular approaches to diagnose Diamond-Blackfan anemia: The EuroDBA experience. Eur J Med Genet. 2017 Oct 26. [Epub ahead of print] Lo Ten Foe JR, Kwee ML, Rooimans MA, et al. Somatic mosaicism in Fanconi anemia: molecular basis and clinical significance.

Eur J Hum Genet. 1997;5(3):137-148.

Henras AK, Plisson-Chastang C, O'Donohue MF, Chakraborty A, Gleizes PE. An overview of pre-ribosomal RNA processing in eukaryotes. Wiley Interdiscip Rev RNA. 2015;6(2):225-242.

Preti M, O'Donohue MF, Montel-Lehry N, Bortolin-Cavaille ML, Choesmel V, Gleizes PE. Gradual processing of the ITS1 from the nucleolus to the cytoplasm during synthesis of the human 18S rRNA. Nucleic Acids Res. 2013;41(8):4709-4723.

Ludwig LS, Gazda HT, Eng JC, et al.

Altered translation of GATA1 in DiamondBlackfan anemia. Nat Med. 2014;20(7):748753.

Horos R, Ijspeert H, Pospisilova D, et al.

Ribosomal deficiencies in DiamondBlackfan anemia impair translation of transcripts essential for differentiation of murine and human erythroblasts. Blood.

2012;119(1):262-272.

haematologica | 2018; 103(6)

ARTICLE

Myelodysplastic Syndromes

Cytokines increase engraftment of human acute myeloid leukemia cells in immunocompromised mice but not engraftment of human myelodysplastic syndrome cells

Maria Krevvata,1,* Xiaochuan Shan,1,* Chenghui Zhou,1 Cedric Dos Santos,1 Georges Habineza Ndikuyeze,1 Anthony Secreto,1 Joshua Glover,1 Winifred Trotman,1 Gisela Brake-Silla,1 Selene Nunez-Cruz,2 Gerald Wertheim,3 Hyun-Jeong Ra,1 Elizabeth Griffiths,4 Charalampos Papachristou,5 Gwenn Danet-Desnoyers,1,# and Martin Carroll1,6,#

Division of Hematology and Oncology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA; 2Center for Cellular Immunotherapies, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA; 3Department of Pathology and Laboratory Medicine, Childrenâ&#x20AC;&#x2122;s Hospital of Philadelphia and The University of Pennsylvania Perelman School of Medicine, Philadelphia, PA; 4Roswell Park Cancer Institute, Buffalo, NY; 5Department of Mathematics, Rowan University, Glassboro, NJ and 6Veterans Administration Hospital, Philadelphia, PA, USA 1

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):959-971

*MK and XS share co-authorship. #GD-D and MC share senior authorship.

ABSTRACT

P

atient-derived xenotransplantation models of human myeloid diseases including acute myeloid leukemia, myelodysplastic syndromes and myeloproliferative neoplasms are essential for studying the biology of the diseases in pre-clinical studies. However, few studies have used these models for comparative purposes. Previous work has shown that acute myeloid leukemia blasts respond to human hematopoietic cytokines whereas myelodysplastic syndrome cells do not. We compared the engraftment of acute myeloid leukemia cells and myelodysplastic syndrome cells in NSG mice to that in NSG-S mice, which have transgene expression of human cytokines. We observed that only 50% of all primary acute myeloid leukemia samples (n=77) transplanted in NSG mice provided useful levels of engraftment (>0.5% human blasts in bone marrow). In contrast, 82% of primary acute myeloid leukemia samples engrafted in NSG-S mice with higher leukemic burden and shortened survival. Additionally, all of 5 injected samples from patients with myelodysplastic syndrome showed persistent engraftment on week 6; however, engraftment was mostly low (<2%), did not increase over time, and was only transiently affected by the use of NSG-S mice. Co-injection of mesenchymal stem cells did not enhance human myelodysplastic syndrome cell engraftment. Overall, we conclude that engraftment of acute myeloid leukemia samples is more robust compared to that of myelodysplastic syndrome samples and unlike those, acute myeloid leukemia cells respond positively to human cytokines, whereas myelodysplastic syndrome cells demonstrate a general unresponsiveness to them.

Introduction Human myeloid neoplasms represent a remarkably diverse array of blood cell diseases. Acute myeloid leukemia (AML) is a clonal hematopoietic disease characterized by an abnormal proliferation of immature leukemic blasts and by a hematopoietic differentiation block.1 Myelodysplastic syndromes (MDS) are characterized by abnormal cell morphology and ineffective blood cell production. MDS mainly affect the elderly and their pathogenesis is not completely understood but haematologica | 2018; 103(6)

Correspondence: carroll2@pennmedicine.upenn.edu

Received: October 23, 2017.

Accepted: February 22, 2018.

Pre-published: March 15, 2018.

doi:10.3324/haematol.2017.183202 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/959 Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

959

M. Krevvata et al.

they are thought to arise from a single transformed hematopoietic cell.2-4 Both AML and MDS are genetically heterogeneous making functional characterization of primary human cells essential for studies of disease pathogenesis. However, primary cells from neither of these diseases survive well in vitro, making the use of xenotransplantation models essential for the study of primary cells.

With the rapid growth in the number of immunocompromised mouse strains modified to express human proteins, studies using mice as bioreactors for human cells to test specific in vitro observations have become feasible.

However, little work has previously been done studying how the recipient mouse affects the biology of the human disease cells. Here we compare the effect of use of NSG versus NSG-S mice on the relative engraftment and growth of human AML and MDS samples.

Collective studies for over three decades have described the contributions of the bone marrow microenvironment to normal hematopoiesis. Since the description of the bone marrow niche by Schofield,5 the regulation of normal hematopoietic stem cell homeostasis by mechanisms involving non-hematopoietic cells has been extensively investigated. It is now well understood that normal stem cell self-renewal is tightly regulated, in part, by cell-extrinsic mechanisms.6-10 Taichman and Emerson have shown that cytokines produced by osteoblasts promote proliferation of hematopoietic cells in culture11 whereas increases in osteoblast numbers in a mouse model with constitutively active osteoblast-specific parathyroid hormone resulted in a simultaneous increase of hematopoietic stem cells.12 As with normal hematopoiesis, several hematopoietic malignancies persist by maintaining a pool of malignant stem cells that may be partly protected by components of the microenvironment.13,14 Conversely, leukemic stem cells induce alterations in hematopoietic regulatory functions to gain growth advantage over normal hematopoietic stem cells.15,16 Schepers et al. have shown that leukemic myeloid cells secrete high levels of pro-inflammatory cytokines, creating a paracrine feedback loop that drives myeloid differentiation. At the same time, myeloid cells stimulate mesenchymal stem cells (MSC) to overproduce functionally altered osteoblastic cells with compromised ability to maintain normal hematopoietic stem cells.17 Thus, cytokine production by the bone marrow microenvironment may modify the phenotype of malignant blood diseases.

We previously demonstrated that the NSG (NOD-ScidIL-2Rgcnull) mouse is a robust recipient for human AML xenotransplantation samples, allowing a better understanding and characterization of AML biology, especially in the context of drug therapy studies.18 However, we observed that a significant proportion of primary AML specimens showed low (0.1 to 1% human blasts in mouse bone marrow) or no (<0.1%) engraftment in NSG mice, suggesting the need for improved xenograft models.18 Transgenic expression of human stem cell factor, granulocyte-macrophage colony-stimulating factor and interleukin-3 in NSG-S mice has been reported to enhance engraftment of primary AML samples, although only a few AML patients were compared between strains.19 These studies did not allow for a rigorous determination of the percentage of patientsâ&#x20AC;&#x2122; samples that engraft (an assessment for stem cell effects) or the bulk of disease burden (reflecting growth after engraftment).

960

A recent study attempting to develop a patient-derived xenotransplantation model for human MDS suggested that patient-derived MSC combined with the use of NSG-S immunodeficient mice, could enhance engraftment levels.20 Indeed, the use of NSG-S mice appeared to improve engraftment levels and also maintained the malignant clone, but these studies were largely done with accompanying injection of MSC so the critical variables for engraftment of MDS samples were largely undetermined.

In this report, we compare the engraftment levels in the two above-mentioned immunodeficient mouse strains as well as the influence of MSC on relative engraftment of human AML and MDS in primary patientsâ&#x20AC;&#x2122; samples. We describe a comprehensive paired analysis of engraftment of primary AML samples in NSG and NSG-S mice.

Consistent with previous studies, the use of the NSG-S strain increased both the percentage of AML samples that engrafted and the level of engraftment. In contrast, MDS engraftment was consistently low and was not influenced significantly by the use of either mouse strain or co-injection of MSC. However, human MSC did not engraft longterm suggesting that a human microenvironment was not established. These results demonstrate that human AML cells respond positively to the three human cytokines as shown by xenografts, while these cytokines appear to be insufficient to enhance the engraftment and expansion of MDS cells. Xenotransplantation models that better mimic the human microenvironment may be necessary to establish robust MDS xenograft models.

Methods Myelodysplastic syndrome and acute myeloid leukemia specimens Peripheral blood, leukapheresis product, or bone marrow from AML patients were collected at the Hospital of the University of Pennsylvania after informed consent. FrenchAmerican-British or World Health Organization classification and cytogenetics were determined at the Hospital of the University of Pennsylvania. For MDS samples, only bone marrow samples were used and were obtained either from the same source or from Roswell Park Cancer Institute.

Mice Mice were used in accordance with a protocol reviewed and approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania. NSG or NSG-S mice were initially purchased from Jackson Laboratories (Bar Harbor, ME, USA) and produced at the University of Pennsylvania Stem Cell and Xenograft Core. Male and female mice 6-8 weeks of age were either sublethally irradiated (250 cGy) or chemically conditioned by intraperitoneal injection of busulfan (30 mg/kg, Otsuka America Pharmaceutical Inc.) 24 h prior to cell injections.

T-cell depleted AML cells (5-10x106 per mouse) were transplanted via tail vein injection into mice.21 Mice were euthanized no later than 16 weeks after AML injection and marrow from femora and tibiae, splenocytes and peripheral blood were harvested.

Human AML engraftment was assessed by flow cytometry and defined as the percentage of human CD45+CD33+ cells in total live mononuclear cells.22-24 For MDS samples, intrafemoral injections with 1x106 human bone mononuclear cells alone or in combination with 5x105 ex vivo-expanded MSC were performed. For MDS-engrafted mice, haematologica | 2018; 103(6)

Cytokines increase AML but not MDS engraftment

engraftment levels were measured in female mice by bone marrow aspiration of either the ipsilateral or contralateral femur at the time points indicated.

The methods are described in more detail in the Online Supplementary Appendix.

Results Increased levels of acute myeloid leukemia engraftment in NSG-S mice We first investigated the engraftment of 77 AML samples in NSG mice, representing all French-AmericanBritish and prognostic groups (Table 1). We observed that only half of the samples (n=39, 51%) were able to engraft at a significant level (>0.5% human blasts in mouse bone marrow) (Figure 1A). In order to compare engraftment between the two mouse strains, 18 of the 39 NSG-engrafting AML samples were also injected in NSG-S mice. All 18 NSG-engrafting samples engrafted in NSG-S mice.

Representative flow cytometry plots of human CD45+CD33+ cells in the bone marrow, spleen and peripheral blood from 2 AML samples that engrafted in both strains are shown in Figure 1B. Importantly, for 44% (8 out of 18) of the NSG-engrafting AML samples, the use of NSG-S mice as recipients was associated with very rapid engraftment, excessive leukemic burden, anemia, weight loss and lethargy, leading to a significantly shorter overall survival (P≤0.005) (Figure 1C). Consequently, a quantitative comparison of engraftment at the same time point for these 8 patients’ samples was not feasible. The enhanced engraftment, and associated mortality, observed in 26% (8/31) of all NSG-S-engrafting samples should be taken into account when using this strain for pre-clinical studies.

For 9 of the 18 patients’ samples tested in both NSG and NSG-S, we were able to sacrifice the mice at the same time (Figure 1D). We observed a significantly higher leukemia burden in bone marrow and spleen from NSG-S compared to NSG mice (32±23% vs. 21±22% of bone marrow blasts, P≤0.035; 8±14% vs. 6±9% splenic blasts, P≤0.0001) (Figure 1E and F, respectively). Interestingly, NSG-S mice showed a dramatic increase in peripheral blast count (2771±7208 vs. 137±166 blasts/mL peripheral blood, P≤0.034) (Figure 1G). This may represent an improvement over NSG mice because it extends the usefulness of peripheral blood sampling to monitor engraftment and response to treatment in pre-clinical studies.

Overall, these data demonstrate that AML engraftment in NSG-S mice is more rapid and yields a higher leukemic burden than in NSG mice.

NSG-S mice could support the engraftment of samples incapable of engrafting in NSG mice To test whether NSG-S mice could support the engraftment of samples incapable of engrafting in NSG mice, we transplanted NSG-S mice with 21 of the 38 NSG nonengrafting samples (Figure 1A). Remarkably, 67% (14/21) of the non-engrafting samples did engraft in NSG-S mice (18±17.5% bone marrow blasts, 9.2±13.7% splenocytes, and 1799±4848 blasts/mL peripheral blood) (Figure 2A).

Thus, overall our results show that 82% (32/39) of all tested AML samples engrafted in NSG-S mice, compared to 51% in NSG mice. The degree of engraftment observed in bone marrow and spleen for 11 representative patients is shown in Figure 2B. These results indicate that the preshaematologica | 2018; 103(6)

ence of systemic human stem cell factor, granulocytemacrophage colony-stimulating factor and interleukin-3 in NSG-S mice contribute to support leukemia-initiating cells for most AML samples. However, 7 out of 39 samples (18%) still failed to engraft in NSG-S mice, indicating that the bone marrow microenvironment in NSG-S mice may remain suboptimal for a minority of AML samples.

We investigated whether engraftment in NSG-S mice was correlated with surface expression of CD116 (granulocyte-macrophage colony-stimulating factor receptor), CD117 (c-kit), and CD123 (interleukin-3 receptor α−chain) on leukemic cells. As shown in Figure 2C, we found no significant difference in the density of cytokine receptor expression or cytogenetic profiles, mutations, and prognosis between NSG-S engrafting and nonengrafting samples. These results indicate that, in a small minority of AML samples, leukemia-initiating cells have requirements beyond the combination of human granulocyte-macrophage colony-stimulating factor, interleukin-3 and stem cell factor capable of supporting the vast majority of primary AML samples in mice.

Inv(16) acute myeloid leukemia shows enhanced engraftment in NSG-S mice Core binding factor (CBF), a heterodimeric transcription factor that plays an essential role in controlling and regulating normal and leukemic differentiation, is a frequent target of gene rearrangements and mutations in AML.25 CBF-AML patients represent 10-15% of all patients with AML and are characterized by two recurrent translocations: t(8;21)(q22;q22) and inv(16)(p13.1; q22) or t(16;16)(p13.1;q22).26,27 These AML samples with favorable karyotypes are known to engraft poorly in NSG mice. We included 10 CBF-AML samples [2 with t(8;21) and 8 with inv(16)] in our strain comparison study, and were able to successfully engraft all 8 inv(16) AML samples in NSG-S mice (Figure 2D). Interestingly, the 2 t(8;21) samples did not show enhanced engraftment in the NSG-S mice suggesting perhaps a specific defect for the particular translocation. The presence of chromosomal abnormalities was confirmed in bone marrow blasts from engrafted NSG-S mice using a reverse transcriptase polymerase chain reaction to amplify the CBFβ-MYH11 fusion transcript and fluorescence in situ hybridization to detect the inv(16) breakpoint region (Figure 2E). Thus, NSG-S mice provide a permissive environment to support leukemia-initiating cells from low risk inv(16) patients. Whether this reflects a particular requirement of cytokines for inv(16) AML stem cells will require further studies.

Characterization of myelodysplastic syndrome cell engraftment in xenotransplantation models We next turned to characterizing MDS cell engraftment in the two mouse strains. MDS engraftment in patientderived xenotransplantation models is less well described than AML. Unlike AML, MDS should provide multi-lineage engraftment which requires further characterization.

It has also been reported that MDS engraftment is enhanced by co-injection of MSC and we designed experiments to study these variables. We initially transplanted 7 MDS patient samples in NSG and/or NSG-S immunocompromised mice (Figure 3A). The patients’ samples used were separately considered as high risk (MDS with excess blasts-1/2), or low risk (MDS/myeloproliferative neoplasm, unclassified low-risk MDS, therapy-related 961

M. Krevvata et al.

Table 1. Clinical characteristics of the patients with acute myeloid leukemia.

Patient Age Sex FAB WBC BM/PB Disease status ID (year) classification AML - myeloid sarcoma AML with 11q23 abnormalities AML with 11q23 abnormalities AML with 11q23 abnormalities AML with inv16(p13;q22) or t AML with inv16(p13;q22) or t AML with inv16(p13;q22) or t Secondary AML with inv16(p13;q22) or t

0.0723 0.1375 0.115 0.0342 0.1136 0.0797 0.0145 0.0218 0.0099

2678 1532 1907 2022 3876 2955 2120 3328 3339

32 69 82 63 49 69 68 68 29

M M M F M F F F M

1780 559 690 1587 3526 3219 3198 3119 2810 2323 2589 3094 3227 3229 3221 2012 3406 3568 3769 1902 2711 1658 3516 2141 1526 2943 1750 1919 3368 3779 1245 2017 2293 1926 1956 2093 2266 2522 2623 3216

65 35 46 70 34 55 76 52 76 75 65 79 65 68 65 65 66 76 35 43 44 70 25 52 78 65 63 27 65 60 58 64 52 58 52 58 56 52 58 51

M AML with inv16(p13;q22) or t M AML with inv16(p13;q22) or t F AML with inv16(p13;q22) or t M AML with t(8;21)(q22;q22) M AML with t(8;21)(q22;q22) M AML-MLD hx prior MDS/MPN F AML-MLD hx prior MDS/MPN M AML-MLD no prior MDS or MPN M AML-MLD with prior MDS M AML-MLD with prior MDS M AML-MLD with prior MDS F AML-MLD with prior MDS M AML-MLD with prior MDS F AML-MLD with prior MDS M AML-MLD with prior MDS M AML-MLD with prior MPN M AML-MLD with prior MPN NA Biphenotypic M CML (myeloid blast crisis) M M0 M M1 F M1 F M1 M M1 M M1 M M1 M M3 (APML) F M4 M M4 M M4 M M4 M M4 M M4 M M4 F M4 F M4 M M4 M M4 F M4 F M4

0.0269 0.15 0.1293 0.1709 0.0315 0.14 0.1446 0.0315 0.1536 0.0753 0.1164 0.0571 0.0793 ND 0.1616 0.1596 0.1704 0.2457 0.5717 ND 0.326 0.0962 0.3045 0.0454 0.0713 0.1928 0.1003 0.1457 0.2516 0.1845 0.1364 0.0766 0.1952 0.0871 0.1391 0.079 0.314 0.113 0.2878 0.1038

3254 3965 3370 1731

63 55 65 76

M M M M

0.1595 0.0577 0.2419 0.129

M4 M4 M4 M5

PB NA PB Treatment related PB De novo PB Treatment related PB De novo PB De novo BM De novo PB Treatment related BM De novo

Cytogenetics

FLT3

Normal NA t(11;22)(q23;q13) WT t(9;11), der(13;14) NA t(11;16)(q23;p13.3) WT del(12), ins(14;12), inv(16),+21 D835 46,XX,inv(16)(p13.2q22) WT inv(16)(p13.2q22) WT inv(16)(p13.1q22)/47,idem,+8 NA 46,XY,del(7)(q32),inv(16), WT (p13.3q22)/47,idem+22/46,XY PB De novo 46,XY,inv(16)(p13.1q22) WT BM De novo 47,XY,inv(16)(p13.3q22),+22/48,XY,idem,+9NA PB De novo 46,XX,inv(16)(p13.3q22)/47,XX,idem+8 NA PB Treatment related t(8;21)(q22;q22),del(13)(q12q22) WT PB NA t(8;21)(q22;q22) NA PB De novo del(7)(p11.2)/46,XY WT PB De novo NA ITD PB De novo Normal WT PB Treatment related add(20)(q11.2) D835 PB De novo ND WT PB Relapsed 47,XY,+11[18]/46,XY[2] ITD PB Relapsed Normal NA PB NA Normal WT PB Refractory Normal WT PB Relapsed Normal WT PB De novo Complex WT PB De novo del(20)(q11.2q13.1) WT PB NA Complex ITD PB NA t(9;22)(q34;q11.2)/48,idem,+8,+19 NA PB Refractory NA NA PB De novo Normal ITD PB De novo Normal ITD PB NA inv(9)(p12q13) ITD PB Relapsed NA ITD PB De novo Normal ITD PB De novo Normal ITD PB NA t(15;17) NA PB Relapsed NA NA PB De novo NA WT PB De novo NA ITD PB De novo normal ITD PB Untreated NA NA PB De novo Normal ITD PB Relapsed NA ITD PB De novo 47,XX,+8/46,XX ITD PB De novo Normal ITD PB De novo Normal ITD PB Refractory Complex ITD PB Relapsed NA ITD PB Relapsed t(3;3)(q21;q26),del(7) WT (p13),del(17)(p12),-21 PB De novo Complex WT PB NA Normal ITD PB De novo Normal WT PB De novo Normal WT

NPM Blasts % Blasts % Engrafter Engrafter (PB) (BM) in NSG in NSG-S NA NA NA NA WT WT NA NA WT

NA 64 75 26 NA 72 NA 50 NA

NA NA NA NA NA NA 60 NA 27

MSC N N N N N N MSC N

ND MSC MSC ND MSC ND MSC MSC MSC

NA NA NA NA NA WT WT MUT WT NA WT NA NA NA NA NA WT WT NA NA WT NA WT NA NA MUT NA NA MUT WT NA NA NA NA NA NA NA NA NA NA

11 NA 61 91 67 10 21 24 40 63 NA 4 89 NA 89 NA 77 NA 66 41 NA NA 91 83 78 94 NA 83 74 NA NA NA 87 89 51 60 90 NA NA 50

NA 77 NA NA 68 NA 60 32 NA NA NA NA NA NA NA NA 70 NA NA 90 NA NA NA NA NA NA NA NA 90 NA NA 51 NA NA NA NA 90 NA NA NA

N N N N N N MSC MSC N N N N MSC MSC MSC MSC MSC MSC N N N N N MSC MSC MSC N N N N N N N MSC MSC MSC MSC MSC MSC MSC

MSC MSC MSC N ND ND MSC ND N ND ND ND ND ND MSC ND MSC MSC N MSC MSC MSC ND ND MSC ND MSC MSC MSC MSC ND ND ND ND ND ND ND ND ND MSC

WT WT MUT NA

NA 89 74 NA

NA NA 90 NA

MSC MSC MSC N

MSC MSC ND MSC

continued on next page

962

haematologica | 2018; 103(6)

Cytokines increase AML but not MDS engraftment continued from the previous page

3918 2258 2698 2844 4090 2750 53 3196 3365 3949 4133 2339 3055 2748 1932 2933 3081 3033 774 2741 2837 3261 1897 2107 2074

73 57 43 73 68 72 59 64 62 59 58 47 54 64 31 56 59 74 37 52 43 39 64 64 78

M M F M M M F M M F M M F F F M M M F M F M F F F

M5 M5 M5 M5 NA NA NA NA NA NA NA ND NOS NOS NOS NOS NOS NOS NOS NOS NOS NOS Secondary Secondary Secondary

0.1397 0.0429 0.1298 0.3246 0.0979 0.0983 0.1836 0.0711 0.0839 0.0778 NA 0.349 0.431 0.1026 0.215 0.0476 0.0224 0.038 0.1245 0.1317 0.0864 0.0955 0.0348 0.0649 0.0237

PB De novo PB De novo PB De novo PB De novo PB NA PB De novo PB NA PB Relapsed PB De novo PB Refractory PB Relapsed PB Relapsed PB De novo PB De novo PB Relapsed PB De novo BM De novo BM De novo PB De novo PB Relapsed PB De novo PB De novo BM Treatment related BM Relapsed PB Treatment related

Normal NA NA 45,X,-MSC/46,XY ND Normal Normal 47,XY,+8, t(3;17) Normal Normal NA NA Normal Normal Normal Normal Normal NA t(7;11)(p15;p15) Complex 47,XX,+8 Normal t(8;16)(p11.2;p13.3) t(8;16)(p11.2;p13.3) Complex

WT NA ITD WT NA WT ITD ITD ITD ITD ITD WT ITD NA NA WT WT WT ITD ND ITD ITD WT WT WT

MUT NA MUT WT NA WT NA NA MUT NA NA NA MUT MUT NA WT MUT WT NA NA WT MUT NA NA NA

40 77 87 51 25 73 NA 87 22 94 72 80 NA 97 NA 76 79 NA NA NA 91 NA NA 94 90

NA NA NA 80 NA NA 70 NA 80 94 NA NA NA NA NA NA 90 68 NA NA NA NA 90 NA NA

N N MSC MSC N N MSC MSC MSC MSC MSC N MSC N N N N N MSC MSC MSC MSC N MSC N

MSC ND MSC MSC MSC N ND ND ND MSC MSC ND ND N ND ND ND ND ND MSC MSC MSC MSC ND ND

M: male; F: female; MSC: engrafted; N: not engrafted; ND: not done; NA: not available; PB: peripheral blood; BM: bone marrow; FAB: French-American-British classification; WBC: white blood cell count (x 109 cells/L); ITD:internal tandem duplication; WT: wild-type; MUT: mutated.

myeloid neoplasms) (Table 2). We performed intrabone injections directly into the femoral cavity at the orthotopic site as it has been described in patient-derived xenotransplantation models for AML, that intrabone cell transplantation results in a higher probability of successful engraftment. This can be advantageous for patient-derived xenotransplantation models of MDS in particular, as the numbers of bone marrow mononuclear cells are commonly modest or low.28 In order to address whether the presence of human stromal cells results in increased engraftment levels, MSC were ex vivo-expanded and co-injected along with the patients’ mononuclear cells. Levels of engraftment were assessed by bone marrow aspiration at different time points throughout the experiment and are expressed as the percentage of human CD45+ cells. To assess the subpopulations of cells engrafted, cells were further analyzed with the lineage-specific antibodies CD19 for B cells and CD3 for T cells as well as for the presence of stem and progenitor cells using the markers CD34 and CD38, CD123 and CD45RA, respectively (Figure 3B). In the majority of engrafted patients’ samples, B and T cells were not detected or were detected at uncommonly low levels at the time points tested. In contrast, the myeloid CD33+ component was present in all mice tested as well as subpopulations of CD34+ and CD38+ human cells. The ability to differentiate into cells of the erythroid lineage was assessed in the low and negative fractions of human CD45+ cells using the erythroid differentiation markers CD71 and glycophorin A. Even though erythroid cells at the initial stages of differentiation were detected in all mice tested, cells lacked the ability to differentiate further haematologica | 2018; 103(6)

into more mature cells (Figure 3C). These results were further confirmed by immunohistochemical analysis of decalcified bone sections, showing a broad presence of human CD33-stained cells, but absence of megakaryocytes expressing GPIIIa and erythroid cells expressing glycophorin C (Figure 3D). Overall, these results confirm that MDS cells with multi-lineage potential can potentially be transferred and maintained in immunocompromised mice.

Most myelodysplastic syndrome samples do not show sustained engraftment The term “engraftment” used in the context of xenotransplantation studies is equated with long-term maintenance and, typically, expansion of transplanted cells. To quantitatively assess whether MDS cells engraft in NSG-S mice, we divided our samples into low- and high-risk MDS. As shown in Figure 4, engraftment of high-risk MDS was heterogeneous with 2 samples never showing robust engraftment but 2 samples showing early and sustained engraftment (Figure 4A). Only one of the 3 low-risk MDS samples demonstrated clear engraftment, whereas levels of engraftment decreased over time for the other 2 samples (Figure 4B). For comparison, a secondary AML sample shows the expected low early engraftment with increase over time (Figure 4A).

The presence of human cytokines marginally improves engraftment Direct comparison of engraftment levels between NSG and NSG-S mice transplanted with mononuclear cells 963

M. Krevvata et al.

A

B

C

D

E

F

G

Figure 1. NSG-S mice showed higher and more rapid engraftment of primary patientsâ&#x20AC;&#x2122; acute myeloid leukemia (AML) samples than did NSG mice. (A) Thirty-nine of the 77 AML samples screened were engrafted (hCD45+CD33+ blasts > 0.5% in mouse bone marrow) in NSG mice. All 39 AML engrafted in NSG mice also engrafted in NSG-S mice. Thirty-eight AML samples did not engraft in NSG, and approximately two-thirds of these engrafted in NSG-S mice. (B) Representative flow cytometry plots of human hCD45+CD33+ leukemia cells from bone marrow (BM), spleen (SPL) and peripheral blood (PB) samples from NSG and NSG-S mice. (C) Kaplan-Meier curves show the time to sacrifice mice for all 77 AML samples. (D) Leukemia burden of 9 AML engrafted in both NSG and NSG-S mice which were sacrificed at the same time was assessed by % of human hCD45+CD33+ cells in mouse BM and SPL. Each symbol represents one mouse. (E and F) Individual mouse BM and SPL leukemia burdens of the 9 AML engrafters are shown. Solid symbols represent data from NSG mice, and empty symbols represent data from NSG-S mice. (G) Elevated levels of leukemia burden (% hCD45+CD33+ cells) were seen in NSG-S mice.

964

haematologica | 2018; 103(6)

Cytokines increase AML but not MDS engraftment

from MDS patients revealed a small increase in the mice expressing the human cytokines at eight weeks after intrabone injections (Figure 4C). This difference was sustained at 16 weeks in all patients’ samples tested but was not statistically significant at this time point (Figure 4D). Thus, in contrast to their effect on AML, human cytokines do not appear to provide a significant engraftment or growth advantage to human MDS in NSG-S mice.

Engraftment levels are independent of the presence and the origin of mesenchymal stem cells To verify that the MSC are functional cells, we performed a phenotypic characterization and also tested the

cells’ ability for trilineage differentiation in vitro (Figure 5AD). Overall, patient-derived and normal donor MSC appear to express all MSC markers (Figure 5A) and are able to differentiate into osteoblasts (Figure 5B), adipocytes (Figure 5C) and chondrocytes (Figure 5D), therefore having features of bona fide MSC. In order to investigate whether the presence of human stromal cells might be used as a better supporting tissue for the engraftment of MDS in immunodeficient mice, in vitroexpanded MSC derived from patients and/or healthy donors were intrafemorally co-transplanted along with the patients’ mononuclear cells. Engraftment levels were measured at eight weeks after transplantation by bone marrow

A

C B

D

E

Figure 2. NSG-S mice enhanced the engraftment of human patients’ acute myeloid leukemia (AML) samples more than NSG mice (non-NSG engrafted patients’ samples were used). (A) Levels of hCD45+CD33+ leukemia blasts in bone marrow (BM), spleen (SPL) and peripheral blood (PB) of NSG or NSG-S mice (n=14) injected with the same number of AML cells. (B) Individual mouse BM and SPL leukemia burdens of 11 out of the 14 mice from (A) are shown. (C) Receptor densities (CD116, CD117 and CD123) on AML cells of the 7 non-NSG-S engrafters and 14 NSG-S engrafters were assessed as number of receptors per cell. (D) Establishment of 8 inv(16) AML patient-derived xenotransplant models in NSG-S mice. (E) The chromosomal abnormalities for inv(16) were confirmed in the BM of engrafted NSG-S mice by fluorescent in situ hybridization (left panels) and breakpoint reverse transcriptase polymerase chain reaction (right panel).

haematologica | 2018; 103(6)

965

M. Krevvata et al.

aspiration. No significant improvement was observed associated with the presence of human MSC in the murine environment. Interestingly, in the majority of cases in which MSC appeared to have a slightly positive effect in engraftment levels on week 8 after transplantation, the effect was lost when engraftment levels were reassessed at week 16 (data not shown). In a direct comparison experiment

between NSG and NSG-S mice, the presence of MSC produced a mild improvement in the engraftment levels in the patientsâ&#x20AC;&#x2122; samples tested, at eight weeks, irrespective of the mouse strain. Similarly to before, the positive effect appeared to fade at 16 weeks in both NSG and NSG-S mice (Online Supplementary Figure S1). Next, to assess whether MSC from various sources might contribute to MDS

A

B

C

D

E

F

Figure 3. Myelodysplastic syndromes (MDS) cells engraft in NSG-S mice. (A) Schematic representation of experimental set up. Patient-derived bone marrow mononuclear cells (MNC) alone or in combination with mesenchymal stem cells (MSC) were intrafemorally injected in NSG-S mice. Characteristic gating strategy for analyzing the (B) myeloid (CD33), lymphoid (CD19, CD3), progenitor (CD45RA, CD123, CD38) and (C) erythroid (GlyA) subpopulations found in the bone marrow of mice engrafted with patientsâ&#x20AC;&#x2122; bone marrow MNC samples. Immunostainings of (D) hCD33, (E) glycophorin C and (F) GPIIIa of decalcified bone marrow sections from sternums of mice engrafted with human MDS.

966

haematologica | 2018; 103(6)

Cytokines increase AML but not MDS engraftment

engraftment, we further compared engraftment of MDS mononuclear cells without MSC to healthy donor-derived MSC (normal), or allogeneic patient-derived MSC (allo) and to patient-derived autologous MSC (auto) (Figure 3A).

Although some mice showed increased engraftment with different MSC samples tested, no consistent pattern of enhanced engraftment was seen that correlated with the source of the MSC (Figure 5E).

To better understand the transient effect of the presence of MSC on the levels of engraftment observed for some of the patientsâ&#x20AC;&#x2122; samples tested, in vitro-expanded MSC were labeled using a lentiviral plasmid expressing green fluorescent protein and luciferase. Those cells were further transplanted via intrabone injections in NSG mice and monitored over time. In vivo imaging revealed gradually decreasing levels of luminescence (Figure 5F).

Luminescence was solely detected in the area of the injected femur and its levels were completely diminished by week 4 after transplantation.

Long-term engraftment of myelodysplastic syndrome cells in NSG-S mice To evaluate whether long-term engraftment of MDS cells can be achieved in NSG-S mice, human CD45 cells, isolated from the bone marrow of well-engrafted primary recipients, were selected and intrafemorally transplanted

into secondary recipients with or without MSC according to the experimental plan (Figure 6A). Thirteen weeks after transplantation the mice were sacrificed. Analysis of the different subpopulations identified the presence of myeloid CD33+ cells, as well as CD34+CD38- cells. As in the primary recipients, no B or T cells were detected (Figure 6B). Small erythroid populations in secondary recipients were also unable to differentiate further to more mature erythroid cells (Figure 6C). All mice showed increased levels of engraftment compared to the engraftment levels of the primary recipients at the same time

Table 2. Clinical samples used for the MDS engraftment studies.

Patient ID# 2970 2381 108 3598 2952 4712 3282

Diagnosis

Risk level

MDS/MPN Therapy related myeloid neoplasm Unclassified MDS MDS-EB-1 MDS-EB-1 MDS-EB-1 MDS-EB-2

Low Low Low High High High High

MDS: myelodysplastic syndrome; MPN: myeloproliferative neoplasm; EB: excess blasts.

A

B

C

D

Figure 4. Most myeloproliferative syndrome (MDS) samples do not show sustained engraftment and human cytokines moderately enhance engraftment. Time course representation of the percentages of hCD45+ cells in the bone marrow of xenografted NSG-S mice injected with (A) high risk and (B) low risk human MDS patientsâ&#x20AC;&#x2122; mononuclear cells. The gray line corresponds to the 0.1% threshold used in the study and the gray zone reflects a broader area of uncertainty of engraftment. Bone marrow mononuclear cells from patients diagnosed with MDS were intrafemorally injected into NSG and NSG-S mice in a strain comparison experiment. Percentage of hCD45+ cells found in the murine bone marrow was evaluated at (C) eight and (D) 16 weeks after transplantation.

haematologica | 2018; 103(6)

967

M. Krevvata et al.

A

B

C

D

E

F Figure 5. Myelodysplastic syndrome (MDS) cell engraftment is not affected by mesenchymal stem cells (MSC). (A) Representative phenotypical characterization of the ex vivo-expanded human MSC used in the transplantation experiments.

Trilineage in vitro differentiation of ex vivo expanded MDS patient MSC into (B) osteoblasts, (C) adipocytes, and (D) chondrocytes. [(1) Differentiated, (2) Undifferentiated, (3) Positive Control] (E) Bone marrow human mononuclear cells were intrafemorally transplanted into NSG-S mice alone or in combination with ex vivo-expanded MSC isolated from various origins. Percentages of hCD45+ cells present in the murine bone marrow at 6-8 weeks post transplantation. Red line at 0.1% represents the positive engraftment threshold level. Comparisons showed no significant statistical differences between the four groups. (F) NSG mice intrafemorally transplanted with labeled MSC-green fluorescent protein-luciferase.

Bioluminescence levels were measured in vivo at the timepoints indicated.

968

haematologica | 2018; 103(6)

Cytokines increase AML but not MDS engraftment

A

B

C

D

Figure 6. Myelodysplastic syndrome-initiating cells can be maintained in NSG-S mice. (A) Schematic representation of experimental set up for secondary intrafemoral transplantations. (Î&#x2019;) Percentages of hCD45+ cells at 13 weeks after transplantation. Characteristic gating strategy for analyzing the (C) myeloid (CD33), lymphoid (CD19, CD3), progenitor (CD45RA, CD123, CD38) and (D) erythroid (GlyA) subpopulations found in the bone marrow of mice engrafted with hCD45+ cells isolated from primary mice.

haematologica | 2018; 103(6)

969

M. Krevvata et al.

point (Figure 6D), whereas the presence of MSC did not appear to improve engraftment regardless of the origin of the MSC (Figure 6D).

Discussion Ongoing development of so-called â&#x20AC;&#x153;humanizedâ&#x20AC;? mouse strains that combine immunodeficiency with expression of human proteins or cells permits increasingly sophisticated modeling of human diseases in xenotransplantation models. Here we have used the NSG and NSG-S strains to study the effects of human cytokines on engraftment and growth of human myeloid diseases, AML and MDS. Our results show that NSG-S mice represent a significantly improved patient-derived xenotransplantation model to accelerate and enhance leukemic engraftment compared to NSG mice, and to support engraftment for the vast majority of all primary AML samples, making this model particularly useful for pre-clinical studies. In contrast, engraftment of human MDS in NSG or NSG-S mice is possible but not robust and in several cases not sustained. High-risk MDS is more likely to achieve long-term engraftment compared to low-risk MDS. The expression of human cytokines in NSG-S mice marginally improves engraftment levels. In developing this model we also note that human MSC do not provide longterm MSC engraftment in NSG-S mice and do not contribute to MDS cell engraftment or expansion. Overall, these results are consistent with in vitro studies suggesting that AML cells are actually more responsive to bone marrow-derived cytokines than MDS cells, at least as measured by engraftment and growth in immunocompromised mouse strains.29-32 Engraftment of AML samples in NSG mice has been the standard measure of assessment of leukemic stem cells for the last ten years. However, our results with AML overall, and particularly with inv(16) AML, demonstrate that severe combined immunodeficiency leukemia engrafting ability, the functional definition of leukemic stem cells, can be strongly affected by the recipient mouse strain and the extent to which the recipient environment contains human cytokines or cytokines that cross-react with human cytokine receptors. More formal exploration of stem cells comparing NSG and NSG-S mice will further define the dependency of severe combined immunodeficiency leukemia-initiating cells on cytokines. Furthermore, these studies raise significant questions about AML biology that should be addressable using the NSG-S model. Previous authors have suggested that AML cells require pathological activation of signaling pathways for full transformation.33 However, our data demonstrate that the majority of AML samples may remain responsive to cytokines produced by the bone marrow as was also shown by Ellegast et al.34 Given the recent results that effective 3rd-generation FLT3 inhibitors are clinically active but not curable it may be of therapeutic value to consider whether cytokines can regulate cell survival allowing AML cells to escape FLT3 inhibition.35 In contrast to AML, human MDS xenotransplantation has not been thoroughly studied. Although we studied a modest number of MDS samples, our results comparing MDS injections into NSG versus NSG-S mice with or without MSC co-injection are similar to other recently described results.36 A critical question is the formal definition of an MDS-engrafting cell, the functional equivalent of the severe 970

combined immunodeficiency leukemia-initiating cell as initially defined by Bonnet and Dick.22 AML stem cells typically traffic to the marrow after xenotransplantation, remain quiescent for some weeks and then initiate expansion. In contrast, MDS cells in our study were injected into the marrow where they appear to remain but do not expand for several months. Some samples, from patients with highrisk MDS, demonstrated expansion and our results show that MDS initiating cells can be maintained in NSG-S mice.

More specifically, secondary recipient NSG-S mice injected with human CD45 cells isolated from primary recipient mice developed abnormal hematopoiesis identical to that of the primary recipients, accompanied by a significant expansion of the phenotypic stem/progenitor cell compartment, which is typical in cases of AML passaged into secondary xenotransplanted animal models. Whereas genetic analysis showed the presence of identical genetic lesions between the patientsâ&#x20AC;&#x2122; samples and the primary recipient animals (data not shown), the selection of an unidentified minor AML clone in the secondary recipient animals is a possibility that requires further exploration. Examination of the human cell populations present in the murine bone marrow revealed signs of abnormal hematopoiesis reflected by the presence of stem/progenitor cells, the prevalence of the myeloid component and, in the majority of cases, the complete absence of lymphoid cells. Erythropoiesis was compromised and even though cells at the initial stages of human erythroid differentiation (ERY1, ERY2) were present in both the bone marrow and the spleen (data not shown), these cells were not capable of further differentiation, suggesting a blockage in the erythroid differentiation process, as has been previously described.37 Our findings suggest that the development of a patient-derived xenotransplantation model for MDS is possible, but not yet robust, and that the definition of MDS stem cells may include long-term engraftment but not expansion. The development of an ectopic human niche might provide a solution since promising results have been reported in several types of leukemia.38,39 The use of MSC to humanize the murine bone marrow niche has been proposed as a method to enhance MDS engraftment.20 However, we demonstrated that human MSC engrafted into murine bone marrow do not establish long-term engraftment. It is theoretically possible that short-term survival of human MSC may promote engraftment of MDS cells but this was not seen in our studies. The reason for this discrepancy from other studies is not clear, but it is notable that AML engraftment is known to be highly variable among animal colonies.

In summary, AML cells demonstrate enhanced engraftment in NSG-S mice compared to NSG mice. MDS cells do not demonstrate a similar response to either the human cytokines produced by the NSG-S strain or by co-injection of human MSC. The findings of these studies are consistent with those of previous in vitro studies of AML and MDS and may suggest that MDS cells are not capable of responding to human cytokines with an increase in survival or growth.

Further work may determine whether this is a fundamental aspect of the ineffective hematopoiesis that defines MDS. Acknowledgments We thank Lucio Castilla (University of Massachusetts, Worcester, MA, USA) for his assistance in analyzing inv(16) and April Schrank-Hacker for performing FISH studies. We also thank Joy Cannon, Rebecca Kotcher and the clinical staff for consenting haematologica | 2018; 103(6)

Cytokines increase AML but not MDS engraftment

and collecting samples from patients, as well as Abigail Smith, Ph.D., Emily Miedel, VMD, and the animal care technicians for their help with our animal colony. We thank Dr Dominique Bonnet for performing the genetic analysis of the MDS cells. We thank the Cancer Histology Core of the University of Pennsylvania for section-

References 1. Tallman MS, Gilliland DG, Rowe JM. Drug therapy for acute myeloid leukemia. Blood.

2005;106(4):1154–1163.

2. Walter MJ, Shen D, Ding L, et al. Clonal architecture of secondary acute myeloid leukemia. N Engl J Med. 2012;366(12):1090– 1098.

3. Will B, Zhou L, Vogler TO, et al. Stem and progenitor cells in myelodysplastic syndromes show aberrant stage-specific expansion and harbor genetic and epigenetic alterations. Blood. 2012;120(10):2076–2086.

4. Woll PS, Kjällquist U, Chowdhury O, et al.

Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo. Cancer Cell. 2014;25(6):794–808.

5. Schofield R. The relationship between the spleen colony-forming cell and the haemopoietic stem cell. Blood Cells.

1978;4(1-2):7–25.

6. Lane SW, Scadden DT, Gilliland DG. The leukemic stem cell niche: current concepts and therapeutic opportunities. Blood.

2009;114(6):1150–1157.

7. Suda T, Arai F, Hirao A. Hematopoietic stem cells and their niche. Trends Immunol.

2005;26(8):426–433.

8. Wilson A, Trumpp A. Bone-marrow haematopoietic-stem-cell niches. 2006;6(2): 93–106.

9. Méndez-Ferrer S, Lucas D, Battista M, Frenette PS. Haematopoietic stem cell release is regulated by circadian oscillations.

Nature. 2008;452(7186):442–447.

10. Yamazaki S, Ema H, Karlsson G, et al.

Nonmyelinating Schwann cells maintain hematopoietic stem cell hibernation in the bone marrow niche. Cell. 2011;147(5):1146– 1158.

11. Taichman RS, Emerson SG. The role of osteoblasts in the hematopoietic microenvironment. Stem Cells. 1998;16(1):7–15.

12. Calvi LM, Adams GB, Weibrecht KW, et al.

Osteoblastic cells regulate the haematopoietic stem cell niche. Nature. 2003;425 (6960):841–846.

13. Kremer KN, Dudakovic A, McGeeLawrence ME, et al. Osteoblasts protect AML cells from SDF-1-induced apoptosis. J Cell Biochem. 2014;115(6):1128–1137.

14. Tavor S, Petit I, Porozov S, et al. CXCR4 regulates migration and development of human acute myelogenous leukemia stem cells in transplanted NOD/SCID mice.

Cancer Res. 2004;64(8):2817–2824.

15. Colmone A, Amorim M, Pontier AL, Wang S, Jablonski E, Sipkins DA. Leukemic cells create

haematologica | 2018; 103(6)

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

ing and staining the micromass MSC and Dr George Dodge for providing us with the primary chondrocytes. Financial support was provided by the Evans Foundation, the Hematologic Malignancies Translational Center of Excellence of the Abramson Cancer Center and NIH grants - R01CA149566 and R01CA198089.

bone marrow niches that disrupt the behavior of normal hematopoietic progenitor cells.

Science. 2008;322(5909):1861–1865.

Zhang Y, Xie R-L, Croce CM, et al. A program of microRNAs controls osteogenic lineage progression by targeting transcription factor Runx2. Proc Natl Acad Sci USA. 2011;108(24):9863–9868.

Schepers K, Pietras EM, Reynaud D, et al.

Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a selfreinforcing leukemic niche. Stem Cell.

2013;13(3):285–299.

Sanchez PV, Perry RL, Sarry JE, et al. A robust xenotransplantation model for acute myeloid leukemia. Leukemia. 2009;23(11): 2109–2117.

Wunderlich M, Chou F-S, Link KA, et al.

AML xenograft efficiency is significantly improved in NOD/SCID-IL2RG mice constitutively expressing human SCF, GM-CSF and IL-3. Leukemia. 2010;24(10):1785–1788.

Medyouf H, Mossner M, Jann J-C, et al.

Myelodysplastic cells in patients reprogram mesenchymal stromal cells to establish a transplantable stem cell niche disease unit.

Cell Stem Cell. 2014;14(6):824–837.

Wunderlich M, Brooks RA, Panchal R, Rhyasen GW, Danet-Desnoyers G, Mulloy JC. OKT3 prevents xenogeneic GVHD and allows reliable xenograft initiation from unfractionated human hematopoietic tissues. Blood. 2014;123(24):e134–144.

Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997;3(7):730–737.

Ishikawa F, Yoshida S, Saito Y, et al.

Chemotherapy-resistant human AML stem cells home to and engraft within the bonemarrow endosteal region. Nat Biotechnol.

2007;25(11):1315–1321.

Sarry J-E, Murphy K, Perry R, et al. Human acute myelogenous leukemia stem cells are rare and heterogeneous when assayed in NOD/SCID/IL2Rgc-deficient mice. J Clin Invest. 2011;121(1):384–395.

Speck NA, Gilliland DG. Core-binding factors in haematopoiesis and leukaemia. Nat Rev Cancer. 2002;2(7):502–513.

Mrózek K, Bloomfield CD. Chromosome aberrations, gene mutations and expression changes, and prognosis in adult acute myeloid leukemia. Hematology Am Soc Hematol Educ Program. 2006;2006(1):169– 177.

Mrózek K, Marcucci G, Paschka P, Whitman SP, Bloomfield CD. Clinical relevance of mutations and gene-expression changes in adult acute myeloid leukemia with normal

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

cytogenetics: are we ready for a prognostically prioritized molecular classification?

Blood. 2007;109(2):431–448.

Kushida T, Inaba M, Hisha H, et al. Intrabone marrow injection of allogeneic bone marrow cells: a powerful new strategy for treatment of intractable autoimmune diseases in MRL/lpr mice.

Blood.

2001;97(10):3292–3299.

Miyauchi J, Kelleher CA, Yang YC, et al. The effects of three recombinant growth factors, IL-3, GM-CSF, and G-CSF, on the blast cells of acute myeloblastic leukemia maintained in short-term suspension culture. Blood.

1987;70(3):657–663.

Hoang T, Haman A, Goncalves O, Wong GG, Clark SC. Interleukin-6 enhances growth factor-dependent proliferation of the blast cells of acute myeloblastic-leukemia.

Blood. 1988;72(2):823–826.

Vellenga E, Ostapovicz D, Orourke B, Griffin JD. Effects of recombinant Il-3, GmCsf, and G-Csf on proliferation of leukemic clonogenic cells in short-term and long-term cultures. Leukemia. 1987;1(8):584–589.

Miyauchi J, Kelleher CA, Wong GG, et al.

The effects of combinations of the recombinant growth-factors Gm-Csf, G-Csf, Il-3, and Csf-1 on leukemic blast cells in suspension-culture. Leukemia. 1988;2(6):382–387.

Kelly LM, Gilliland DG. Genetics of myeloid leukemias. Annu Rev Genomics Hum Genet. 2002;3:179–198.

Ellegast JM, Rauch PJ, Kovtonyuk LV, et al. inv(16) and NPM1mut AMLs engraft human cytokine knock-in mice. Blood. 2016;128 (17):2130–2134.

Perl AE, Altman JK, Cortes J, et al. Selective inhibition of FLT3 by gilteritinib in relapsed or refractory acute myeloid leukaemia: a multicentre, first-in-human, open-label, phase 1-2 study. Lancet Oncol. 2017;18(8): 1061–1075.

Rouault-Pierre K, Mian SA, Goulard M, et al.

Preclinical modeling of myelodysplastic syndromes. Leukemia. 2017;31(12):2702-2708.

Shi H, Yamamoto S, Sheng M, , et al. ASXL1 plays an important role in erythropoiesis. Sci Rep. 2016;6:28789.

Reinisch A, Thomas D, Corces MR, et al. A humanized bone marrow ossicle xenotransplantation model enables improved engraftment of healthy and leukemic human hematopoietic cells. Nat Med. 2016;22(7): 812–821.

Abarrategi A, Foster K, Hamilton A, et al.

Versatile humanized niche model enables study of normal and malignant human hematopoiesis. J Clin Invest. 2017;127(2): 543–548.

971

ARTICLE

Myeloproliferative Neoplasms

Ferrata Storti Foundation

Haematologica 2018 Volume 103(2):972-981

Impact of hydroxycarbamide and interferon-α on red cell adhesion and membrane protein expression in polycythemia vera Mégane Brusson,1,2,3* Maria De Grandis,1,2,3*† Sylvie Cochet,1,2,3 Sylvain Bigot,1,2,3¶ Mickaël Marin,1,2,3 Marjorie Leduc,4 François Guillonneau,4 Patrick Mayeux,4 Thierry Peyrard,1,2,3 Christine Chomienne,5,6 Caroline Le Van Kim,1,2,3 Bruno Cassinat,6 Jean-Jacques Kiladjian7 and Wassim El Nemer1,2,3

Biologie Intégrée du Globule Rouge UMR_S1134, Inserm, Université Paris Diderot, Sorbonne Paris Cité, Université de la Réunion, Université des Antilles; 2Institut National de la Transfusion Sanguine, F-75015 Paris; 3Laboratoire d’Excellence GR-Ex, Paris; 4 Plateforme de Protéomique de l’Université Paris Descartes (3P5), Institut Cochin, INSERM U1016, CNRS UMR 8104, Université Sorbonne Paris Cité, Laboratoire d’Excellence GR-Ex, Paris; 5Université Sorbonne Paris Cité, Université Paris Diderot, Inserm UMR-S1131, Hôpital Saint Louis, Institut Universitaire d'Hématologie, Laboratoire de Biologie Cellulaire, Paris; 6AP-HP, Hôpital Saint-Louis, Laboratoire de Biologie Cellulaire, Paris and 7Centre d'Investigations Cliniques, Hôpital Saint-Louis, Université Paris Diderot, Paris, France 1

*MB and MDG contributed equally to this work. †Current address: Etablissement Français du Sang PACA Corse, Biologie des Groupes Sanguins, Aix Marseille Université, CNRS, EFS, ADES, Marseille, France; ¶Current address: Centre d'Immunologie de Marseille-Luminy, Aix Marseille Université, Inserm, CNRS, Marseille, France

ABSTRACT

P

Correspondence: wassim.el-nemer@inserm.fr

Received: October 11, 2017.

Accepted: March 21, 2018.

Pre-published: March 29, 2018. doi:10.3324/haematol.2017.182303 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/972 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

972

olycythemia vera is a chronic myeloproliferative neoplasm characterized by the JAK2V617F mutation, elevated blood cell counts and a high risk of thrombosis. Although the red cell lineage is primarily affected by JAK2V617F, the impact of mutated JAK2 on circulating red blood cells is poorly documented. Recently, we showed that in polycythemia vera, erythrocytes had abnormal expression of several proteins including Lu/BCAM adhesion molecule and proteins from the endoplasmic reticulum, mainly calreticulin and calnexin. Here we investigated the effects of hydroxycarbamide and interferon-α treatments on the expression of erythroid membrane proteins in a cohort of 53 patients.

Surprisingly, while both drugs tended to normalize calreticulin expression, proteomics analysis showed that hydroxycarbamide deregulated the expression of 53 proteins in red cell ghosts, with overexpression and downregulation of 37 and 16 proteins, respectively. Within overexpressed proteins, hydroxycarbamide was found to enhance the expression of adhesion molecules such as Lu/BCAM and CD147, while interferon-α did not. In addition, we found that hydroxycarbamide increased Lu/BCAM phosphorylation and exacerbated red cell adhesion to its ligand laminin. Our study reveals unexpected adverse effects of hydroxycarbamide on red cell physiology in polycythemia vera and provides new insights into the effects of this molecule on gene regulation and protein recycling or maturation during erythroid differentiation.

Furthermore, our study shows deregulation of Lu/BCAM and CD147 that are two ubiquitously expressed proteins linked to progression of solid tumors, paving the way for future studies to address the role of hydroxycarbamide in tissues other than blood cells in myeloproliferative neoplasms.

Introduction Polycythemia vera (PV) is a chronic myeloproliferative neoplasm (MPN) characterized by clonal expansion of an abnormal hematopoietic stem cell due in most cases to the V617F activating mutation in the tyrosine kinase JAK2.1-4 PV is marked haematologica | 2018; 103(6)

Effects of HC and IFN-α on PV RBC adhesion

A

B

Figure 1. Calreticulin expression is decreased in polycythemia vera (PV) patients under hydroxycarbamide (HC) and interferon-α (IFN) treatments. Quantification of Calr expression normalized by actin from (A) 4 PV patients before (UT) and after (HC) HC treatment and (B) 11 control (CT), 19 UT, 11 HC, and 7 IFN patients. Horizontal lines represent medians: 0.9883, 2.023, 0.8760 and 0.8760, respectively.

by thrombo-hemorrhagic complications and a propensity to transform into myelofibrosis and acute leukemia.5 Reducing the vascular risk and preventing transformation are the main purpose of PV treatment. According to current guidelines, low-risk PV patients are managed with phlebotomy and aspirin, whereas high-risk patients receive cytoreduction with hydroxycarbamide (HC) or interferonα (IFN).6 New therapeutic approaches aim at developing small molecules as specific inhibitors of JAK2 mutated forms.7 HC, also known as hydroxyurea, is commonly used to treat PV patients requiring cytoreductive therapy. HC is a cytostatic non-alkylating hydroxylated urea analog that inactivates ribonucleotide reductase, the enzyme that catalyses the conversion of ribonucleotides to deoxyribonucleotides during de novo DNA synthesis,8 leading to the interruption of DNA synthesis and cell death at the Sphase.9 HC does not target the malignant clone and therefore appears unable to eradicate it. The efficacy of HC in preventing thrombosis was suggested in several randomized clinical trials,10 but is still not proven. HC alone or in combination with other therapies may increase the risk of acute leukemia in MPN.11 Interferon-α is a non-leukemogenic drug that, in some cases, induces cytogenetic remissions or reversion from monoclonal to polyclonal patterns of hematopoiesis, and may specifically target the malignant clone.12 Its widespread use was hampered by its parenteral administration, cost and toxicity. The development of pegylated forms (PEG-IFN-α) increased tolerance and efficacy in IFN-treated patients.13 Increased thrombotic risk in PV patients is associated with high levels of hemoglobin, impaired rheology, and increased viscosity resulting from erythrocytosis.14-16 Targeting a hematocrit below 45% seems to decrease the thrombotic risk of these patients.10 An additional parameter that might contribute to this risk was brought to light by our work showing abnormal activation of adhesion proteins in PV red blood cells (RBC).17,18 More recently, using a proteomic approach, we also found that membranes of PV RBCs had abnormal expression of several proteins including the adhesion protein Lu/BCAM (Lutheran/Basal Cell Adhesion Molecule) and proteins from the endoplasmic reticulum, such as calreticulin (Calr) and calnexin.19 These findings indicate that JAK2V617F not only impacts cell proliferation, but also induces changes in the repertoire of erythroid proteins expressed in circulating RBCs, which might contribute to the circulatory complications described in PV. haematologica | 2018; 103(6)

In this study, we investigated the effects of HC and IFN treatments on the expression of membrane proteins in circulating RBCs of PV patients. We show that HC and IFN each have a different impact on the expression and function of erythroid membrane proteins. While both drugs seem to normalize Calr expression, HC (but not IFN) enhances the expression of several proteins, including Lu/BCAM and CD147 adhesion proteins, and further exacerbates RBC adhesion to laminin.

Methods Patients and blood samples The study obtained ethical approval from Comité de Protection des Personnes Ile de France VII (PP 14-035); blood samples from 53 PV patients, followed at Saint Louis Hospital, Paris, were obtained with informed consent. Patients are divided in 3 groups: 25 patients (16 men and 9 women; mean age: 63±14 years) were treated with phlebotomy in addition to low-dose aspirin; 20 patients (11 men and 9 women; mean age: 64±16 years) were treated with HC [mean dose 0.8 g/day (0.3-1.5g/day)]; 12 patients (9 men and 3 women; mean age: 50±13 years) were treated with PEG-IFN-α according to international and local guidelines20 [mean dose 115 mg/week (45-180 mg/week)]. The patient groups are small, and not necessarily matched for age or sex, partially due to the treatment guidelines of IFN and HC. Four patients from the first category were treated with HC during the study; blood samples from these 4 patients were collected before and during HCtreatment, at least six months after starting the treatment. All HCtreated patients were compliant to the treatment as they had complete hematologic response and increased mean corpuscular volume (MCV). JAK2V617F mutation was observed in all patients.

Blood samples from 16 regular healthy donors at the Etablissement Français du Sang were also analyzed in this study (9 women and 7 men; mean age: 40±12 years, range 25-67 years).

The 2 control (CT) blood samples used for the proteomic analysis, CT1 and CT2, were chosen to match as well as possible with the PV patient group of the study; they were from 67 and 57 year-old donors, respectively.

Blood samples and ghost preparation Blood samples were collected on sodium heparin tubes. Buffy coat was removed and RBCs were cryopreserved at Centre National de Référence pour les Groupes Sanguins, Paris. RBC membranes were prepared by hypotonic lysis and washed with 5 mM sodium phosphate, pH 8.0 containing 0.2 mM phenylmethylsulfonylfluoride.

973

M. Brusson et al.

A

B

C

Figure 2. Lu/BCAM and CD147 erythroid expression are increased under hydroxycarbamide (HC) treatment. Flow cytometry analyses of (A and B) Lu/BCAM and (C) CD147 expression on red blood cells (RBC) from 17 untreated (UT), 16 HC-treated (HC), and 11 interferon-α-treated (IFN) polycythemia vera patients. MFI: mean fluorescence intensity. Horizontal lines represent medians.

ITRAQ labeling and quantification by nano-liquid chromatography and mass spectrometry Isobaric tag for relative and absolute quantitation (ITRAQ) multiplex analysis was carried out at the 3P5 proteomics facility as previously described,21 with some modifications (Online Supplementary File 1). Trypsin digested peptides from red cell ghosts of 3 PV patients (before and during HC treatment: total 6 samples) and 2 healthy donors (CT) were processed. (For further details see Online Supplementary File 1). Data were analyzed using Protein Pilot 4 with the Uniprot human database. The analysis yielded 12,459 peptide spectrum matches corresponding to 2664 non-redundant peptides assigned to 375 proteins. Proteins were considered over-expressed or down-regulated when they showed PV/CT or PVHC/PV ratios ≥1.3 or ≤0.7, respectively, in at least 2 out of 3 PV samples, with no ratios lower than 1.2 or higher than 0.8 for over-expressed or down-regulated proteins, respectively.

(GlutaMAX™ I, 4500 mg/L D-glucose) without sodium pyruvate for 2 hours at 37°C, 0.5% CO2, centrifuged at 1500 rpm for 5 min, and suspended in 1 mL of the same DMEM containing 32P (160 mCi) overnight at 37°C, 0.5% CO2. RBCs were lysed for 45 min at 4°C with lysis buffer containing: 20mM Tris, 150mM NaCl, 5mM EDTA, 0.002% NaN3, 1% Triton X-100, 0.2% BSA, phosphatase (Sigma-Aldrich), and protease inhibitor cocktails (Roche Diagnostics). Lu/BCAM was immunopurified with F241 mAb and protein A-sepharose CL4B beads (Roche Diagnostics) overnight at 4°C. After electrophoresis and protein transfer, phosphorylated proteins were detected and quantified with a FujiFilm BAS-1800 II PhosphorImager, using Image Reader BAS-1800 II v.1.8 and Multi Gauge, v.3.0 software, respectively (Fuji). Total Lu/BCAM was then revealed on the same membrane using biotinylated anti-Lu/BCAM antibody (R&D Systems) and ECL. Proteins were quantified using Chemidoc and Quantity One software.

JAK2V617F allele burden quantification The percentage of the JAK2V617F allele was determined in DNA extracted from whole blood using the Mutaquant® (Ipsogen) method as described.22

Statistical analysis Statistical analysis was performed with GraphPad Prism using Mann-Whitney test (Figures 1-3) and paired t-test (Figure 4): *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

Flow adhesion assays Red blood cell adhesion to laminin 521 was measured under flow conditions using a capillary flow chamber. Recombinant laminin 521 (BioLamina) at 5 ng/mL was immobilized in Vena8 Endothelial+TM biochips (internal channel dimensions: length 20 mm, width 0.8 mm, height 0.12 mm). RBCs were perfused at 5.107 RBCs/mL for 5 minutes (min) at 0.5 dyn/cm² and 5-min washouts were performed at 0.5, 1, 2, 3, 5, 7 dyn/cm² using the ExiGo™ pump (Cellix Ltd., Dublin, Republic of Ireland). After each wash, adherent RBCs were counted in 6 representative areas along the centerline of the biochip using the AxioObserver Z1 microscope (10X objective) and AxioVision 4 analysis software (Carl Zeiss).

Images of the same 6 areas were obtained throughout each experiment using the “Mark and Find” module of AxioVision analysis software.

Flow cytometry Cell surface expression of Lu/BCAM and CD147 and percentage of reticulocytes were determined using specific antibodies and Retic-CountTM (thiazole orange) reagent, respectively, using a BD FACScanto II flow cytometer (Becton-Dickinson), as described.23,24

Phosphorylation assays and western blot Phosphorylation of Lu/BCAM was assessed in PV RBCs, as described.23,24 Briefly, RBCs were incubated in DMEM 974

Results Altered membrane protein expression in RBCs of HC-treated patients In our recent work, we showed abnormal expression of several proteins at the membrane of PV RBCs, with a high number of proteins from the endoplasmic reticulum, including calreticulin (Calr).19 To verify if HC treatment restores normal expression of these proteins, we performed a proteomic analysis with RBC ghosts from 3 patients before and during their treatment with HC (prepost patients). Three hundred and seventy-five proteins were confidently identified, with a number of peptides that allowed 358 distinct protein quantification (Online Supplementary File 2). Among the deregulated proteins initially reported,19 HC decreased the expression of several over-expressed proteins and increased the expression of those down-regulated (Table 1). Nevertheless, the treatment did not seem to restore a normal expression pattern of these proteins for all 3 patients when compared to control (Table 1 and Online Supplementary Table S1). In addition, the comparative analysis showed that HC deregulated the expression of 53 proteins that had normal expression before the treatment started, with overexpression of haematologica | 2018; 103(6)

Effects of HC and IFN-α on PV RBC adhesion Table 1. ITRAQ ratios of proteins with increased (i) or decreased (d) expression at the membrane of polycythemia vera (PV) red blood cells, and effect of hydroxycarbamide (HC) treatment. Proteins from endoplasmic reticulum are in bold.

CT: control; PVHC: blood sample from HC-treated patient.

37 proteins (ratio ≥1.3) and downregulation of 16 proteins (ratio ≤0.7) (Table 2).

Expression of Calr was diminished during HC treatment but did not reach control levels (Table 1). We analyzed Calr by western blot in RBC ghosts from a total of 4 prepost patients, including the 3 patients of the proteomics study, and found that its expression was decreased during the treatment without reaching significance (Figure 1A), confirming the proteomics results. This was probably due to the variability and relatively short duration of the treatment in this group (mean duration 1.4 years, range 0.5-3 years). Therefore, we tested for the presence and expression level of Calr by western blot in a group of 11 patients treated with HC for a longer period of time (mean duration 5.9 years, range 0.5-20 years), a group of 19 untreated (UT) patients (i.e. with no cytotoxic antiproliferative treatment), and a group of 11 healthy donors (Figure 1B). We found no significant difference between the HC and healthy donor groups indicating that long-term HC treatment restores the expression of Calr. Similarly, we investigated Calr in a group of 7 patients treated with IFN and found no difference with the healthy donors group, strongly suggesting that IFN restores normal expression of Calr in PV RBCs (Figure 1B).

Overexpression of Lu/BCAM and CD147 in HC-treated patients In the group of membrane proteins abnormally expressed in PV RBCs,19 Lu/BCAM was the only protein whose overexpression was further exacerbated by HC in all 3 patients, with a fold increase comprised between 6.6 and 11.8 (Table 1). Lu/BCAM is a low abundance surface protein that is expressed on a subpopulation of circulating RBCs. To determine whether HC increases the percentage haematologica | 2018; 103(6)

of Lu/BCAM-positive RBCs or Lu/BCAM expression level per RBC, or both, we performed flow cytometry assays with 18 UT and 17 HC blood samples, using a specific mouse monoclonal anti-Lu/BCAM antibody. The percentage of Lu/BCAM-expressing cells and the number of Lu/BCAM molecules per RBC, estimated by the mean fluorescence intensity (MFI) of Lu/BCAM-positive RBCs, were significantly higher in the HC group (median 94.35%, MFI=4570) than in the UT group (median 73%, MFI=1819) (P<0.001, Mann-Whitney test) (Figure 2A and B, and Online Supplementary Figure S1A). As Lu/BCAM expression is known to be higher in young (reticulocytes) than mature RBCs,25 we determined the percentage of reticulocytes in all blood samples. There was no significant difference between both patient groups (mean UT:0.6 ±0.05; HC: 0.9±0.17, P=0.097) (Online Supplementary Figure S1B) indicating that the increase of Lu/BCAM was not due to an HC-induced imbalance between reticulocytes and mature RBCs. Flow cytometry analysis was conducted with 11 blood samples from IFN-treated patients and no significant difference was found with the UT group (Lu/BCAM-positive RBCs: 81.4% vs. 73%, P=0.3722; MFI: 2479 vs. 1819, P=0.5588, respectively) (Figure 2A and B), indicating that IFN did not influence Lu/BCAM protein expression.

In order to explore the potential effect of HC and IFN on the expression of other erythroid adhesion proteins, we performed flow cytometry analysis of four additional adhesion markers: CD44, CD47, CD147 and CD242. All 4 markers are expressed on almost 100% of circulating RBCs and there was no effect of HC or IFN on this percentage (data not shown). Likewise, there was no effect of either HC or IFN treatment on the expression level per RBC of CD44, CD47 and CD242, estimated by the MFI, 975

M. Brusson et al.

A

B

C

D

Figure 3. Increased adhesion to laminin of red blood cells (RBCs) from hydroxycarbamide (HC)-treated patients. (A) Adhesion to laminin at 3 dyn/cm2 of RBCs from 11 control (CT), 17 untreated (UT), 16 HC-treated (HC) and 11 IFN-treated (IFN) polycythemia vera patients. Horizontal lines represent medians. (B) The JAK2V617F allele burden (%JAK2V617F). RBC adhesion as a function of (C) %JAK2V617F (R²=0.048) and (D) Lu/BCAM mean fluorescence intensity (MFI) (R²=0.67) for 16 HCtreated patients.

when compared to UT patients. However, we observed increased expression of CD147 in the HC group (median MFI=8472) when compared to the UT group (median MFI=5406) (P<0.0001, Mann-Whitney test) (Figure 2C) suggesting a common upregulation of Lu/BCAM and CD147 by HC. This increase was confirmed by proteomics in the 3 pre-post patients who showed increased CD147 expression during HC treatment (Table 2).

Increased RBC adhesion to laminin in HC-treated patients As Lu/BCAM mediates abnormal adhesion of PV RBCs to laminin,17 we examined the effect of HC treatment on PV RBC adhesion by performing adhesion assays under flow conditions. RBCs from HC-treated patients adhered much more than those from UT patients, with respective medians of 969 and 231 RBCs/mm2 at 3 dyn/cm2 (P=0.0047, Mann-Whitney test) whereas no significant difference was found between the IFN (271 RBCs/mm2) and UT groups (P=0.8508, Mann-Whitney test) (Figure 3A). Adhesion of RBCs from 6 control donors was minor (median=37 RBCs/mm2), and significantly lower than in the UT, HC and IFN groups (P<0.01, Mann-Whitney test).

The impact of HC and IFN on cell adhesion was also analyzed by comparing the UT, HC and IFN groups in terms of RBC adhesion value distributions. We found that these values were broadly dispersed in the UT and HC groups but were less scattered in the IFN group (Figure 3A), with a significant difference between the variance of the IFN group as compared to the UT (P<0.05) or the HC 976

group (P<0.05) (Fisher exact test). This was probably linked to the JAK2V617F allele burden of each group. As a matter of fact, we have previously shown that PV RBC adhesion to laminin in UT patients was correlated with the JAK2V617F allele burden, defined as the percentage of circulating JAK2 alleles with the V617F mutation (%V617F).26 We determined the %V617F of all patients and, as expected and as reported in previous studies,13 the median was significantly lower in the IFN than in the UT and HC groups (Figure 3B). Moreover, we found no significant difference between the UT and the HC groups, despite a lower median in the latter (Figure 3B), indicating that increased RBC adhesion in the HC group was independent of the %V617F and was most probably due to the expression level of Lu/BCAM. This was demonstrated by plotting RBC adhesion values against the percentage of JAK2V617F (Figure 3C) and the MFI of Lu/BCAM (Figure 3D).

Longitudinal study of HC treatment in 4 PV patients As only HC had activating effects on RBC adhesion, and because of wide interindividual variability in PV, we further investigated the effects of HC treatment in the 4 prepost HC patients of our cohort by performing adhesion and flow cytometry assays. This longitudinal analysis showed a great increase of adhesion after HC treatment for all 4 patients (Figure 4A and B), confirming the results obtained with the UT and HC groups. As seen in the UT and HC groups, we found that the increased adhesion was associated with increased Lu/BCAM expression, both in haematologica | 2018; 103(6)

Effects of HC and IFN-α on PV RBC adhesion

Table 2. ITRAQ ratios of 53 proteins modulated by hydroxycarbamide (HC) in polycythemia vera (PV) red blood cell ghosts.

Protein Hemoglobin subunit gamma-1 Lu/BCAM Estradiol 17-β-dehydrogenase 11 Vacuolar protein sorting-associated protein 4A cAMP-dependent protein kinase type I-α regulatory subunit Atlastin-3 ATP-binding cassette sub-family G member 2 Kell blood group. Metallo-endopeptidase Monocarboxylate transporter 1 Membrane-associated progesterone receptor component 2 RAB1B protein Gamma-adducin Membrane transport protein XK Phosphatidylinositide phosphatase SAC1 Rh-associated glycoprotein Extended synaptotagmin-2 Solute carrier family 2. Facilitated glucose transporter member 1 Vesicle-trafficking protein SEC22b Heat shock cognate 71 kDa protein Basigin (CD147) Plasma membrane calcium-transporting ATPase 4 Heat shock 70 kDa protein 1A/1B Lysophospholipid acyltransferase 2 Dolichyl-diphosphooligosaccharide--protein glycosyltransferase subunit 1 Vacuolar protein sorting-associated protein 13A Multidrug resistance-associated protein 5 Ras-related protein Rab-8A Putative uncharacterized protein DKFZp686H20196 ADP-ribosylation factor-like protein 6-interacting protein 1 Neutral cholesterol ester hydrolase 1 Calcium and integrin-binding protein 1 Abhydrolase domain-containing protein 16A Mannose-P-dolichol utilization defect 1 protein Dolichyl-phosphate β-glucosyltransferase LEM domain-containing protein 2 F-actin-capping protein subunit α-1 Large neutral amino acids transporter small subunit 3 ATP-citrate synthase TXNDC5 protein Phosphatidylinositol 5-phosphate 4-kinase type-2 α AP complex subunit β Raichu404X Tropomyosin α-3 chain Galectin-3 Small conductance potassium channel type 4 transcript variant 4 ATP-dependent 6-phosphofructokinase Ras-related C3 botulinum toxin substrate 1 AP-2 complex subunit α-1 Cell division control protein 42 homolog AP-2 complex subunit mu Epididymis secretory protein Li 109 26S proteasome non-ATPase regulatory subunit 2 Apolipoprotein A-I haematologica | 2018; 103(6)

Gene

PV1HC PV1

PVHC PV PV2HC PV2

HBG1 BCAM HSD17B11 VPS4A DKFZp779L0468 ATL3 ABCG2 KEL SLC16A1 PGRC2 RAB1B ADD3 XK SACM1L RHAG ESYT2 SLC2A1 SEC22B HSPA8 hEMMPRIN ATP2B4 HEL-S-103 MBOAT2 RPN1 VPS13A ABCC5 RAB8A DKFZp686H20196 ARL6IP1 NCEH1 CIB1 ABHD16A MPDU1 ALG5 LEMD2 CAPZA1 SLC43A1 ACLY TXNDC5 PIP4K2A AP2B1 Raichu404X TPM3 LGALS3 KCNN4 PFKM RAC1 AP2A1 CDC42 AP2M1 HEL-S-109 PSMD2 APOA1

15.0 11.8 9.0 8.7 8.6 5.6 4.7 4.4 3.6 3.0 2.8 2.6 2.6 2.3 2.2 2.2 2.1 2.1 2.1 1.9 1.9 1.9 1.8 1.8 1.7 1.5 1.5 1.5 1.4 1.4 1.4 1.3 1.3 1.3 1.3 1.2 1.2 0.7 0.7 0.6 0.4 0.4 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.1 0.0 0.0

6.9 6.6 2.4 2.1 1.3 1.2 2.5 1.5 2.8 2.6 1.2 1.2 1.5 2.4 1.2 6.2 1.3 1.4 3.2 2.0 1.3 4.4 1.2 5.1 6.9 1.2 2.5 15.7 1.4 1.4 1.3 2.3 2.4 1.2 1.6 1.7 1.7 0.7 0.7 0.6 0.6 0.7 0.1 0.7 0.3 0.2 0.4 0.4 0.6 0.4 0.1 0.0 0.2

PV3HC PV3 1.7 9.5 8.9 4.4 8.7 2.2 99.1 6.5 12.9 4.4 2.2 1.5 4.4 4.4 5.5 6.2 3.9 6.8 10.4 4.8 2.8 4.4 1.3 2.1 2.9 2.2 1.8 2.0 1.3 1.4 1.3 4.5 14.7 2.9 1.3 1.4 1.5 0.3 0.4 0.5 0.6 0.6 0.1 0.0 0.5 0.2 0.8 0.5 0.7 0.7 0.1 0.0 0.1 977

M. Brusson et al.

terms of percentage of Lu/BCAM-positive RBCs and associated mean fluorescence intensity (Figure 4C), confirming that HC increases both the percentage of Lu/BCAM-positive RBCs and the expression of this protein within each positive RBC. This increase was reflected by higher amounts of Lu/BCAM during HC treatment detected by western blot (Figure 4D).

We have shown that Lu/BCAM long isoform was constitutively phosphorylated in PV RBCs18 and that inhibi-

A

B

C

D

E

978

tion of this phosphorylation was associated with less PV RBC adhesion to laminin.17 In order to determine whether the increased adhesion induced by HC was associated with Lu/BCAM activation, we tested the phosphorylation state of Lu/BCAM long isoform. Radiophospholabeling experiments confirmed that Lu/BCAM was phosphorylated in PV RBCs and showed increased phosphorylation during HC treatment (Figure 4E, top panel). This increased phosphorylation was not only due to increased Lu/BCAM

Figure 4. Longitudinal analyses of red blood cell (RBC) adhesion and Lu/BCAM expression and activation in 4 polycythemia vera (PV) patients during hydroxycarbamide (HC) treatment. All results were obtained with RBCs from 4 PV patients before (UT) and during (HC) HC treatment. (A) Typical images of RBCs adhering to laminin 521 at 3 dyn/cm². (B) Mean number of RBCs/mm2. (C) Flow cytometry analysis of Lu/BCAM expression: (top) histograms (before: dotted; during: solid); (bottom) percentage of RBCs expressing Lu/BCAM and mean fluorescence intensity (MFI). (D) Western blot analysis of Lu/BCAM expression; the upper band corresponds to the long isoform Lu and the lower one to the short isoform Lu(v13). (E) Lu/BCAM phosphorylation rate.

The top (P) and bottom (T) panels show the phosphorylation and the total amount of the immunopurified proteins, respectively. The phosphorylated fraction is determined by the P/T ratio.

haematologica | 2018; 103(6)

Effects of HC and IFN-Îą on PV RBC adhesion

expression, as the phosphorylation ratio (P/T), in which P is the measured phosphorylation and T the expression of total Lu/BCAM, was higher during than before HC treatment (Figure 4E, bottom panel), indicating that HC both increases the expression of Lu/BCAM and the proportion of phosphorylated molecules.

Discussion Our recent work showed abnormal expression of several proteins in RBCs of PV patients,19 revealing qualitative alterations in these cells in addition to their increased number in the circulation. The aim of the current study was to address the effects of HC treatment on the expression of erythroid membrane proteins of PV patients using a global proteomic approach.

Our proteomic results obtained with ghosts from 3 prepost patients showed that HC treatment tends to restore the expression of deregulated membrane proteins.

Nevertheless, normal expression of these membrane proteins was not reached, suggesting that the duration of the treatment for these patients was not sufficient, which is supported by the results of Calr expression in the group of 11 patients treated with HC for a longer duration.

Since we have shown that Calr overexpression was associated with the presence of JAK2V617F,19 the decreased Calr expression in the HC group could have been the consequence of fewer JAK2V617F-positive clones in the bone marrow of these patients. This was reflected by their lower %V617F when compared to the UT group (median=42.5 and 60% for HC and UT, respectively), although the difference was not significant, which is in accordance with a study showing that HC does not appreciably reduce the JAK2V617F allele burden.27 The IFN group also showed lower expression levels of Calr than the UT group, along with a significant decrease of the %V617F (median=15%). This decrease was expected as IFN treatment is known to reduce the JAK2V617F allele burden, with some cases of complete remission.13 Calr is a calcium-binding chaperone that promotes efficient folding of glycoproteins but whose role in circulating RBCs has not been elucidated. Because calcium regulates several erythrocyte functions,28 we believe that the decrease in Calr expression triggered by HC and IFN might be beneficial to the calcium homeostasis and subsequent calciumregulated functions of PV RBCs. However, despite the tendency to normalize Calr expression, the proteomic analysis showed that HC treatment clearly affects the expression of several membrane proteins. This is not the first example of such deregulation, as HC is known to induce the neosynthesis of fetal hemoglobin in the erythroid lineage and is used to this purpose in sickle cell disease (SCD) patients.29,30 Among the proteins over-expressed by HC in PV patients, Lu/BCAM and CD147 were also reported to be up-regulated in RBCs of sickle cell disease patients treated with HC,31 indicating that the observed effect of HC is not specific to PV. This suggests the triggering of common pathways by HC independently of the underlying illness, most likely through the activation of gene transcription. This is supported by the findings of Odievre et al.31 showing increased expression of Lu/BCAM in erythroid progenitors from SCD patients and healthy donors differentiated in vitro in the presence of HC.31 Moreover, we have previhaematologica | 2018; 103(6)

ously shown activation of Lu/BCAM gene transcription

by HC in endothelial cells grown in vitro.32 Similar to our previous reports in sSCD, we found that HC significantly increases erythroid Lu/BCAM expression by enhancing both the percentage of Lu/BCAM-positive RBCs and the Lu/BCAM copy number per RBC. The increase in the percentage of Lu/BCAM-positive RBCs indicates that the observed higher expression is not due to the documented increase of the RBC volume consecutive to the HC treatment. Opposite to RBCs from HC-treated SCD patients, those from treated PV patients had increased adhesion to laminin. In SCD, increased Lu/BCAM phosphorylation is achieved in a PKA/cAMPdependent manner, with patients showing high cAMP levels that are decreased during HC treatment.23 In PV, Lu/BCAM phosphorylation is driven by a JAK2V617Fdependent pathway involving Akt, and not PKA, which could explain the difference of HC effects between SCD and PV. HC is a nitric oxide (NO) donor in vivo33-35 and its effect in PV RBCs might be due to Akt activation through an NO-dependent pathway. Indeed, RBCs have been shown to express an active NO synthase (NOS)36 and are thus capable of generating NO and activating downstream effectors like Akt. Such activation of Akt by NO donors has been reported in endothelial cells37 and chick retinal neurons;38 it occurs through the activation of soluble guanylyl cyclase (sGC) and protein kinase G. Interestingly, HC has also been shown to activate sGC in erythroid cells,39 such activation might trigger Akt activity in PV RBCs and lead to Lu/BCAM activation and increased RBC adhesion.

Activation of leukocytes, platelets and endothelial cells is known to promote a prothrombotic state. Markers for such activation have been reported in PV patients, including activation of polymorphonuclear leukocytes and endothelial cells,40 and increased levels of leukocyteplatelet aggregates.41 A previous report showed that treatment of PV patients with HC did not influence these marker levels.42 Our study reveals a new pro-adhesive effect of HC in PV patients that acts by increasing the expression of two adhesion molecules, Lu/BCAM and CD147, and by activating RBC adhesion to laminin.

CD147, also named basigin or neurothelin, is a member of the immunoglobulin superfamily that is expressed in various tissues, including brain, leukocytes, endothelial cells, and most tumor cell lines. CD147 is expressed during erythroid differentiation43 and is carrier molecule for the blood group antigen Oka.44 Its erythroid adhesive function is important during the circulation of RBCs in the spleen.45 It was shown to bind endothelial cells and fibroblasts,46 and to be a receptor essential for erythrocyte invasion by Plasmodium falciparum.47 It is noteworthy that Lu/BCAM and CD147 are adhesion markers linked to progression of solid tumors. Lu/BCAM is over-expressed in carcinomas in vivo and up-regulated following malignant transformation in some cell types.48-51 Likewise, CD147 plays a central role in the progression of many cancers by inducing the secretion of matrix metalloproteinases and various cytokines (reviewed by Xiong, Edwards and Zhou52). Several studies have indicated that CD147 is a multifunctional glycoprotein that inhibits tumor cell anoikis,53 enhances tumor angiogenesis,54 and promotes invasion, metastasis,55 and glycolytic energy metabolism.56 Considering the wide tissue distribution of both proteins, and the fact that HC distributes throughout the body reaching approximately all 979

M. Brusson et al.

tissues, one would expect that Lu/BCAM and CD147 would be also over-expressed in cell types other than RBCs, such as endothelial and epithelial cells. This is supported by our study showing that HC increases the endogenous expression of Lu/BCAM in endothelial cells ex vivo.32 Consequently, overexpression of Lu/BCAM and CD147 in HC-treated patients might possibly have a negative impact both in the vascular territory by promoting abnormal cellular interactions, and in the higher incidence of skin cancer reported in HC-treated PV patients.57 Although HC and IFN inhibit the proliferation of progenitor cells, nothing was known about their impact on PV RBCs once they exit the bone marrow. Our study is the first to address the effects of these molecules on PV RBCs, and to show that HC and IFN have a different impact on RBC protein expression and adhesive function.

Leukocytosis and high hematocrit are two parameters involved in thrombotic events in MPN patients, and are both decreased during HC treatment.10 A PV Study Group non-randomized trial showed that HC was associated with a lower incidence of early thrombosis compared to a historical cohort treated with phlebotomy alone (6.6% vs.

14% at 2 years).58 Nevertheless, thrombosis is not totally abrogated in HC-treated patients. Patients treated with IFN also encounter less thrombotic events,59 but whether this happens more or less than those treated with HC is still unknown because there had been no clinical trials comparing the effects of both drugs in the same cohort of MPN patients. Such trials are currently ongoing and might show differences between HC and IFN regarding circulatory complications.

Altogether, our study shows that HC and IFN each have a different impact on RBCs and reveals unexpected

References 1. Baxter EJ, Scott LM, Campbell PJ, et al.

Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet. 2005;365(9464):1054-1061.

2. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera.

Nature.

2005; 434(7037):1144-1148.

3. Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med. 2005;352(17):1779-1790.

4. Levine RL, Wadleigh M, Cools J, et al.

Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Cancer Cell. 2005; 7(4):387-397.

5. Schafer AI. Bleeding and thrombosis in the myeloproliferative disorders. Blood. 1984; 64(1):1-12.

6. Tefferi A, Barbui T. Essential Thrombocythemia and Polycythemia Vera: Focus on Clinical Practice. Mayo Clin Proc.

2015;90(9):1283-1293.

7. Nazha A, Khoury JD, Verstovsek S, Daver N. Second line therapies in polycythemia vera: What is the optimal strategy after hydroxyurea failure? Crit Rev Oncol Hematol. 2016;105:112-117.

8. Gwilt PR, Tracewell WG. Pharmacokinetics and pharmacodynamics

980

9.

10.

11.

12.

13.

14.

15.

16.

adverse effects of HC on RBC physiology in PV. Our findings show that HC deregulates the expression of several proteins at the red cell membrane providing new insights into the effects of this molecule on gene regulation and protein recycling or maturation during erythroid differentiation. Furthermore, our study shows that HC increases the expression of two ubiquitously expressed proteins that are linked to progression of solid tumors.

Investigating such overexpression in tissues other than blood cells will be of interest in MPNs. Acknowledgments We thank Ms. Dominique Gien, Sirandou Tounkara and Eliane Véra at the Centre National de Référence pour les Groupes Sanguins for the management of blood samples, EmilieFleur Gautier and Morgane Le Gall for assistance in proteomics data analyses, and the France Intergroupe Syndromes Myéloprolifératifs (FIM) for helpful discussions.

Funding The research leading to these results has received funding from the Institut National de la Santé et de la Recherche Médicale (Inserm), the Institut National de la Transfusion Sanguine, the European Union’s Horizon 2020 research and innovation program through the RELEVANCE project under the Marie Skłodowska-Curie grant agreement n. 675115, and the Laboratory of Excellence GR-Ex, reference ANR-11-LABX0051. GR-Ex is funded by the program “Investissements d’Avenir” of the French National Research Agency, reference ANR-11-IDEX-0005-02. MB and MDG were funded by the Ministère de l’Enseignement Supérieur et de la Recherche (Ecole Doctorale BioSPC). They received a financial support from: Club du Globule Rouge et du Fer and Société Française d’Hématologie.

of hydroxyurea. Clin Pharmacokinet. 1998; 34(5):347-358.

Yarbro JW. Mechanism of action of hydroxyurea. Semin Oncol. 1992;19(3 Suppl 9):1-10.

Marchioli R, Finazzi G, Specchia G, et al.

Cardiovascular events and intensity of treatment in polycythemia vera. N Engl J Med. 2013;368(1):22-33.

Kiladjian JJ, Chevret S, Dosquet C, Chomienne C, Rain JD. Treatment of polycythemia vera with hydroxyurea and pipobroman: final results of a randomized trial initiated in 1980. J Clin Oncol. 2011; 29(29):3907-3913.

Kiladjian JJ, Masse A, Cassinat B, et al.

Clonal analysis of erythroid progenitors suggests that pegylated interferon alpha-2a treatment targets JAK2V617F clones without affecting TET2 mutant cells. Leukemia.

2010;24(8):1519-1523.

Kiladjian JJ, Cassinat B, Chevret S, et al.

Pegylated interferon-alfa-2a induces complete hematologic and molecular responses with low toxicity in polycythemia vera.

Blood. 2008;112(8):3065-3072.

Kwaan HC, Wang J. Hyperviscosity in polycythemia vera and other red cell abnormalities. Semin Thromb Hemost.

2003;29(5):451-458.

Pearson TC, Wetherley-Mein G. Vascular occlusive episodes and venous haematocrit in primary proliferative polycythaemia.

Lancet. 1978;2(8102):1219-1222.

Spivak JL. Polycythemia vera: myths,

17.

18.

19.

20.

21.

22.

23.

mechanisms, and management. Blood.

2002;100(13):4272-4290.

De Grandis M, Cambot M, Wautier MP, et al. JAK2V617F activates Lu/BCAM-mediated red cell adhesion in polycythemia vera through an EpoR-independent Rap1/Akt pathway. Blood. 2013;121(4):658-665.

Wautier MP, El Nemer W, Gane P, et al.

Increased adhesion to endothelial cells of erythrocytes from patients with polycythemia vera is mediated by laminin alpha5 chain and Lu/BCAM. Blood. 2007; 110(3):894-901.

Brusson M, Cochet S, Leduc M, et al.

Enhanced calreticulin expression in red cells of polycythemia vera patients harboring the JAK2V617F mutation.

Haematologica. 2017;102(7):e241-e244.

Barbui T, Barosi G, Birgegard G, et al.

Philadelphia-negative classical myeloproliferative neoplasms: critical concepts and management recommendations from European LeukemiaNet. J Clin Oncol.

2011;29(6):761-770.

Gautier EF, Ducamp S, Leduc M, et al.

Comprehensive Proteomic Analysis of Human Erythropoiesis. Cell Rep. 2016; 16(5):1470-1484.

Kouroupi E, Kiladjian JJ, Dosquet C, et al.

Does increasing the JAK2V617F assay sensitivity allow to identify more patients with MPN? Blood Cancer J. 2012;2(5):e70.

Bartolucci P, Chaar V, Picot J, et al.

Decreased sickle red blood cell adhesion to laminin by hydroxyurea is associated with

haematologica | 2018; 103(6)

Effects of HC and IFN-Îą on PV RBC adhesion

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

inhibition of Lu/BCAM protein phosphorylation. Blood. 2010;116(12):2152-2159.

Gauthier E, Rahuel C, Wautier MP, et al.

Protein kinase A-dependent phosphorylation of Lutheran/basal cell adhesion molecule glycoprotein regulates cell adhesion to laminin alpha5.

J Biol Chem.

2005;280(34):30055-30062.

El Nemer W, Gane P, Colin Y, et al. The Lutheran blood group glycoproteins, the erythroid receptors for laminin, are adhesion molecules. J Biol Chem. 1998; 273(27):16686-16693.

De Grandis M, Cassinat B, Kiladjian JJ, Chomienne C, El Nemer W. Lu/BCAMmediated cell adhesion as biological marker of JAK2V617F activity in erythrocytes of polycythemia vera patients. Am J Hematol.

2015;90(7):E137-138.

Antonioli E, Carobbio A, Pieri L, et al.

Hydroxyurea does not appreciably reduce JAK2 V617F allele burden in patients with polycythemia vera or essential thrombocythemia. Haematologica. 2010;95(8):14351438.

Bogdanova A, Makhro A, Wang J, Lipp P, Kaestner L. Calcium in red blood cells-a perilous balance. Int J Mol Sci. 2013; 14(5):9848-9872.

Charache S. Fetal hemoglobin, sickling, and sickle cell disease. Adv Pediatr. 1990;37:1-31.

Platt OS, Orkin SH, Dover G, Beardsley GP, Miller B, Nathan DG. Hydroxyurea enhances fetal hemoglobin production in sickle cell anemia. J Clin Invest.

1984;74(2):652-656.

Odievre MH, Bony V, Benkerrou M, et al.

Modulation of erythroid adhesion receptor expression by hydroxyurea in children with sickle cell disease. Haematologica.

2008;93(4):502-510.

Chaar V, Laurance S, Lapoumeroulie C, et al. Hydroxycarbamide decreases sickle reticulocyte adhesion to resting endothelium by inhibiting endothelial lutheran/basal cell adhesion molecule (Lu/BCAM) through phosphodiesterase 4A activation. J Biol Chem. 2014;289(16):11512-11521.

Glover RE, Ivy ED, Orringer EP, Maeda H, Mason RP. Detection of nitrosyl hemoglobin in venous blood in the treatment of sickle cell anemia with hydroxyurea. Mol Pharmacol. 1999;55(6):1006-1010.

Jiang J, Jordan SJ, Barr DP, Gunther MR, Maeda H, Mason RP. In vivo production of nitric oxide in rats after administration of hydroxyurea.

Mol Pharmacol.

1997;52(6):1081-1086.

Nahavandi M, Wyche MQ, Perlin E, Tavakkoli F, Castro O. Nitric Oxide Metabolites in Sickle Cell Anemia Patients after Oral Administration of Hydroxyurea;

haematologica | 2018; 103(6)

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

Hemoglobinopathy. Hematology. 2000; 5(4):335-339.

Kleinbongard P, Schulz R, Rassaf T, et al.

Red blood cells express a functional endothelial nitric oxide synthase. Blood.

2006;107(7):2943-2951.

Kawasaki K, Smith RS Jr, Hsieh CM, Sun J, Chao J, Liao JK. Activation of the phosphatidylinositol 3-kinase/protein kinase Akt pathway mediates nitric oxide-induced endothelial cell migration and angiogenesis.

Mol Cell Biol. 2003;23(16):5726-5737.

Mejia-Garcia TA, Portugal CC, Encarnacao TG, Prado MA, Paes-de-Carvalho R. Nitric oxide regulates AKT phosphorylation and nuclear translocation in cultured retinal cells. Cell Signal. 2013;25(12):2424-2439.

Cokic VP, Andric SA, Stojilkovic SS, Noguchi CT, Schechter AN. Hydroxyurea nitrosylates and activates soluble guanylyl cyclase in human erythroid cells. Blood.

2008;111(3):1117-1123.

Falanga A, Marchetti M, Evangelista V, et al. Polymorphonuclear leukocyte activation and hemostasis in patients with essential thrombocythemia and polycythemia vera.

Blood. 2000;96(13):4261-4266.

Falanga A, Marchetti M, Vignoli A, Balducci D, Barbui T. Leukocyte-platelet interaction in patients with essential thrombocythemia and polycythemia vera.

Exp Hematol. 2005;33(5):523-530.

Cerletti C, Tamburrelli C, Izzi B, Gianfagna F, de Gaetano G. Platelet-leukocyte interactions in thrombosis. Thromb Res. 2012;129(3):263-266.

Papayannopoulou T, Brice M. Integrin expression profiles during erythroid differentiation. Blood. 1992;79(7):1686-1694.

Spring FA, Holmes CH, Simpson KL, et al.

The Oka blood group antigen is a marker for the M6 leukocyte activation antigen, the human homolog of OX-47 antigen, basigin and neurothelin, an immunoglobulin superfamily molecule that is widely expressed in human cells and tissues. Eur J Immunol. 1997;27(4):891-897.

Coste I, Gauchat JF, Wilson A, et al.

Unavailability of CD147 leads to selective erythrocyte trapping in the spleen. Blood.

2001;97(12):3984-3988.

Mutin M, Dignat-George F, Sampol J. Immunologic phenotype of cultured endothelial cells: quantitative analysis of cell surface molecules. Tissue Antigens.

1997;50(5):449-458.

Crosnier C, Bustamante LY, Bartholdson SJ, et al. Basigin is a receptor essential for erythrocyte invasion by Plasmodium falciparum. Nature. 2011;480(7378):534-537.

Bernemann TM, Podda M, Wolter M, Boehncke WH. Expression of the basal cell

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

adhesion molecule (B-CAM) in normal and diseased human skin. J Cutan Pathol.

2000;27(3):108-111.

Garinchesa P, Sanzmoncasi M, Campbell I, Rettig W. Non-polarized expression of Basal-cell adhesion molecule B-cam in epithelial ovarian cancers. Int J Oncol.

1994;5(6):1261-1266.

Rettig WJ, Garin-Chesa P, Beresford HR, Oettgen HF, Melamed MR, Old LJ. Cell-surface glycoproteins of human sarcomas: differential expression in normal and malignant tissues and cultured cells. Proc Natl Acad Sci USA. 1988;85(9):3110-3114.

Schon M, Klein CE, Hogenkamp V, Kaufmann R, Wienrich BG, Schon MP. Basal-cell adhesion molecule (B-CAM) is induced in epithelial skin tumors and inflammatory epidermis, and is expressed at cell-cell and cell-substrate contact sites. J Invest Dermatol. 2000;115(6):1047-1053.

Xiong L, Edwards CK 3rd, Zhou L. The biological function and clinical utilization of CD147 in human diseases: a review of the current scientific literature. Int J Mol Sci. 2014;15(10):17411-17441.

Yang JM, O'Neill P, Jin W, et al. Extracellular matrix metalloproteinase inducer (CD147) confers resistance of breast cancer cells to Anoikis through inhibition of Bim. J Biol Chem. 2006;281(14):9719-9727.

Chen Y, Zhang H, Gou X, Horikawa Y, Xing J, Chen Z. Upregulation of HAb18G/CD147 in activated human umbilical vein endothelial cells enhances the angiogenesis.

Cancer Lett.

2009;278(1):113-121.

Dai JY, Dou KF, Wang CH, et al. The interaction of HAb18G/CD147 with integrin alpha6beta1 and its implications for the invasion potential of human hepatoma cells. BMC Cancer. 2009;9:337.

Le Floch R, Chiche J, Marchiq I, et al.

CD147 subunit of lactate/H+ symporters MCT1 and hypoxia-inducible MCT4 is critical for energetics and growth of glycolytic tumors. Proc Natl Acad Sci USA. 2011;108(40):16663-16668.

Najean Y, Rain JD. Treatment of polycythemia vera: the use of hydroxyurea and pipobroman in 292 patients under the age of 65 years. Blood. 1997;90(9):3370-3377.

Fruchtman SM, Mack K, Kaplan ME, Peterson P, Berk PD, Wasserman LR. From efficacy to safety: a Polycythemia Vera Study group report on hydroxyurea in patients with polycythemia vera. Semin Hematol. 1997;34(1):17-23.

Silver RT. Long-term effects of the treatment of polycythemia vera with recombinant interferon-alpha. Cancer. 2006; 107(3):451-458.

981

ARTICLE

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):982-987

A novel regimen for relapsed/refractory adult acute myeloid leukemia using a KMT2A partial tandem duplication targeted therapy: results of phase 1 study NCI 8485 Alice S. Mims,1 Anjali Mishra,2 Shelley Orwick,2 James Blachly,1 Rebecca B. Klisovic,3 Ramiro Garzon,1 Alison R. Walker,1 Steven M. Devine,1 Katherine J. Walsh,1 Sumithira Vasu,1 Susan Whitman,2 Guido Marcucci,4 Daniel Jones,5 Nyla A. Heerema,5 Gerard Lozanski,5 Michael A. Caligiuri,2 Clara D. Bloomfield,1 John C. Byrd,1 Richard Piekarz,6 Michael R. Grever1 and William Blum3

Division of Hematology, Department of Medicine, The Ohio State University and The Ohio State University Comprehensive Cancer Center, Columbus, OH; 1Department of Medicine, The Ohio State University and The Ohio State University Comprehensive Cancer Center, Columbus, OH; 3Department of Hematology and Medical Oncology, Emory University School of Medicine, Winship Cancer Institute, Atlanta, GA; 4 Department of Hematology and Hematopoietic Cell Transplantation, City of Hope National Medical Center, Duarte, CA; 5Department of Pathology, The Ohio State University and The Ohio State University Comprehensive Cancer Center, Columbus, OH and 6Investigational Drug Branch of CTEP, National Cancer Institute, Bethesda, MD, USA 1

ABSTRACT

K

Correspondence: alice.mims@osumc.edu

Received: December 27, 2017.

Accepted: March 21, 2017.

Pre-published: March 22, 2017. doi:10.3324/haematol.2017.186890 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/982 Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

982

MT2A partial tandem duplication occurs in approximately 5-10% of patients with acute myeloid leukemia and is associated with adverse prognosis. KMT2A wild type is epigenetically silenced in KMT2A partial tandem duplication; re-expression can be induced with DNA methyltransferase and/or histone deacetylase inhibitors in vitro, sensitizing myeloid blasts to chemotherapy. We hypothesized that epigenetic silencing of KMT2A wildtype contributes to KMT2A partial tandem duplication-associated leukemogenesis and pharmacologic reexpression activates apoptotic mechanisms important for chemoresponse. We developed a regimen for this unique molecular subset, but due to relatively low frequency of KMT2A partial tandem duplication, this dose finding study was conducted in relapsed/refractory disease regardless of molecular subtype. Seventeen adults (< age 60) with relapsed/refractory acute myeloid leukemia were treated on study.

Patients received decitabine 20 milligrams/meter2 daily on days 1-10 and vorinostat 400 milligrams daily on days 5-10. Cytarabine was doseescalated from 1.5 grams/meter2 every 12 hours to 3 grams/meter2 every 12 hours on days 12, 14 and 16. Two patients experienced dose limiting toxicities at dose level 1 due to prolonged myelosuppression. However, as both patients achieved complete remission after Day 42, the protocol was amended to adjust the definition of hematologic dose limiting toxicity. No further dose limiting toxicities were found. Six of 17 patients achieved complete remission including 2 of 4 patients with KMT2A partial tandem duplication. Combination therapy with decitabine, vorinostat and cytarabine was tolerated in younger relapsed/refractory acute myeloid leukemia and should be explored further focusing on the KMT2A partial tandem duplication subset. (clinicaltrials.gov identifier 01130506).

Introduction Though acute myeloid leukemia (AML) is considered a curable disease, the majority of patients will succumb to their diagnosis. Prognosis has been based primarily on age and cytogenetic/molecular mutations at diagnosis with younger patients (<60 years) faring better, in particular those with European LeukemiaNet haematologica | 2018; 103(6)

A novel regimen for relapsed/refractory adult acute myeloid leukemia Table 1. Treatment Dose and Schedule.

Dose level

1 2 3 4

Decitabine (mg/m2/day) Days 1-10

Vorinostat (mg/day) Days 5-10

Cytarabine (g/m2/q12hr) Days 12,14,16

Number treated

Number of DLTs

20 20 20 20

400 400 400 400

1.5 2 2.5 3

6 3 3 5

2 0 0 0

DLTs: dose limiting toxicities.

(ELN) favorable risk subtypes.1 Despite recent advances in understanding of leukemogenesis, the initial treatment for most AML patients remains largely unchanged over the past 30 years and salvage regimens also remain similar in their use of a cytarabine backbone. However, targeted therapeutics for specific molecular subsets of AML are beginning to emerge including inhibitors for patients with FLT3-ITD/TKD, IDH1 or IDH2 mutations.2,3 Another potential targetable population in AML includes patients with partial tandem duplication (PTD) in the lysine methyltransferase 2A (KMT2A) gene which was formerly known as mixed lineage leukemia 1 (MLL1). The KMT2A gene is located on Chromosome 11q23 and KMT2A PTD occurs in a single allele of this gene. This alteration occurs more commonly in AML with normal cytogenetics and trisomy 11 and is associated with an adverse prognosis.4-6 Multiple mechanisms are attributed to these adverse outcomes including hypermethylation of gene promoters leading to the silencing of potential tumor suppressors.7,8 Our published data show that the KMT2A wild type (WT) allele is epigenetically silenced in AML with KMT2A PTD.9 We have shown that re-expression of the KMT2A WT allele can be induced with DNA methyltransferase (DNMT) and/or histone deacetylase (HDAC) inhibitors.10,11 Indeed, we demonstrated that epigenetic silencing of KMT2A WT contributes to KMT2A PTDassociated leukemogenesis and that pharmacologic reexpression of this gene with DNMT and HDAC inhibitors attenuates the KMT2A PTD leukemogenic potential and activates apoptotic mechanism important to enhance chemosensitivity, in vitro. Re-expression of KMT2A WT following exposure to decitabine, followed by an HDAC inhibitor, was associated with a lower apoptotic threshold and sensitized KMT2A PTD cells to chemotherapyinduced cytotoxicity. In order to develop a regimen that might be effective in the subset of patients with AML and KMT2A PTD, we conducted a phase 1 study of a novel regimen of combined epigenetic and chemotherapies in relapsed and refractory AML patients. Because of the relatively low frequency of KMT2A PTD AML, the initial dose finding portion of this study was conducted in any patient with relapsed/refractory AML regardless of their molecular subtype but was enriched for KMT2A PTD.

Methods

history of neurological toxicity with cytarabine or vorinostat were ineligible (See Appendix 1 for full Eligibility Criteria). Informed written consent approved by The Ohio State University Humans Studies Committee was obtained on all patients prior to enrollment, in accordance with the Declaration of Helsinki.

Treatment The regimen consisted of epigenetic priming with decitabine followed by vorinostat, then high dose cytarabine (which was dose-escalated). The dosing regimen was based on pre-clinical data showing the myeloid apoptotic threshold decreased most significantly compared to other therapeutic sequences. Decitabine was given intravenously over 1 hour at a dose of 20mg/m2/day on Days 1-10. Vorinostat was given orally at a dose of 400mg/day on Days 5-10. Cytarabine was administered intravenously over 2 hours every 12 hours on Days 12, 14, and 16 for 6 doses total.

Cytarabine was dose escalated as follows: dose level 1, 1.5g/m2/q12hr; dose level 2, 2g/m2/q12hr; dose level 3, 2.5g/m2/q12hr; and dose level 4, 3g/m2/q12hr (Table 1). The study was designed in classic 3+3 phase I design schema to determine the maximum tolerated dose (MTD) and define dose limiting toxicity (DLT). Adverse events were graded according to National Cancer Institute Common Toxicity Criteria for Adverse Events, version 4.0. Responses were defined according to International Working Group (IWG) Criteria for AML, including complete remission (CR) and CR with incomplete count recovery (CRi), partial remission (PR), and treatment failure.12 Next generation sequencing using MiSeq platform assessed over 80 AML-associated gene mutations as previously described.13 KMT2A PTD and FLT3-ITD mutations were performed by PCR testing.14,15

Definition of dose limiting toxicity Grade 4 non-hematological toxicity attributable to any of the therapeutic agents, with exception of line-associated venous thrombosis, infection, fatigue, or nausea and vomiting controllable with anti-emetic therapy were defined as DLT. Hematologic toxicity was initially defined as failure to recover peripheral blood counts by Day 42 in patients with <5% blasts in the bone marrow, absence of myelodysplastic changes, and/or absence of disease by flow cytometry in the bone marrow. However, 2 patients at dose level 1 experienced delayed count recovery (beyond day 42) meeting the hematological DLT definition but both patients achieved CR with no long-term sequelae. It was felt disadvantageous to reduce chemotherapy doses due to high risk nature of the disease, and the protocol was modified to extend duration of hematologic DLT to Day 56 with G-CSF permitted to hasten neutrophil recovery in patients with hypoplastic bone marrow after treatment.

Eligibility criteria Eligible patients were adults (â&#x2030;Ľ18 and <60 years) with relapsed/refractory non-M3 AML with adequate organ function and ECOG performance status â&#x2030;¤2. Patients with previous exposure to high-dose cytarabine were eligible. Patients with previous haematologica | 2018; 103(6)

Statistical analysis A standard method 3 + 3 phase I design of dose escalation using 3 patients per dose level cohort and a minimum of 6 patients at the MTD was performed. As an exploratory, phase I study, no infer983

A.S. Mims et al.

ential statistical tests of hypotheses were planned. Data collected are descriptive and provide limited estimates of variability given the small patient sample size at each dose level.

Results Patient characteristics and treatment groups Seventeen adults with relapsed/refractory AML were treated on this phase I study. The median age of patients

was 46 years (range, 21-59 years). Median white blood cell count and bone marrow blast percentage was 3.5 x 103/mL (range, 0.4-75.8) and 66% (range, 4-87%), respectively. The median number of prior induction therapies was 2 (range, 1-4). All patients had previous anthracycline exposure, and 15 patients had previous high-dose cytarabine exposure. Twelve patients had relapsed disease, with 8 patients having prior CR1 duration of less than 1 year and 4 patients experiencing CR1 ranging from 16-38 months. Five patients had primary refractory disease;

Table 2. Patient Characteristics.

Patient Age/ # of prior Pretreatment karyotype Sex inductions 1

58/F

1

2 3 4

45/F 46/M 25/M

3 2 4

5

58/F

1

6 7

54/M 50/M

1 2

8 9

42/F 26/F

2 2

10

59/M

4

11

24/F

3

12 13

42/M 27/F

2 4

14

21/F

3

.

15 16

58/M 58/F

1 2

17

57/M

2

KMT2A PTD

46,XX,t(12;22)(q13;q13)[15]/42-45,XX,-2,add(3) n/a (p25),del(3)(p24),-11,-17,add (17)(p11.2),der(19)t(17;19)(q21;p12),+mar1,+mar2[4]/46,XX[1] 46, XX [19]/nonclonal[1] n/e 47,XY,+11 [2]/47, idem, t(16;21)(p11.2;q22)[18] n/e 45,XY,der n/a dic(13;18)(q21;p11.3)ins(13;?)(q21;?)[13]/45,idem,t(10;12) (p13;q13)[cp2]/46,XY[5]/nonclonal[2] 47,XX,t(16;16)(p13.1;q22),+22[14]/nonclonal PTD+ w/clonal abnormalities [1]/46,XX[7] 46,XY[19]/nonclonal[1] neg 43,XY,del(4)(q31),-5,-12,del(13)(q12q14),n/e 15,add(17)(q25),add(18)(q12),add(19)(p13.3),add(20) (q13.3),add(21)(q22)[1]/ 42-44,sl,-Y,add(6)(p21), +i(8)(q10)[cp3]/ 43,sdl1,-add(6),+del(6)(p23) [cp6]/ 42-44,sl,+8,add(22)(q22)[cp4]/ 43,sl,-Y,+5,45,XX,-7[20] neg 46,XX,t(1;16)(p32;p13.1),der(3)t(3;17)(q29;q11.2), PTD+ der(5)t(5;9)(q11.2;q21),add(8)(q22),9,add(12)(q13),der(12)t(12;12)(p13;q13)ins(12;17) (p13;q11.2q21),+16,der(16)t(1;16),-17,-17,-18,add(22)(q13),+mar1, +mar2,+mar3[12]/45,sl,-der(16)t(1;16)(p32;p13.1),+der(16)t(1;16) (p32;p13.1)add(1) (p36.3)[2]/46,sl,del(6)(q13q25) [1]/nonclonal w/clonal abnormalities[1]/46,XX[4] 45,XY,t(3;12)(q26;p13),t(4;5)(q21;q31),del(5) neg (q15q35),-7,t(10;21)(q22;q22)[cp1] 46,XX,del(11)(p13p15.1)[16]/nonclonal neg w/clonal abnormalities[4] 46,XY,der(12)t(12;21)(p12;q11.2)[19]/nonclonal[1] neg 46,XX,dup(2)(q21q31),t(8;21)(q22;q22)[18]/nonclonal PTD+ w/clonal abnormalities[1]/46,XX[1] 47,XX,der(10)t(10;11)(p11.2;q13)inv(11)(q13q23), PTD+ der(11)t(10;11)(p11.2;q13),+mar1[cp7]/47,idem,add(X) (q26),add(8)(q24.1),add(13)(p11.2),-15,add(18) (p11 2),+21,-mar1,+mar2[cp15]/nonclonal w/clonal abnormalities[2] 46,XY,t(3;15)(p13;q13)[16]/47,idem,+13[2]/46,XY[1]/nonclonal[1] n/a 46,XX,del(12)(p11.2p13)[1]/45,sl,del(19)(q13.3), neg -20,add(21)(q22)[3]/45,sdl1,-7[2]/45,sdl2,-20[2]/46,XX[12] 45,X,-Y[20] neg

Initial WBC

Initial blast count

Dose Level

Length of 1st CR (mos)

40.9

81

1

6

3.5 3.3 3.2

79 77 40

1 1 1

17 10 n/a

2.2

22

1

18

1.3 5.9

36 8

1 2

16 n/a

22.1 2.0

66 51

2 2

n/a n/a

3.2

4

3

38

75.8

82

3

2

4.0 0.4

87 66

3 4

10 11

17.8

82

4

3

30.5 2.1

83 69

4 4

10 n/a

36.6

66

4

6

CR: cytogenetic remission; F: female; M: male; mos: month; n/a: not applicable; n/e: not evaluable; WBC: white blood.

984

haematologica | 2018; 103(6)

A novel regimen for relapsed/refractory adult acute myeloid leukemia each one entered the study after failure of at least 2 conventional regimens. One patient had undergone prior allogeneic stem cell transplant and 2 patients had undergone autologous transplant in CR1 (on study protocols).

Two patients had secondary or therapy-related AML, 14 patients had abnormal karyotypes and 4 patients were found to have KMT2A PTD molecular subtype. Clinical as well as pre-treatment cytogenetic and molecular characteristics of enrolled patients are summarized in Tables 2 and 3.

Dose escalation and treatment Six patients were treated at dose level 1 due to prolonged myelosuppression in 2 patients requiring dose expansion. However, both patients achieved CR after toxicity â&#x20AC;&#x153;cut-offâ&#x20AC;? and the protocol was amended to allow further time for count recovery as well as G-CSF to hasten count recovery. Three patients were treated on dose levels 2 and 3; 5 patients were treated on dose level 4. No other DLTs were observed.

Toxicities As this treatment approach was intensive, patients experienced universal pancytopenia and toxicities as expected in this poor risk cohort of patients. Treatment overall was well tolerated. Diarrhea, nausea, fatigue, febrile neutropenia and elevated alanine aminotransferase were the most common occurrences for all grade toxicities occurring in 41%, 29%, 29%, 35%, and 35% of patients respectively. With regards to Grade 3 or greater toxicities, febrile neutropenia and catheter-related infections were most common at 35% and 24% of patients and are common complications that occur in this patient population. A summary of all Grade toxicities and Grade 3 or greater non-hematological toxicities possibly attributable to the treatment are listed in Table 4.

Clinical responses The overall response rate was 35% (6/17 patients) as seen in Table 3. All 6 responses by IWG criteria were CR; 3 of these 6 patients had abnormal cytogenetics and all 3 achieved a cytogenetic CR (Patients 5, 7 and 9). The median number of prior therapies for patients with CR was 2 (range 1-3). Response duration assessment is compromised due to 4 patients subsequently receiving allogeneic transplantation (Patients 2, 5, 7, and 9). However, all patients with CR on study except one (Patient 5) eventually relapsed and succumbed to complications of their underlying disease including patients who underwent allogeneic transplantation.

Count recovery for the 6 patients who achieved CR was prolonged with average absolute neutrophil count (ANC) recovery of 45 days (range Day 39-55) and average platelet recovery of 52 days (range Day 34-67). Four of the 6 patients received G-CSF to aid in count recovery with 3 patients receiving this therapy on dose level 1 and 1 patient on dose level 2. No patients had a serious adverse event (SAE) felt related to prolonged count recovery.

None of the patients who were considered treatment failures received any G-CSF support.

Of the 4 patients known to have KMT2A PTD mutations, 2 patients responded achieving cytogenetic CRs. It is interesting to note one of the KMT2A PTD responders also had a TP53 mutation with a high variant allele frequency (VAF), but this response may not be due to the regimen examined, considering recent findings with a 10-day decitabine schedule in TP53 mutated AML patients.13 It is also of interest that 2 KMT2A PTD patients (1 responder and 1 non-responder) were associated with favorable karyotypes as this has not been commonly reported. It was difficult to make any other definitive conclusions about responders and non-responders with other concurrent mutations with such a small number of patients.

Table 3. Patient molecular mutations and responses.

Responders Patient 2 Patient 5 Patient 6 Patient 7 Patient 9 Patient 17 Non-Responders Patient 1 Patient 3 Patient 4 Patient 8 Patient 10 Patient 11 Patient 12 Patient 13 Patient 14 Patient 15 Patient 16

KMT2A PTD +

+

+ +

Other gene mutations (VAF) FLT3-TKD, PTPN11(0.3), U2AF1(0.23) None DDX41(0.5), ASXL1(0.19) None TP53(0.53) FLT3-ITD (heterozygous), NRAS(0.36)

Best response CR CRc CR CRc CRc CR

NPM1(.54), FLT3-ITD (hemizygous), FLT3-TKD, TET2(0.47) None NPM1(0.35), FLT3-ITD (heterozygous) PTPN11(0.52), RUNX1(0.41), NRAS(0.2) PTPN11(0.14), SF1(0.13) NPM1(0.42), FLT3-ITD (hemizygous), WT1(0.64), SF3A1(0.49) IDH2(0.21), KRAS(0.19) FLT3-ITD (heterozygous) RUNX1(0.39) None TP53(0.5)

CR: morphologic Complete Remission; CRc: morphologic and cytogenetic Complete Remission; VAF: variant allele frequency.

haematologica | 2018; 103(6)

985

A.S. Mims et al.

Discussion

Table 4. Non-hematological toxicities possibly attributable.

Toxicity Epigenetic deregulation is felt to contribute to the underlying pathobiology of AML and both aberrant DNA methylation and histone acetylation have been explored as potential therapeutic targets in this disease.16 Monotherapy with hypomethylating agents, azacitidine and decitabine, that target DNMTs have shown some success in regards to clinical activity and are currently used in the treatment of AML, though neither are considered curative therapies in this patient population.17 These agents have become the standard of care for elderly patients who are unfit to undergo intensive induction chemotherapy.

With regards to HDAC inhibition, though preclinical findings in AML have been exciting, these results have yet to translate into significant clinical responses in AML both as monotherapy and in combination with hypomethylating agents or with cytotoxic chemotherapy.18-21 However, priming with both hypomethylating agents and HDAC inhibitors prior to salvage chemotherapy has shown some promise.22 A dose-finding study of decitabine and valproic acid in 25 patients in both relapsed/refractory and untreated, unfit AML populations has been previously performed.

Due to the major DLT of the study of encephalopathy in older patients attributed to valproic acid, this current study of combination therapy with HDAC inhibitors was limited to those <60 years old. It was also felt a different HDAC inhibitor, such as vorinostat, might be better tolerated.23 As mentioned, vorinostat has also been studied in combination with other agents in AML, including hypomethylating agents and cytotoxic chemotherapies. Kirschbaum, et al. randomized patients with untreated (n=31) and relapsed/refractory AML (n=29) along with myelodysplastic syndrome (MDS) (n=31) to concurrent decitabine plus vorinostat versus sequential decitabine followed by vorinostat. Both schedules were felt to be well-tolerated but more objective responses (CRs plus PRs) were seen in the concurrent schedule (46% vs. 14% in untreated AML, 15% vs. 0% in relapsed/refractory AML, and 60% vs. 0% in MDS).24 Gojo et al. assessed the ability to add vorinostat prior to etoposide and cytarabine in relapsed/refractory AML or acute lymphoblastic leukemia, newly diagnosed secondary AML, or chronic myeloid leukemia in accelerated or blastic phase failing or intolerant of tyrosine kinase inhibitor therapy. The MTD was found to be vorinostat 200mg orally twice a day (Days 1-7) followed by cytarabine and etoposide on Days 11-14 with DLTs of hyperbilirubinemia/septic death and anorexia/fatigue at higher dosing. In the 21 patients treated, there were 7 CRs (n=5) or CRiâ&#x20AC;&#x2122;s (n=2) all in the AML population (2 newly diagnosed high-risk and 5 relapsed/refractory patients).25 A phase II trial combined vorinostat 500mg orally 3 times a day (Days 1-3) followed by idarubicin and cytarabine induction chemotherapy in younger untreated AML and higher-risk MDS patients and saw no excess in vorinostatrelated toxicities with an overall response rate of 85% including 76% CR and 9% CRi rates.20 A Phase III randomized intergroup study assessed the benefit of vorinostat in addition to high dosed cytarabine-based induction chemotherapy vs. induction chemotherapy alone and did not show any differences in CR, event-free survival, or overall survival with the addition of vorinostat. Toxicity was considered similar between groups.26 However, the 986

Diarrhea Nausea Fatigue Catheter-related infection Lung infection Hypoxia Febrile neutropenia Blood bilirubin increased Peripheral sensory neuropathy Dyspnea Lymph node pain Abdominal pain Colitis Dry mouth Mucositis oral Proctitis Rectal pain Vomiting Sinusitis Prolonged PTT Increased ALT Increased AST Creatinine increased Anorexia Urinary incontinence Cough Weight loss Headache Proteinuria

All grades number of events

Grade 3 or greater number of events

7 5 5 4 2 1 6 1 1 2 1 2 1 1 1 1 1 1 1 1 6 3 1 4 1 2 1 1 1

2

4 2 1 6

1

1 1 1

ALT: alanine aminotransferase; AST: aspartate aminotransferase; PTT: partial thromboplastin time.

effect of epigenetic priming, particular for subsets of disease that may be sensitive to the approach, remains an area of interest for clinical development.

In this phase I study, we evaluated the combination of decitabine and vorinostat priming prior to high-dose cytarabine in a cohort of younger AML patients with relapsed or refractory AML. We developed a regimen for subsequent phase 2 testing in the select population that may be sensitive to the approach (KMT2A PTD). The treatment was generally well-tolerated with exception of the initial prolonged myelosuppression identified in dose level 1 and there were otherwise no DLTs. The most common toxicities seen including febrile neutropenia, nausea, and diarrhea are all common side effects that can be seen with high-dose cytarabine alone. Neurotoxicity that was seen previously with valproic acid in combination with decitabine was not seen in this patient population.23 This combination therapy resulted in CR in 6 patients with an overall response rate of 35%. Although the small number of patients limits the interpretation of these findings, prior haematologica | 2018; 103(6)

A novel regimen for relapsed/refractory adult acute myeloid leukemia high dose cytarabine exposure does not appear to preclude a response to this combination therapy. With regards to patients with KMT2A PTD mutations, as noted, 2 patients were able to obtain response to this combination therapy while the other 2 were refractory to this treatment with no definitive features to explain difference in outcomes. In conclusion, we successfully determined the recommended phase 2 dose for this novel treatment regimen. The regimen had modest toxicities beyond uncomplicated (though prolonged) myelosuppression, and we propose that the study provides a framework for larger

References 1. Dรถhner H, Etsey E, Grimwade D, et al.

Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel. Blood. 2017; 129(4):424-447.

2. Stone RM, Mandrekar SJ, Sanford BL, et al.

Midostaurin plus chemotherapy for acute myeloid leukemia with a FLT3 mutation. N Engl J Med. 2017;377(5):454-464.

3. Stein EM, DiNardo CD, Pollyea DA, et al.

Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood.

2017;130(6):722-731.

4. Caligiuri MA, Strout MP, Lawrence D, et al.

Rearrangement of ALL1 (MLL) in acute myeloid leukemia with normal cytogenetics. Cancer Res. 1998;58(1):55-59.

5. Dรถhner K, Tobis K, Ulrich R, et al.

Prognostic significance of partial tandem duplications of the MLL gene in adult patients 16 to 60 years old with acute myeloid leukemia and normal cytogenetics: a study of the Acute Myeloid Leukemia Study Group Ulm. J Clin Oncol. 2002; 20(15):3254-3261.

6. Schnittger S, Kinkelin U, Schoch C, et al.

Screening for MLL tandem duplication in 387 unselected patients with AML identify a prognostically unfavorable subset of AML. Leukemia. 2000;14(5):796-804.

7. Dorrance AM, Liu S, Chong A, et al. The Mll partial tandem duplication: differential, tissue-specific activity in the presence or absence of the wild-type allele. Blood.

2008;112(6): 2508-2511.

8. Dorrance AM, Liu S, Yuan W, et al. Mll partial tandem duplication induces aberrant Hox expression in vivo via specific epigenetic alterations. J Clin Invest. 2006; 116(10):2707-2716.

9. Whitman SP, Liu S, Vukosavljevic T, et al.

The MLL partial tandem duplication: evidence for recessive gain-of-function in acute myeloid leukemia identifies a novel patient subgroup for molecular-targeted therapy. Blood. 2005;106(1):345-352.

haematologica | 2018; 103(6)

efficacy studies for AML patients with the uncommon but biologically distinct molecular feature of KMT2A PTD. Acknowledgments The authors thank all of the patients who participated in this trial as well as the dedicated teams who cared for them in the James Cancer Hospital inpatient/outpatient Leukemia Units.

Funding This work was supported by: U01 CA76576-05, P30 CA016058 and SPORE P50-CA140158.

10. Whitman SP, Hackanson B, Liyanarachchi S, et al. DNA hypermethylation and epigenetic silencing of the tumor suppressor gene, SLC5A8, in acute myeloid leukemia with the MLL partial tandem duplication.

Blood. 2008;112(5):2013-2016.

11. Bernot KM, Siebenaler RF, Whitman SP, et al. Toward personalized therapy in AML: in vivo benefit of targeting aberrant epigenetics in MLL-PTD-associated AML. Leukemia. 2013;27(12):2379-2382.

12. Cheson BD, Bennett JM, Kopecky KJ, et al.

Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol.

2003;21(24):4642-4649.

13. Eisfeld AK, Mr zek K, Kohlschmidt J, et al.

The mutational oncoprint of recurrent cytogenetic abnormalities in adult patients with de novo acute myeloid leukemia.

Leukemia. 2017;31(10):2211-2218.

14. Caligiuri MA, Strout MP, Schichman SA, et al. Partial tandem duplication of ALL1 as a recurrent molecular defect in acute myeloid leukemia with Trisomy 11. Cancer Res. 1996;56(6):1418-1425.

15. Thiede C, Steudel C, Mohr B, et al.

Analysis of FLT3-activating mutations in 979 patients with acute myelogenous leukemia: association with FAB subtypes and identification of subgroups with poor prognosis. Blood. 2002;99(12):4326-4335.

16. Welch JS, Petti AA, Miller CA, et al. TP53 and decitabine in acute myeloid leukemia and myelodysplastic syndromes. N Engl J Med. 2016;375(21):2023-2036.

17. Marks P, Rifkind RA, Richon VM, et al.

Histone deacetylases and cancer: causes and therapies. Nat Rev Cancer. 2001; 1(3):194-202.

18. Blum W, Garzon R, Klisovic RB, et al.

Clinical response and miR-29b predictive significance in older AML patients treated with a 10-day schedule of decitabine. Proc Natl Acad Sci USA. 2010;107(16):7473-7478.

19. Kirschbaum M, Gojo I, Goldberg SL, et al.

20.

21.

22.

23.

24.

25.

26.

A phase 1 clinical trial of vorinostat in combination with decitabine in patients with acute myeloid leukaemia or myelodysplastic syndrome. Br J Haematol. 2014; 167(2):185-193.

Garcia-Manero G, Yang H, Bueso-Ramos C, et al. Phase 1 study of the histone deacetylase inhibitor vorinostat (suberoylanilide hydroxamic acid [SAHA]) in patients with advanced leukemias and myelodysplastic syndromes. Blood. 2008; 111(3):1060-1066.

Garcia-Manero G, Tambaro FP, Bekele NB, et al. Phase II trial of vorinostat with idarubicin and cytarabine for patients with newly diagnosed acute myelogenous leukemia or myelodysplastic syndrome. J Clin Oncol. 2012;30(18):2204-2210.

Issa, JP, Garcia-Manero G, Huang X, et al.

Results of phase 2 randomized study of low-dose decitabine with or without valproic acid in patients with myelodysplastic syndrome and acute myelogenous leukemia. Cancer. 2015;121(4):556-561.

Blum W, Klisovic RB, Hackanson B, et al.

Phase I study of decitabine alone or in combination with valproic acid in acute myeloid leukemia. J Clin Oncol.

2007;25(25):3884-3891.

Scandura JM, Roboz GJ, Moh M, et al.

Phase I study of epigenetic priming with decitabine prior to standard induction chemotherapy for patients with AML. Blood. 2011;118(6):1472-1480.

Gojo I, Tan M, Fang HB, et al. Translational phase I trial of vorinostat (suberoylanilide hydroxamic acid) combined with cytarabine and etoposide in patients with relapsed, refractory, or high-risk acute myeloid leukemia. Clin Cancer Res. 2013;19(7):1838-1851.

Garcia-Manero G, Othus M, Pagel JM, et al.

SWOG S1203: A randomized phase III study of standard cytarabine plus daunorubicin (7+3) therapy versus idarubicin with high dose cytarabine (IA) with or without vorinostat (IA+V) in younger patients with previously untreated acute myeloid leukemia (AML). Blood. 2016;128(22):901.

987

ARTICLE

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):988-998

Dexamethasone in hyperleukocytic acute myeloid leukemia

Sarah Bertoli,1,2,3* Muriel Picard,4* Emilie Bérard,5,6* Emmanuel Griessinger,7 Clément Larrue,3 Pierre Luc Mouchel,1,3 François Vergez,2,3,8 Suzanne Tavitian,1 Edwige Yon,5 Jean Ruiz,4 Eric Delabesse,2,3,8 Isabelle Luquet,8 Laetitia Karine Linares,9,10,11 Estelle Saland,3 Martin Carroll,12 Gwenn Danet-Desnoyers,12 Audrey Sarry,1 Françoise Huguet,1 Jean Emmanuel Sarry3 and Christian Récher1,2,3

Service d'Hématologie, Centre Hospitalier Universitaire de Toulouse, Institut Universitaire du Cancer de Toulouse Oncopole, France; 2Université Toulouse III Paul Sabatier, France; 3 Cancer Research Center of Toulouse, UMR1037-INSERM, ERL5294 CNRS, France; 4 Service de Réanimation Polyvalente, Centre Hospitalier Universitaire de Toulouse, Institut Universitaire du Cancer de Toulouse Oncopole, France; 5Service d'Epidémiologie, Centre Hospitalier Universitaire de Toulouse, France; 6UMR 1027, INSERM-Université de Toulouse III, France; 7Université Côte d'Azur, INSERM U1065, Centre Méditerranéen de Médecine Moléculaire, Nice, France; 8Laboratoire d’Hématologie, Centre Hospitalier Universitaire de Toulouse, Institut Universitaire du Cancer de Toulouse Oncopole, France; 9IRCM, Institut de Recherche en Cancérologie de Montpellier-INSERM, U1194, France; 10Université Montpellier, F-34298, France; 11Institut Régional du Cancer Montpellier, France and 12Stem Cell and Xenograft Core, University of Pennsylvania, Perelman School of Medicine, Philadelphia, PA, USA 1

*SB, MP and EB contributed equally to this work.

ABSTRACT

Correspondence: recher.christian@iuct-oncopole.fr

Received: November 8, 2017.

Accepted: March 2, 2018.

Pre-published: March 8, 2018.

doi:10.3324/haematol.2017.184267 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/988 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

988

P

atients with acute myeloid leukemia and a high white blood cell count are at increased risk of early death and relapse. Because mediators of inflammation contribute to leukostasis and chemoresistance, dexamethasone added to chemotherapy could improve outcomes.

This retrospective study evaluated the impact of adding or not adding dexamethasone to chemotherapy in a cohort of 160 patients with at least 50×109 white blood cells. In silico studies, primary samples, leukemic cell lines, and xenograft mouse models were used to explore the antileukemic activity of dexamethasone. There was no difference with respect to induction death rate, response, and infections between the 60 patients in the dexamethasone group and the 100 patients in the no dexamethasone group. Multivariate analysis showed that dexamethasone was significantly associated with improved relapse incidence (adjusted sub-HR: 0.30; 95% CI: 0.14-0.62; P=0.001), disease-free survival (adjusted HR: 0.50; 95% CI: 0.29-0.84; P=0.010), event-free survival (adjusted HR: 0.35; 95% CI: 0.21-0.58; P<0.001), and overall survival (adjusted HR: 0.41; 95% CI: 0.22-0.79; P=0.007). In a co-culture system, dexamethasone reduced the frequency of leukemic long-term culture initiating cells by 38% and enhanced the cytotoxicity of doxorubicin and cytarabine. In a patient-derived xenograft model treated with cytarabine, chemoresistant cells were enriched in genes of the inflammatory response modulated by dexamethasone. Dexamethasone also demonstrated antileukemic activity in NPM1-mutated samples. Dexamethasone may improve the outcome of acute myeloid leukemia patients receiving intensive chemotherapy. This effect could be due to the modulation of inflammatory chemoresistance pathways and to a specific activity in acute myeloid leukemia with NPM1 mutation.

haematologica | 2018; 103(6)

Dexamethasone in AML

Introduction Acute myeloid leukemias (AML) are myeloid malignancies induced by the oncogenic transformation of hematopoietic progenitors in the bone marrow leading to the destruction of blood tissue and, therefore, to profound pancytopenia, severe bleeding, and infection.1 Approximately 20% of patients present at diagnosis with high white blood cell (WBC) counts (i.e. >50x109/L).2 In this high-risk situation, the probability of severe complications is increased because of leukemic organ infiltration, severe hemorrhage, or metabolic disorders, including tumor lysis syndrome, renal failure, and disseminated intravascular coagulopathy, which is further worsened by the induction of antileukemic treatment.

Hyperleukocytosis is also associated with leukostasis syndrome within the lung or brain, which can potentially lead to acute respiratory distress syndrome or stroke. Thus, patients with a high WBC count share an increased risk of death during the initial phase of the disease.

Hyperleukocytosis is also independently associated with shorter relapse-free survival in patients treated by intensive chemotherapy, indicating a potential link with chemoresistance.2 Dexamethasone is an anti-inflammatory drug widely used in acute lymphoblastic leukemia and other lymphoid malignancies.3 Much less frequently used in myeloid disorders, this drug is often offered to prevent or treat a severe inflammatory status, so-called differentiation syndrome in patients with acute promyelocytic leukemia treated with all trans-retinoic acid and/or arsenic trioxide.4,5 Mediators of inflammation induced by leukemic blasts and endothelial cells contribute to the pathogenesis of leukostasis.6 Studies on the molecular mechanisms of leukostasis and leukemic cell invasion have shown that leukemic blasts use integrins and selectins to attach to cytokine-activated endothelium and directly activate endothelial cells by secreting inflammatory cytokines, such as tumor necrosis factor-α, interleukin-1β, and interleukin-6, which induce the conditions necessary for their adhesion to vascular endothelium, migration to tissues, proliferation, and chemoresistance.6,7 The central role of the inflammatory response prompted us to assess the impact of dexamethasone in this setting because this drug exerts a potent inhibitory effect on cytokine production.8 We hypothesized that introducing a short course of dexamethasone into routine practice during the early phase of induction chemotherapy would improve the outcome of hyperleukocytic AML patients.

with the Declaration of Helsinki, the study was reviewed and approved by the research ethics committee at Toulouse University Hospital.

Treatment Study patients received induction chemotherapy that included daunorubicin at a daily dose of 60–90 mg/m2 of body surface area daily for 3 days, or idarubicin at a daily dose of 8-9 mg/m2 daily for 5 days, together with a continuous intravenous infusion of cytarabine at a daily dose of 100–200 mg/m2 daily for 7 days.10 No patient received an FLT3 inhibitor in combination with chemotherapy during first-line induction. Lomustine was added in patients aged over 60 years.11 Hydroxyurea could be started promptly at diagnosis for leukocytic reduction. Leukapheresis was not performed. Starting in January 2010, dexamethasone (10 mg b.i.d. given for 3 days) was systematically added to induction chemotherapy in all patients who had a WBC count of at least 100 x 109/L or in patients with a WBC count over 50 x 109/L and clinical symptoms of leukostasis. This dexamethasone schema was used based on our previous experience in patients with acute promyelocytic leukemia.4 Supportive care, which included prevention of invasive fungal infections with voriconazole from 2004 to 2008 then posaconazole, treatment of febrile neutropenia and disseminated intravascular coagulopathy, and blood-product transfusions were given according to standard guidelines that did not change over the study period.12,13 Patients who achieved complete remission proceeded to subsequent treatment steps. Postremission therapy was based on relapse risk and whether an HLAidentical donor had been identified or not. Patients at low risk of relapse (i.e. patients with a core-binding factor AML, NPM1, or CEBPA mutation without FLT3-ITD) received only chemotherapy as post-remission therapy. All other patients with an HLAmatched donor underwent allogeneic stem-cell transplantation, whereas those without such a donor received chemotherapy.

Response criteria and end points Complete response was defined according to standard criteria.14 Relapse was defined as leukemia recurrence after a first complete remission. Disease-free survival was measured from the time of complete remission evaluation to the date of relapse or death, whatever the cause. Event-free survival was measured from the date of diagnosis to the date of failure to enter complete remission, relapse, or death, whichever came first. Overall survival was defined as the time interval from diagnosis until death, whatever the cause.

The statistical analyses and exploratory analyses are described in the Online Supplementary Appendix.

Results Methods Patients Between January 2004 and December 2015, 802 patients aged between 18 and 75 years with cytologically confirmed AML were consecutively treated with intensive chemotherapy at Toulouse University Hospital. Patients with acute promyelocytic leukemia were not considered. Patients were classified into three prognostic categories based on cytogenetics.9 FLT3-ITD and NPM1 mutations were assessed in patients with intermediate-risk cytogenetics.

Data were collected from the patients' files and certified by the Data Management Committee of the AML database of Toulouse University Hospital registered at the Commission Nationale de l’Informatique et des Libertés (CNIL, #1778920).10 In accordance haematologica | 2018; 103(6)

Study population The flowchart of the 160 patients with a WBC count of at least 50 x 109/L included in this retrospective study is shown in Figure 1 and the patients’ characteristics are summarized in Table 1. The median follow-up period of patients still alive at the date of last contact was 52.2 months [inter-quartile range (IQR); 23.7-72.9 months]; the median periods in the dexamethasone and the no dexamethasone groups were 44.1 months (IQR; 19.6-55.8) and 65.7 months (IQR 52.0-79.7), respectively. The median age of the patients was 60.1 years (IQR: 49.2-67.3); 50% of patients were aged ≥60 years. Compared to the 100 patients in the no dexamethasone group, the 60 patients in the dexamethasone group were more likely to have a poor 989

S. Bertoli et al.

performance status, features of leukostasis syndrome, and higher WBC count. Hydroxyurea treatment was given to 49 patients in the dexamethasone group and to 59 patients in the no dexamethasone group. Allogeneic stem cell transplantation was given to 19 patients in the dexamethasone group and to 25 patients in the no dexamethasone group. Of note, overall survival of patients (aged 18– 75 years) with a WBC count <50 x 109/L who were treated with intensive chemotherapy between 2004 and 2009 (336 patients with 232 deaths; median overall survival, 21.5 months; IQR: 7.6–158.8) did not differ significantly from that of patients treated between 2010 and 2015 (295 patients with 164 deaths, median overall survival, 25.8 months; IQR: 9.1–not achieved) (hazard ratio for 2010– 2015 vs. 2004–2009=0.95; 95% CI: 0.77–1.16; P=0.595).

Impact of dexamethasone during the induction phase Fifty patients (83.3%) from the dexamethasone group and 74 (74%) from the no dexamethasone group achieved a complete response (P=0.171) (Table 2). At day 60 of induction chemotherapy, 7 patients (11.7%) in the dexamethasone group had died compared to 20 patients (20%) in the no dexamethasone group (P=0.173). There were no significant differences between the two groups in terms of

fungal (P=0.710) or bacterial (P=0.192) infections.

However, grade 3-4 bleeding events were more frequent in the dexamethasone group compared to the no dexamethasone group [13 (21.7%) vs. 6 (6.2%); P=0.038] as were admissions to the intensive-care unit by day 90 [29 (48.3%) vs. 17 (17.0%); P<0.0001].

Impact of dexamethasone on relapse and survival In the univariate analyses, the use of dexamethasone was associated with an improved outcome, with the improvement reaching statistical significance for relapse incidence (sub-HR: 0.43; 95% CI: 0.25-0.74; P=0.003), disease-free survival (HR: 0.48; 95% CI: 0.29-0.80; P=0.005), event-free survival (HR: 0.52; 95% CI: 0.34-0.79; P=0.002), and overall survival (HR: 0.55; 95% CI: 0.35-0.85; P=0.005) (Figure 2). In a Fine and Gray competing risks model, the use of dexamethasone was associated with a significantly lower risk of relapse (adjusted sub-HR: 0.30; 95% CI: 0.14-0.62; P=0.001) (Online Supplementary Table S1). In multivariate analyses, the use of dexamethasone was associated with significantly better outcomes when considering disease-free survival (adjusted HR: 0.50; 95% CI: 0.29-0.84; P=0.010) (Online Supplementary Table S2), event-free survival (adjusted HR: 0.35; 95% CI: 0.21-0.58;

Figure 1. Study flowchart. y: years; WBC: white blood cell; OS: overall survival; CR: complete response.

990

haematologica | 2018; 103(6)

Dexamethasone in AML

Table 1. Characteristics of the 160 acute myeloid leukemia patients with hyperleukocytosis.

Characteristics Sex – n. (%) Male Female Age – years Median (IQR) <60 years – n. (%) ≥60 years – n. (%) ECOG performance status – n. (%) 0-1 2-4 Extramedullary involvement – n. (%) No Yes AML status – n. (%) De novo Secondary Leukostasis – n. (%) No Central nervous system Lung Central nervous system and lung Infection at diagnosis – n. (%) No Yes White blood cell count – x109/L Median (IQR) <100 – n. (%) ≥100 – n. (%) Platelet count – x109/L Median (IQR) <50 – n. (%) ≥50 – n. (%) CD56 – n. (%) ≤20% >20% Cytogenetic risk – n. (%) Favorable Intermediate Adverse ELN classification – n. (%) Favorable Intermediate-1 Intermediate-2 Adverse Unknown$ NPM1 mutations¶ No Yes FLT3-ITD mutations¶ No Yes Creatinine - mg/dL Median (IQR) <1.36 – n. (%) >1.36– n. (%) Bilirubin - mg/dL Median (IQR) ≤1.47 – n. (%) >1.47 – n. (%)

No dexamethasone n=100 (62.5%)

Dexamethasone n=60 (37.5%)

P

All patients n=160 (100%)

56 (56.0) 44 (44.0)

27 (45.0) 33 (55.0)

0.178

83 (51.9) 77 (48.1)

60.7 (48.3-68.5) 47 (47.0) 53 (53.0)

58.8 (50.5-66.5) 33 (55.0) 27 (45.0)

0.244

60.1 (49.2-67.3) 80 (50.0) 80 (50.0)

60 (77.9) 17 (22.1)

31 (56.4) 24 (43.6)

0.008

91 (68.9) 41 (31.1)

38 (46.3) 44 (53.7)

25(42.4) 34 (57.6)

0.640

63 (44.7) 78 (55.3)

81 (81.0) 19 (19.0)

55 (91.8) 5 (8.3)

0.067

136 (85.0) 24 (15.0)

84 (84.0) 2 (2.0) 10 (10.0) 4 (4.0)

34 (56.7) 9 (15.0) 12 (20.0) 5 (8.3)

<0.001

118 (73.8) 11 (6.9) 22 (13.8) 9 (5.6)

78 (80.4) 19 (19.6)

46 (76.7) 14 (23.3)

0.576

124 (70.9) 33 (21.0)

86.3 (66.1-115.5) 62 (62.0) 38 (38.0)

119 (90.7-181.4) 21 (35.0) 39 (65.0)

<0.001

97.6 (71.0-142.6) 83 (51.9) 77 (48.1)

54.5 (35.0-85.5) 44 (44.0) 56 (56.0)

49.5 (24.5-72.0) 30 (50.0) 30 (50.0)

0.062

71 (78.9) 19 (21.1)

44 (75.9) 14 (24.1)

0.666

115 (77.7) 33 (22.3)

8 (8.0) 79 (79.0) 13 (13.0)

7 (11.7) 48 (80.0) 5 (8.3)

0.530

15 (9.4) 127 (79.4) 18 (11.3)

23 (23.0) 32 (32.0) 21 (21.0) 13 (13.0) 11 (11.0)

17 (28.3) 25 (41.7) 12 (20.0) 5 (8.3) 1 (1.7)

0.155

40 (25.0) 57 (35.6) 33 (20.6) 18 (11.3) 12 (7.5)

27 (39.7) 41 (60.3)

17 (35.4) 31 (64.6)

0.639

44 (37.9) 72 (62.1)

33 (48.5) 35 (51.5)

20 (41.7) 28 (58.3)

0.465

53 (45.7) 63 (54.3)

1.05 (0.89-1.25) 82 (82.0) 18 (18.0)

0.91 (0.71-1.24) 50 (83.3) 10 (16.7)

0.029

1.01 (0.83-1.24) 132 (82.5) 28 (17.5)

0.54 (0.38-0.79) 95 (95.0) 5 (5.0)

0.53 (0.35-0.79) 55 (94.8) 3 (5.2)

0.327

<0.001

0.461

0.830 0.649 1.000

52.0 (31.0-77.0) 74 (46.3) 86 (53.8)

0.53 (0.36-0.79) 150 (94.9) 8 (5.1) continued on the next page

haematologica | 2018; 103(6)

991

S. Bertoli et al. continued from the previous page

Characteristics Albumin - g/dL Median (IQR) <3.5 – n. (%) ≥3.50 – n. (%) Ferritin - ng/mL Median (IQR) ≤1000 – n. (%) >1000 – n. (%) Lactate dehydrogenase – IU/L Median (IQR) ≤1550 – n. (%) >1550 – n. (%) Fibrinogen – g/L Median (IQR) ≤1.5 – n. (%) >1.5 – n. (%) Study period – n. (%) 2004-2009 2010-2015 Hydroxyurea – n. (%) No Yes Chemotherapy – n. (%) Daunorubicin-cytarabine Idarubicin-cytarabine Idarubicin-cytarabine-lomustine Time sequential induction Other Allo-SCT – n. (%) No Yes

No dexamethasone n=100 (62.5%)

Dexamethasone n=60 (37.5%)

P

All patients n=160 (100%)

3.50 (3.30-3.90) 41 (41.4) 58 (58.6)

3.50 (3.20-4.00) 29 (49.2) 30 (50.8)

0.500

3.50 (3.20-4.00) 70 (44.3) 88 (55.7)

928.5 (570.0-1342.0) 24 (52.2) 22 (47.8)

1103.0 (606.0-2249.0) 25 (46.3) 29 (53.7)

0.287

1618.0 (913.0-2506.0) 48 (48.0) 52 (52.0)

1498.5 (790.0-2369.0) 32 (53.3) 28 (46.7)

0.455

4.0 (2.8-4.8) 8 (8.0) 92 (92.0)

3.7 (2.6-4.7) 7 (11.7) 53 (88.3)

0.197

67 (67.0) 33 (33.0)*

0 (0.0) 60 (100.0)

<0.001 93 (58.1)

67 (41.9)

41 (41.0) 59 (59.0)

11 (18.3) 49 (81.7)

0.003

52 (32.5) 108 (67.5)

32 (32.0) 15 (15.0) 45 (45.0) 5 (5.0) 3 (3.0)

16 (26.7) 13 (21.7) 21 (35.0) 5 (8.3) 5 (8.3)

0.281

48 (30.0) 28 (17.5) 66 (41.3) 10 (6.3) 8 (5.0)

75 (75.0) 25 (25.0)

41 (68.3) 19 (31.7)

0.361

116 (72.5) 44 (27.5)

0.344

0.558

0.514

0.441

1064.0 (599.5-1904.0) 49 (49.0) 51 (51.0) 1551.5 (840.5-2476.0) 80 (50.0) 80 (50.0) 3.9 (2.7-4.8) 15 (9.4) 145 (90.6)

IQR: interquartile range; ECOG: Eastern Cooperative Oncology Group; ELN: European LeukemiaNet; Allo-SCT: allogeneic stem cell transplantation (16 from sibling and 28 from HLA 9/10 or 10/10 matched unrelated donors). $ELN is unknown if FLT3-ITD or NPM1 mutation was missing for normal karyotypes or if karyotype is missing. ¶NPM1 and FLT3ITD mutations in patients with intermediate-risk cytogenetics. * Among the 33 patients who did not receive dexamethasone, only 3 patients had symptoms of pulmonary leukostasis and more than 100 x109/L white blood cell count (WBC), whereas 3 additional patients had more than 100 x109/L WBC without leukostasis.

P<0.001) (Online Supplementary Table S3), and overall survival (adjusted HR: 0.41; 95% CI: 0.22-0.79; P=0.007) (Table 3). Of note, when put in the multivariate model, FLT3-ITD mutations had no significant impact. Among patients who had undergone allogeneic stem cell transplantation in first complete response, the outcome of the dexamethasone group was still better than that of the no dexamethasone group (Online Supplementary Figure S1).

compared to untreated primary AML cells (Figure 3A).

Interestingly, primary AML cells treated with dexamethasone presented a higher expression profile of the CD38 marker after one week, and a higher percentage of myeloid and lymphoid lineage positive cells as well as monocytic CD11b/CD14-positive cells within the longterm culture weeks suggesting that dexamethasone may have a differentiation effect on AML (Figure 3B and C).

Impact of dexamethasone on leukemia-initiating cells in a co-culture system

Impact of dexamethasone on chemoresistance

The use of dexamethasone was unexpectedly associated with a lower relapse rate suggesting that this drug could display potent antileukemic activity against AML cells at the origin of relapse and/or by enhancing the cytotoxicity of chemotherapy. Leukemic long-term culture initiating cells have been shown to be a reliable functional readout for monitoring the activity of leukemia-initiating cells, an AML subpopulation of cells thought to be at the origin of relapse.15,16 Using an optimized niche-like co-culture system capable of maintaining leukemia-initiating cells ex vivo, dexamethasone reduced the frequency of leukemic long-term culture initiating cells by 38±14% as 992

We next sought to determine whether dexamethasone could improve the cytotoxicity of genotoxic drugs used in AML. In liquid culture, short-term dexamethasone treatment with or without cytarabine or doxorubicin did not show synergy or an additive effect in a panel of genetically diverse AML cell lines (Online Supplementary Figure S2 and Online Supplementary Table S4). However, in a co-culture system, one week of exposure to dexamethasone significantly enhanced cytarabine activity in most AML cell lines (Figure 3D). Recently, it has been shown that cytarabine resistance of AML cells is associated with increased sensitivity to glucocorticoids.17,18 We thus wondered whether AML cells resistant to cytarabine display specific tranhaematologica | 2018; 103(6)

Dexamethasone in AML

Table 2. Patients’ outcomes during induction chemotherapy according to study group.

Admission in intensive care unit*– n (%) No Yes Bacterial infections - n (%) No Yes Fungal infections - n (%) No Yes Bleeding events (grade 3-4) - n (%) No Yes Day-60 deaths - n (%) No Yes Induction failure - n (%) No Yes Complete response - n (%) No Yes

No dexamethasone n=100 (62.5%)

Dexamethasone n=60 (37.5%)

P

All patients n=160

83 (83.0) 17 (17.0)

31 (51.7) 29 (48.3)

<0.0001

114 (71.3) 46 (28.8)

71 (73.2) 26 (26.8)

38 (63.3) 22 (36.7)

0.192

109 (69.4) 48 (30.6)

86 (88.7) 11 (11.3)

52 (86.7) 8 (13.3)

0.710

138 (87.9) 19 (12.1)

91 (93.8) 6 (6.2)

47 (78.3) 13 (21.7)

0.004

138 (87.9) 19 (12.1)

80 (80.0) 20 (20.0)

53 (88.3) 7 (11.7)

0.173

133 (83.1) 27 (16.9)

92 (92.0) 8 (8.0)

57 (95.0) 3 (5.0)

0.538

149 (93.1) 11 (6.9)

26 (26.0) 74 (74.0)

10 (16.7) 50 (83.3)

0.171

36 (22.5) 124 (77.5)

*During the first three months following chemotherapy.

B

A

C

D

Figure 2. Estimates of survival end points and incidence of relapse. (A) Kaplan–Meier curves for overall survival in patients treated with dexamethasone (black line) or not (solid line in red). (B) Show event-free survival, (C) disease-free survival and (D) cumulative incidence of relapse. It is worth nothing that the follow up concerning relapse seems to be equal in the dexamethasone and no dexamethasone groups. This is due to the competing risk analyses. Indeed, in the competing risk analyses, a subject having a non-relapse death was not censored at the date of death and was still virtually considered to be at risk of having a relapse (and was still in follow up).

haematologica | 2018; 103(6)

993

S. Bertoli et al.

scriptomic characteristics in vivo. To test this, we used a patient-derived xenograft model of chemoresistance (Online Supplementary Figure S3A).19 Eight to 18 weeks after transplantation of the AML sample, the mice were given daily intraperitoneal injections of 60 mg/kg cytarabine or vehicle for 5 days. Three days after the last dose of cytarabine or vehicle, viable human AML blasts were collected from the bone marrow, then purified and processed for transcriptomic analysis. The transcriptome of residual human AML cells exhibiting in vivo resistance to cytarabine treatment displayed a strong upregulation of the genes involved in immune and inflammatory responses, including the nuclear factor-κB network (Figure 3E).

Furthermore, this gene signature of chemoresistance displayed a highly significant interaction with the dexamethasone gene signature (Figure 3F and Online Supplementary Table S5). Similarly, interrogation of a publicly available transcriptomic data set established from AML patients in first relapse and a data mining algorithm (Genomatix) revealed that the dexamethasone signature was also enriched within AML cells collected at relapse (Figure 3G and Online Supplementary Table S6).20 Moreover, in two patient-derived xenograft models, treatment of

NSG mice with the dexamethasone-cytarabine combination induced a deeper therapeutic response compared to that achieved with cytarabine alone (Figure 3H). All together, these data strongly suggest that the impact of dexamethasone with intensive chemotherapy observed in the clinic could result from the targeting of chemoresistant AML cells.

Pre-clinical antileukemic activity of dexamethasone in acute myeloid leukemia with NPM1 mutations A recent pre-clinical study demonstrated that AML cells with RUNX1 mutations were sensitive to glucocorticoids while earlier in vitro studies suggested an antileukemic activity in AML with the t(8;21)/RUNX1-RUNX1T1 translocation.21,22 To find out whether other molecular subgroups could benefit from glucocorticoids, we first tested the in vitro activity of dexamethasone against AML cell lines with various genetic backgrounds. As expected, dexamethasone had no significant activity as a single agent in most AML cell lines cultured in suspension (Figure 4A).

Only two out of seven AML cell lines were moderately sensitive to the growth inhibition effect of dexamethasone, including OCI-AML3, an NPM1-mutated cell line.

Table 3. Multivariate analysis for overall survival.

Dexamethasone No Yes AML status De novo Secondary Infection at diagnosis No Yes Albumin- g/dL > 3.5 ≥ 3.5 Lactate dehydrogenase – UI/L ≤1550 >1550 Fibrinogen – g/L ≤ 1.5 > 1.5 Cytogenetic risk Favorable Intermediate Unfavorable Hydroxyurea No Yes Admission to intensive care unit* No Yes Study period 2004-2009 2010-2015 Allogeneic stem cell transplantation No Yes

Numbers

Events

Adjusted HR

95% CI

P

100 60

74 27

1 0.41

0.22-0.79

0.007

136 24

80 21

1 2.44

1.45-4.11

0.001

124 33

74 24

1 1.76

1.06-2.93

0.029

70 88

47 52

1 0.61

0.39-0.94

0.027

80 80

41 60

1 1.76

1.14-2.70

0.010

15 145

12 89

1 0.32

0.17-0.62

0.001

15 127 18

3 87 11

1 4.18 7.07

1.28-13.62 1.83-27.29

0.018 0.005

52 108

37 64

1 0.61

0.39-0.97

0.037

114 46

70 31

1 3.76

2.21-6.38

<0.001

67 93

54 47

1 0.62

0.36-1.07

0.087

116 44

79 22

1 0.51

0.29-0.92

0.026

*During the first three months following chemotherapy. HR: Hazard Ratio; CI: Confidence Interval.

994

haematologica | 2018; 103(6)

Dexamethasone in AML

We thus explored the impact of dexamethasone treatment in the OCI-AML3 xenotransplantation model (Online Supplementary Figure S3B). Dexamethasone treatment resulted in a significant survival advantage compared to vehicle. Moreover, the combination of dexamethasone

A

B

E

plus cytarabine significantly improved mouse survival compared to that following cytarabine treatment alone (Figure 4B). We then tested the in vitro activity of dexamethasone against primary samples from patients with or without an NPM1 mutation. Primary AML samples with

C

F

D

G

H

Figure 3. Impact of dexamethasone on chemoresistance. (A) Leukemia long-term culture initiating cell (L-LTC-IC) frequency in acute myeloid leukemia (AML) samples upon dexamethasone treatment (#38 to 49, Online Supplementary Table S8). (B) Expression of CD34 and CD38 upon dexamethasone treatment in co-culture with AML samples and MS-5 stromal cells. (C) Expression of lineage and CD14/11b markers upon dexamethasone treatment in co-culture with AML samples and MS-5 stromal cells. (D) Seven AML cell lines incubated in a co-culture system with MS-5 stromal cells were treated for 1 week with vehicle, dexamethasone, cytarabine, or dexamethasone plus cytarabine. (E) Gene ontology enrichment analysis of down-regulated and up-regulated genes from RNA expression profiles of viable AML cells following cytarabine versus vehicle-treated AML-patient-derived xenograft (PDX) mice, by 1.5-fold or more. (F and G) Gene-to-small molecule associations that are significantly enriched within residual post-cytarabine AML cells (Figure 3F and Online Supplementary Table S5) or in relapse (compared to pairwise diagnosis, Figure 3G and Online Supplementary Table S6) using a data-mining algorithm (Genomatix) from GSE9763119 or GSE6652520 publicly accessible transcriptomic databases, respectively. These two graphs shown a gene signature ranking assessed by the number of observed versus expected genes significantly modulated in transcriptomes after treatment with diverse small molecules and significantly enriched in AML transcriptomes of residual post-cytarabine AML cells or of relapse. (H) Treatment of PDX models from 2 AML samples collected at diagnosis (black dots: normal karyotype, NPM1 mutation, wild-type for FLT3-ITD, DNMT3A-exon23, CEBPA, IDH1, and IDH2, red dots: normal karyotype, NPM1 mutation, DNMT3A-exon23 mutation, FLT3 wild type) with vehicle, dexamethasone (10 mg/kg/day, 5 days), cytarabine (30 mg/kg/day, 5 days), or dexamethasone plus cytarabine. At day 8, the reduction of the total AML cell burden was assessed by the absolute quantification of the hCD45+hCD33+mCD45.1- cell population in bone marrow and spleen. Mann Whitney test: ****P<0.0001; ***P<0.001; **P<0.01; *P<0.05; ns: not significant.

haematologica | 2018; 103(6)

995

S. Bertoli et al.

NPM1 mutations were more sensitive to dexamethasoneinduced apoptosis than samples without NPM1 mutations (Figure 4C).

Overlap between mutant NPM1 up-regulated target genes and a dexamethasone-associated gene expression signature In line with these results, interrogation of a transcriptomic data set from a series of AML patients with NPM1 mutations and a data mining algorithm revealed that the NPM1 mutation gene signature was highly enriched in

A

C

E

996

genes responsive to several small molecules, including dexamethasone as well as all-trans retinoic acid and dactinomycin, which have previously demonstrated therapeutic activity in this subgroup of AML (Figure 4D and E, and Online Supplementary Table S7).23 To address the genetic heterogeneity of AML, further transcriptomic analyses from two independent data sets revealed more complex molecular interactions between dexamethasone and some AML subgroups, such as those with RUNX1 mutations or the CBF-MYH11 rearrangement (Online Supplementary Figure S4A).23,24 Overall, enrichment in the dexamethasone

B

D

Figure 4. Activity of dexamethasone against acute myeloid leukemia (AML). (A) Anti-leukemic effect of dexamethasone on AML cell lines cultured in minimum essential medium alpha whose cytogenetic and molecular characteristics are shown in Online Supplementary Table S4 (two-way ANOVA test, P<0.001).

OCI-AML3 and KG1a were significantly more sensitive compared to the other AML cell lines (Tukey multiple comparison test, P<0.001). (B) Kaplanâ&#x20AC;&#x201C;Meier curves for survival in a xenotransplantation mouse model of AML using the OCI-AML3 cell line (NPM1 and DNMT3A-R882C mutated) and NSG mice treated with vehicle, dexamethasone (10 mg/kg/day, 5 days), cytarabine (30 mg/kg/day, 5 days), or dexamethasone plus cytarabine. P-values of the log-rank test are vehicle versus cytarabine (P=0.0105), vehicle versus dexamethasone (P=0.0014), vehicle versus dexamethasone plus cytarabine (P<0.0001), cytarabine versus dexamethasone (P=0.1474), cytarabine versus dexamethasone plus cytarabine (P=0.0001), and dexamethasone versus dexamethasone plus cytarabine (P=0.0943). (C) Apoptosis induction by dexamethasone in primary AML samples (#1 to 37, Online Supplementary Table S8) according to NPM1 mutational status (16 samples with NPM1 mutations and 18 samples that were wild-type for NPM1). AML samples were incubated with or without 100 and 300 nM of dexamethasone for 24 hours and then processed for apoptosis studies using annexin-V/propidium iodide staining.

Results are presented as mean percentages of annexin V-positive cells in treated versus untreated cells (P=0.049 and P=0.04 for 100 and 300 nM when comparing NPM1 mutated versus NPM1 wildtype samples, respectively). (D) Gene ontology enrichment analysis using the data set of Verhaak et al. in AML with NPM1 mutations.23 (E) Gene-gene associations between mutant NPM1 up-regulated target genes and small molecule gene signatures and a data-mining algorithm (Genomatix). Gene expression analyses revealed that the NPM1 mutation gene signature was highly enriched in genes responsive to treatment using small molecules, such as all trans-retinoic acid (ATRA), retinoic acid (RA), dexamethasone, and to a lesser extent dactinomycin (Online Supplementary Table S7).

haematologica | 2018; 103(6)

Dexamethasone in AML

gene signature was observed in 45-60% of AML patients (Online Supplementary Figure S4B).

All together, these results demonstrate that dexamethasone has significant activity against NPM1-mutated AML cells, corresponding to ~65% of the patients treated by dexamethasone in the clinical study.

Discussion Three very recent pre-clinical studies have demonstrated that glucocorticoids could be of real interest in AML. Malani et al. and Kurata et al. have shown that the development of cytarabine resistance in AML cells is associated with increased sensitivity to glucocorticoids.17,18 Using a chemogenomic approach, Simon et al. demonstrated that AML samples bearing inactivating RUNX1 mutations are particularly sensitive to glucocorticoids.21 Our study is the first to detect a significant clinical correlation between dexamethasone treatment and outcome in adult patients with AML. Indeed, we showed that the addition of dexamethasone to intensive chemotherapy was associated with significantly better disease-free and overall survival in hyperleukocytic AML patients. We must, however, acknowledge that Turkish investigators previously reported their long-lasting experience on the potential impact of high-dose methylprednisolone in pediatric AML.25 The strong and unexpected impact of dexamethasone in preventing relapses prompted us to undertake in silico, in vitro, and in vivo exploratory analyses. The gene signatures of some molecular subgroups of AML were highly enriched in genes responsive to dexamethasone, including AML with NPM1 mutations, which were particularly sensitive to the antileukemic activity of dexamethasone both in vitro and in vivo. Moreover, using a xenotransplantation model of chemoresistance, we demonstrated that the transcriptome of viable AML cells in xenograft NSG mice following cytarabine exposure is highly enriched in inflammatory response genes as well as in genes responsive to dexamethasone.19 Although inflammation is a hallmark of cancer, its role has been neglected in AML in which other oncogenic pathways, including transcriptional dysregulation, sustained proliferative signaling, epigenetic or metabolic alterations, as well as deregulated splicing have been more deeply assessed. Yet, several aspects of inflammation could be explored to increase our knowledge of AML pathophysiology and to expand therapeutic opportunities or prognostic markers.26 The results of our study led us to speculate that dexamethasone, by affecting specific transcriptomic programs and/or by modulating the early inflammatory response which is associated with chemoresistance, might sensitize AML cells to chemotherapy-induced cell death and thereby limit the risk of leukemic regrowth and relapse. Thus, although there was only a trend for a higher complete response rate in the dexamethasone group, this effect was translated in our study into a reduction of cumulative incidence of relapse, which is a better end point than the complete response rate for assessing the quality of response and impact on chemoresistant disease. Our findings also suggest that dexamethasone, used as a chemosensitizer in combination with intensive chemotherapy, should be assessed in prospective trials regardless of the WBC count.

Dexamethasone has both cytoplasmic and nuclear activities that interfere with signal transducers or tranhaematologica | 2018; 103(6)

scription factors such as PI3-kinase, activating protein-1, and nuclear factor-κB, which are both involved in leukemic stem-cell biology.8,27 Inflammatory cytokines can induce both nuclear factor-κB and activating protein-1 to support leukemic stem-cell survival in a synergistic manner.28 Thus, by suppressing cytokine release and targeting specific intracellular pathways, dexamethasone could make leukemic stem cells more susceptible to chemotherapy-induced cell death. The mechanisms of action underpinning dexamethasone activity in AML are likely to be multiple as leukemic stem cells are subject to different levels of regulation which are either cell autonomous or driven by interactions with the microenvironment.29,30 In most studies that have focused on hyperleukocytosis in AML, the early mortality rate is about 20–30%, which is similar to that in the no dexamethasone group in our study.31 Early mortality has remained very high compared to that of patients without hyperleukocytosis even in recent series, and therapeutic strategies, aimed at reducing leukocytosis through the use of leukapheresis or lowdose chemotherapy, failed to demonstrate any benefit.2,31 Our results show that dexamethasone treatment was associated with a lower rate of early mortality following induction chemotherapy despite a higher rate of admission to the intensive care unit. There was, however, no significant difference compared to patients in the no dexamethasone group, which may reflect the low number of events. The criteria for intensive care unit admission in our center changed in 2015 when AML patients with a WBC count >100x109/L or leukostasis were admitted directly into the unit. However, we analyzed the 20042013 period and found the same difference with more intensive care unit admissions in the dexamethasone group (31% vs. 15%; P=0.028). Thus, it is likely that this difference is related to the selection criteria for giving dexamethasone, including higher WBC count and leukostasis syndrome, which are the main risk factors for transfer into the intensive care unit. Furthermore, although many physicians may be reluctant to use steroids in AML because of the potential risk of invasive fungal infections, we did not identify such adverse effects in dexamethasone-treated patients who also received antifungal prophylaxis.

Because this study is retrospective and included only a relatively low number of patients, there are some limitations. However, it also reflects some real-life aspects of the care of AML patients with a high WBC count, a difficultto-treat population requiring immediate medical treatment, which means that these patients are often excluded from prospective trials. In our study, the impact of dexamethasone was adjusted for several clinical and biological factors to limit the potential biases inherent to non-randomized studies, and a biological rationale was further provided to strengthen the clinical findings. Although prospective randomized clinical trials are needed to confirm the results of this study, our findings argue for a repositioning of dexamethasone use within the backbone of intensive chemotherapy in AML patients.

Acknowledgments This work was supported by grants from the French government under the "Investissement d'Avenir" program (ANR-11PHUC-001), the Institut National du Cancer (PLBIO 2015143), the InnaBioSanté Foundation (RESISTAML project), the Toulouse Cancer Santé Foundation, the Laboratoire d'Excellence 997

S. Bertoli et al.

TOUCAN and La Ligue Contre le Cancer. EG is supported by a postdoctoral grant from La Fondation de France.

The authors would like to thank the data management unit of Toulouse University Hospital for its support enabling e-CRF. We thank Dr. Véronique De Mas for the management of the

References 1. Dohner H, Weisdorf DJ, Bloomfield CD. Acute myeloid leukemia. N Engl J Med. 2015;373(12):1136-1152.

2. Rollig C, Ehninger G. How I treat hyperleukocytosis in acute myeloid leukemia.

Blood. 2015;125(21):3246-3252.

3. Inaba H, Pui CH. Glucocorticoid use in acute lymphoblastic leukaemia. Lancet Oncol. 2010;11(11):1096-1106.

4. Kelaidi C, Chevret S, De Botton S, et al.

Improved outcome of acute promyelocytic leukemia with high WBC counts over the last 15 years: the European APL Group experience. J Clin Oncol. 2009;27(16):26682676.

5. Sanz MA, Montesinos P. How we prevent and treat differentiation syndrome in patients with acute promyelocytic leukemia. Blood. 2014;123(18):2777-2782.

6. Stucki A, Rivier AS, Gikic M, Monai N, Schapira M, Spertini O. Endothelial cell activation by myeloblasts: molecular mechanisms of leukostasis and leukemic cell dissemination. Blood. 2001;97(7):2121-2129.

7. Griffin JD, Rambaldi A, Vellenga E, Young DC, Ostapovicz D, Cannistra SA. Secretion of interleukin-1 by acute myeloblastic leukemia cells in vitro induces endothelial cells to secrete colony stimulating factors.

Blood. 1987;70(4):1218-1221.

8. Baschant U, Tuckermann J. The role of the glucocorticoid receptor in inflammation and immunity. J Steroid Biochem Mol Biol.

2010;120(2-3):69-75.

9. Grimwade D, Hills RK, Moorman AV, et al.

Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876 younger adult patients treated in the United Kingdom Medical Research Council trials.

Blood. 2010;116(3):354-365.

10. Bertoli S, Berard E, Huguet F, et al. Time from diagnosis to intensive chemotherapy initiation does not adversely impact the outcome of patients with acute myeloid leukemia. Blood. 2013;121(14):2618-2626.

11. Pigneux A, Harousseau JL, Witz F, et al.

Addition of lomustine to idarubicin and cytarabine improves the outcome of elderly

998

12.

13.

14.

15.

16.

17.

18.

19.

20.

Biobank BRC-HIMIP (Biological Resources Centres-INSERM Midi-Pyrénées “Cytothèque des Hémopathies Malignes”). We also thank all the members of the Gaël Adolescent Espoir Leucémie (GAEL). Finally, we are indebted to Dr. Mary Selak for the critical reading of the manuscript.

patients with de novo acute myeloid leukemia: a report from the GOELAMS. J Clin Oncol. 2010;28(18):3028-3034.

Dohner H, Estey EH, Amadori S, et al.

Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood. 2010;115(3):453-474.

Chabrol A, Cuzin L, Huguet F, et al.

Prophylaxis of invasive aspergillosis with voriconazole or caspofungin during building work in patients with acute leukemia.

Haematologica. 2010;95(6):996-1003.

Cheson BD, Bennett JM, Kopecky KJ, et al.

Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol.

2003;21(24):4642-4649.

Griessinger E, Anjos-Afonso F, Vargaftig J, et al. Frequency and dynamics of leukemiaInitiating cells during short-term ex vivo culture informs outcomes in acute myeloid leukemia patients. Cancer Res. 2016;76(8): 2082-2086.

Griessinger E, Anjos-Afonso F, Pizzitola I, et al. A niche-like culture system allowing the maintenance of primary human acute myeloid leukemia-initiating cells: a new tool to decipher their chemoresistance and self-renewal mechanisms. Stem Cells Transl Med. 2014;3(4):520-529.

Kurata M, Rathe SK, Bailey NJ, et al. Using genome-wide CRISPR library screening with library resistant DCK to find new sources of Ara-C drug resistance in AML. Sci Rep. 2016;6:36199.

Malani D, Murumagi A, Yadav B, et al.

Enhanced sensitivity to glucocorticoids in cytarabine-resistant AML. Leukemia.

2017;31(5):1187-1195.

Farge T, Saland E, de Toni F, et al.

Chemotherapy-resistant human acute myeloid leukemia cells are not enriched for leukemic stem cells but require oxidative metabolism. Cancer Discov. 2017;7(7):716735.

Hackl H, Steinleitner K, Lind K, et al. A gene expression profile associated with relapse of cytogenetically normal acute myeloid

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

leukemia is enriched for leukemia stem cell genes. Leuk Lymphoma. 2015;56(4):11261128.

Simon L, Lavallee VP, Bordeleau ME, et al.

Chemogenomic landscape of RUNX1mutated AML reveals importance of RUNX1 allele dosage in genetics and glucocorticoid sensitivity. Clin Cancer Res. 2017;23(22):6969-6981.

Miyoshi H, Ohki M, Nakagawa T, Honma Y. Glucocorticoids induce apoptosis in acute myeloid leukemia cell lines with a t(8;21) chromosome translocation. Leuk Res. 1997;21(1):45-50.

Verhaak RG, Wouters BJ, Erpelinck CA, et al. Prediction of molecular subtypes in acute myeloid leukemia based on gene expression profiling. Haematologica. 2009;94(1):131134.

Cancer Genome Atlas Research N. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368(22):2059-2074.

Hicsonmez G. The effect of steroid on myeloid leukemic cells: the potential of short-course high-dose methylprednisolone treatment in inducing differentiation, apoptosis and in stimulating myelopoiesis. Leuk Res. 2006;30(1):60-68.

Pietras EM. Inflammation: a key regulator of hematopoietic stem cell fate in health and disease. Blood. 2017;130(15):16931698.

Rosen JM, Jordan CT. The increasing complexity of the cancer stem cell paradigm.

Science. 2009;324(5935):1670-1673.

Volk A, Li J, Xin J, et al. Co-inhibition of NFkappaB and JNK is synergistic in TNFexpressing human AML. J Exp Med. 2014;211(6):1093-1108.

Schepers K, Campbell TB, Passegue E. Normal and leukemic stem cell niches: insights and therapeutic opportunities. Cell Stem Cell. 2015;16(3):254-267.

Korn C, Mendez-Ferrer S. Myeloid malignancies and the microenvironment. Blood.

2017;129(7):811-822.

Oberoi S, Lehrnbecher T, Phillips B, et al.

Leukapheresis and low-dose chemotherapy do not reduce early mortality in acute myeloid leukemia hyperleukocytosis: a systematic review and meta-analysis. Leuk Res. 2014;38(4):460-468.

haematologica | 2018; 103(6)

ARTICLE

Acute Lymphoblastic Leukemia

Fit αβ T-cell receptor suppresses leukemogenesis of Pten-deficient thymocytes

Ferrata Storti Foundation

Stéphanie Gon,1* Marie Loosveld,1,2* Thomas Crouzet,1 Delphine Potier,1 Mélanie Bonnet,1 Stéphanie O. Morin,3 Gérard Michel,4 Norbert Vey,3,5 Jacques A. Nunès,3 Bernard Malissen,1 Romain Roncagalli,1 Bertrand Nadel,1 and Dominique Payet-Bornet1

1 Aix-Marseille Université, CNRS, INSERM, CIML; 2APHM, Hôpital La Timone, Laboratoire d’Hématologie; 3Aix-Marseille Université, CNRS, INSERM, Institut Paoli-Calmettes, CRCM; 4 APHM, Hôpital La Timone, Service d’Hématologie et d’Oncologie Pédiatrique and 5 Institut Paoli-Calmettes, Hematology Department, Marseille, France

Haematologica 2018 Volume 103(6):999-1007

*These authors contributed equally to this work.

ABSTRACT

ignaling through the αβT cell receptor (TCR) is a crucial determinant of T-cell fate and can induce two opposite outcomes during thymocyte development: cell death or survival and differentiation.

To date, the role played by T-cell receptor in the oncogenic transformation of developing T cells remains unclear. Here we show that human primary T-cell acute lymphoblastic leukemias expressing an αβT cell receptor are frequently deficient for phosphatase and tensin homolog protein (PTEN), and fail to respond strongly to T-cell receptor activation.

Using Pten-deficient T-cell acute lymphoblastic leukemia mouse models, we confirm that T-cell receptor signaling is involved in leukemogenesis.

We show that abrogation of T-cell receptor expression accelerated tumor onset, while enforced expression of a fit transgenic T-cell receptor led to the development of T-cell receptor-negative lymphoma and delayed tumorigenesis. We further demonstrate that pre-tumoral Pten-deficient thymocytes harboring fit T-cell receptors undergo early clonal deletion, thus preventing their malignant transformation, while cells with unfit Tcell receptors that should normally be deleted during positive selection, pass selection and develop T-cell acute lymphoblastic leukemias.

Altogether, our data show that fit T-cell receptor signaling suppresses tumor development mediated by Pten loss-of-function and point towards a role of Pten in positive selection.

S

Correspondence: payet@ciml.univ-mrs.fr/nadel@ciml.univ-mrs.fr

Received: January 11, 2018.

Accepted: March 15, 2018.

Pre-published: March 22, 2018.

Introduction doi:10.3324/haematol.2018.188359 Thymopoiesis aims to create a repertoire of mature T cells equipped with a diverse array of functional αβ or δg T-cell receptors (TCR) able to recognize the broadest possible range of foreign antigens. This large receptor diversity is mainly due to the V(D)J recombination process, in which few of a large pool of V, D and J gene segments are somatically rearranged with imprecise joining. The price to pay for this strategy of random generation of diversity is the creation of a high load of CD4+ CD8+ DP cortical thymocytes bearing no, or “unfit” receptors, i.e. displaying a too low or too high affinity for self peptide-major histocompatibility complex (pMHC), and which will have to be eliminated through death by neglect (>90% thymocytes) and negative selection (Online Supplementary Figure S1A).1 Only the small pool of thymocytes expressing TCR with intermediate affinity and/or avidity for pMHC (denoted here as fit TCR) will be induced to further differentiate into mature CD4 or CD8 single-positive (SP) thymocytes, a transition known as positive selection. Therefore, TCR signaling can induce two opposite outcomes during thymocyte development: cell death or survival and differentiation.

Somatic rearrangement, proliferation and selection provide a propitious environment for major derailments of thymocyte ontogeny. T-cell acute lymphoblastic leukemias (T-ALL) are malignant proliferation of such T-cell progenitors abnormally arrested at various stages of their maturation process (Online Supplementary Figure S1B).2-4 They constitute a particularly heterogeneous group of diseases, resulting haematologica | 2018; 103(6)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/999 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

999

S. Gon et al.

from a large array of genetic and epigenetic alterations in oncogenes and tumor suppressors.5-7 A wealth of information has emerged from recent pan-(epi)genomic analysis of T-ALLs, and this has allowed mapping of cell-intrinsic genetic defaults in these cells to be expanded. However, much less is known about the potential cell-extrinsic cues that may impact on the leukemia genesis process, including the role that the TCR might play in malignant transformation throughout thymocyte selection, survival and proliferation.8-10 In this study, we sought to address how TCR signaling can interfere or, on the contrary, can be integrated in T-ALL oncogenic networks.

Methods Patients’ samples Diagnostic specimens (peripheral blood or bone marrow) collected from patients treated at the Timone Children’s Hospital or Paoli Calmettes Institute (Marseille, France) or from Necker Hospital (Paris, France) were used to generate xenografts.

Diagnosis and classification were defined by expression of specific

T-cell markers and negativity for B cells and myeloid markers.

Healthy human thymus were obtained from Timone Children’s Hospital. Samples were purified by Ficoll-Hypaque centrifugation.

T-ALLs were included within FRALLE-2000 or GRAALL-2005 protocols, and informed consent for use of diagnostic specimens for future research was obtained from the patients or relatives in accordance with the Declaration of Helsinki. This study was approved by institutional review boards of all hospitals involved.

Mice Mice were bred and housed in specific pathogen-free conditions in CIML animal facilities and were handled in accordance with French and European guidelines. Mice strains and oligonucleotides used for mice genotyping are listed in the Online Supplementary Methods and Online Supplementary Table S1. Xenotransplantation of primary human T-ALL samples in immunodeficient NSG mice was performed as previously described.11

TCR-signaling ability assays TCR-signaling ability assays were performed with 2x107 wild-

A

B

C

D

1000

E

Figure 1. Fit TCRαβ signaling functions as a tumor suppressor. (A) Thymus (left) and spleen (right) of typical wild-type (WT) and tumoral Ptendel, [Rag1–/– x Ptendel] and [OT-II x Rag1–/– x Ptendel] mice. (B) Phenotypes of typical tumors generated by Ptendel in the spleen and by [Rag1–/– x Ptendel] and [OT-II x Rag1–/– x Ptendel] in the thymus. WT and [OT-II x Rag1–/–] controls are shown. CD4 SP or DP gates (top plots, bold squares) were further analyzed for CD3/TCRβ expression (bottom plots).

Two typical thymi of [OT-II x Rag1–/– x Ptendel] mice are shown (1 representative of 10). (C) Thymi of indicated mice were analyzed by immunoblotting with antibodies specific for Pten, Myc, cleaved Notch1, Bcl2 and Actin as a loading control. #Identification number of analyzed mice. *Mice that did not display T-cell acute lymphoblastic leukemias (T-ALL) or T-cell lymphoblastic lymphomas (T-LBL) symptoms at the time of sacrifice. (D) Transcriptional downderegulation of transgenic TCRβ chain in [OT-II x Rag1–/– x Ptendel] tumor thymocytes (n=6). Transcripts levels were normalized to ABL. Error bars show means with Standard Deviation. Statistical significance was assessed using Mann-Whitney test (**P<0.01). (E) Ptendel mice survival curve was compared to [Rag1–/– x Ptendel] or [OT-II x Rag1–/– x Ptendel] mice survival curves using log-rank (Mantle-Cox) test (**P<0.01; ***P<0.001); median weeks of survival are 11 (as previously observed15), 9.95 and 17.9, respectively.

haematologica | 2018; 103(6)

αβTCR signaling acts as a tumor suppressor type or leukemic cells resuspended in RPMI medium (200 mL) and stimulated for 2 minutes (min) at 37°C with avidin (14 mg) and biotinylated anti-CD3 (10 mg; clone 2C11, BD Pharmingen, for mouse cells and clone UCHT1, eBiosciences for human cells) and biotinylated anti-CD28 (10 mg; clone 37.51, BD Pharmingen, for mouse cells and clone CD28.2, eBiosciences for human cells).

Unstimulated (control) cells were incubated with avidin alone.

After lysis in 2X TNE buffer (100 mM Tris, 2% Nonidet P-40, 40 mM EDTA) supplemented with protease and phosphatase inhibitors, protein extracts (approx. 60 mg) were analyzed by immunoblot (see Online Supplementary Methods and Online Supplementary Table S2). P-Tyr levels were quantified on the entire lane and normalized to ACTIN. P-AKT protein levels were normalized to AKT.

was evaluated by two-tailed Mann-Whitney U-test. P<0.05 was considered significant.

Further details are provided in the Online Supplementary Methods and antibodies used for flow cytometry are listed in Online Supplementary Tables S3 and S4.

Results Ptendel thymocytes expressing transgenic TCR are counter-selected during leukemogenesis Phosphatase and tensin homolog protein (PTEN) is a well-known tumor suppressor involved in numerous types of cancers, and represents the main negative regulator of PI3K/AKT signaling pathway.12 To investigate the role of the TCR in leukemogenesis, we used a Pten-deficient mouse model of T-ALL which has previously been shown to induce TCRαβ+ tumors.13,14 PtenFlox/Flox mice were crossed into a CD4-Cre background (hereafter referred to as Ptendel) in which Cre is fully active at the CD4+CD8+ double positive (DP) stage of thymocyte development.15 As previously described,16 Ptendel mice developed leukemia characterized by malignant proliferation of mono/oligoclonal T cells (Online Supplementary Table S5), and enlarged thymus and spleen (Figure 1A). Peripheral leukemic blasts from Ptendel mice were typically CD4 SP and expressed αβTCR at their surface (Figure 1B and Online Supplementary Table S5), in line with previous reports.14,16,17 To analyze the impact of positive selection, we used the OT-II mouse model which expresses a transgenic Vα2/Vβ5.1 TCR recognizing the chicken ovalbumin antigen in the context of MHC-II molecules.18 Ptendel mice were crossed with OT-II Rag1-deficient mice, in which all developing T cells do express an OT-II fit TCR, designed to trigger positive selection and give rise to mature SP T

Proliferation and apoptosis assays following CD3/CD28 stimulation For comparison of the TCR-signaling ability to mediate proliferation or apoptosis of normal or leukemic cells, 1.105 of non-purified or CD4 SP purified T cells were mixed (ratio 1:1) with Dynabeads Mouse T-Activator CD3/CD28 (Life Technologies) or Dynabeads Human T-Activator CD3/CD28 (Life Technologies) in 96-well flat bottom plates and incubated for 24 or 72 hours (h) (37°C, 5% CO2). For unstimulated controls, thymocytes were incubated in the same conditions but without anti-CD3/CD28 coated beads. Proliferation was measured using CFSE labeling (CellTrace™ CFSE Cell Proliferation Kit, Life Technologies) and apoptosis was followed using AnnexinV labeling (BD Pharmingen) and 7-AAD (BD Pharmingen) according to the manufacturer’s instructions.

Statistical analysis Kaplan-Meier survival curves and statistical analyses were performed using GraphPad Prism software. Survival curves were compared using log-rank (Mantle-Cox) test. Statistical significance

Table 1. Immuno-phenotypes of T-cell acute lymphoblastic leukemia developed by OT-II x Ptendel mice in I-Ab/b or I-Ab/d backgrounds.

Mouse #

I-A

CD4/CD8

TCRαβ

TCR OT-II

TCRVβ

TCRV α

25 26 28 70 83 114 140 141 350 2 11 12 17 19 278 281 291 321 354

I-Ab/b I-Ab/b I-Ab/b I-Ab/b I-Ab/b I-Ab/b I-Ab/b I-Ab/b I-Ab/b I-Ab/d I-Ab/d I-Ab/d I-Ab/d I-Ab/d I-Ab/d I-Ab/d I-Ab/d I-Ab/d I-Ab/d

CD4+ CD4+/DN CD4+ CD4+ CD4+ CD4+ CD4+ CD4+/DN CD4+ CD4+ CD4+ CD4+ CD4+ CD4+ CD4+ CD4+ CD4+/DN CD4+/DN CD4+/DP

+ + + + + + + + + + + + + + + + + + +

Neg Neg Neg Neg Neg Neg Neg Neg Neg + + + Neg + + + + + +

Vβ14 Vβ11 Vβ14 Vβ14 Vβ14 Vβ6 Vβ8.1/8.2 Vβ5 Vβ14 Vβ5 Vβ5 Vβ5 Vβ6 Vβ5 Vβ5 Vβ5 Vβ5 Vβ5 Vβ5

Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Nd* Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Vα2 Vα2

*Negativity for TCRVa2 and the TCRVα expressed was not determined.

haematologica | 2018; 103(6)

1001

S. Gon et al.

cells. Both [OT-II x Rag1–/– x Ptendel] and the control TCRdeficient mice [Rag1–/– x Ptendel] (Online Supplementary Figure S2) developed T-cell lymphoblastic lymphomas (T-LBL) which were mostly restricted to the thymus (Figure 1A), over-expressed Bcl219 and, as previously described,17 were recurrently Notch1-dependent (Figure 1C). Indeed, we found that all [Rag1–/– x Ptendel] tumors tested (n=7) and 3 out of 7 [OT-II x Rag1–/– x Ptendel] tumors

A

were Notch1 activated. Of note, Notch1 activation does not impact latency of [OT-II x Rag1–/– x Ptendel] tumors (Online Supplementary Figure S3). Strikingly, in the [OT-II x Rag1–/– x Ptendel] model, the examined tumors (n=15) had lost surface expression of the OT-II transgene and were either TCRneg or TCRαβlow (Figure 1B), consistent with previous observations.17 Molecular analysis of the transgenic β chain mRNA expression in the tumors revealed down-

B

C

D

1002

Figure 2. Counter-selection of T cells harboring H-Y TCR. (A) Spleen (left) and thymus (right) of typical H-Y and tumoral [H-Y x Ptendel] female mice.

[H-Y x Ptendel] mice developed T-cell acute lymphoblastic leukemias (T-ALL) in approximately ten weeks (n =5). (B) Flow cytometry analysis of typical spleens from H-Y and tumoral [H-Y x Ptendel] female mice. (C) Flow cytometry analysis of typical diseasefree thymus (left panels) and spleens (right panels) from young (4-week old) H-Y and [H-Y x Ptendel] female mice.

Percentages of cells in depicted gates are indicated. Representative data of at least 3 experiments are shown.

Dot plots show percentages of CD8 SP, H-Y+ CD8 SP or H-Y+ DP thymic cells and HY+ CD8 T cells from spleens of control H-Y (n= 7) and [H-Y x Ptendel] (n=7) female mice. (D) Pretumoral single positive (SP) thymocytes are partially blocked at the immature CD69+CD62Llow stage.

Analysis of 4-week old disease-free (pre-tumoral) Ptendel and [HY x Ptendel] mice and their respective control counterparts is shown (representative data of at least 3 experiments).

CD8 M: mature CD8 SP (CD69–CD62LHi); CD8 IM: immature + Lo CD8 SP (CD69 CD62L ). Arrows indicate the stage of differentiation arrest (SP2 and CD8 IM). Dot plots show percentages of CD4 SP2 and CD8 IM T cells in the indicated backgrounds (n=5 or 6). Error bars show means with Standard Deviation.

Statistical significance was assessed using Mann-Whitney test (**P<0.01; ***P<0.001).

haematologica | 2018; 103(6)

αβTCR signaling acts as a tumor suppressor

regulation of the Vβ5.1 transcript (Figure 1D). This suggests an active counter-selection of leukemic (or preleukemic) thymocytes bearing the transgenic TCR. The latency of tumor onset was significantly increased in [OTII x Rag1–/– x Ptendel] mice compared to Ptendel mice, consistent with the time that is likely required for selection of TCRneg cells, while in the enforced absence of TCR ([ Rag1–/– x Ptendel] mice), latency was significantly reduced (Figure 1E), evoking a potential tumor suppressor role of TCR signaling in leukemogenesis.

To rule out transgenic-specific effect, we also tested the impact of selection in the H-Y mouse model, which expresses a transgenic TCR recognizing the male H-Y antigen in the context of MHC-I molecules. In female HY mice, negative selection is not operating, and positively selected mature H-Y+ TCR T cells differentiate as CD8 SP.20 H-Y mice were crossed with Ptendel mice, on a Ragproficient background to allow non-transgenic TCR competitive formation and development. [H-Y x Ptendel] females developed TCRαβ+ T-ALL (Figure 2A and B).

Remarkably, tumors were typically CD4 SP (never CD8 SP), and none of them expressed the transgenic H-Y TCR (Figure 2B). Thus, regardless of the model used, the expression by (pre-)tumoral thymocytes of a fit TCR is counter-selected during T-ALL development. In diseasefree thymi and spleens of young female [H-Y x Ptendel] mice (and thus before clinical tumor manifestation), we detected a severe reduction of H-Y+ TCR cells at the CD8 SP stage compared to control H-Y mice, and this was already apparent at the DP stage of thymocyte development (Figure 2C). In addition, this counter selection of HY+ thymocytes occurs post β-selection, after immature single positive (ISP) stage (Online Supplementary Figure S4).

These data suggest that Pten-deficient H-Y+ TCR thymocytes are eliminated instead of positively selected, and that this counter-selection occurs before full malignant transformation, preventing leukemia development.

A

Disruption of final maturation of Pten-deficient SP cells Since most mature thymocytes with fit H-Y TCR are eliminated in young female disease-free [H-Y x Ptendel] mice, we analyzed the remaining H-Y TCR negative SP cells, using the CD69, CD62L and CCR9 markers of T-cell differentiation. Developmental sequence of CD4 SP thymocyte maturation is usually described as SP1 (CD69+,CD62LLow/med, CCR9+), SP2 (CD69+, CD62LLow/med, CCR9Neg) and SP3 (CD69Neg, CD62LHigh, CCR9Neg).21,22 For CD8 SP cells, the most immature cells are CD69+ and CD62LLow/med while the more mature are CD69Neg and CD62LHigh. We observed for both [H-Y x Ptendel] and Ptendel models a partial block of positively selected cells at the immature CD69+CD62Llow stage (Figure 2D). Such differentiation arrest could provide an additional opportunity for malignant transformation.

Fit TCR signaling acts as a tumor suppressor With the premise that Pten-deficient cells with fit TCR are counter-selected, we next assessed whether cells carrying low affinity (unfit) TCR were prone to leukemia development. OT-II TCR originates from CD4+ I-Ab-restricted T-cell hybridoma,18 thus positive selection is optimal in IAb/b background and sub-optimal in I-Ab/d background.

C57BL/6 (I-Ab) [OT-II x Rag1–/– x Ptendel] mice were crossed with BALB/C (I-Ad) mice to generate Rag-proficient [OT-II x Ptendel] mice on I-Ab/d background. In the spleens from OT-II control mice, percentages of CD4+ T cells dropped from approximately 78% in syngeneic IAb/b background to approximately 7% in allogenic I-Ab/d background (Figure 3A). [OT-II x Ptendel] Rag-proficient mice developed T-ALL with a similar latency as Ptendel mice (approx. 11 weeks), irrespective of the backgrounds (I-Ab/b or I-Ab/d). Leukemic blasts in the spleen were mostly CD4+ (Table 1). On the I-Ab/b background, while all T-ALL analyzed (n=9) expressed a TCRαβ, none of them

B

Figure 3. Thymocytes harboring unfit TCRαβ signaling develop T-cell acute lymphoblastic leukemias (T-ALL). (A) Flow cytometry analysis of typical spleens from [OT-II x Ptenflox] (Control) and tumoral [OT-II x Ptendel] mice bred either on I-Ab/b (left) or I-Ab/d (right) backgrounds (1 representative of n=9). Leukemic cells from [OT-II x Ptendel] I-Ab/b mouse was further screened by cytometry using a Vβ panel and (B) the result indicates that this TALL expresses a TCRVα2Vβ14 receptor (see also Table 1).

haematologica | 2018; 103(6)

1003

S. Gon et al.

expressed the full transgenic OT-II TCR (usually TCRVα2 chain was expressed, but not associated with TCRVβ5 chain) (Table 1 and Figure 3). In line with the H-Y model above, this indicates that (pre)leukemic clones harboring OT-II fit TCR are counter-selected during oncogenesis. In striking contrast, most of TCRαβ+ T-ALL in the allogenic I-Ab/d background (9 of 10) expressed OT-II (Table 1 and Figure 3A). This indicates that in a context of sub-optimal positive selection, Ptendel OT-II+ blasts are not counterselected, but rather bypass death-by-neglect during posi-

A

tive selection allowing further leukemia development.

Altogether our data indicate that, in Pten-deficient TALL mouse models, fit TCR functions as a tumor suppressor impeding thymocytes to develop leukemia, while thymocytes expressing no or unfit TCR are prone to leukemogenesis.

TCRαβ signaling is disabled in Pten-deficient T-ALL We next asked whether endogenous TCRαβ+ from Ptendel mice, which presumably passed positive selection

B

C

D E

F

1004

G

Figure 4. TCRαβ signaling is disabled in Ptendel T-ALL blasts. (A) Indicated cells labeled with CFSE were either left unstimulated (US) or stimulated with anti-CD3/28 beads (CD3/28) for 24 or 72 hours (h), and then analyzed by flow cytometry. SSC/FSC dot plots (left) and CFSE histograms (right) are shown. Numbers indicate percentage of living cells. Representative data of Ptendel T-cell acute lymphoblastic leukemias (T-ALL) (n=5) and Cdkn2a–/– T-ALL (n=5). (B) Analysis of early TCR signaling by immunoblots. Two representative cases of Ptendel T-ALL (n=5) and Cdkn2a–/– T-ALL (n=5), and wild-type (WT) thymocytes are shown.

Cells were untreated (-) or stimulated (+) with anti-CD3/CD28 antibodies for 2 minutes and analyzed by immunoblotting with antibodies specific for phosphorylated tyrosine (P-Tyr), phosphorylated AKT S473 (P-AKT), AKT and Actin. (C) Levels of P-Tyr species normalized to Actin (top) and of P-Akt normalized to Akt (bottom) in unstimulated (US) and in CD3/CD28-stimulated (S) of indicated cells: WT thymus, Ptendel T-ALL (n=5) and Cdkn2a–/– (n=5) assayed in duplicate for P-Tyr. (D and E) Impact of TCR stimulation on human T-ALL cell survival. Cells were either left unstimulated or stimulated with beads coated with anti-CD3 and anti-CD28 antibodies (CD3/CD28) during 72 h and then stained with Annexin V and 7-AAD to monitor cell death by flow cytometry. (D) Typical dot plots for Xg9 and Xg35 are shown. Percentage of live cells (gated) is indicated. (E) Survival index as determined by ratio of live cells in treated (CD3/28) versus unstimulated conditions 72 h post induction.

Each dot corresponds to the mean survival index (obtained from at least 2 assays) of one T-ALL xenograft. Xg35 sample is depicted as a white square. (F and G) Human thymus and T-ALL were analyzed as described in (B). Thymus NP and SP CD4+ correspond to total non-purified (NP) and purified CD4 SP cells, respectively, from healthy human thymus. (F) Two representative samples of human T-ALL: Xg13 (TCRneg) and Xg8 (TCRαβ+) are shown (see also Online Supplementary Figure S9). (G) P-Tyr activation index which corresponds to the ratio of P-Tyr species levels in stimulated (CD3/28) versus unstimulated samples (top); P-AKT normalized to AKT (bottom) in unstimulated (US) and in CD3/CD28-stimulated (S) (Bottom). Statistical significance of P-AKT levels between unstimulated TCR+ PDX versus TCRneg PDX, NP or SP thymocytes are indicated with blue asterisks. TCRneg (n=5: Xg3, 13, 20, 23 & 40) and TCRαβ+ (n=5: Xg8, 9, 35, 38 & 47). P-Tyr species levels were normalized to ACTIN. (C, E and G) Error bars represent means + Standard Deviation. Statistical significance was assessed using Mann-Whitney test; ns: non-significant P>0.05; *P<0.05; **P<0.01.

haematologica | 2018; 103(6)

αβTCR signaling acts as a tumor suppressor

and developed T-ALL, acquired a deficiency in TCR signaling. Unstimulated primary mouse T-ALLs tend to quickly undergo apoptosis in liquid culture in vitro and can be rescued through anti-CD3/anti-CD28 TCR activation, as shown for the TCRαβ+ Cdkn2a–/– T-ALL mouse model (Figure 4A, Online Supplementary Table S6 and Online Supplementary Figures S5 and S6A). By contrast, no rescue could be obtained for T-ALL blasts from the Ptendel T-ALL model, and TCRαβ+ Ptendel leukemic cells quickly died with or without stimulation. To further investigate the impact of TCR triggering at the molecular level, freshly harvested tumors were CD3/CD28 stimulated for 2 min and analyzed by immunoblotting with antibodies specific for phosphorylated tyrosine species (P-Tyr) (Figure 4B). In control purified DP and non-purified (NP) mouse WT thymocytes, activation of the TCR pathway triggered various intracellular signaling molecules23 leading to a marked increase in the pattern of global tyrosine phosphorylation (Figure 4B and C), as previously described.24 Similar results were obtained with Cdkn2a–/– T-ALLs, in line with proliferation data described above. By contrast, global tyrosine phosphorylation was significantly dampened in murine TCRαβ+ Ptendel T-ALLs (Figure 4B and C). However, P-Tyr antibody does not detect the activation of AKT which is the main downstream target of Pten.12 Thus, we specifically monitored phosphorylation of Akt at Ser473. It was previously showed that Akt phosphorylation was very high in non-tumoral Pten-deficient thymocytes compared to Pten-proficient thymocytes.13 However, we found that P-Akt levels in Ptendel T-ALL are similar to the one detected for WT thymocytes and Cdkn2a–/– T-ALL (Figure 4B and C) and thus are lower than one might expect from Ptendeficient thymocytes.13 It is noteworthy that when we induced inactivation of Pten in Cdkn2a–/– T-ALL cells, the ability of those cells to proliferate upon stimulation was conserved (Online Supplementary Figure S7), suggesting that once pre-tumoral thymocytes have passed selection and the tumor is established, late deletion of Pten no longer interferes with TCRmediated activation. In the same line, T cells from diseasefree Ptendel spleen were able to proliferate upon antiCD3/28 stimulation (Online Supplementary Figure S8), confirming that, per se, Pten loss is not directly responsible for the TCR signaling inhibition observed in Ptendel T-ALL. Together with the fact that human late cortical T-ALLs are frequently carrying PTEN loss-of-function alterations,5 this pointed to a possible role of Pten loss in the dysregulation of the selection process. We thus assessed whether the above observation indicating that TCR signaling is impaired in Pten-deficient T-ALL was relevant in human primary T-ALL samples. To obtain adequate quantities of viable human leukemia cells devoid of contaminating residual physiological mature T cells, samples from T-ALL patients were engrafted into immunodeficient NSG mice.

In our patient-derived xenograft (PDX) collection, PTEN was present in all TCRneg T-ALL, while it was not expressed in 5 out of 6 TCRαβ+ samples. We investigated the 5 PTEN-deficient TCRαβ+ T-ALL (T-ALL 8, 9, 35, 38 and 47; their corresponding xenografts were denoted Xg8, Xg9, Xg35, Xg38 and Xg47), and 5 TCRneg T-ALLs were used as controls (Xg3, Xg13, Xg20, Xg23 and Xg40) (Online Supplementary Table S7 and Online Supplementary Figure S6B). Leukemic grafts were harvested from mice and stimulated with anti-CD3 and anti-CD28. In contrast to mouse, human T-ALL do not die quickly in liquid culhaematologica | 2018; 103(6)

Figure 5. Model for integration of Pten loss-of-function and TCR signaling-mediated tumor suppression. In the context of Pten loss, thymocytes bearing fit or high affinity TCR would be eliminated while those bearing no/low affinity TCR would be rescued from death-by-neglect. However, harboring TCR complex that does not signal properly would prevent further thymocyte differentiation, providing an additional opportunity for malignant transformation.

ture; thus we assessed the impact of TCR stimulation at the cellular level. We observed that activation-induced cell death (AICD) was triggered for only 1 TCRαβ+ T-ALL (Xg35), the remaining 4 TCRαβ+ T-ALL, as well as control TCRneg T-ALLs, being resistant to AICD (Figure 4D and E). To investigate signaling downstream of the TCR, as described above for mouse T-ALL, PDX cells were lysed 2 min post activation with anti-CD3/CD28, and analyzed by immunoblotting. In control (disease-free) non-purified (NP) or purified CD4 SP human thymocytes, activation of the TCR pathway led to a marked increase in the pattern of global tyrosine phosphorylation. In contrast, the TCRαβ+ T-ALLs samples showed a reduced and somewhat intermediate activation of tyrosine phosphorylated species compared to NP or CD4 SP thymocytes and TCRαβneg controls (Figure 4F). Globally, the P-Tyr activation index of TCRαβ+ T-ALL was significantly lower than physiological controls (NP or CD4 SP; P<0.05) but not significantly higher than TCRαβneg T-ALL (Figure 4G). In contrast, AKT phosphorylation was detected in PTEN-deficient TCRαβ+ T-ALLs in unstimulated conditions, then following stimulation, it was slightly increased (approx. 3.5fold) and ended up equivalent to CD4 SP positive control (Figure 4G).

1005

S. Gon et al.

Taken together, these data indicate that both in mouse and human, TCRαβ signaling network is largely disabled in PTEN-deficient TCRαβ+ T-ALLs.

Discussion Here we undertook to investigate the impact of TCR signaling during T-ALL leukemogenesis. We show that in a mouse model of Pten loss-of-function, a frequent event among human TCRαβ+ T-ALLs (approx. 70%) (Online Supplementary Table S8), early counter-selection of fit TCRαβ+ thymocytes, occurs before the onset of leukemia development. Furthermore, we show that established TCRαβ+ T-ALLs are carrying TCRs which are unfit and/or impaired in signaling.

T-ALLs represent the malignant counterparts of most thymocyte stages of development. Analysis of T-ALL subtype distribution based on a human T-ALL cohort of 230 subjects25 showed that early immature (n=52), cortical T-ALL (n=103), mature TCRgδ+ T-ALL (n=36) and mature TCRαβ+ T-ALL (n=39) represent 22.6%, 44.8%, 15.6% and 17%, respectively, of T-ALL cases.

Intriguingly, among the TCR+ T-ALL, TCRαβ+ T-ALLs were under-represented (52%) relative to TCRgδ+ (48%) compared to physiological counterparts in which TCRαβ largely dominate the fraction of TCR + thymocytes (approx. 95% vs. approx. 5% TCRgδ).26 Unlike αβ T cells, gδ T cells are not restrained to MHC and do not undergo conventional MHC-mediated positive and negative selections.26,27 Conversely to pre-tumoral αβ T cells, pretumoral gδ T cells might thus not be counter-selected, possibly explaining their over-representation compared to TCRαβ+ T-ALL. Our in vivo data showed that pre-tumoral Ptendel cells with unfit TCR signaling (OT-II in I-Ab/d background) are positively selected for leukemogenesis, while thymocytes with fit TCR signaling (H-Y or OT-II in I-Ab/b background) are counter-selected and never developed leukemia. Thus this study points to a role of Pten during the positive selection process. Yet the specific molecular mechanism allowing positive selection of pre-tumoral cells with unfit TCR (and counter-selection of cells with fit TCR) in the absence of Pten remains to be determined. A possible scenario would be that Pten loss merely shifts the window of positive and negative selection thresholds (Figure 5). On one hand, PTEN loss might substitute for missing TCR signaling, allowing cells with no TCR or low affinity TCR to be rescued from death by neglect and bypass positive selection. Accordingly, an increase in positively selected T cells was observed in mouse models in which AKT was hyperactive.28 Herein, we found that, following TCR-stimulation, AKT activation is similar in Pten-deficient T-ALL and in WT thymocytes. Therefore, while most of the TCR signaling network is disabled (Figure 4), AKT pathway appears ‘normal’ and, in the context of our scenario (Figure 5), is likely to be the main element contributing to the bypass of positive selection by Pten-deficient thymocytes harboring unfit/low TCR. On the other hand, integration of signals resulting from both PTEN loss and a fit TCR (passing positive selection) might reach over-thethreshold signaling and trigger negative selection, eliminating thymocytes carrying fit TCRs even before malignant transformation. This would involve multiple pathways downstream of PTEN loss,12,29 since a mere AKT 1006

hyperactivation was insufficient to recapitulate the loss of fit H-Y+ thymocytes.28 A challenging perspective would be to decipher the molecular mechanism underlying the counter-selection of Pten-deficient fit αβ TCR+ thymocytes and then to assess the possibility of activating this apoptotic program in tumoral cells.

Our data indicate that TCRαβ signaling pathway is actively involved in T-ALL oncogenesis. We show that TCRαβ signaling can impede the development of Ptendeficient tumors and thus acts as a bona fide tumor suppressor. However, given the diametrically opposed effect of TCR activation on discrete stages of T-cell development, the TCR might also have pro-oncogenic effects in other contexts and/or developmental stages. For example, we show that in a Pten-proficient Cdkn2a–/– T-ALL model, TCRαβ+ tumors are sensitive to TCR activation.

Likewise, thymocytes harboring fit H-Y TCR are not counter-selected in female TEL-JAK2 mouse model and develop leukemia.8 In the same line, Pten-deficient TCRα–/– or SLP76–/– mice, in which TCR signaling is abrogated, display delayed tumor onset.17 Yet, and in contrast to the Ptendel model, TCRα–/– and SLP76–/– thymocytes are blocked before (and therefore not subjected to) positive selection.30,31 Pre-TCR might also be directly involved in oncogenesis. For instance, in dominant active NOTCH1 (ICN1) model, pre-TCR signaling is required for tumorigenesis.32 Conversely [Rag1–/– x Ptendel] thymocytes that are devoid of pre-TCR bypass β-selection and develop DP T-cell lymphoma in short latency (Figure 1); this is also in striking contrast to Pten-deficient TCRα–/– or SLP76–/– thymocytes (described above) that express a pre-TCR and for which leukemogenesis is impaired.17 In normal β-selected cells, the exit of proliferation is induced by preTCR signals that inhibit Notch1 pathway leading to Myc downregulation.33,34 In [Rag1–/– x Ptendel] DP lymphoma, Notch1 pathway is systematically activated leading to sustained expression of Myc (Figure 1C) it might be that pre-TCR signaling exerts a tumor suppressor role by shutting down Notch1 and Myc pathways.

A recent study indicated that TCRαβ+ T-ALL were prone to activation-induced cell death (AICD), and antiCD3 stimulation of TCR signaling was proposed as a therapeutic strategy to eliminate leukemic blasts.8 By contrast, most of our TCRαβ+ T-ALL samples (4 of 5) were resistant to AICD (Figure 4E). This discrepancy could be due to the stage of arrest (before or after positive/negative selections) of T-ALL samples, as in both studies sensitive TCRαβ+ samples (4 of 5 in the in vivo analysis of Trinquand et al.8 and 1 of 5 in our study) were CD1a+, consistent with an arrest at cortical DP stage, during which positive and negative selections occur.35 In contrast, AICD-resistant samples in both studies represented true late-cortical CD1aneg SP T-ALLs. This cautions that anti-CD3 therapeutic strategies might be restrained to a subgroup of sensitive TCR+T-ALL (such as TCRαβ+CD1a+, eventually the rare TCRαβ+PTEN+ cases, or TCRgδ+ 8), and thus for mature TCRαβ+PTENneg T-ALL alternative options should be considered. Here we have showed that integration of Pten loss and fit TCR signaling promotes a deletional program. Thus, an attractive perspective would be to decipher the mechanism underlying this apoptotic program in order to uncover an actionable target inducing cell-death, which might open new therapeutic avenues for this poor prognosis PTEN-deficient TCRαβ+ subgroup.36 haematologica | 2018; 103(6)

αβTCR signaling acts as a tumor suppressor

Funding This work was supported by grants from LYSARC-Institut CARNOT CALYM-ANR (#R14-2014, #R20-2015, #R292016, #R31-2017), CNRS and INSERM. TC was supported by grants from Canceropôle PACA and la Fondation pour la Recherche Médicale (ING20121226364). ML was a recipient of a fellowship from INCa (#ASC12035ASA). SOM is supported by a fellowship from Aix-Marseille Université. JAN is supported by a grant from Fondation pour la Recherche Médicale (Equipe FRM DEQ20140329534).

References 1. Klein L, Kyewski B, Allen PM, Hogquist KA. Positive and negative selection of the T cell repertoire: what thymocytes see (and don't see).

Nat Rev Immunol.

2014;14(6):377-391.

2. Asnafi V, Beldjord K, Boulanger E, et al.

Analysis of TCR, pT alpha, and RAG-1 in T-acute lymphoblastic leukemias improves understanding of early human T-lymphoid lineage commitment. Blood. 2003; 101(7):2693-2703.

3. Ferrando AA, Look AT. Gene expression profiling in T-cell acute lymphoblastic leukemia. Semin Hematol. 2003;40(4):274280.

4. Soulier J, Clappier E, Cayuela JM, et al.

HOXA genes are included in genetic and biologic networks defining human acute Tcell leukemia (T-ALL). Blood. 2005; 106(1):274-286.

5. Belver L, Ferrando A. The genetics and mechanisms of T cell acute lymphoblastic leukaemia. Nat Rev Cancer. 2016; 16(8):494-507.

6. Meijerink JPP. Genetic rearrangements in relation to immunophenotype and outcome in T-cell acute lymphoblastic leukaemia. Best Pract Res Clin Haematol.

2010;23(3):307-318.

7. Durinck K, Goossens S, Peirs S, et al. Novel biological insights in T-cell acute lymphoblastic leukemia. Exp Hematol.

2015;43(8):625-639.

8. Trinquand A, Dos Santos NR, Tran Quang C, et al. Triggering the TCR Developmental Checkpoint Activates a Therapeutically Targetable Tumor Suppressive Pathway in T-cell Leukemia.

Cancer Discov.

2016;6(9):972-985.

9. Cui Y, Onozawa M, Garber HR, et al.

Thymic expression of a T-cell receptor targeting a tumor-associated antigen coexpressed in the thymus induces T-ALL. Blood. 2015;125(19):2958-2967.

10. Klinger MB, Guilbault B, Goulding RE, Kay RJ. Deregulated expression of RasGRP1 initiates thymic lymphomagenesis independently of T-cell receptors. Oncogene.

2005;24(16):2695-2704.

11. Loosveld M, Castellano R, Gon S, et al.

Therapeutic Targeting of c-Myc in T-Cell Acute Lymphoblastic Leukemia, T-ALL. Oncotarget. 2014;5(10):3168-3172.

haematologica | 2018; 103(6)

Acknowledgments The authors thank the European Mouse Mutant Archives (EMMA) for providing Ptenflox/flox mice, Charles Pignon and Maeva Patry for mice genotyping, Virginie Fouilloux (Cardiac Surgery Department, APHM, Marseille, France) for providing human thymus, Marie Malissen for providing CD4-Cre and HY mice, the CRCM cytometry platform for phosphoflow FACS analysis and Philippe Naquet for the critical reading of the manuscript.

12. Milella M, Falcone I, Conciatori F, et al.

PTEN: Multiple Functions in Human Malignant Tumors. Front Oncol. 2015;5:24.

13. Newton RH, Lu Y, Papa A, et al.

Suppression of T-cell lymphomagenesis in mice requires PTEN phosphatase activity.

Blood. 2015;125(5):852-855.

14. Suzuki A, Yamaguchi MT, Ohteki T, et al.

T cell-specific loss of Pten leads to defects in central and peripheral tolerance.

Immunity. 2001;14(5):523-534.

15. Guo W, Lasky JL, Chang CJ, et al. Multigenetic events collaboratively contribute to Pten-null leukaemia stem-cell formation.

Nature. 2008;453(7194):529-533.

16. Hagenbeek TJ, Spits H. T-cell lymphomas in T-cell-specific Pten-deficient mice originate in the thymus. Leukemia. 2008; 22(3):608-619.

17. Liu X, Karnell JL, Yin B, et al. Distinct roles for PTEN in prevention of T cell lymphoma and autoimmunity in mice. J Clin Invest.

2010;120(7):2497-2507.

18. Barnden MJ, Allison J, Heath WR, Carbone FR. Defective TCR expression in transgenic mice constructed using cDNA-based alphaand beta-chain genes under the control of heterologous regulatory elements.

Immunol Cell Biol. 1998;76(1):34-40.

19. Feng H, Stachura DL, White RM, et al. Tlymphoblastic lymphoma cells express high levels of BCL2, S1P1, and ICAM1, leading to a blockade of tumor cell intravasation.

Cancer Cell.

2010;18(4):353-366.

20. Kisielow P, Teh HS, Bluthmann H, von Boehmer H. Positive selection of antigenspecific T cells in thymus by restricting MHC molecules. Nature. 1988;335(6192): 730-733.

21. Hogquist KA, Xing Y, Hsu FC, Shapiro VS. T Cell Adolescence: Maturation Events Beyond Positive Selection. J Immunol.

2015;195(4):1351-1357.

22. Cowan JE, Parnell SM, Nakamura K, et al.

The thymic medulla is required for Foxp3+ regulatory but not conventional CD4+ thymocyte development. J Exp Med. 2013;210(4):675-681.

23. Malissen B, Bongrand P. Early T cell activation: integrating biochemical, structural, and biophysical cues. Annu Rev Immunol.

2015;33:539-561.

24. Poltorak M, Arndt B, Kowtharapu BS, et al.

TCR activation kinetics and feedback regulation in primary human T cells. Cell

Commun Signal. 2013;11:4.

25. Dadi S, Le Noir S, Payet-Bornet D, et al.

TLX homeodomain oncogenes mediate T cell maturation arrest in T-ALL via interaction with ETS1 and suppression of TCRalpha gene expression. Cancer Cell.

2012;21(4):563-576.

26. Chien YH, Meyer C, Bonneville M. gammadelta T cells: first line of defense and beyond. Annu Rev Immunol. 2014;32:121155.

27. Born WK, Huang Y, Reinhardt RL, Huang H, Sun D, O'Brien RL. gammadelta T Cells and B Cells. Adv Immunol. 2017;134:1-45.

28. Na SY, Patra A, Scheuring Y, et al.

Constitutively active protein kinase B enhances Lck and Erk activities and influences thymocyte selection and activation. J Immunol. 2003;171(3):1285-1296.

29. Newton RH, Turka LA. Regulation of T cell homeostasis and responses by pten. Front Immunol. 2012;3:151.

30. Maltzman JS, Kovoor L, Clements JL, Koretzky GA. Conditional deletion reveals a cell-autonomous requirement of SLP-76 for thymocyte selection. J Exp Med. 2005;202(7):893-900.

31. Mombaerts P, Clarke AR, Rudnicki MA, et al. Mutations in T-cell antigen receptor genes alpha and beta block thymocyte development at different stages. Nature.

1992;360(6401):225-231.

32. Allman D, Karnell FG, Punt JA, et al.

Separation of Notch1 promoted lineage commitment and expansion/transformation in developing T cells. J Exp Med. 2001;194(1):99-106.

33. Mingueneau M, Kreslavsky T, Gray D, et al. The transcriptional landscape of alphabeta T cell differentiation. Nat Immunol.

2013;14(6):619-632.

34. Yashiro-Ohtani Y, He Y, Ohtani T, et al.

Pre-TCR signaling inactivates Notch1 transcription by antagonizing E2A. Genes Dev. 2009;23(14):1665-1676.

35. Spits H. Development of alphabeta T cells in the human thymus. Nat Rev Immunol.

2002;2(10):760-772.

36. Trinquand A, Tanguy-Schmidt A, Ben Abdelali R, et al.

Toward a NOTCH1/FBXW7/RAS/PTEN-Based Oncogenetic Risk Classification of Adult TCell Acute Lymphoblastic Leukemia: A Group for Research in Adult Acute Lymphoblastic Leukemia Study. J Clin Oncol. 2013;31(34):4333-4342.

1007

ARTICLE

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

Leukemia reconstitution in vivo is driven by cells in early cell cycle and low metabolic state Luca Trentin,1* Manon Queudeville,1** Sarah Mirjam Eckhoff,1 Nabiul Hasan,1*** Vera Münch,1,2 Elena Boldrin,1,2 Felix Seyfried,1 Stefanie Enzenmüller,1 Klaus-Michael Debatin1# and Lüder Hinrich Meyer1# Department of Pediatrics and Adolescent Medicine, Ulm University Medical Center and International Graduate School in Molecular Medicine, Ulm University, Germany

1 2

Haematologica 2018 Volume 103(6):1008-1017

*Present address: Department of Woman and Child Health, University of Padova, Italy. **Present address: University Medical Center Tübingen, Department of Pediatrics and Adolescent Medicine, Tübingen, Germany. ***Present address: University of Pittsburgh, Department of Neurology, Pittsburgh, PA, USA. #KMD and LHM contributed equally to this work.

ABSTRACT

I

Correspondence: lueder-hinrich.meyer@uniklinik-ulm.de or klaus-michael.debatin@uniklinik-ulm.de; Received: February 24, 2017.

Accepted: March 1, 2018.

Pre-published: March 8, 2018.

n contrast to well-established hierarchical concepts of tumor stem cells, leukemia-initiating cells in B-cell precursor acute lymphoblastic leukemia have not yet been phenotypically identified. Different subpopulations, as defined by surface markers, have shown equal abilities to reconstitute leukemia upon transplantation into immunodeficient mice.

Using a non-obese diabetes/severe combined immunodeficiency human acute lymphoblastic leukemia mouse model and cell cycle analysis annotating cells to distinct cycle phases, we functionally characterized leukemia-initiating cells and found that cells in all stages of the cell cycle are able to reconstitute leukemia in vivo, with early cycling cells (G1blow population) exhibiting the highest leukemia-initiating potential.

Interestingly, cells of the G2/M compartment, i.e. dividing cells, were less effective in leukemia reconstitution. Moreover, G1blow cells were more resistant to spontaneous or drug-induced cell death in vitro, were enriched for stem cell signatures and were less metabolically active, as determined by lower levels of reactive oxygen species, compared to G2/M stage cells.

Our data provide new information on the biological properties of leukemia-initiating cells in B-cell precursor acute lymphoblastic leukemia and underline the concept of a stochastic model of leukemogenesis.

doi:10.3324/haematol.2017.167502

Introduction Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1008 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1008

According to a hierarchical stem cell model, continuous proliferation of progenitor and mature blood cells is sustained by hematopoietic stem cells (HSC) characterized by their ability to differentiate into all hematopoietic blood lineages while retaining self-renewal potential.1 In leukemia, the presence of a leukemia-initiating cell (LIC) was proven by studies showing that only CD34+/CD38– cells were able to transfer acute myeloid leukemia when transplanted into non-obese diabetes/severe combined immunodeficiency (NOD/SCID) mice, suggesting a hierarchical stem cell model in acute myeloid leukemia.2,3 Recent work challenged this view and showed that CD34–, CD34+/CD38+ and CD123+ cells were also able to reconstitute acute myeloid leukemia.4-8 In B-cell precursor acute lymphoblastic leukemia (BCP-ALL), several studies have addressed the identification of LIC, demonstrating that more or less immature cells expressing different surface markers were all able to reconstitute leukemia in immunodeficient mice, supporting a stochastic stem cell model.9-13 Thus, the capacity to propagate BCP-ALL clearly appeared not to be restricted to a phenotypically defined subpopulation, but may rather be defined at a functional level. Previously, we identified two distinct engraftment phenotypes, i.e. "Time To Leukemia short" (TTLshort) and "Time To Leukemia long" (TTLlong), reflecting early and late engrafthaematologica | 2018; 103(6)

Leukemia initiating cells in ALL

ment of primary cells in a BCP-huALL model.14 These two phenotypes were strongly associated with patients’ outcome and the TTLshort/high-risk phenotype involved increased activation of the mammalian target or rapamycin (mTOR) pathway14,15 and deficient apoptosis signaling.16 In this study, we identified annotations to distinct cell cycle compartments as a biological discriminator identifying a cellular subfraction with higher LIC activity capable of driving leukemia reconstitution in a NOD/SCID/huALL mouse model.

Methods Twenty patient-derived xenograft samples established by transplantation of patients’ ALL cells into NOD/SCID mice (NOD.CB17-Prkdcscid/J, Charles River) as previously described14,16 were used in this study. The patients’ samples were obtained with informed consent in accordance with the institution’s (Ulm University) ethical review board; all animal experiments were approved by the appropriate authority (Regierungspräsidium Tübingen) and carried out following institutional and national guidelines on the care and use of laboratory animals. The characteristics of the patients and their malignancies are summarized in Online Supplementary Table S1.

For gene expression analysis, RNA was prepared from sorted cells and analyzed using Affymetrix Human Genome-U133 Plus 2.0 arrays. Gene-expression data were deposited in the National Center for Biotechnology Information Gene Expression Omnibus (GEO series accession #GSE71836). Cell cycle staining was performed as previously described17-19 labeling DNA/RNA with Hoechst (Molecular Probes, ThermoFisher Scientific) and pyronin Y (Polysciences, Hirschberg, Germany). Post-sorting analysis confirmed the positions of the sorted populations within the sorting gates. Viable cells (identified by trypan blue exclusion) were transplanted into NOD/SCID mice (105 viable cells/recipient); group sizes were chosen based on the availability of cells after sorting.

Secondary recipients were transplanted with unsorted cells (1x105) isolated from mice with full-blown leukemia, which had been transplanted with G1blow or G2/M sorted cells. For analysis of reactive oxygen species (ROS), cells were stained with CMH2DCFDA (Invitrogen/ThermoFisher Scientific) and analyzed by flow cytometry. DNA was labeled with Hoechst. Post-sorting analysis confirmed high/low ROS levels. Viable cells (identified by trypan blue exclusion) were transplanted (105 viable cells/recipient); cells sorted based on FSC-A/SSC-A and FSC-A/FSC-H gates were used as control. Drug sensitivities (cell death upon prednisolone or cytarabine) were investigated in sorted cellular subfractions.

Statistical analyses were carried out using the Mann-Whitney test, the unpaired t-test with Welch correction (two-tailed), the one sample t-test or the log-rank test (two-tailed) as indicated.

P≤0.05 was considered statistically significant.

Additional and detailed information on the methods used can be found in the Online Supplementary Data.

Results Leukemia-initiating cell potential in B-cell precursor acute lymphoblastic leukemia is associated with cell cycle activity We previously characterized the engraftment potential of primary BCP-ALL cells transplanted into NOD/SCID mice and found that a rapid engraftment and a short time to leukemia (TTLshort) are associated with poor patients’ outcome.14 The capacity of a cell to give rise to a phenotypically equal progeny in vivo is considered a stem cell’s hallmark feature.20 We evaluated LIC activities by transplanting decreasing numbers of four primograft samples (TTLshort, n=2; TTLlong, n=2) in limiting dilutions (105 to 101 cells) assessing leukemia engraftment 25 weeks after transplantation. Interestingly, higher LIC frequencies (ID04, LIC: 1/329, TTL: 9 weeks; ID05, LIC: 1/739, TTL: 10 weeks) were observed in TTLshort/poor prognosis

Table 1. High leukemia-initiating cell frequencies in TTLshort /poor prognosis acute lymphoblastic leukemia. Estimated leukemia-initiating cell (LIC) frequencies of 2 TTLshort and 2 TTLlong ALL samples. Limiting dilution analysis.

Sample

TTL

N. cells transplanted 5

ID04

short

ID05

short

ID12

long

ID11

long

10 104 103 102 101 105 104 103 102 101 105 104 103 102 101 105 104 103 102 101

Recipients transplanted

Recipients engrafted

8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8

8 8 7 4 0 8 8 6 1 0 8 8 3 0 0 6 1 0 0 0

LIC frequency

1/329

1/739

1/2159

1/74028

TTL: time to leukemia; N: number.

haematologica | 2018; 103(6)

1009

L. Trentin et al.

leukemias as compared to decreasing frequencies along with prolonged engraftment (ID012, LIC: 1/2159, TTL: 19 weeks ID11, LIC: 1/74028, TTL: 22 weeks) in TTLlong leukemias (Table 1). Accordingly, only TTLshort cells led to engraftment upon transplantation of 102 cells.

Next, we analyzed expression of the lineage and stem cell markers CD19, CD10, CD34 and CD38, previously described to be characteristic of cells with stem or initiating cell potential.5,9-12 Altogether, 50 patients’ ALL samples, which had been transplanted and characterized for their engraftment phenotype, were analyzed. No differences in marker expression were observed between the two phenotypes (Figure 1A); however, a trend of higher proportions of CD34+ cells in TTLlong/good prognosis samples was seen, in line with earlier reports.21,22 In order to look for stem cell features, which are different from expression of surface markers, we analyzed our previously obtained gene expression data14 using gene set enrichment analysis.

We identified 23 gene sets significantly enriched in the TTLshort/high risk profile (false discovery rate q-value ≤ 0.05), of which 17 were annotated to cell cycle functions, pointing to an association of cell cycle regulation with the TTL phenotype and, therefore, LIC activity in ALL (Figure 1B and Online Supplementary Table S2).

To further investigate these findings functionally, we analyzed the proportions of cells in active mitosis in all 20 samples (n=10 TTLshort and n=10 TTLlong), (Online Supplementary Table S1) by staining for phosphorylated histone H3 (Ser10). Significantly higher proportions of mitotic ALL cells were identified in TTLshort compared to TTLlong leukemias (Figure 2A), in line with our gene expression analysis results. Moreover, we investigated cellular proliferation of leukemia cells in vivo in one leukemia of each TTL phenotype. Dividing cells were marked with bromodeoxyuridine and huCD19/bromodeoxyuridinepositive cells were analyzed after labeling/pulse and dur-

A

ing follow up/chase. At the end of the labeling (day 0), significantly higher percentages of huCD19/bromodeoxyuridine-positive cells were detected in spleen and bone marrow of TTLshort mice than in TTLlong mice (Figure 2B).

Moreover, a clear reduction of bromodeoxyuridine positivity in human ALL cells was observed during chase in TTLshort in contrast to similar or slowly decreasing levels in TTLlong leukemias (Figure 2C). During the experiment, all animals showed similarly high leukemia loads (Figure 2D).

These findings indicate that the LIC frequency is related to a higher in vivo proliferation capacity. Moreover, despite variation in frequencies between different samples, we did not find that LIC in BCP-ALL are extremely rare, which further supports recent observations suggestive of a stochastic stem cell concept in ALL in which many cells possess leukemia-initiating potential.

Cells in early G1-S transition possess higher leukemia-initiating cell potential Since we found that differences in LIC frequencies and cell cycle progression are associated with distinct engraftment capabilities, we hypothesized that leukemia cells in different cell cycle phases are characterized by a specific repopulating potential. We used a cell cycle “live” staining with simultaneous staining of DNA and RNA17,19 distinguishing G0/G1, S and G2/M phases. In particular, cells in G0/G1 were further divided based on increasing RNA intensity, reflecting transition from G to S phases.23,24 Early cell cycle annotated cells (G0/G1) were subdivided into G1a, G1blow and G1bhigh fractions according to progressively increasing RNA fluorescence. A fourth gate was placed on cells in G2/M (Figure 3A). Staining of G1a/G1blow-sorted cells with the proliferation marker Ki-67 revealed that Ki-67-negative resting G0 cells are part of the G1a fraction in this analysis (Online Supplementary Figure S1).

All sorted subpopulations showed leukemia-initiating

B

Figure 1. Leukemia engraftment is associated with cell cycle activity. (A) No difference in surface marker expression (CD34, CD10 and CD19 n=50; CD38 n=40) as measured by flow cytometry on primary patients’ ALL cells with either TTLshort or TTLlong phenotype. Single and median values are indicated. Mann-Whitney test; P= statistical significance. (B) Representative gene set enrichment analysis plots of cell cycle annotated gene sets in the TTLshort/high LIC activity profile False Discovery Rate (FDR) (q-value ≤0.05).

1010

haematologica | 2018; 103(6)

Leukemia initiating cells in ALL

activity upon transplantation into subsequent mice, irrespective of the originating cell cycle compartment and TTL phenotype (Online Supplementary Table S3), leading to full-blown leukemia of the initial common-ALL immunophenotype. Interestingly, of the four cell cycle fractions transplanted, G1blow cells showed quickest engraftment and were associated with the shortest leukemia-free survival. This feature was observed in samples with both short and long TTL phenotypes, suggesting that the high leukemia repopulating and initiating capacity of G1blow cells is a general feature of this early cell cycle leukemia cell subfraction. Importantly, despite these distinct repopulating activities of ALL cells from different cell cycle subgroups, the overall short or long leukemia engraftment of unfractionated leukemia cells was recapitulated in the sorted subfractions. Moreover, G2/M cells were always the last to engraft, leading to longer leukemia free survival of the recipient animals (Figure 3B). In addition, no differences in the expression of lineage/stem cell markers huCD19, huCD38, huCD10 or huCD34 were observed on ALL cells of either G1blow or G2/M cell fractions (Online Supplementary Figure S2). Most interestingly, the distinct engraftment of G1blow versus G2/M cells was also retained upon secondary transplantation of G1blowand G2/M-derived, unsorted bulk leukemia cells, indicating maintenance of LIC-capacities in the bulk of G1blowand G2/M-derived cells (Figure 3C and Online Supplementary Table S3).

A

G1blow acute lymphoblastic leukemia cells are characterized by a stem cell signature To further corroborate our observation of higher LIC activity in G1blow leukemia cells as suggested by our functional in vivo data, we investigated transcriptional signatures of sorted G1blow and G2/M leukemia cells (pairs of 4 samples; TTLshort= 1; TTLlong= 3). We identified 865 genes (1122 probe sets, false discovery rate q-value <0.05) as being differentially regulated between G1blow and G2/M subpopulations irrespective of their TTL phenotype, with 330 up- and 535 down-regulated genes in G1blow, which were mainly attributed to cell cycle regulation (Figure 4A and Online Supplementary Tables S4 and S5). Interestingly, gene set enrichment analysis identified a positive enrichment of 4 out of 16 gene sets previously associated with stem cell activity with the G1blow/high LIC-enriched cells (Figure 4B and Online Supplementary Table S6A), whereas genes sets from short-term HSC or mature cells were found to be positively enriched in G2/M/low LIC8,25-29 (Figure 4B and Online Supplementary Table S6B). Thus, these gene expression data further support the higher LIC activity of G1blow cells observed upon transplantation.

Different cell death sensitivities of cell cycle annotated acute lymphoblastic leukemia cells Cancer initiating or stem cells have been described to be characterized by an increased resistance to anti-tumor therapies. To investigate this issue, we studied the intrin-

B

C D

Figure 2. High leukemia-initiating cell activity is associated with increased cell cycle activity. (A) Higher phosphorylated histone 3 (P-H3; Ser10)-positive cells in active mitosis in TTLshort (n=10) as compared to TTLlong leukemia samples (n=10), Mann-Whitney U-test; the line represents the median; P=statistical significance. (B and C) Higher bromodeoxyuridine (BrdU) uptake (B) and increased decline (C) after in vivo labeling as detected by flow cytometry of ALL cells in TTLshort/high LIC frequency compared to TTLlong/low LIC frequency ALL bearing recipients (n=3/time point; biological replicates). Percentages of huCD19+/BrdU+ cells in bone marrow (BM) and spleen of ALL bearing recipients (mean ±SD). Unpaired t-test with Welch correction (two-tailed); P= statistical significance; *≤0.05; n.s.: not significant. (D) Similar high leukemia load in recipients used for in vivo proliferation analysis; percentages of huCD19+ ALL cells in spleen and BM over time in recipients (n=3 per group; biological replicates) bearing a TTLshort or TTLlong leukemia (mean ± Standard Deviation).

haematologica | 2018; 103(6)

1011

L. Trentin et al.

A

B

C

Figure 3. Higher leukemia-initiating cell potential of cells in early G1-S transition. (A) Cell cycle “live” staining: simultaneous DNA (Hoechst 33342) and RNA (pyronin Y) labeling of BCP-ALL cells and cell cycle annotated cellular subfractions G1a, G1blow, G1bhigh and G2/M analyzed by FACS. Sorted fractions were transplanted into NOD/SCID mice (105 cells/mouse) and leukemia engraftment was analyzed as weeks from transplantation until appearance of ≥1% huCD19+ ALL cells in peripheral blood of the recipients. (B) Increased engraftment activity upon transplantation of G1blow (red) compared to G2/M (blue) cells; (i, ii) short time to leukemia (TTLshort) ALL, two independent experiments, n=3 and n=4 mice/group; (iii) long time to leukemia (TTLlong) ALL, n=3 mice/group. (C) Maintained distinct engraftment in secondary recipients transplanted with unsorted bulk primograft ALL derived from primary recipients transplanted with G1blow or G2/M sorted fractions, n=3 and n=2 mice/group, respectively. Log-rank test; P=statistical significance.

A

B

Figure 4. G1blow acute lymphoblastic leukemia cells are characterized by a transcriptional stem cell program. (A) Unsupervised cluster analysis of 865 genes (1122 probe sets) differentially regulated (green: down-regulated; red: up-regulated) between G1blow and G2/M sorted cellular subfractions False discovery rate (FDR) qval<0.05 (B) Positive (left) enrichment of gene sets attributed to stemness or negative enrichment of sets annotated to mature or short-term stem cells (right) with the G1blow-profile (gene set enrichment analysis, NOM P-value ≤0.05 and FDR q-value ≤0.05).

1012

haematologica | 2018; 103(6)

Leukemia initiating cells in ALL

sic propensity of primograft ALL cells of G1blow and G2/M subfractions to undergo cell death, both upon ex vivo culture and in response to prednisone and cytarabine, two drugs used in protocols to treat pediatric ALL patients.

Interestingly, all 4 primograft samples analyzed showed lower rates of spontaneous and drug-induced cell death in G1blow as compared to G2/M cells (Figure 5A and B).

Leukemia-initiating activity in acute lymphoblastic leukemia is characterized by distinct cellular oxidative states Recently, low levels of ROS were described in cellular subfractions associated with stem cell properties.30-32 We, therefore, investigated ROS activity in xenograft samples of TTLshort or TTLlong phenotypes. Lower ROS activities were observed in rapidly engrafting, TTLshort/poor prognosis leukemia cells (Figure 6A) and in G1blow-sorted cells compared to G2/M-sorted cells (Figure 6B).

We observed that cells with low ROS activity (ROSlow) were almost exclusively allocated to G0/G1 cell cycle phases, whereas cells with high levels of ROS activity (ROShigh) included those in later S and G2/M phases (Figure 7A and B), suggesting that the ALL cell’s oxidative state is indicative of its leukemia-initiating activity. To further address this hypothesis on a functional level, we investigated 3 primograft ALL samples and sorted cellular subfractions according to high or low ROS levels (upper or lower 15% fluorescence intensity, ROShigh and ROSlow

cells, respectively). Upon transplantation into recipient animals, both sorted subfractions led to leukemia engraftment. However, in all three leukemias, ROSlow cells displayed a higher repopulating activity and were associated with significantly shorter leukemia-free survival in contrast to prolonged engraftment and survival in mice transplanted with ALL cells with a high oxidative state (ROShigh) (Figure 7C). In HSC, a ROS-MAP kinase axis has been implicated in negative regulation of the life span of the cells.30,33 A high LIC potential of the ROSlow subtype was also found in T-ALL and functionally linked to expression of PKC-θ as a consequence of deregulated NOTCH signaling.34 However, the analysis of MAP kinase p38α/β-expression did not reveal significant differences between the engraftment phenotypes (Online Supplementary Figure S3A and B) and, in contrast to the TALL data reported, we did not detect PKC-θ expression in our BCP-ALL cells (Online Supplementary Figure S3C).

Taken together, we observed that LIC are not rare in BCP-ALL, pointing to a stochastic stem cell concept in this type of leukemia. We identified distinct LIC activities in cell cycle annotated cellular subfractions, with early cycling (G1blow) cells possessing the highest LIC potential independently of the overall engraftment phenotype and high/low relapse risk. Moreover, early cycling (G1blow) cells with high LIC potential are characterized by a transcriptional stem cell profile, cell death resistance, and a low oxidative state, which in turn results in higher LIC

A

B

Figure 5. G1blow acute lymphoblastic leukemia (ALL) cells are characterized by cell death insensitivity. (A and B) Low sensitivity for G1blow sorted ALL cells for spontaneous (A) and drug induced (B) cell death measured by flow cytometry. Cell death upon culture or exposure (24 h) to prednisolone (PRED, 250, 500 or 1000 mg/mL) or arabinosylcytosine (ARA-C, 10, 100, 500 mg/mL) in cell cycle annotated, sorted subfractions, (mean ± Standard Deviation). Data points represent technical replicates within 4 individual leukemia samples, unpaired t-test with Welch correction; *P≤ 0.05; **P≤ 0.01; ***P≤ 0.001; n.s.: not significant.

haematologica | 2018; 103(6)

1013

L. Trentin et al.

A

B

Figure 6. Reactive oxygen species (ROS) activity is associated with different cell cycle phases and cycling potential. (A) Low ROS activity in TTLshort/high proliferating (n=8) compared to TTLlong/slow proliferating (n=8) acute lymphoblastic leukemia samples: fold-change difference, (mean Âą Standard Deviation) measured by flow cytometry. (B) Low ROS activity in sorted G1blow annotated primograft leukemia cells compared to G2/M cells measured by flow cytometry. TTLshort: Time To Leukemia short; TTLlong: Time To Leukemia long; ctrl: control.

A

B

C

Figure 7. High leukemia initiating-cell activity in acute lymphoblastic leukemia (ALL) is associated with low reactive oxygen species (ROS) activity. (A) ALL cells with low ROS levels (ROSlow) are predominantly in early G0/G1 cell cycle phases, whereas cells with high ROS (ROShigh) include later S-G2/M phases. Simultaneous analysis of ROS activity and cell cycle distribution; cell cycle analysis in gated subpopulations of low or high (lower or upper 15%) ROS levels; one representative example of six analyses is shown. (B) Cell cycle distribution according to flow cytometry analysis in ROSlow and ROShigh subpopulations of 6 primograft ALL samples. (C) Increased engraftment activity of ROSlow ALL cells. Sorted ROShigh or ROSlow subfractions were transplanted (105 cells/mouse) and leukemia engraftment was analyzed as weeks from transplantation until appearance of â&#x2030;Ľ1% huCD19+ ALL cells in peripheral blood of the recipients. N=4 (ID03, ID04), n=6 (ID06) mice/group, log-rank test; P= statistical significance.

1014

haematologica | 2018; 103(6)

Leukemia initiating cells in ALL

activity in functional repopulation experiments, opening up ROS modulation as a perspective for treatment.

Discussion In this study, we used engraftment phenotypes identified in our NOD/SCID/huALL model to investigate characteristic features of LIC in BCP-ALL. First, we did not obtain any evidence of preferential accumulation of LIC in subpopulations defined by surface markers. Since highrisk leukemia was characterized by higher LIC frequencies and higher cell cycle progression in vivo, we analyzed leukemia-initiating capacities of cells in different cell cycle compartments. Using a vital DNA/RNA staining method, we found that: (i) all cells possess LIC potential in vivo independently of the cell cycle phase of origin (i.e. G1a, G1blow, G1bhigh and G2/M); (ii) early G1b cells entering into S phase (G1blow) lead to the most rapid engraftment, suggesting that G1blow cells are the first engrafting leukemiainitiating cells; (iii) G2/M cells possess significantly lower engraftment potential; (iv) these features were maintained independently of the leukemia cell engraftment phenotypes; and (v) G1blow LICs exhibit low metabolic activity and ROS potential, probably associated with increased cell death resistance. These findings have implications for the characterization of the putative leukemic stem cell in ALL, the heterogeneity of leukemic clones and sensitivity and/or resistance of functionally defined subpopulations for treatment approaches using targeted or conventional therapy.

Two models have been proposed for cancer stem cells or LIC. According to the hierarchical concept, a few immature cells harboring stem cell properties are considered to be able to generate their progeny and give rise to leukemia. While a number of data have supported this hierarchical model in acute myeloid leukemia,2,3 for ALL and particularly for BCP-ALL a more stochastic model, in which literally every cell, including more mature and differentiated phenotypes, is considered to be able to initiate leukemia, appears to be valid. This implies that the establishment of the leukemia phenotype in patients and upon transplantation in mice may be an intrinsic capacity of individual cells independently of surface phenotype and maturation stage.9-13 This concept corresponds to the findings of leukemia initiation by low cell numbers and by all subfractions seen in our huALL mouse model. Another issue along this line is whether or not LIC are derived from quiescent, cycling or dividing subpopulations. The data from our in vivo labeling experiments indicate that increased proliferation is associated with rapid engraftment and rapid development of full-blown leukemia resulting in the TTLshort phenotype defining a poor prognosis subgroup. Thus, one may have expected that actually dividing cells (G2/M) harbor the highest leukemogenic potential upon transplantation. However, the resting/early recruitment compartment (G0, G1) contained similar reconstitution potential and the early recruitment phenotype (G1blow) exhibited the highest leukemogenic potential in the transplantation experiments.

The potential of both G0/G1 and G2/M cells to reconstitute the leukemia phenotype in vivo is a new finding compared to previous data on human and murine HSC. Indeed, human and murine HSC have been reported to be heterogeneous with respect to the cell cycle: while only haematologica | 2018; 103(6)

cells in G0/G1 sufficiently reconstituted hematopoiesis in sublethally irradiated mice,18,35-38 cells in S-G2/M did not repopulate the bone marrow at all or had only a minimal engraftment potential.35,39 The higher in vivo leukemogenic activity of G1blow cells suggests that these cells are likely the first out of the leukemic bulk to engraft in recipients. The higher â&#x20AC;&#x153;stem cellnessâ&#x20AC;? of this subpopulation is also supported by gene signatures previously assigned to stem cell activity. In contrast, G2/M cells with lower LIC potential were negatively associated with stem cell-like profiles or were enriched for genes characteristic of short-term HSC or mature cells.8,2527,29 Importantly, the G1blow and G2/M features are maintained irrespective of the engraftment phenotype suggesting that these characteristics are conserved features of BCP-ALL LIC. However, the engraftment potential of both subpopulations, irrespective of stem cell signatures, emphasizes the stochastic nature of LIC in ALL, in line with data recently reported on similar engraftment activities of slowly or rapidly dividing BCP-ALL cells.40 LIC frequencies calculated in our NOD/SCID/huALL model showed variations associated with the patientsâ&#x20AC;&#x2122; outcome and engraftment phenotype, suggesting that speed of ALL repopulation is a measure of LIC activity, as suggested before.41 Thus, we analyzed repopulation times to evaluate LIC activities of sorted cells. In both TTLshort/poor prognosis ALL samples, engraftment was observed upon transplantation of down to 100 cells, in line with reported high LIC frequencies in studies including poor outcome BCP-ALL, even in the more immunodeficient NSG mouse strain.9,40,42 However, higher minimum cell numbers of up to 103 cells were required to initiate leukemia in both TTLlong samples, of which one also showed hyperdiploidy, similar to numbers observed in studies including favorable prognosis ALL.10-13,43,44 Accordingly, TTLlong phenotypes were always observed in ALL with the favorable prognostic features hyperdiploidy or ETV6/RUNX1 rearrangements,14 suggesting lower LIC frequencies in good outcome BCP-ALL. In TTLshort leukemia, gene expression, more cells in active mitosis and the in vivo bromodeoxyuridine labeling data indicate a higher proliferation rate, including activated mTOR signaling.15 Along this line, effects of the mTOR pathway on cell cycle progression and particularly regulation of the G1 phase have already been described.45,46 Accordingly, in addition to a distinctive transcriptional program,14 different basal mTOR activation15 and deficient apoptosis signaling16 were found in the TTLshort versus TTLlong phenotypes.

The higher LIC capacity of G1blow appears to result from a favorable functional status. G1blow cells were less prone to undergo spontaneous and drug-induced cell death ex vivo whereas G2/M slowly engrafting cells showed a greater predisposition to both intrinsic and induced cell death. Increased intrinsic cell death resistance may be a consequence of a block in cell death pathways or alteration of the metabolic state. Recent work showed that different levels of ROS and a lower mitochondrial mass distinguished cells with higher LIC activity.31,32 We found that G1blow cells were characterized by lower ROS activity compared to G2/M cells. Moreover, when analyzing ROS levels in cell cycle compartments, we observed that ROSlow cells were almost exclusively found in G0/G1, while ROShigh cells were progressing not only through the G0/G1 but also the S and G2/M phases of the cell cycle.

1015

L. Trentin et al.

Interestingly, ROSlow cells possessed an engraftment advantage with the transplanted mice having a shorter leukemia-free survival. Therefore, as already reported for HSC,30,33 our data suggest that increased ROS levels are associated with reduced repopulating stem cell activity in BCP-ALL. Along this line, it is interesting to note that higher endogenous ROS levels were detected in TTLlong/good prognosis samples. While an increased ROS level in HSC was directly linked to MAP kinase activation, we could not detect differential expression of p38α/β in the TTLlong/TTLshort subgroups characterized by different engraftment properties and ROS levels. In T-ALL, a different pathway for modulating the ROS status involving PKC-θ was described. However, we could not detect PKC expression at all in our leukemia cells. Thus, the molecular mechanism and/or association of specific pathways with different ROS states is unclear, but may be linked to different activation of survival pathways, as indicated by the gene profile.

In addition to cell intrinsic mechanisms, neighboring cells in the bone marrow environment interact with residing hematopoietic or leukemia cells and contribute to regulation of cellular programs including proliferation and cell cycle,47 and homing and interaction of leukemia cells with the environment are modulated by the expression of adhesion molecules.48 Topographically, in ALL a rare subfraction of cells was described to preferentially reside close to the endosteum,40 the assumed site of the hematopoietic niche in the bone marrow.49 Similar to our findings, these cells were characterized by low proliferation and insensitivity to chemotherapy, but did not show higher LIC activity compared to the corresponding nondormant bulk leukemia cells.40 It appears very likely, that the ability of a cell to initiate leukemia is determined by different factors including niche interactions contributing to the phenotype of early cycling cells with low metabolic activity identified in our study. Interestingly, recently presented data showed that HSC are able to transfer mitochondria into adjacent stromal cells, thereby lower-

References 1. Bryder D, Rossi DJ, Weissman IL. Hematopoietic stem cells: the paradigmatic tissue-specific stem cell. Am J Pathol.

2006;169(2):338-346.

2. Lapidot T, Sirard C, Vormoor J, et al. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice.

Nature. 1994;367(6464):645-648.

3. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997;3(7):730-737.

4. Sarry JE, Murphy K, Perry R, et al. Human acute myelogenous leukemia stem cells are rare and heterogeneous when assayed in NOD/SCID/IL2Rgammac-deficient mice. J Clin Invest. 2011;121(1):384-395.

5. Taussig DC, Miraki-Moud F, Anjos-Afonso F, et al. Anti-CD38 antibody-mediated clearance of human repopulating cells masks the heterogeneity of leukemia-initiating cells.

Blood. 2008;112(3):568-575.

6. Taussig DC, Vargaftig J, Miraki-Moud F, et

1016

7.

8.

9.

10.

11.

ing their intracellular ROS levels.50 High mTOR expression,15 lower cellular propensity to undergo apoptosis,16 and the higher cycling capacity are cell intrinsic characteristics leading to the high proliferative potential of leukemia cells in poor prognosis patients, independently of the capacity of all leukemia cells to engraft in mice as described in this work, and the even higher LIC potential distinguishing G1blow leukemia cells.

Our work adds new information in the search for LIC in BCP-ALL, at least in experimental models of human BCP-ALL. The ability to initiate leukemia seems not to be associated with particular subtypes of the heterogeneous leukemia subfractions and also appears to be independent of the cell cycle compartment from which the leukemogenic cells are transplanted. This argues strongly against the concept that leukemogenesis is restricted to quiescent clones and absent or reduced in cycling cells or vice versa. Thus, this finding has consequences for therapeutic strategies, which aim to target exclusively quiescent or dividing cells. Interestingly however, cells in quiescence or the early phase of the cell cycle appear to be more resistant to cell death induction than cells in other compartments of the cell cycle. The high LIC potential and low metabolic activity associated with cell death resistance may also indicate that strategies to activate these cells to exit quiescence or the early phase compartment may decrease intrinsic cell death resistance with resulting consequences for treatment sensitivity.

Acknowledgments The authors would like to thank Sandra Lutz and Manuel Herrmann for excellent technical assistance and the Sorting and Chip Core Facilities, Ulm University. This work was supported by the German Research Foundation (KFO 167, K-MD and MQ; SFB1074 B6, LHM and K-MD), the Helmholtz Association/Preclinical Comprehensive Cancer Center (K-MD), the International Graduate School in Molecular Medicine Ulm (MNH, VM and EB), and the “Förderkreis für Tumor- und Leukämiekranke Kinder Ulm”.

al. Leukemia-initiating cells from some acute myeloid leukemia patients with mutated nucleophosmin reside in the CD34(-) fraction. Blood. 2010;115(10):1976-1984.

Jin L, Lee EM, Ramshaw HS, et al.

Monoclonal antibody-mediated targeting of CD123, IL-3 receptor alpha chain, eliminates human acute myeloid leukemic stem cells.

Cell Stem Cell. 2009;5(1):31-42.

Eppert K, Takenaka K, Lechman ER, et al.

Stem cell gene expression programs influence clinical outcome in human leukemia.

Nat Med. 2011;17(9):1086-1093.

Rehe K, Wilson K, Bomken S, et al. Acute B lymphoblastic leukaemia-propagating cells are present at high frequency in diverse lymphoblast populations. EMBO Mol Med. 2012;5(1):38-51.

Cobaleda C, Gutierrez-Cianca N, PerezLosada J, et al. A primitive hematopoietic cell is the target for the leukemic transformation in human Philadelphia-positive acute lymphoblastic leukemia. Blood. 2000;95(3): 1007-1013.

Hong D, Gupta R, Ancliff P, et al. Initiating and cancer-propagating cells in TEL-AML1-

12.

13.

14.

15.

16.

associated childhood leukemia. Science.

2008;319(5861):336-339. le Viseur C, Hotfilder M, Bomken S, et al. In childhood acute lymphoblastic leukemia, blasts at different stages of immunophenotypic maturation have stem cell properties.

Cancer Cell. 2008;14(1):47-58.

Cox CV, Diamanti P, Evely RS, Kearns PR, Blair A. Expression of CD133 on leukemiainitiating cells in childhood ALL. Blood.

2009;113(14):3287-3296.

Meyer LH, Eckhoff SM, Queudeville M, et al. Early relapse in ALL is identified by time to leukemia in NOD/SCID mice and is characterized by a gene signature involving survival pathways. Cancer Cell. 2011;19(2): 206-217.

Hasan MN, Queudeville M, Trentin L, et al.

Targeting of hyperactivated mTOR signaling in high-risk acute lymphoblastic leukemia in a pre-clinical model.

Oncotarget. 2015;6(3):1382-1395.

Queudeville M, Seyfried F, Eckhoff SM, et al.

Rapid engraftment of human ALL in NOD/SCID mice involves deficient apoptosis signaling. Cell Death Dis. 2012;3(e364.

haematologica | 2018; 103(6)

Leukemia initiating cells in ALL

17. Shapiro HM. Flow cytometric estimation of DNA and RNA content in intact cells stained with Hoechst 33342 and pyronin Y. Cytometry. 1981;2(3):143-150.

18. Wilpshaar J, Falkenburg JH, Tong X, et al.

Similar repopulating capacity of mitotically active and resting umbilical cord blood CD34(+) cells in NOD/SCID mice. Blood.

2000;96(6):2100-2107.

19. Darzynkiewicz Z, Juan G, Srour EF. Differential staining of DNA and RNA. Curr Protoc Cytom. 2004;Chapter 7:Unit 7.3.

20. Dick JE, Bhatia M, Gan O, Kapp U, Wang JC. Assay of human stem cells by repopulation of NOD/SCID mice. Stem Cells. 1997;15 (Suppl 1):199-203; discussion 204-197.

21. Borowitz MJ, Shuster JJ, Civin CI, et al.

Prognostic significance of CD34 expression in childhood B-precursor acute lymphocytic leukemia: a Pediatric Oncology Group study. J Clin Oncol. 1990;8(8):1389-1398.

22. Pui CH, Hancock ML, Head DR, et al.

Clinical significance of CD34 expression in childhood acute lymphoblastic leukemia.

Blood. 1993;82(3):889-894.

23. Darzynkiewicz Z, Sharpless T, StaianoCoico L, Melamed MR. Subcompartments of the G1 phase of cell cycle detected by flow cytometry. Proc Natl Acad Sci USA. 1980;77(11):6696-6699.

24. Darzynkiewicz Z, Traganos F, Melamed MR. New cell cycle compartments identified by multiparameter flow cytometry.

Cytometry. 1980;1(2):98-108.

25. Ivanova NB, Dimos JT, Schaniel C, Hackney JA, Moore KA, Lemischka IR. A stem cell molecular signature. Science. 2002;298 (5593):601-604.

26. Georgantas RW 3rd, Tanadve V, Malehorn M, et al. Microarray and serial analysis of gene expression analyses identify known and novel transcripts overexpressed in hematopoietic stem cells. Cancer Res. 2004;64(13):4434-4441.

27. Gal H, Amariglio N, Trakhtenbrot L, et al.

Gene expression profiles of AML derived stem cells; similarity to hematopoietic stem cells. Leukemia. 2006;20(12):2147-2154.

28. Jaatinen T, Hemmoranta H, Hautaniemi S, et al. Global gene expression profile of human cord blood-derived CD133+ cells.

Stem Cells. 2006;24(3):631-641.

haematologica | 2018; 103(6)

29. Gentles AJ, Plevritis SK, Majeti R, Alizadeh AA. Association of a leukemic stem cell gene expression signature with clinical outcomes in acute myeloid leukemia. JAMA. 2010;304(24):2706-2715.

30. Ito K, Hirao A, Arai F, et al. Reactive oxygen species act through p38 MAPK to limit the lifespan of hematopoietic stem cells. Nat Med. 2006;12(4):446-451.

31. Lagadinou ED, Sach A, Callahan K, et al.

BCL-2 inhibition targets oxidative phosphorylation and selectively eradicates quiescent human leukemia stem cells. Cell Stem Cell.

2013;12(3):329-341.

32. Romero-Moya D, Bueno C, Montes R, et al.

Cord blood-derived CD34+ hematopoietic cells with low mitochondrial mass are enriched in hematopoietic repopulating stem cell function. Haematologica.

2013;98(7):1022-1029.

33. Ito K, Hirao A, Arai F, et al. Regulation of oxidative stress by ATM is required for selfrenewal of haematopoietic stem cells.

Nature. 2004;431(7011):997-1002.

34. Giambra V, Jenkins CR, Wang H, et al.

NOTCH1 promotes T cell leukemia-initiating activity by RUNX-mediated regulation of PKC-theta and reactive oxygen species.

Nat Med. 2012;18(11):1693-1698.

35. Fleming WH, Alpern EJ, Uchida N, Ikuta K, Spangrude GJ, Weissman IL. Functional heterogeneity is associated with the cell cycle status of murine hematopoietic stem cells. J Cell Biol. 1993;122(4):897-902.

36. Morrison SJ, Weissman IL. The long-term repopulating subset of hematopoietic stem cells is deterministic and isolatable by phenotype. Immunity. 1994;1(8):661-673.

37. Gothot A, Pyatt R, McMahel J, Rice S, Srour EF. Functional heterogeneity of human CD34(+) cells isolated in subcompartments of the G0 /G1 phase of the cell cycle. Blood.

1997;90(11):4384-4393.

38. Gothot A, van der Loo JC, Clapp DW, Srour EF. Cell cycle-related changes in repopulating capacity of human mobilized peripheral blood CD34(+) cells in non-obese diabetic/severe combined immune-deficient mice. Blood. 1998;92(8):2641-2649.

39. Glimm H, Oh IH, Eaves CJ. Human hematopoietic stem cells stimulated to proliferate in vitro lose engraftment potential

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

during their S/G(2)/M transit and do not reenter G(0). Blood. 2000;96(13):4185-4193.

Ebinger S, Ozdemir EZ, Ziegenhain C, et al.

Characterization of rare, dormant, and therapy-resistant cells in acute lymphoblastic leukemia. Cancer Cell. 2016;30(6):849-862.

Greaves M. Cancer stem cells renew their impact. Nat Med. 2011;17(9):1046-1048.

Morisot S, Wayne AS, Bohana-Kashtan O, et al. High frequencies of leukemia stem cells in poor-outcome childhood precursor-B acute lymphoblastic leukemias. Leukemia.

2010;24(11):1859-1866.

Cox CV, Evely RS, Oakhill A, Pamphilon DH, Goulden NJ, Blair A. Characterization of acute lymphoblastic leukemia progenitor cells. Blood. 2004;104(9):2919-2925.

Castor A, Nilsson L, Astrand-Grundstrom I, et al. Distinct patterns of hematopoietic stem cell involvement in acute lymphoblastic leukemia. Nat Med. 2005;11(6):630-637.

Fingar DC, Richardson CJ, Tee AR, Cheatham L, Tsou C, Blenis J. mTOR controls cell cycle progression through its cell growth effectors S6K1 and 4E-BP1/eukaryotic translation initiation factor 4E. Mol Cell Biol. 2004;24(1):200-216.

Gao N, Flynn DC, Zhang Z, et al. G1 cell cycle progression and the expression of G1 cyclins are regulated by PI3K/AKT/mTOR/ p70S6K1 signaling in human ovarian cancer cells. Am J Physiol Cell Physiol. 2004;287(2): C281-291.

Walkley CR, Shea JM, Sims NA, Purton LE, Orkin SH. Rb regulates interactions between hematopoietic stem cells and their bone marrow microenvironment.

Cell.

2007;129(6):1081-1095.

Churchman ML, Evans K, Richmond J, et al.

Synergism of FAK and tyrosine kinase inhibition in Ph+ B-ALL. JCI. 2016;1(4): Morrison SJ, Spradling AC. Stem cells and niches: mechanisms that promote stem cell maintenance throughout life. Cell. 2008;132 (4):598-611.

Golan K, Wellendorf A, Takihara Y, et al.

Mitochondria transfer from hematopoietic stem and progenitor cells to Pdgfr +/Sca-1/CD48dim BM stromal cells via CX43 gap junctions and AMPK signaling inversely regulate ROS generation in both cell populations. Blood. 2016;128(22):5

1017

ARTICLE

Non-Hodgkin Lymphoma

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):1018-1028

A distinct subtype of Epstein-Barr virus-positive T/NK-cell lymphoproliferative disorder: adult patients with chronic active Epstein-Barr virus infection-like features Keisuke Kawamoto,1,2 Hiroaki Miyoshi,1 Takaharu Suzuki,1,2 Yasuji Kozai,3 Koji Kato,4 Masaharu Miyahara,5 Toshiaki Yujiri,6 Ilseung Choi,7 Katsumichi Fujimaki,8 Tsuyoshi Muta,9 Masaaki Kume,10 Sayaka Moriguchi,11 Shinobu Tamura,12 Takeharu Kato,13 Hiroyuki Tagawa,14 Junya Makiyama,15 Yuji Kanisawa,16 Yuya Sasaki,1 Daisuke Kurita,1 Kyohei Yamada,1 Joji Shimono,1 Hirohito Sone,2 Jun Takizawa,2 Masao Seto,1 Hiroshi Kimura17* and Koichi Ohshima1*

Department of Pathology, Kurume University School of Medicine; 2Department of Hematology, Endocrinology and Metabolism, Faculty of Medicine, Niigata University; 3 Department of Hematology, Tokyo Metropolitan Tama Medical Center, Fuchu; 4Department of Medicine and Biosystemic Science, Kyushu University Graduate School of Medical Sciences, Fukuoka; 5Department of Hematology, Karatsu Red Cross Hospital; 6Third Department of Internal Medicine, Yamaguchi University School of Medicine, Ube; 7 Department of Hematology, National Hospital Organization Kyushu Cancer Center, Fukuoka; 8 Department of Hematology, Fujisawa City Hospital; 9Department of Hematology, Japan Community Health Care Organization Kyushu Hospital; 10Department of Hematology, Hiraka General Hospital, Yokote; 11Department of Pathology, Miyazaki University; 12Department of Hematology and Oncology, Wakayama Medical University; 13Department of Hematology, Sasebo City General Hospital; 14Department of Hematology, Nephrology, and Rheumatology, Akita University Graduate School of Medicine; 15Department of Hematology, National Hospital Organization Nagasaki Medical Center, Omura; 16Department of Hematology and Oncology, Oji General Hospital, Tomakomai and 17Department of Virology, Faculty of Medicine, Nagoya University, Japan 1

*HK and KO share senior authorship.

Correspondence: miyoshi_hiroaki@med.kurume-u.ac.jp

Received: June 6, 2017.

Accepted: December 6, 2017.

Pre-published: December 14, 2017.

doi:10.3324/haematol.2017.174177 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1018 Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1018

ABSTRACT

T

he characteristics of adult patients with chronic active EpsteinBarr virus infection are poorly recognized, hindering early diagnosis and an improved prognosis. We studied 54 patients with adult-onset chronic active Epstein-Barr virus infection diagnosed between 2005 and 2015. Adult onset was defined as an estimated age of onset of 15 years or older. To characterize the clinical features of these adults, we compared them to those of 75 pediatric cases (estimated age of onset <15 years). We compared the prognosis of adultonset chronic active Epstein-Barr virus infection with that of patients with nasal-type (n=37) and non-nasal-type (n=45) extranodal NK/Tcell lymphoma. The median estimated age of onset of these lymphomas was 39 years (range, 16â&#x20AC;&#x201C;86 years). Compared to patients with pediatric-onset disease, those in whom the chronic active Epstein-Barr virus infection developed in adulthood had a significantly decreased incidence of fever (P=0.005), but greater frequency of skin lesions (P<0.001). Moreover, hypersensitivity to mosquito bites and the occurrence of hydroa vacciniforme were less frequent in patients with adultonset disease (P<0.001 and P=0.0238, respectively).

Thrombocytopenia, high Epstein-Barr virus nuclear antigen antibody titer, and the presence of hemophagocytic syndrome were associated with a poor prognosis (P=0.0087, P=0.0236, and P=0.0149, respectively). Allogeneic hematopoietic stem cell transplantation may improve survival (P=0.0289). Compared to pediatric-onset chronic active Epstein-Barr virus infection and extranodal NK/T-cell lymphoma, haematologica | 2018; 103(6)

Adult patients with chronic active EBV-like features

adult-onset chronic active Epstein-Barr virus infection had a poorer prognosis (P<0.001 and P=0.0484, respectively). Chronic active Epstein-Barr virus infection can develop in a wide age range, with clinical differences between adult-onset and pediatric-onset disease. Adult-onset chronic active Epstein-Barr virus infection is a disease with a poor prognosis. Further research will be needed.

Introduction Epstein-Barr virus (EBV) generally infects almost all people by early adulthood. Although EBV infection in childhood is asymptomatic in most people, some develop infectious mononucleosis (the so-called "kissing disease"). Almost all of them recover spontaneously after EBV-specific immunity is established.1 In addition, EBV has been reported to be involved in transformation from reactive to neoplastic or abnormal proliferation in some immunocompetent hosts after it infects B cells, T cells, and natural killer (NK) cells. This results in a wide range of conditions known as EBV-associated lymphoproliferative diseases (EBV-LPD),1,2 including chronic active Epstein-Barr virus infection (CAEBV), which produces infectious mononucleosis-like symptoms such as chronic persistent or recurrent fever, lymphadenopathy, skin rash, liver dysfunction, and hepatosplenomegaly, and has been reported to have a high mortality rate.3-5 In CAEBV, EBV-infected T cells and NK cells play important roles. Hypersensitivity to mosquito bites, the presence of hydroa vacciniforme (HV), high EBV-related antibody titers, and a high EBV-deoxyribonucleic acid (DNA) copy number in peripheral blood have been reported to be characteristic of CAEBV; however, there is often no discernable immunological abnormality, or history thereof until the patient has been diagnosed, at which point symptoms have rapidly progressed.6-8 Although the mechanism of onset is still unknown, CAEBV is thought to progress to an EBV-associated T/NK-cell lymphoproliferative disorder by clonal expansion of T cells or NK cells infected with EBV.9 In addition, CAEBV can have potentially fatal complications such as hemophagocytic syndrome, interstitial pneumonia, malignant lymphoma, myocarditis, and central nervous system infiltration.4,7,10,11 This entity is currently defined as systemic EBV-positive T-LPD of childhood because almost all diagnoses are made in pediatric patients.12,13 However, similar clinical conditions have also been reported in adults (EBV-T/NK-LPD with CAEBVlike features in adult; adult-onset CAEBV).14,15 Most pediatric patients with EBV-T/NK-LPD show symptoms of CAEBV, although lymphomas are rarely observed.16 Conversely, adult-onset EBV-T/NK-LPD shows very diverse clinical manifestations, such as extranodal NK/T-cell lymphoma (ENKTL) and aggressive NKcell leukemia (ANKL), which both follow a relatively rapid clinical course. CAEBV, however, has a chronic clinical course. In cases in which there is clonal expansion due to CAEBV, it may be difficult to distinguish it pathologically from ENKTL without detailed clinical information. A thorough medical interview and careful examination by physicians are, therefore, very important for the diagnosis of adult-onset CAEBV. However, because of its rarity, the clinical features of adult-onset CAEBV have been poorly detailed, and there is no consensus as to haematologica | 2018; 103(6)

whether there are clinical differences between adultonset and pediatric-onset CAEBV. The purpose of this study was to characterize the clinicopathological features of EBV-T/NK-LPD with CAEBVlike features in adults (adult-onset CAEBV) by comparing the features to those of patients with pediatric-onset CAEBV and ENKTL. In addition, we compared the prognosis of adult patients with CAEBV to that of patients with ENKTL.

Methods Patients We enrolled 54 patients who were diagnosed with adultonset CAEBV at the Department of Pathology, Kurume University between January 2005 and December 2015. Patients with adult-onset CAEBV were defined as those whose estimated age at onset was 15 years or older, and who met the criteria for the diagnosis of systemic EBV-T-LPD of childhood according to the 2008 and 2016 World Health Organization (WHO) classifications of lymphoid neoplasms.12 Consequently, patients with pediatric-onset CAEBV were defined as those with an estimated age at onset of less than 15 years. All patients with CAEBV satisfied the following diagnostic criteria, based on the previous report by Kimura et al.17 (with the exception of EBV-DNA viral load, which was measured in plasma in this study as previously reported18): (i) sustained or recurrent infectious mononucleosislike symptoms lasting more than three months: fever (≥38.3ºC or ≥101ºF), liver dysfunction (elevated liver enzymes), lymphadenopathy, hepatosplenomegaly, cytopenia, interstitial pneumonia, hydroa vacciniforme, and hypersensitivity to mosquito bites; (ii) increased quantities of EBV in affected tissues [i.e. detection of EBV-DNA in tissues or peripheral blood by Southern blot hybridization, or EB-encoded small RNA 1 (EBER)-positive cells detected in affected tissues by microscopy (≥10 cells/high power field)], or in peripheral blood [i.e. EBVDNA detected in plasma (≥2×102 copies/mL in plasma)]; and (iii) no evidence of any previous immunological abnormalities or any other infections that could otherwise explain the condition.

We confirmed negativity for human immunodeficiency virus antibody and human T-cell lymphoma virus 1 antibody in all patients. Hemophagocytic syndrome was diagnosed according to the HLH 2004 guidelines;19 all patients with hemophagocytic syndrome met the study criteria.

This study was carried out in accordance with the recommendations of the Declaration of Helsinki and was approved by the ethics review committee of Kurume University (approval number: 291).

The methodological details regarding determination of the EBV-DNA viral load in peripheral blood,18 Southern blot hybridization,8,20 T-cell receptor gamma gene rearrangement,21,22 in situ hybridization for EBER,21 histology and immunophenotyping,9 determination of EBV-infected cell type,6,22 and statistical analysis23 were based on previous reports, as described in the Online Supplementary Appendix.

1019

K. Kawamoto et al.

Epstein-Barr virus-DNA viral load in peripheral blood A peripheral blood sample was obtained from each patient at diagnosis, in order to investigate the viral load by real-time quantitative polymerase chain reaction. A positive result was defined as an EBV-DNA viral load of ≥200 copies/mL, as previously reported.18

Comparison of clinical features with pediatric-onset chronic active Epstein-Barr virus infection and extranodal NK/T-cell lymphoma To compare the clinical symptoms and prognosis of adult-, and pediatric-onset CAEBV, data from 75 patients younger than 15 years who developed CAEBV were extracted from the previously

Table 1. Characteristics of adult-onset chronic active Epstein-Barr virus infection patients.

All patients

Sex Male, n (%)/female, n (%) Median age, years (range) ≥ 50 years, n (%) Symptoms and involvement sites Fever, n (%) Bone marrow, n (%) Splenomegaly, n (%) Hepatomegaly, n (%) Lymphadenopathy, n (%) Skin rash, n (%) Abdominal disturbance, n (%) Laryngopharynx, n (%) Lung, n (%) Gastrointestinal tract, n (%) Oral lesion, n (%) Central nervous system, n (%) Myocarditis, n (%) Past medical history Hypersensitivity to mosquito bites, n (%) Hydroa vacciniforme, n (%) ECOG PS high (2-4), n (%) Laboratory test at initial diagnosis Anemia (Hb <10.5 g/dL), n (%) Thrombocytopenia (< 100×109/L), n (%) LDH elevation, n (%) Transaminase, elevation, n (%) Hemophagocytic syndrome, n (%) EBV-related antibody to VCA-IgG, median titer (range) VCA-IgM, median titer (range) EBNA, median titer (range) Unknown, n (%) EBV-DNA, in plasma (copies/mL) Median (range) Unknown, n (%) EBV monoclonality by Southern blot, n (%) EBER+cell counts /HPF, median (range) Histological classification9 A1, n (%) A2, n (%) A3, n (%) Immunophenotype CD4 CD8 CD56 TCRab (n = 7) Allogeneic HSCT

Infected-cell type NK-cell (n = 32)

P§

(n = 54)

T-cell (n = 22)

31 (57.4)/23 (42.6) 39 (16-86) 22 (40.7)

12 (54.5)/10 (45.5) 37 (19-86) 10 (0.455)

19 (59.4)/13 (40.6) 41 (16-78) 12 (0.545)

0.784 0.672 0.585

35 (64.8) 31 (57.4) 28 (51.9) 22 (40.7) 21 (38.9) 21 (38.9) 11 (20.4) 11 (20.4) 8 (14.8) 5 (9.3) 2 (3.7) 1 (1.9) 1 (1.9)

15 (68.2) 13 (59.1) 13 (59.1) 12 (54.5) 13 (59.1) 4 (18.2) 5 (22.7) 4 (18.2) 2 (9.1) 2 (9.1) 0 (0) 1 (4.5) 0 (0)

20 (62.5) 18 (56.3) 15 (46.9) 10 (31.3) 8 (25.0) 17 (53.1) 6 (18.8) 7 (21.9) 6 (18.8) 3 (9.4) 2 (6.3) 0 (0) 1 (3.1)

0.775 0.783 0.418 0.101 0.0202* 0.0119* 0.743 1 0.449 1 0.508 0.407 1

4 (7.4) 2 (3.7) 17 (31.5)

1 (4.5) 0 (0) 10 (45.5)

3 (9.4) 2 (6.3) 7 (21.9)

0.638 0.508 0.0814

15 (27.8) 26 (48.1) 40 (74.1) 22 (40.7) 25 (46.3)

7 (31.8) 13 (59.1) 16 (72.7) 12 (54.5) 12 (54.5)

8 (25.0) 13 (40.6) 24 (75.0) 10 (31.3) 13 (40.6)

0.758 0.268 1 0.101 0.407

320 (10-5120) <10 (<10-60) 40 (<10-320) 13 (24.1)

160 (10-2560) <10 (<10-40) 10 (<10-80) 4 (18.2)

160 (40-5120) <10 (<10-60) 20 (<10-320) 9 (28.1)

na

1×104(nd-1×106) 17 (31.5) 25/30 (81.3) 58 (2-487)

1×104(2×102-1.1×106) 8 (36.4) 16/17 (94.1) 32 (2-435)

1×104(nd-3.7×105) 9 (28.1) 9/13 (69.2) 61(3-487)

na

16 (29.6) 20 (37.0) 18 (33.3)

10 (45.5) 6 (27.2) 6 (27.2)

6 (18.7) 14 (43.8) 12 (37.5)

0.127

19 (35.2) 20 (37.0) 34 (61.8) 7/7 (100) 9 (16.7)

12 (63.2) 11 (55.0) 8 (23.5) 7/7 (100) 3 (13.6)

7 (36.8) 9 (45.0) 26 (76.5) na 6 (18.8)

na

Patients’ characteristics

na 0.180

0.723

CAEBV: chronic active EBV infection; EBV: Eptein-Barr virus; EBNA: Epstein–Barr virus nuclear antigen 1; EBV-DNA: EBV-deoxyribonucleic acid; ECOG PS: Eastern Cooperative Oncology Group Performance Status; EBER: Epstein-Barr virus-encoded RNA; nd: not detected (EBV-DNA < 2×102 copies/mL); HPF: high power field; HSCT: hematopoietic stem cell transplantation; Hb: hemoglobin; LDH: lactate dehydrogenase; TCR: T-cell receptor; VCA: viral capsid antigen.*Statistically significant; na: not avalilable; P§: T-cell type versus NK-cell type.

1020

haematologica | 2018; 103(6)

Adult patients with chronic active EBV-like features

published papers by Kimura et al.17 The clinical features of all 75 patients are shown in Online Supplementary Table S2.

Furthermore, 82 patients, who were diagnosed with ENKTL at our institution according to the WHO classification were used for the prognostic comparison.12 These patients’ clinical features are shown in Online Supplementary Table S3. ENKTL cases were divided into “nasal-type” (n=35) and non-nasal-type (n=47), according to the anatomical sites of the lesions

Results Clinical characteristics of patients with adult-onset chronic active Epstein-Barr virus infection The clinical characteristics of all patients with CAEBV are summarized in Table 1. Detailed characteristics including estimated age at onset, EBV-related antibody titers, EBV-DNA copy number in plasma, T-cell receptor gamma

rearrangement status, EBV Southern blot analysis, EBERpositive cell counts, histological classification, treatment, and outcome of the 54 patients with CAEBV are presented in Online Supplementary Tables S5-S8.

The median period from estimated onset to diagnosis was approximately 12-24 months (range, 3- >120 months). Figure 1 shows the periods from the estimated onset to the diagnosis of CAEBV and discontinuation of observation, which was more than one year in most cases.

The median age at diagnosis was 39 years (range, 16–86 years). The number of cases by age at onset is shown in Online Supplementary Figure S2. The onset of CAEBV was observed at all ages. However, as the age distribution of adult-onset CAEBV appeared to be bimodal, we also investigated the clinicopathological factors of CAEBV patients diagnosed at ≥50 years of age. In EBV-infected cells, lymphadenopathy was significantly more frequent in patients with the T-cell type (P=0.0202), whereas those

Figure 1. Approximate time from estimated onset and survival. The figure shows the periods from estimated onset to diagnosis of chronic active EpsteinBarr virus infection/discontinuation of observation. In most cases more than one year elapsed from the estimated onset to the diagnosis.

haematologica | 2018; 103(6)

1021

K. Kawamoto et al.

with the NK-cell type had significantly more skin lesions (P=0.0119). A comparison of the clinicopathological features of CAEBV with nodal lesions versus CAEBV with extranodal lesions is shown in Online Supplementary Table S2. There were no clinicopathological differences between adult-onset CAEBV with nodal lesions and those with extranodal lesions. In addition, the anti-viral capsid antigen

(VCA)-IgM antibody tended to be low or less than the level of detection. The median count of EBER-positive cells was 53 per high power field (range, 2–487), and 86.3% (44/51) of the cases showed ten or more positive cells per high power field. Furthermore, 97.3% (36/37) of the cases had ≥102 EBV copies/mL, and only 2.7% (1/37) had levels below that of the level of detection (2×102 copies/mL).

Table 2. Characteristics of patients aged over 50 years with adult-onset chronic active Epstein-Barr virus infection.

All patients

Sex Male, n (%)/female, n (%) Symptoms and involved sites Fever, n (%) Bone marrow, n (%) Splenomegaly, n (%) Hepatomegaly, n (%) Lymphadenopathy, n (%) Skin rash, n (%) Abdominal disturbance, n (%) Laryngopharynx, n (%) Lung, n (%) Gastrointestinal tract, n (%) Oral lesion, n (%) Central nervous system, n (%) Myocarditis, n (%) Past medical history Hypersensitivity to mosquito bites, n (%) Hydroa vacciniforme, n (%) ECOG PS high (2-4), n (%) Laboratory test at initial diagnosis Anemia (Hb <10.5 g/dL), n (%) Thrombocytopenia (< 100×109/L), n (%) LDH elevation, n (%) Transaminase, elevation, n (%) Hemophagocytic syndrome, n (%) EBV-related antibody to: VCA-IgG, median titer (range) VCA-IgM, median titer (range) EBNA, median titer (range) Unknown, n (%) EBV-DNA, in plasma (copy/mL) Median (range) Unknown, n (%) EBV monoclonality by Southern blot, n (%) EBER+cell counts /HPF, median (range) Histological classification9 A1, n (%) A2, n (%) A3, n (%) Immunophenotype CD4 CD8 CD56 TCRab (n = 3) Allogeneic HSCT

Infected-cell type NK-cell (n = 12)

P§

(n = 22)

T-cell (n = 10)

12 (54.5)/10 (45.5)

5 (50.0)/5 (50.0)

7 (58.3)/5 (41.7)

1

9 (40.9) 8 (36.4) 7 (31.8) 5 (22.7) 10 (45.5) 11 (50.0) 3 (13.6) 5 (22.7) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

5 (55.6) 4 (50.0) 4 (57.1) 4 (80.0) 7 (70.0) 2 (18.2) 1 (33.3) 2 (40.0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

4 (44.4) 4 (50.0) 3 (42.9) 1 (20.0) 3 (30.0) 9 (81.8) 2 (66.7) 3 (60.0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

0.666 1 0.652 0.135 0.0212 0.03 1 1 na na na na na

2 (9.1) 1 (3.7) 6 (27.3)

0 (0) 0 (0) 4 (66.7)

2 (100) 1 (100) 2 (33.3)

0.481 1 0.348

6 (27.3) 16 (72.7) 13 (74.1) 4 (18.2) 5 (22.7)

3 (50.0) 5 (31.3) 6 (72.7) 3 (75.0) 4 (80.0)

3 (50.0) 11 (68.7) 7 (75.0) 1 (25.0) 1 (20.0)

1 0.0557 1 0.293 0.135

160 (80-5120) <10 (<10-10) 40 (<10-80) 9 (40.9)

160 (10-1280) <10 (<10-10) 10 (<10-40) 3 (33.3)

160 (80-5120) <10 (<10-10) 40 (<10-80) 6 (66.7)

na

1×104(nd-1×106) 13 (59.1) 8/9 (88.9) 87 (2-487)

2×105(5×104-1×106) 7 (53.8) 3/4 (75.0) 58 (2-435)

1.1×103(nd-7.3×104) 6 (46.2) 5/5 (100) 168 (15-487)

na

5 (22.7) 7 (31.8) 10 (45.5)

3 (60.0) 3 (42.9) 4 (40.0)

2 (40.0) 4 (57.1) 6 (60.0)

0.864

10 (45.5) 10 (45.5) 10 (45.5) 3/3 (100) 2 (9.1)

7 (70.0) 4 (40.0) 5 (50.0) 3/3 (100) 1 (50.0)

3 (30.0) 6 (60.0) 5 (50.0) na 1 (50.0)

na

Patients’ characteristics

na 0.308

na

CAEBV: chronic active EBV infection; EBV: Eptein-Barr virus; EBNA: Epstein–Barr virus nuclear antigen 1; EBV-DNA: EBV-deoxyribonucleic acid; ECOG PS: Eastern Cooperative Oncology Group Performance Status; EBER: Epstein-Barr virus-encoded RNA; nd: not detected (EBV-DNA < 2×102 copies/mL); HPF: high power field; HSCT: hematopoietic stem cell transplantation; Hb: hemoglobin; LDH: lactate dehydrogenase; TCR: T-cell receptor; VCA: viral capsid antigen. *Statistically significant; na: not avalilable; P§: T-cell type versus NK-cell type.

1022

haematologica | 2018; 103(6)

Adult patients with chronic active EBV-like features

Clinicopathological analysis of patients diagnosed at over 50 years of age with chronic active Epstein-Barr virus infection The clinical characteristics of CAEBV patients diagnosed at over 50 years of age are presented in Table 2.

Eight out of 9 cases (88.9%) showed EBV monocolonality

by Southern blot analysis. However, in our study, there

was no difference in clinical features between all CAEBV cases and those aged over 50 years. Furthermore, as shown in Online Supplementary Figure S4, we performed a prognostic analysis based on the log-rank test by comparing CAEBV in patients aged 50 years or older and those under 50 years. However, there was no difference in overall survival between these two groups (P=0.922).

A

B

C

D

E

F

Figure 2. Indicators for predicting prognosis in terms of overall survival in adult-onset chronic active Epstein-Barr virus (EBV) infection patients. Although infectedcell type (A) and histological classification (B) were not prognostic factors for overall survival (P=0.587 and P=0.822, respectively), thrombocytopenia (B), platelet count < 100×109/L), EBNA antibody titer ≥ 40 (C), the presence of hemophagocytosis syndrome (HPS) (D) at the initial diagnosis were poor prognostic indicators for overall survival (P=0.0087, P=0.0236, and P=0.0149, respectively). With regards to treatment, allogeneic HSCT improved survival (F) (P=0.0289). CAEBV: chronic active EBV infection; EBNA: Epstein–Barr virus nuclear antigen 1; EBV: Epstein-Barr virus; HPS: hemophagocytic syndrome; HSCT: hematopoietic stem cell transplantation.

haematologica | 2018; 103(6)

1023

K. Kawamoto et al.

Comparative analysis of clinical features of adult-onset and pediatric-onset patients

Discussion

Comparison of clinical features of adult-, and pediatriconset patients is shown in Table 3. Patients with adultonset CAEBV had a significantly lower frequency of fever, and more frequent occurrence of skin lesions (erythema), compared to pediatric-onset patients (P=0.005 and P<0.001, respectively). Hypersensitivity to mosquito bites and hydroa vacciniforme were also statistically less frequent in patients with adult-onset CAEBV (P<0.001 and P=0.0238, respectively). As regards laboratory results at initial diagnosis, while elevated liver enzymes were more frequently observed in patients with pediatric-onset type CAEBV (P<0.001), hemophagocytic syndrome was observed in bone marrow biopsies in patients with adultonset CAEBV (P=0.0073).

In the present study, we analyzed 54 patients with adult-onset CAEBV meeting the diagnostic criteria outlined in the Methods section. Non-nasal-type ENKTL, ANKL, and cytotoxic-type and EBV-positive peripheral Tcell lymphomas not otherwise specified (PTCL-NOS) did not meet the diagnostic criteria. As the clinical stage progresses, adult-onset CAEBV may eventually show findings similar to those of malignant lymphomas such as ENKTL, ANKL, and cytotoxic-type and EBV-positive PTCL-NOS. However, it is critical to diagnose CAEBV as early as possible, before fatal complications, such as hemophagocytic syndrome and malignant lymphoma, have developed. The present study showed that adultonset CAEBV is more weakly associated with some characteristics, such as hypersensitivity to mosquito bites and hydroa vacciniforme, compared to pediatric-onset CAEBV. Although it appears that adult-onset CAEBV overlaps clinically with ENKTL, ANKL, and PTCL-NOS (cytotoxic-type and EBV-positive), many CAEBV cases were diagnosed only after progression to malignant lymphomas. In addition, histopathological analysis alone may make it difficult to differentiate among these diseases; however, CAEBV may be considered symptomatically and clinically completely different because of its unique symptoms.

Of the patients diagnosed with adult-onset CAEBV in this study, 18 (33.3%) were diagnosed with malignant lymphoma (ANKL, ENKTL, and EBV + PTCL-NOS) at the time of diagnosis of the CAEBV. When the clinical course of the disease showed the presentation of CAEBV symptoms for diagnosis, the cases were considered to have developed malignant lymphomas during the clinical course of CAEBV. In our analysis, adult-onset CAEBV had a poor prognosis even in cases in which malignant lymphoma has not developed at the time of diagnosis.

Simple diagnostic criteria are considered to be necessary

Indicators for predicting prognosis of patients with adult-onset chronic active Epstein-Barr virus infection We searched for indicators to predict prognosis at initial diagnosis because at that stage, there is no indication of the severity of CAEBV disease progression.

In log-rank test analysis (Figure 2A-E), thrombocytopenia (platelet count <100×109/L), EBNA antibody titer ≥40, and the presence of hemophagocytic syndrome at initial diagnosis were associated with a poor prognosis (i.e. decreased overall survival; P=0.0087, P=0.0236, and P=0.0149, respectively); however, type of infected cell and histological classification were not prognostic factors for overall survival (P=0.587 and P=0.822, respectively). In terms of treatment for CAEBV, although many cases were initially treated with various chemotherapeutic regimens (Online Supplementary Table S8), allogeneic hematopoietic stem cell transplantation (HSCT) was found to be the most effective treatment for improving survival (P=0.0289) (Figure 2F and Online Supplementary Figure S3).

In both univariate and multivariate analyses for predicting overall survival, log-rank tests yielded similar results (Table 4). Age (> 60 years), high-risk Performance Status (2-4), type of infected cell, elevated lactate dehydrogenase level, number of EBV-DNA copies in peripheral blood, EBER-positive cell counts per high power field, and EBV detected by Southern blot using a terminal repeat probe were not prognostic factors in univariate analysis.

Conversely, thrombocytopenia (platelet count <100×109/L; hazard ratio=6.157, 95% confidence interval: 2.433–15.58; P<0.001), high EBNA titer (≥40; hazard ratio=2.815, 95% confidence interval: 1.225-2.497, P=0.0148), and not receiving HSCT (hazard ratio=5.410, 95% confidence interval: 1.892–15.47, P=0.0016) were independent poor prognostic factors.

Statistical comparison of overall survival We compared the overall survival between pediatriconset CAEBV and ENKTL. Overall survival of patients with adult-onset CAEBV (n=54), pediatric-onset CAEBV (n=75), and ENKTL (n=82) is depicted in Figure 3. Adult-onset CAEBV had a poorer prognosis compared to both pediatriconset CAEBV and ENKTL (P<0.001 and P=0.0484, respectively). Even when survival rate was stratified by allogeneic HSCT, significant differences in prognosis were observed between adult-onset and pediatric-onset CAEBV (P<0.001) (Online Supplementary Figure S3). Furthermore, the prognosis for non-nasal-type ENKTL and adult-onset CAEBV appeared to be comparable (P=0.972) (Figure 4).

1024

Table 3. Comparison of adult-onset and pediatric-onset chronic active EpsteinBarr virus (EBV) infection patients.

Patients’ characteristics

Adult onset (n = 54)

Pediatric onset (n = 75)

P

Sex Male, n (%)/female, n (%) 31 (57.4)/23 (42.6) 39 (52.0)/36 (48.0) 0.593 Symptoms and involved sites Fever, n (%) 35 (64.8) 65 (86.7) 0.005* Splenomegaly, n (%) 28 (51.9) 44 (58.7) 0.476 Lymphadenopathy, n (%) 21 (38.9) 30 (40.0) 1 Skin rash, n (%) 21 (38.9) 9 (12.0) < 0.001* Lung, n (%) 8 (14.8) 9 (12.0) 0.793 Oral lesion, n (%) 2 (3.7) 4 (5.3) 1 Central nervous system, n (%) 1 (1.9) 4 (5.3) 0.399 Myocarditis, n (%) 1 (1.9) 6 (8.0) 0.238 Past medical history Hypersensitivity to mosquito bites, n (%) 4 (7.4) 27 (36.0) < 0.001* Hydroa vacciniforme, n (%) 2 (3.7) 13 (17.3) 0.0238* Laboratory test at initial diagnosis Thrombocytopenia, n (%) 26 (48.1) 34 (45.3) 0.481 Transaminase elevation, n (%) 22 (40.7) 54 (72.0) < 0.001* Hemophagocytic syndrome, n (%) 25 (46.3) 17 (22.7) 0.0073* T-cell type/NK-cell type, n (%) 22 (40.7)/32 (59.3) 34 (45.3)/41 (54.7) 0.153 CAEBV: chronic active EBV infection, *Statistically significant difference. Thrombocytepenia: platelet count < 100×109 /L.

haematologica | 2018; 103(6)

Adult patients with chronic active EBV-like features

because adult-onset CAEBV requires active treatments including HSCT. Epstein-Barr virus infection has been considered to be a cause of fever of unknown origin in adults (apart from infectious mononucleosis) and there are some reports describing adult-onset CAEBV not only in Asia, but also in the USA and other regions.24,25 This suggests that adultonset CAEBV is an important disease entity to consider in the differential diagnosis of fever of unknown origin, in addition to previously known diseases. Although the clinical features of adult-onset CAEBV were unclear, we found that CAEBV can develop at any age. Although EBVpositive, nodal, and cytotoxic-type PTCL-NOS have been reported to have a poor prognosis,26 there were no clinicopathological differences between adult-onset CAEBV with nodal lesions and those with extranodal lesions in the present study (Online Supplementary Table S4). This

result suggests that CAEBV has a different background from PTCL-NOS. Furthermore, although we could not compare the clinical features in detail, we did compare the prognosis between patients with nodular EBV+PTCLNOS and CAEBV with nodular lesions and found no statistical difference in prognosis (P=0.143) (Online Supplementary Figure S5). Patients with CAEBV with nodal lesions tended to have a poor prognosis. In the future, we intend to clarify the concepts in these diseases by accumulating more cases.

In this study, EBV-DNA copy number in peripheral blood plasma was detectable (≥2×102 copies/mL) in 97.3% of the cases, and ten EBER-positive cells per high power field were observed in 86.3%, suggesting that these tests may also be useful for diagnosing CAEBV in both adult and pediatric patients.17 Since many cases with low EBVrelated antibody titers were observed, it was suggested

Figure 3. Comparison of overall survival. Adult-onset chronic active Epstein-Barr virus infection may be a distinct entity with a poorer prognosis compared to pediatric-onset CAEBV and extranodal NK/T-cell lymphoma, nasal type (P<0.001 and P=0.0484, respectively).

Figure 4. Comparison of survival for overall survival.

When extranodal NK/T-cell lymphoma, nasal type (ENKTL) was divided into “nasal type” and “non-nasal type” according to the anatomical sites of development, there was no statistical difference in prognosis between non-nasal type ENKTL and adult-onset chronic active Epstein-Barr virus infection (P=0.922).

haematologica | 2018; 103(6)

1025

K. Kawamoto et al.

Table 4. Univariate and multivariate analyses for predicting overall survival of adult-onset chronic active Epstein-Barr virus (EBV) infection patients.

Age ≥ 61 years PS High (2-4) T-cell type Thrombocytopenia (< 100×109 /L) Elevated LDH EBNA antibody titer ≥ 40 EBV-DNA in PB ≥ 103 copies/mL EBV monoclonality by Southern blot EBER-positive cells ≥30 /HPF Treatment without HSCT

Univariate analysis Hazard ratio (95% CI) P 1.340 (0.698-2.574) 1.719 (0.885-3.342) 1.190 (0.623-2.274) 2.277 (1.187-4.368) 1.610 (0.762-3.402) 2.341 (0.329-2.497) 0.866 (0.361-2.078) 0.789 (0.229-2.719) 0.637 (0.329-1.235) 2.524 (1.044-6.104)

0.3793 0.1101 0.5989 0.0133* 0.2125 0.0351* 0.7470 0.7069 0.1821 0.0398*

Multivariate analysis Hazard ratio (95% CI) P

6.157 (2.433-15.58)

<0.001*

2.815 (1.225-8.468)

0.0148*

5.410 (1.892-15.47)

0.0016*

CAEBV: chronic active EBV infection; EBNA: Epstein–Barr virus nuclear antigen 1; EBV-DNA: EBV-deoxyribonucleic acid; PS: Performance Status; EBER: Epstein-Barr virus-encoded RNA; LDH: lactate dehydrogenase; HSCT: hematopoietic stem cell transplantation. *Statistically significant difference; PB: peripheral blood (plasma); HPF: high power field.

that direct verification of increased quantitative EBV values may be important for diagnosis. If EBV infection in T cells or NK cells and an increase in plasma EBV-DNA level are proven, it may be necessary to diagnose CAEBV. Many differences in the clinical features between adultonset, and pediatric-onset type CAEBV were elucidated.

As for other infectious diseases, there was a lower frequency of fever in adult-onset CAEBV, than in pediatriconset CAEBV.27 The frequency of skin lesions was higher in the adult-onset type, while hypersensitivity to mosquito bites and hydroa vacciniforme were significantly much less frequent in adult-onset CAEBV than in pediatric-onset CAEBV. Although there was a difference in the appearance of skin lesions between adult-onset and pediatriconset CAEBV, it is not known whether this could be explained only by the difference in age at onset. It should be noted that many patients with adult-onset CAEBV do not have a history of hypersensitivity to mosquito bites and hydroa vacciniforme, despite these symptoms having been thought to be clues to the diagnosis of CAEBV. Adult- and pediatric-onset CAEBV may constitute a continuous spectrum because the diagnostic criteria for adult-onset CAEBV in this analysis were the WHO criteria for “Systemic EBV positive T-cell lymphoproliferative disorders of childhood”. Comparing clinical features, CAEBV patients over 50 years old were considered to share the pathogenesis with that of their young and adult-onset counterparts. However, the clinical and molecular details are still unknown. In this analysis, we did not perform molecular biology to investigate the common features. In future, we would like to investigate whether the molecular background of adult- and pediatric-onset CAEBV is the same.

This study suggests that thrombocytopenia, high EBNA antibody titer (≥40), and the presence of hemophagocytic syndrome at the initial diagnosis of CAEBV are prognostic factors in adult patients, as previously reported in children and young adults, and further suggested to be indicators for aggressive therapeutic intervention including allogeneic HSCT.17,28 It has been reported that patients with clinically aggressive CAEBV have a high level of expression of EBNA-1;29 this report may support our results. It has also been reported that treatment with EBNA-1-specific T cells 1026

may be effective.30 This may be a treatment option for CAEBV patients with high EBNA antibody levels.

Conversely, no difference in prognosis was detected depending on the type of cell infected by EBV in patients with adult-onset CAEBV, which contradicts previous findings in children and young adults.6,17 Precursor T cells are reported to have the potential to differentiate into NK cells, suggesting that phenotype could be changed.31 Moreover, it has been proven that a single EBV clonotype can infect multiple NK-cell and T-cell subsets.32 The identification of infected cells in patients with adult-onset disease may not be very important. In this study, there were 2 patients who did not express cytotoxic molecules such as TIA-1 and granzyme B from their T cells, as determined by immunohistochemistry. Although we could not further characterize the clinical features of these cases because of their small number, further studies are necessary to determine the significance of EBV-infected cells. In addition, there were no differences in prognosis among the three histological classifications (A1, A2, and A3).9 Regardless of histological category, this study suggests that treatment including allogeneic HSCT is the cure for CAEBV. Although allogeneic HSCT has been reported to be effective,6,15,17 there is still controversy about when the transplant should be performed and further research is, therefore, needed.

Genetic analysis by next-generation sequencing is a very important issue in order to evaluate subtypes and different genetic abnormalities which result in CAEBV. According to the results of genetic analysis of terminal repeats of EBV, since it was suggested that CAEBV in the early stage has a polyclonal state, genetic analysis by sequencing was not necessarily valid.9 Although other research groups have detected genetic abnormalities, such as those of the T-cell receptor β repertoire and perforin,33,34 these genetic abnormalities are not useful for explaining the development and mechanism of progression of CAEBV. In situations in which the actual condition of the disease is unclear, we believe it is important to recognize that there are various types of disease. Next-generation sequencing to analyze the genetic landscape of CAEBV will be necessary and important in the future.

It has been suggested recently that the development of haematologica | 2018; 103(6)

Adult patients with chronic active EBV-like features

EBV-related lymphoproliferative diseases could be related to an immune escape mechanism in the programmed death 1 (PD-1)/programmed death ligand 1 (PD-L1) pathway.35,36 In fact, it has been reported that an anti-PD-1 monoclonal inhibitor is effective in Hodgkin lymphoma and ENKTL.37-39 The existence of such an immune escape mechanism in CAEBV is presumed. Unfortunately, it was impossible to investigate the immune background of the EBV infection in the present study because of a lack of specimens. With further accumulation of cases, we hope to investigate the immune response against EBV infection in the future, and examine the differences in immunological background between patients with CAEBV and healthy individuals.

Almost all of the Asian cases have a T-cell or NK-cell origin.17 In contrast, the most common type of CAEBV in the USA originates from B cells.28 There may be racial differences in susceptibility to EBV infection and host immunity. In the future, it will be necessary to clarify these genetic backgrounds. In the present study we confirmed that adult-onset CAEBV has a poorer prognosis than ENKTL. Although adult EBV-positive T/NK-cell LPD includes various lymphomas, such as ENKTL and EBV-positive PTCLNOS,12 we found that CAEBV needs to be distinguished by detailed interview and medical history because of the differences in prognosis and treatment strategies.

However, at present, CAEBV and ENKTL can only be distinguished by differences in their clinical courses, given that differences in biological mechanisms of action are not known. Although it is thought that a simple prognostic comparison should not be performed because the treatments are quite different, the present study showed there was no statistical difference in the prognosis between non-nasal type ENKTL and adult-onset CAEBV. In fact, there was a report in which it was difficult to distinguish between the two diseases.40 In future, further research is necessary to establish novel testing methods to improve the differential diagnosis of these two diseases.

References 1. Cohen JI. Epstein-Barr virus infection. N Engl J Med. 2000;343(7):481-492.

2. Williams H, Crawford DH. Epstein-Barr virus: the impact of scientific advances on clinical practice. Blood. 2006;107(3):862-869.

3. Straus SE. The chronic mononucleosis syndrome. J Infect Dis. 1988;157(3):405-412.

4. Jones JF, Shurin S, Abramowsky C, et al. Tcell lymphomas containing Epstein-Barr viral DNA in patients with chronic EpsteinBarr virus infections. N Engl J Med. 1988;318(12):733-741.

5. Okano M, Kawa K, Kimura H, et al.

Proposed guidelines for diagnosing chronic active Epstein-Barr virus infection. Am J Hematol. 2005;80(1):64-69.

6. Kimura H, Hoshino Y, Kanegane H, et al.

Clinical and virologic characteristics of chronic active Epstein-Barr virus infection.

Blood. 2001;98(2):280-286.

7. Schooley RT, Carey RW, Miller G, et al.

Chronic Epstein-Barr virus infection associated with fever and interstitial pneumonitis. Clinical and serologic features and response to antiviral chemotherapy. Ann Intern Med. 1986;104(5):636-643.

8. Tsuge I, Morishima T, Morita M, Kimura

haematologica | 2018; 103(6)

9.

10.

11.

12.

The reasons why adult-onset CAEBV has a poor prognosis may be as follows: (i) recognition of adult-onset CAEBV by physicians is poor, so the condition is often regarded as an unknown fever; (ii) remissions and exacerbations recur for a long period; (iii) treatments are performed after systemic conditions worsen and/or lifethreatening complications develop, including hemophagocytic syndrome; (iv) not much is known about the pathobiology; and (v) there is no fundamental treatment.

The accumulation of more cases may help in the recognition of adult-onset CAEBV by revealing the clinical features and elucidating the mechanisms of the molecular pathogenesis.

The following proposed diagnostic criteria for adultonset CAEBV are very simple and are based on the pediatric-onset disease: (i) several symptoms of infectious mononucleosis are present, such as fever, lymphadenopathies, and hepatosplenomegaly; (ii) exacerbation and remission of symptoms repeat within a certain period; (iii) proven EBV infection in T cells or NK cells in the affected lesions. Hydroa vacciniforme or hypersensitivity to mosquito bites are not essential for the diagnosis, as these symptoms are present in only approximately 30% of cases, even in pediatric-onset CAEBV. The foregoing criteria may help the timely diagnosis of adult-onset CAEBV. Although there were clinical differences between adultand pediatric-onset CAEBV, we confirmed that CAEBV is a disease with a varying age of onset. In addition, the prognosis of adult-onset CAEBV appears to be very poor.

Therefore, a prescise, early diagnosis and appropriate treatment strategies are critical for adult patients.

Acknowledgments The authors thank Fumiko Arakawa, Kazutaka Nakashima, Mayumi Miura, Kanoko Miyazaki, Yuki Morotomi, Chie Kuroki, and Kaoruko Nagatomo for their technical assistance.

This work was partly supported by The Tsukada Medical Foundation and The Yasuda Medical Foundation.

H, Kuzushima K, Matsuoka H. Characterization of Epstein-Barr virus (EBV)-infected natural killer (NK) cell proliferation in patients with severe mosquito allergy; establishment of an IL-2-dependent NK-like cell line. Clin Exp Immunol.

1999;115(3):385-392.

Ohshima K, Kimura H, Yoshino T, et al.

Proposed categorization of pathological states of EBV-associated T/natural killer-cell lymphoproliferative disorder (LPD) in children and young adults: overlap with chronic active EBV infection and infantile fulminant EBV T-LPD. Pathol Int. 2008;58(4):209-217.

Kikuta H, Taguchi Y, Tomizawa K, et al.

Epstein-Barr virus genome-positive T lymphocytes in a boy with chronic active EBV infection associated with Kawasaki-like disease. Nature. 1988;333(6172):455-457.

Okano M, Matsumoto S, Osato T, Sakiyama Y, Thiele GM, Purtilo DT. Severe chronic active Epstein-Barr virus infection syndrome.

Clin Microbiol Rev. 1991;4(1):129-135.

Swerdlow SH, Campo E, Pileri SA, et al.

The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016;127(20):2375-2390.

13. Hong M, Ko YH, Yoo KH, et al. EBVPositive T/NK-cell lymphoproliferative disease of childhood. Korean J Pathol.

2013;47(2):137-147.

14. Yamada H, Kohno S, Koga H, et al. An adult case of severe chronic active EpsteinBarr virus infection syndrome. Intern Med. 1992;31(12):1381-1386.

15. Saburi M, Ogata M, Satou T, et al.

Successful cord blood stem cell transplantation for an adult case of chronic active Epstein-Barr virus infection. Intern Med. 2016;55(23):3499-3504.

16. Huang Y, Xie J, Ding Y, Zhou X. Extranodal Natural killer/T-cell lymphoma in children and adolescents: a report of 17 cases in China. Am J Clin Pathol. 2016;145(1):46-54.

17. Kimura H, Ito Y, Kawabe S, et al. EBV-associated T/NK-cell lymphoproliferative diseases in nonimmunocompromised hosts: prospective analysis of 108 cases. Blood.

2012;119(3):673-686.

18. Ohga S, Kimura N, Takada H, et al.

Restricted diversification of T-cells in chronic active Epstein-Barr virus infection: potential inclination to T-lymphoproliferative disease. Am J Hematol. 1999;61(1):2633.

19. Henter JI, Horne A, Arico M, et al. HLH-

1027

K. Kawamoto et al.

20.

21.

22.

23.

24.

25.

26.

1028

2004: Diagnostic and therapeutic guidelines for hemophagocytic lymphohistiocytosis.

Pediatr Blood Cancer. 2007;48(2):124-131.

Imai S, Sugiura M, Oikawa O, et al.

Epstein-Barr virus (EBV) - carrying and expressing T-cell lines established from severe chronic active EBV infection. Blood.

1996;87(4):1446-1457.

Kawamoto K, Miyoshi H, Yanagida E, et al.

Comparison of clinicopathological characteristics between T-PLL and PTCL, NOS. Eur J Haematol. 201;98(5):459-466.

Kawamoto K, Miyoshi H, Yoshida N, Takizawa J, Sone H, Ohshima K. Clinicopathological, cytogenetic, and prognostic analysis of 131 myeloid sarcoma patients. Am J Surg Pathol. 2016;40(11): 1473-1483.

Kanda Y. Investigation of the freely available easy-to-use software 'EZR' for medical statistics. Bone Marrow Transplant.

2013;48(3):452-458.

Cunha BA, Petelin A, George S. Fever of unknown origin (FUO) in an elderly adult due to Epstein-Barr virus (EBV) presenting as "typhoidal mononucleosis," mimicking a lymphoma. Heart Lung. 2013;42(1):79-81.

Levy L, Nasereddin A, Rav-Acha M, Kedmi M, Rund D, Gatt ME. Prolonged fever, hepatosplenomegaly, and pancytopenia in a 46-year-old woman. PLoS Med. 2009;6(4):e1000053.

Kato S, Takahashi E, Asano N, et al. Nodal cytotoxic molecule (CM)-positive EpsteinBarr virus (EBV)-associated peripheral T cell lymphoma (PTCL): a clinicopathological

27.

28.

29.

30.

31.

32.

33.

34.

study of 26 cases. Histopathology.

2012;61(2):186-199.

Roghmann MC, Warner J, Mackowiak PA. The relationship between age and fever magnitude. Am J Med Sci. 2001;322(2):6870.

Cohen JI, Jaffe ES, Dale JK, et al.

Characterization and treatment of chronic active Epstein-Barr virus disease: a 28-year experience in the United States. Blood.

2011;117(22):5835-5849.

Iwata S, Wada K, Tobita S, et al.

Quantitative analysis of Epstein-Barr virus (EBV)-related gene expression in patients with chronic active EBV infection. J Gen Virol. 2010;91(Pt 1):42-50.

Bollard CM. Improving T-cell therapy for Epstein-Barr virus lymphoproliferative disorders. J Clin Oncol. 2013;31(1):5-7.

Shen HQ, Lu M, Ikawa T, et al. T/NK bipotent progenitors in the thymus retain the potential to generate dendritic cells. J Immunol. 2003;171(7):3401-3406.

Ohga S, Ishimura M, Yoshimoto G, et al.

Clonal origin of Epstein-Barr virus (EBV)infected T/NK-cell subpopulations in EBVpositive T/NK-cell lymphoproliferative disorders of childhood. J Clin Virol.

2011;51(1):31-37.

Liu S, Zhang Q, Huang D, et al.

Comprehensive assessment of peripheral blood TCRbeta repertoire in infectious mononucleosis and chronic active EBV infection patients.

Ann Hematol.

2017;96(4):665-680.

Katano H, Ali MA, Patera AC, et al.

35.

36.

37.

38.

39.

40.

Chronic active Epstein-Barr virus infection associated with mutations in perforin that impair its maturation. Blood. 2004;103 (4):1244-1252.

Kiyasu J, Miyoshi H, Hirata A, et al.

Expression of programmed cell death ligand 1 is associated with poor overall survival in patients with diffuse large B-cell lymphoma.

Blood. 2015;126(19):2193-2201.

Jo JC, Kim M, Choi Y, et al. Expression of programmed cell death 1 and programmed cell death ligand 1 in extranodal NK/T-cell lymphoma, nasal type. Ann Hematol.

2017;96(1):25-31.

Ansell SM, Lesokhin AM, Borrello I, et al.

PD-1 blockade with nivolumab in relapsed or refractory Hodgkin's lymphoma. N Engl J Med. 2015;372(4):311-319.

Younes A, Santoro A, Shipp M, et al.

Nivolumab for classical Hodgkin's lymphoma after failure of both autologous stem-cell transplantation and brentuximab vedotin: a multicentre, multicohort, singlearm phase 2 trial. Lancet Oncol.

2016;17(9):1283-1294.

Kwong YL, Chan TSY, Tan D, et al. PD1 blockade with pembrolizumab is highly effective in relapsed or refractory NK/T-cell lymphoma failing L-asparaginase. Blood.

2017;129(17):2437-2442.

Ohtsuka R, Abe Y, Sada E, et al. Adult patient with Epstein-Barr virus (EBV)-associated lymphoproliferative disorder: chronic active EBV infection or de novo extranodal natural killer (NK)/T-cell lymphoma, nasal type? Intern Med. 2009;48(6):471-474.

haematologica | 2018; 103(6)

ARTICLE

Non-Hodgkin Lymphoma

Highly sensitive MYD88L265P mutation detection by droplet digital polymerase chain reaction in Waldenström macroglobulinemia

Ferrata Storti Foundation

Daniela Drandi,1 Elisa Genuardi,1 Irene Dogliotti,1,2 Martina Ferrante,1 Cristina Jiménez,3 Francesca Guerrini,4 Mariella Lo Schirico,1 Barbara Mantoan,1 Vittorio Muccio,2 Giuseppe Lia,1 Gian Maria Zaccaria,1,5 Paola Omedè,2 Roberto Passera,6 Lorella Orsucci,7 Giulia Benevolo,7 Federica Cavallo,1,2 Sara Galimberti,4 Ramón García Sanz,3 Mario Boccadoro,1,2 Marco Ladetto8 and Simone Ferrero1,2

Department of Molecular Biotechnologies and Health Sciences, Hematology Division, University of Torino, Italy; 2Division of Hematology 1U, AOU Città della Salute e della Scienza di Torino, Italy; 3Hematology Department, University Hospital of Salamanca and Research Biomedical Institute of Salamanca, Spain; 4Division of Hematology, Department of Oncology, Santa Chiara Hospital, Pisa, Italy; 5Biolab, Department of Electronics and Telecommunications, Politecnico di Torino, Italy; 6Biostatistics Unit, Division of Nuclear Medicine, AOU Città della Salute e della Scienza di Torino, Italy; 7Division of Hematology 2, AOU Città della Salute e della Scienza di Torino, Italy and 8Division of Hematology, AO SS Antonio e Biagio e Cesare Arrigo, Alessandria, Italy 1

Haematologica 2018 Volume 103(6):1029-1037

ABSTRACT

W

e here describe a novel method for MYD88L265P mutation detection and minimal residual disease monitoring in Waldenström macroglobulinemia, by droplet digital polymerase chain reaction, in bone marrow and peripheral blood cells, as well as in circulating cell-free DNA. Our method shows a sensitivity of 5.00x10-5, which is far superior to the widely used allele-specific polymerase chain reaction (1.00x10-3). Overall, 291 unsorted samples from 148 patients (133 with Waldenström macroglobulinemia, 11 with IgG lymphoplasmacytic lymphoma and 4 with IgM monoclonal gammopathy of undetermined significance) were analyzed: 194 were baseline samples and 97 were followup samples. One hundred and twenty-two of 128 (95.3%) bone marrow and 47/66 (71.2%) baseline peripheral blood samples scored positive for MYD88L265P. To investigate whether MYD88L265P detection by droplet digital polymerase chain reaction could be used for minimal residual disease monitoring, mutation levels were compared with IGH-based minimal residual disease analysis in 10 patients, and was found to be as informative as the classical, standardized, but not yet validated in Waldenström macroglobulinemia, IGH-based minimal residual disease assay (r2=0.64).

Finally, MYD88L265P detection by droplet digital polymerase chain reaction on plasma circulating tumor DNA from 60 patients showed a good correlation with bone marrow findings (bone marrow median mutational value 1.92x10-2, plasma circulating tumor DNA value: 1.4x10-2, peripheral blood value: 1.03x10-3). This study indicates that droplet digital polymerase chain reaction assay of MYD88L265P is a feasible and sensitive tool for mutation screening and minimal residual disease monitoring in Waldenström macroglobulinemia. Both unsorted bone marrow and peripheral blood samples can be reliably tested, as can circulating tumor DNA, which represents an attractive, less invasive alternative to bone marrow for MYD88L265P detection.

haematologica | 2018; 103(6)

Correspondence: daniela.drandi@unito.it

Received: December 18, 2017.

Accepted: February 23, 2018.

Pre-published: March 22, 2018.

doi:10.3324/haematol.2017.186528 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1029 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1029

D. Drandi et al.

Introduction In recent years, the MYD88L265P mutation has been recurrently identified in Waldenström macroglobulinemia (WM),1,2 an indolent lymphoplasmacytic lymphoma (LPL) characterized by the accumulation in the bone marrow (BM) of monoclonal lymphocytes, lymphoplasmacytic cells and plasma cells, responsible for monoclonal IgM protein secretion.3 Several studies using different techniques such as Sanger sequencing, polymerase chain reaction (PCR) and allele-specific quantitative PCR (ASqPCR), on CD19-sorted BM samples, found that about 90% of WM patients carry the MYD88L265P mutation, while it is present in 14-29% of “activated B-cell” diffuse large B-cell lymphomas,4,5 6-10% of marginal zone lymphomas, 3-5% of chronic lymphocytic leukemia and is absent in multiple myeloma and in non-IgM monoclonal gammopathies of uncertain significance (MGUS).6–8 Therefore, MYD88L265P is now considered a hallmark of WM and may be helpful in the differential diagnosis from other lymphoproliferative neoplasms with overlapping clinical features, such as multiple myeloma.9,10 Furthermore, it might represent an ideal marker for minimal residual disease (MRD) monitoring in a disease whose therapeutic scenario is rapidly changing, with many newly available and highly effective drugs.11–17 Moreover, MYD88L265P has been demonstrated in CD19sorted BM samples from 50-80% of patients with IgMMGUS, an asymptomatic phase which represents a preneoplastic condition, suggesting a potential role in disease progression.6,8,18–20 BM biopsy is mandatory for the differential diagnosis between WM and IgM-MGUS, but patients with an asymptomatic M component do not readily agree to undergo such an invasive procedure. The availability of accurate diagnostic tools based on the use of peripheral blood (PB), or even urine samples, would overcome this problem and avoid the risk of misclassification of patients.

Additionally, the current MYD88L265P ASqPCR method lacks sensitivity (1.00x10-3) and is not, therefore, suitable for MRD.6,21 Indeed, ASqPCR is suboptimal for testing specimens such as unsorted BM or even PB, which contains low concentrations of circulating tumor cells (especially after immunochemotherapy), or for assessing cellfree tumor DNA (ctDNA), usually present in only very small amounts in plasma, including cerebrospinal fluid and pleural effusions.22,23 Recently, digital PCR has been shown to be a powerful technique that provides improved sensitivity, precision and reproducibility, overcoming some of the pitfalls of qPCR.24,25 We here describe a newly developed, highly sensitive, droplet digital PCR (ddPCR) assay for the identification of the MYD88L265P mutation in patients affected by WM or LPL, suitable for screening and MRD monitoring on BM and PB cells, as well as on cellfree DNA.

Methods Patients and samples collection BM, PB, plasma and urine samples were collected at baseline and during follow up from patients affected by WM, IgM-MGUS or IgG-secreting LPL (Online Supplementary Figure S1). Consecutive patients were included in this study and were classified based on the 2008 World Health Organization classification criteria.3 PB from 40 healthy subjects and BM from 20 patients with multiple myeloma were used as negative controls. Sample collection and 1030

storage as well as nucleic acid extraction procedures are described in the Online Supplementary Appendix.

All patients provided written informed consent to the use of their biological samples for research, in accordance with Institutional Review Board provisions and the Declaration of Helsinki. This study was approved by the local ethics committee.

Droplet digital polymerase chain reaction assays for detection of the MYD88L265P mutation The mutation detection assay was designed as reported in Online Supplementary Figure S2A. ddPCR was performed using the QX100 Droplet Digital PCR system (Bio-Rad Laboratories) as detailed in the Online Supplementary Appendix. Samples were tested in triplicate and results are expressed as merge of wells. The cut-off for defining the presence of the mutation was set based on the highest MYD88L265P level detected within the control group and is indicated in figures as a red dashed line. Samples with a ratio value below the red dashed line are considered MYD88L265P wildtype (WT). Each experiment included a no template control, a known highly mutated positive control sample (mutation rate = 70%, mutated/WT ratio 7x10-1), previously tested by Sanger sequencing, as reported by Treon et al.,1 and a negative control (healthy donor or multiple myeloma gDNA). dMIQE guidelines (Minimum Information for Publication of Quantitative Digital PCR Experiments) for ddPCR experiments are listed in Online Supplementary Table S1.25

Allele-specific quantitative polymerase chain reaction assay for detection of the MYD88L265P mutation The sensitivity of MYD88L265P ddPCR was compared to that of the ASqPCR assay previously described by Xu et al.6 on a standard curve of 10-fold serial dilutions. In parallel, 100 WM samples from the Torino series and 23 patients from the Salamanca series were analyzed, following the strategy described by Jiménez et al.21 Additionally, MYD88L265P mutation detection by ddPCR was compared to that of the qBiomarker™ Somatic Mutation Assay kit for MYD88_85940 (Qiagen) on 15 samples from the University of Pisa.

Allele-specific quantitative polymerase chain reaction assay for tumor-specific IGH-VDJ rearrangement In order to perform a comparison to the worldwide standardized ASqPCR technique for immunoglobulin-VDJ rearrangement (IGH-VDJ) MRD analysis, patient-specific IGH-VDJ was amplified and directly sequenced26 (Online Supplementary Figure S2B). This IGH-based MRD analysis was performed according to the EuroMRD guidelines.27

Statistical analysis Associations between categorical variables were analyzed by the Fisher exact test, while Mann-Whitney and Kruskal-Wallis tests were used for the inference on continuous variables. Results of the analyses of continuous variables are expressed as the median (range); ddPCR and ASqPCR results are expressed as mutated:WT ratio. The interrater agreement on categorical data was estimated by computing the Fleiss kappa (k) index. All reported P-values were estimated by the two-sided exact method with the conventional 5% significance level. Data were analyzed as of July 2017 using R 3.4.1 (R Foundation for Statistical Computing, Vienna, Austria).28

Results Patients and samples Overall, 291 samples from 148 patients from the three series were analyzed (133 WM, 1 amyloid-associated haematologica | 2018; 103(6)

MYD88L265P mutation detection by ddPCR on BM, PB and ctDNA

IgM-LPL, 10 IgG-LPL, 4 IgM-MGUS): 194 samples were taken at baseline in active disease (128 BM, 66 PB) and 97 samples during follow up (43 BM and 54 PB). The baseline samples were defined either “treatment-naïve” (143/194, 73.7%, 99 BM and 44 PB) or “relapsed” (51/194, 26.2%, 29 BM and 22 PB), according to previous exposure to chemoimmunotherapy. All the follow-up samples were taken for the purpose of studying MRD at different time points after specific treatment. The distribution of the study population is summarized in Online Supplementary Figure S1, while the patients’ main clinical features are detailed in Table 1.

Table 1. Patients’ clinical and biological features at baseline.

Patients’ chacteristics Baseline samples (n= 148) Waldenström macroglobulinemia IgG-lymphoplasmacytic lymphoma IgM-MGUS IgM-LPL amyloid tumor Sex, female Median age (years, range) Median hemoglobin (range), g/dL

133 (89.8%) 10 (6.7%) 4 (2.8%) 1 (0.7%) 51 (34%) 67 (24-88) 11.5 (7.5-16.8)

Median IgM (range), g/dL

Detection limit of droplet digital polymerase chain reaction The detection limit of ddPCR was determined using a serial dilution of MYD88L265P mutated gDNA in WT DNA at levels of 35, 3.5, 0.35, 0.035 and 0.0035%, corresponding to 10500, 1050, 105, 10.5 and 1 mutated copy present in 100 ng of gDNA (30,000 copies) which is the overall quantity of gDNA loaded per well. We identified the limit of detection using a statistical method based on binomial distribution, as previously reported.29 This analysis indicated that we were able to detect the mutation, with a good degree of confidence, when the level of mutated copies was more than 0.035% (10 mutated copies in 30000 WT). This value mirrored the cut-off mutated:WT ratio we identified based on the control group (40 healthy subjects) (Figure 1A). Additional reproducibility tests confirmed that the above calculated limit of detection and the experimentally detected cut-off, emerging from the control group, are equally reliable (Online Supplementary Table S2).

MYD88L265P screening by droplet digital polymerase chain reaction in baseline samples Overall 142/148 (96%) patients were identified as having the MYD88L265P mutation. We observed a 91.6% mutation rate (33/36) among relapsed patients and 97.3% rate (109/112) among treatment-naïve patients (P=0.15).

Notably, no one of the 6 MYD88L265P WT patients (4 WM with histological BM invasion of 20%, 30%, 30% and 60%) or 2 patients with IgG LPL (with 20% and 30% histological BM invasion) showed either alternative MYD88 or CXCR4 mutations, as investigated by Sanger sequencing on unselected cells.10,12,30 All BM samples from 20 patients with multiple myeloma used as negative controls were below the limit of detection (defined as previously described) and below the mutation cut-off ratio established based on 40 PB samples from healthy individuals (mutated:WT ratio <3.4x10-4) (Figure 1A). Moreover, to confirm the specificity of our assay for mutational screening, 15 patients with mantle cell lymphoma, 10 with follicular lymphoma and 10 with chronic lymphocytic leukemia were tested for the MYD88L265P mutation. All samples from the patients with follicular lymphoma or mantle cell lymphoma were WT, whereas one of the 10 patients with chronic lymphocytic leukemia showed a mutation ratio of 4.4x10-4, as already described in this disease.6 Looking at the single tissues, 95.3% (122/128) of baseline BM and 71.2% (47/66) of PB samples scored positive for MYD88L265P (BM median mutation burden 3.60x10-2, range: 2.00x10-4 – 7.30 x10-1; PB median 5.00x10-3, range: 1.00x10-4 – 2.80 x10-1) (Figure 1A). Notably, among the PB haematologica | 2018; 103(6)

Waldenström macroglobulinemia MGUS

2.397 (0.233-12.5) 0.785 (0.492-2.252)

Median IgG (range), g/dL IgG-lymphoplasmacytic lymphoma Median β2 microglobulin (range), mg/L

1.958 (0.916-3.447) 2.62 (1-8.83)

Median bone marrow infiltration Bone marrow biopsy Bone marrow flow cytometry

40% (0-95%) 12% (0-87%)

Organomegaly Splenomegaly Adenopathies

18 (12%) 29 (19.5%)

LPL: lymphoplasmacytic lymphoma; MGUS: monoclonal gammopathy of undetermined significance.

samples there was a statistically significant difference in mutation rate between samples from relapsed patients (median burden 4.00x10-4, range: 1.00x10-4 – 1.00x10-3) and treatment-naïve patents (median burden 2.80x10-3, range: 2.00x10-4 – 1.00x10-2), supporting the possible negative impact on mutation detection of previous treatment on PB samples (P<0.0001) (Figure 1B).

Seventy-four patients in this series had paired baseline BM and PB samples. Overall, in this subgroup of patients, the rate of MYD88L265P mutation detection by ddPCR was 93% (69/74) on BM and 72% (53/74) on PB samples.

Accordingly, the detection rates were higher among treatment-naïve patients than among relapsed patients: 95% (52/55) on BM and 82% (45/55) on PB versus 89% (17/19) on BM and 42% (8/19) on PB, respectively (P=0.014). In addition, 15 of these 74 patients (20%) showed BMmutated/PB-WT discordance, such discordance being more frequent among relapsed cases than among treatment-naïve cases (8/19, 42% vs. 7/55, 13%; P=0.02) (Figure 1C).

Finally, a test of within-run reproducibility showed uniform results across the experiments, with an inter-assay Standard Deviation (SD) of 1.3% for the MYD88L265P positive control and 0.01% for WT gDNA samples.

Comparison of MYD88L265P droplet digital polymerase chain reaction versus allele-specific quantitative polymerase chain reaction assays Once the ddPCR assay had been optimized, the sensitivity of the MYD88L265P ddPCR was compared to that of ASqPCR on a standard curve of 10-fold serial dilutions constructed with a highly MYD88L265P mutated WM sample (70%, mutated/WT ratio 7.00x10-1, diluted to 35%), 1031

D. Drandi et al.

previously identified by Sanger sequencing.1 Whereas the ASqPCR standard curve confirmed the reported sensitivity of 1.00x10-3 (0.25% out of 6000 WT),6 ddPCR reached a 1.5 log higher sensitivity (0.0035% out of 30000 WT) also because of the larger amount of gDNA used in ddPCR than in ASqPCR (100 ng vs. 20 ng) (Figure 2 and Online Supplementary Table S2). Overall, 100 WM samples (48 BM, 52 PB, 40 baseline and 60 follow up) from 48 patients, as well as 20 control samples (15 healthy subjects and 5 with multiple myeloma) were tested by both methods (Online Supplementary Figure S3). An example of mutation analysis performed by both methods on 2 patients is presented in Online Supplementary Figure S4. Of the 40 baseline samples tested by both methods, 35 (87.5%) scored positive and 4 (10%) scored negative for MYD88L265P by both ddPCR and ASqPCR (with there being only one discordant ddPCR+/ASqPCR– case), while higher number of discordances were observed among follow-up samples: 13/60 (21.7%) ddPCR+/ASqPCR–, and 11/60 (18.3%) ddPCR–/ASqPCR+ (Online Supplementary Table S3). Indeed, the strength of agreement was very good in the baseline cases (Cohen κ=0.8) but poor in follow-up samples (Cohen κ=0.2), resembling what had been previously shown for low burden infiltrated samples.31 All control samples scored negative by both methods (median: ddPCR 1.75x10-4 (range 3.10x10-4 – 2.70x10-5) and ASqPCR DCT=10 (range DCT=8.4-10.7, setting DCT=8 as the cutoff for negativity).

A

Comparison of MYD88L265P and IGH-based digital droplet polymerase chain reaction assays for the purposes of minimal residual disease detection To investigate whether ddPCR of MYD88L265P could be used for MRD detection, we compared it to the highly sensitive IGH-based MRD ddPCR assay. Only patients with available follow-up samples were screened for IGH rearrangements. A clonal VDJ rearrangement was detected in 34/52 (65%) patients, as expected in WM.32 All these patients scored positive for MYD88L265P by ddPCR. We, therefore, tested 10 informative patients, at baseline and during clinical follow up, with both techniques. Overall, there was good concordance (r2=0.64) between the two methods (P<0.0001) in the 23 samples (18 BM, 5 PB) (Figure 3).

MYD88L265P digital droplet polymerase chain reaction on circulating tumor DNA In order to investigate the feasibility and the sensitivity advantages of ddPCR-based MYD88L265P mutation detection on plasma ctDNA vs. PB gDNA, paired samples from 60 patients were analyzed. Interestingly, a higher median MYD88L265P mutated / WT ratio was detected in plasma ctDNA (1.4x10-2) than in PB (1.03 x10-3) (P<0.001) (Online Supplementary Figure S5), while no statistically significant difference was observed between ctDNA and BM samples from 32 patients (1.92x10-2 vs. 1.4x10-2; P=0.2). Figure 4 shows the matches among BM, plasma and PB MYD88L265P

B

C

Figure 1. MYD88L265P mutation at baseline is lower in peripheral blood (PB) than in bone marrow (BM). (A) Waldenström macroglobulinemia (WM) patients show statistically significant differences in mutated/wild-type ratio (MUT/WT) between BM and PB samples (P<0.0001). The control group of healthy subjects and multiple myeloma (MM) patients show MUT/WT ratios below the limit of detection. The dashed red line shows the cut-off for mutational status. Symbols below the red line represent MYD88WT samples. (B) MUT/WT ratio differences between relapsed (RE) and Naïve to Treatment (NT) patients, at baseline, is higher in PB than in BM samples. (C) MYD88L265P MUT/WT ratio at baseline in pared BM/PB samples from 55 NT and 19 RE patients, highlight the differences between biological specimens suggesting that for RE patients, PB is less reliable than BM.

1032

haematologica | 2018; 103(6)

MYD88L265P mutation detection by ddPCR on BM, PB and ctDNA

mutation quantification. Of note, 2 samples scored WT in all three compartments (Figure 4 and Online Supplementary Figure S6), while in 3 other patients MYD88L265P was detected in BM and PB, but scored WT in plasma ctDNA; these samples were excluded from the analysis since negativity, confirmed by RNAseP analysis, was ascribed to

pre-analytical issues, related to sample collection and storage. Two cases were erroneously collected in lithiumheparin tubes instead of K3EDTA, while one case was collected correctly but processed more than 6 h after the blood had been drawn, confirming the importance of collection tubes and processing timing for good quality ctDNA, as already reported.33–35

Minimal residual disease monitoring in Waldenström macroglobulinemia by MYD88L265P droplet digital polymerase chain reaction

A

To explore the potential role of MYD88L265P mutation detection by ddPCR for MRD monitoring and therapy response evaluation, we focused our analysis on 52 patients who had at least one available follow-up sample.

These patients’ features and the treatment they received

B

C

Figure 2. Digital droplet polymerase chain reaction sensitivity test and example plots of controls (MUT vs. WT). (A) ddPCR assay measured on a 10-fold dilution standard curve, shows a sensitivity of 0.0035% MUT (1 mutated copy in 30000 WT). (B) Example of ddPCR result plot for the MYD88L265P control. (C) Example of ddPCR result plot for the MYD88WT control. MUT: mutated; WT: wildtype.

Figure 3. MYD88L265P assay and IGH-based approaches are superimposable.

Comparison of MYD88 and IGH-based minimal residual disease monitoring (expressed as copies of mutation in 1x105 cells) in 23 samples from 10 patients shows a good degree of correlation (r2=0.64).

are presented in Online

Figure 4. Cell tumor DNA from plasma mirrors the bone marrow mutation level. MYD88L265P mutated/wild-type (MUT/WT) ratio in bone marrow (BM), peripheral blood (PB) and plasma (PL) paired samples from 32 patients show no statistical differences between BM and PL. The dashed red line shows the cut-off for mutational status. Symbols below the red line represent MYD88WT samples. RE: relapsed; NT: Naïve-to-Treatment patients.

haematologica | 2018; 103(6)

1033

D. Drandi et al.

Supplementary Table S4. MRD monitoring of this series is graphically depicted in Figure 5. All the treatment-naïve patients were MYD88L265P mutated at baseline (either on BM or PB or both), while 2 relapsed patients (one BM and one PB) scored WT (as false negative results, due to previous rituximab exposure, see below). After therapy, 6/22 informative cases showed MRD negativization on BM (27%) and 17/35 (49%) on PB, further suggesting a faster clearance of MRD from PB than from BM. At subsequent time points, the majority of patients remained MRD-positive (namely, patients 26, 28-31 and 33-35), some remained MRD-negative (patients 14-16, 37, 46 and 48), while a few show alternating results (patients 32, 40 and 50). Given the limited series of patients and the retrospective nature of the analysis, it is difficult to draw conclusions on the efficacy of single therapeutic schedules in clearing MRD in WM. However, the high potential of MRD shrinkage of regimens containing either fludarabine or bendamustine (each combined with rituximab) is highlighted in the Online Supplementary Methods (Online Supplementary Figure S7A-C).

Of note, there were 2 false-negative cases among relapsed patients in this series. Patient 44, who had a baseline BM sample depicted as WT (meaning “MRD negative”), but with a borderline mutated:WT ratio, followed

by PB positivity in the first follow-up sample (taken after 3 years). The other false-negative case was patient 45, whose PB relapsed sample was negative at baseline, but

then reverted to positive during the subsequent follow up (“MRD reappearance”).

Additionally, 27 paired PB/plasma follow-up samples highlight the role that ctDNA might have for MYD88L265P mutation detection in pre-treated cases, as well as for MRD monitoring. Of note, 14/27 were concordantly positive and 3/27 concordantly negative, while 9/27 scored positive only in plasma and 1/27 positive on PB but not on plasma. Interestingly, the 9 plasma positive cases “rescued” by ctDNA analysis showed MYD88L265P levels in plasma comparable to those detected in paired BM samples (see Figure 5: patients 52, 50 and 37).

Discussion ASqPCR has been widely used for MYD88L265P detection in WM, providing a higher level of sensitivity than Sanger sequencing.6,21 However, the recently developed ddPCR simplifies and improves the accuracy of MRD monitoring.24,31 This study describes a new ddPCR tool for MYD88L265P screening and MRD monitoring in WM. Our results can be summarized as follows: 1. ddPCR is a sensitive tool for MYD88L265P detection in WM and provides higher resolution compared to the canonical ASqPCR; 2. MYD88L265P ddPCR assay is particularly useful for reli-

Figure 5. MYD88L265P can be used for disease monitoring. MYD88L265P monitored in 52 patients with at least one follow-up (FU) sample. ctDNA: cell tumor DNA; RE: relapsed at baseline; NT: Naïve to Treatment at baseline; MRD: minimal residual disease; PL: plasma; BM: bone marrow; PB: peripheral blood.

1034

haematologica | 2018; 103(6)

MYD88L265P mutation detection by ddPCR on BM, PB and ctDNA

able screening of low infiltrated WM specimens, such as unsorted BM or PB (routinely used in clinical practice); 3. both BM and PB are informative in treatment-naïve patients when tested for MYD88L265P by ddPCR; however, in relapsed patients (mainly exposed to rituximab), PB has a 1 log lower median mutated:WT ratio compared to BM, likely related to the efficacy of the anti-CD20 antibody in clearing the circulating tumor cells: therefore, BM analysis should be preferred for these patients; 4. the MYD88L265P ddPCR assay can be reliably used for MRD purposes, being as informative as the classical, standardized but less applicable, IGH-based MRD assay, not yet validated in WM; 5. the MYD88L265P ddPCR assay targeting ctDNA is very promising for identifying the mutation in less invasively collected tissues, such as plasma or urine, and in samples from pre-treated patients.

The here described MYD88L265P ddPCR assay showed an overall mutation detection rate on baseline unselected mononuclear cells samples of 95.3% in BM and 71.2% in PB. These data might help to solve the still open debate on whether to use sorted versus unsorted BM mononuclear cells to assess the MYD88L265P mutation36 and to extend the implementation of the ddPCR mutation assay to general diagnostic laboratories that do not routinely perform cell selection. At present, only 40% of MYD88L265P cases can be detected by ASqPCR using unselected PB cells, with an overall sensitivity of 39.5% and specificity of 100%.37 In our study we observed that the ddPCR assay on unselected cells greatly improved the rate of MYD88L265P detection compared to that achieved by ASqPCR. In fact, an analysis of 74 paired BM/PB baseline samples showed an overall detection rate of 93% (69/74) in BM and 72% (53/74) in PB. Among 69 patients with detectable mutations in BM samples, the sensitivity for MYD88L265P mutation detection by ddPCR on paired PB samples was 77% (53/69). The highly sensitive results of MYD88L265P ddPCR assay on unsorted samples makes this assay ideal for diagnostic use in clinical routine, avoiding the costs and technical requirements of cell sorting. In addition, our data confirmed, as also reported by Xu et al.,37 that PB samples are suboptimal for mutational screening in previously treated patients, evidence confirmed by the high rate of false-negative results. The sensitivity of our assay in PB samples dropped from 85% in treatment-naïve patients to 47% in relapsed patients. Thus, a BM sample is essential to accurately identify MYD88L265P WT status in pre-treated patients (strongly suggested, for example, before starting an expensive therapy with ibrutinib).

This study shows that, besides its potential diagnostic role, MYD88L265P can effectively and easily be used for MRD monitoring in WM, achieving similar results to the lessapplicable IGH-based MRD assay. In fact, monitoring allele levels can also provide insights into treatment effectiveness in a disease whose therapeutic scenario is rapidly changing, with many new and highly effective drugs.13 Of note, a report claimed the achievement of the first “molecular remission” in WM treated with carfilzomib.38 However, the deepness of molecular response matters, as does the sensitivity of the assay used for MYD88L265P detection, which was quite modest in that case (1.00x10-3)6 compared to the newly developed flow cytometry assays39 and with the ddPCR strategy described in this manuscript.

We observed that the agreement between ddPCR and haematologica | 2018; 103(6)

ASqPCR results was weaker in follow-up samples than in baseline ones. Further evaluations on large, prospective series of patients are needed to assess whether ddPCR could be useful for the identification of false-positive ASqPCR results, as was recently demonstrated in a nextgeneration sequencing/real-time quantitative PCR comparison performed in acute lymphoblastic leukemia.40 Moreover, a methodological validation against IGHbased MRD detection and multiparametric flow cytometry, as well as correlations with clinical impact, are eagerly awaited in this setting and are currently ongoing in series of external samples.

The most innovative aspect of our study is the impact that ctDNA might have on MYD88L265P mutation detection and MRD monitoring. We showed that ctDNA mirrors the BM mutational burden much better than gDNA from PB at baseline. Moreover, our data suggest that ctDNA might be able to reflect the dynamic changes in tumor burden in response to treatment, with higher sensitivity than PB, which is particularly promising in pretreated cases (Figures 3 and 4). Similar promising data have been obtained with ctDNA extracted from exosomes from plasma and urine in a pivotal series of patients (D Drandi et al., 2018, unpublished data), although further investigations, as well as technical optimization, are needed. Nevertheless, these data add to the previously published experiences on the promising role of ctDNA analysis in lymphoproliferative disorders other than WM.41–50 Indeed, ctDNA represents an alternative, less invasive and “patient-friendly” tissue source for mutational analysis, and eventually MRD, which is especially attractive for screening and monitoring asymptomatic patients, such as those with MGUS. (MYD88L265P ddPCR screening on a large group of IgM-MGUS patients is currently ongoing). However, plasma collection needs particular care since blood collection tubes and drawing procedures, as well known, can affect the stability of ctDNA.33,34 Pre-analytical standardized procedures are a prerequisite for clinical application and for consolidation of the promising potential of ctDNA analysis.

Finally, we described the impact of different therapies on MRD clearance in WM (i.e. chlorambucil, rituximab monotherapy), RCD (rituximab, cyclophosphamide, dexamethasone), FCR (fludarabine, cyclophosphamide, rituximab) and bendamustine-containing regimens (Online Supplementary Appendix). We acknowledge that the limitations of the present study include the composite and retrospective nature of the series of patients characterized, moreover, by little outcome information.

One of the main goals of this manuscript was the technical description of the MYD88L265P ddPCR assay on unsorted cells. The illustration of the broad range of possible applications, including those on liquid biopsy, makes this approach potentially practical for implementation in routine diagnostic laboratories. We are aware that this assay represents exploratory research, whose real advantages and predictive values need to be further validated in the context of prospective clinical trials. For this purpose, we are currently investigating MRD assessment by ddPCR on gDNA and ctDNA in the context of a phase II prospective study of relapsed WM patients, sponsored by the Fondazione Italiana Linfomi (EudraCT number: 2013-005129-22). Moreover, further efforts to establish uniform guidelines regarding standardization procedures for sample collection and 1035

D. Drandi et al.

methodological validation are currently ongoing, both at national and European levels.

In conclusion, our study shows that ddPCR is a feasible and highly sensitive assay for MYD88L265P mutational screening and MRD monitoring in WM, particularly in samples harboring low concentrations of circulating tumor cells. For this reason, plasma ctDNA represents a promising tissue source, and might be an attractive, less invasive alternative to BM for MYD88L265P detection, also beyond the scenario of WM. Acknowledgments The authors would like to thank all the patients who participated in the study. We are grateful to Luca Arcaini, Alfredo Benso, Stefano Di Carlo, Angelo Fama, Paola Ghione, Idanna Innocenti, Luca Laurenti, Giacomo Loseto, Gianfranco Politano,

References 11.

1. Treon SP, Xu L, Yang G, et al. MYD88 L265P somatic mutation in Waldenström’s macroglobulinemia. N Engl J Med. 2012;367(9):826–833.

2. Yang G, Zhou Y, Liu X, et al. A mutation in MYD88 (L265P) supports the survival of lymphoplasmacytic cells by activation of Bruton tyrosine kinase in Waldenström macroglobulinemia. Blood. 2013;122(7): 1222–1232.

3. Campo E, Swerdlow SH, Harris NL, Pileri S, Stein H, Jaffe ES. The 2008 WHO classification of lymphoid neoplasms and beyond: evolving concepts and practical applications. Blood. 2011;117(19):5019– 5032.

4. Jiménez C, Sebastián E, Chillón MC, et al.

MYD88 L265P is a marker highly characteristic of, but not restricted to, Waldenström’s macroglobulinemia.

Leukemia. 2013;27(8): 1722–1728.

5. Poulain S, Roumier C, Decambron A, et al.

MYD88 L265P mutation in Waldenstrom macroglobulinemia. Blood. 2013;121(22): 4504–4511.

6. Xu L, Hunter ZR, Yang G, et al. MYD88 L265P in Waldenstrom macroglobulinemia, immunoglobulin M monoclonal gammopathy, and other B-cell lymphoproliferative disorders using conventional and quantitative allele-specific polymerase chain reaction. Blood. 2013;121(11):2051– 2058.

7. Ngo VN, Young RM, Schmitz R, et al.

Oncogenically active MYD88 mutations in human lymphoma. Nature. 2011;470 (7332):115–119.

8. Varettoni M, Arcaini L, Zibellini S, et al.

Prevalence and clinical significance of the MYD88 (L265P) somatic mutation in Waldenstrom’s macroglobulinemia and related lymphoid neoplasms. Blood.

2013;121(13):2522–2528.

9. Swerdlow SH, Campo E, Pileri SA, et al.

The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016;127(20):2375– 2390.

10. Treon SP, Cao Y, Xu L, Yang G, Liu X, Hunter ZR. Somatic mutations in MYD88 and CXCR4 are determinants of clinical presentation and overall survival in

1036

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Marzia Varettoni and Silvia Zibellini for their scientific advice and to Antonella Fiorillo and Sonia Perticone for administrative support.

This work was supported by: Progetto di Rilevante Interesse Nazionale (PRIN2009) from Ministero Italiano dell'Università e della Ricerca (MIUR), Roma, Italy [7.07.02.60 AE01]; Progetto di Ricerca Sanitaria Finalizzata 2009 [RF-2009-1469205] and 2010 [RF-2010-2307262 to S.C.], A.O.S. Maurizio, Bolzano/Bozen, Italy; Fondi di Ricerca Locale, Università degli Studi di Torino, Italy; Fondazione Neoplasie Del Sangue (Fo.Ne.Sa), Torino, Italy; Associazione Da Rosa, Torino, Italy; Comitato Regionale Piemontese (Gigi Ghirotti),Italy and Ricerca biomedica condotta da giovani ricercatori - Fondazione Cariplo (project codes: 2016-0476), Dr.Jiménez was supported by a grant from the Sociedad Espanola de Hematologia y Hemoterapia (SEHH) 2016.

Waldenstrom macroglobulinemia. Blood.

2014;123(18):2791–2796.

Treon SP. How I treat Waldenström macroglobulinemia. Blood. 2009;114(12): 2375–2385.

Treon SP, Xu L, Hunter Z. MYD88 Mutations and response to ibrutinib in Waldenström’s macroglobulinemia. N Engl J Med. 2015;373(6):584–586.

Leblond V, Kastritis E, Advani R, et al.

Treatment recommendations from the Eighth International Workshop on Waldenström’s Macroglobulinemia. Blood.

2016;128(10):1321–1328.

Castillo JJ, Hunter ZR, Yang G, Argyropoulos K, Palomba ML, Treon SP. Future therapeutic options for patients with Waldenström macroglobulinemia.

Best Pract Res Clin Haematol. 2016;29 (2):206–215.

Abeykoon JP, Yanamandra U, Kapoor P. New developments in the management of Waldenström macroglobulinemia. Cancer Manag Res. 2017;973–983.

Hunter ZR, Yang G, Xu L, Liu X, Castillo JJ, Treon SP. Genomics, signaling, and Treatment of Waldenström macroglobulinemia. J Clin Oncol. 2017;35(9): 994–1001.

Tedeschi A, Picardi P, Ferrero S, et al.

Bendamustine and rituximab combination is safe and effective as salvage regimen in Waldenström macroglobulinemia. Leuk Lymphoma. 2015;56(9):2637–2642.

Kyle RA. Long-term follow-up of IgM monoclonal gammopathy of undetermined significance. Blood. 2003;102(10): 3759–3764.

Landgren O, Staudt L. MYD88 L265P somatic mutation in IgM MGUS. N Engl J Med. 2012;367(23):2255-2257.

Varettoni M, Zibellini S, Arcaini L, et al.

MYD88 (L265P) mutation is an independent risk factor for progression in patients with IgM monoclonal gammopathy of undetermined significance. Blood. 2013; 122(13):2284–2285.

Jiménez C, Chillón Mdel C, Balanzategui A, et al. Detection of MYD88 L265P mutation by real-time sllele-specific oligonucleotide polymerase chain reaction. Appl Immunohistochem Mol Morphol. 2014;22 (10):768–773.

Gustine JN, Meid K, Hunter ZR, Xu L, Treon SP, Castillo JJ. MYD88 mutations can be

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

used to identify malignant pleural effusions in Waldenström macroglobulinaemia. Br J Haematol. 2018;180(4):578-581.

Komatsubara KM, Sacher AG. Circulating tumor DNA as a liquid biopsy: current clinical applications and future directions.

Oncology. 2017;31(8):618–627.

Hindson BJ, Ness KD, Masquelier DA, et al. High-throughput droplet digital PCR system for absolute quantitation of DNA copy number. Anal Chem. 2011;83 (22):8604–8610.

Huggett JF, Whale A. Digital PCR as a novel technology and its potential implications for molecular diagnostics. Clin Chem. 2013;59(12):1691–1693.

Ladetto M, Donovan JW, Harig S, et al.

Real-Time polymerase chain reaction of immunoglobulin rearrangements for quantitative evaluation of minimal residual disease in multiple myeloma. Biol Blood Marrow Transplant. 2000;6(3):241–253. van der Velden VHJ, Cazzaniga G, Schrauder A, et al. Analysis of minimal residual disease by Ig/TCR gene rearrangements: guidelines for interpretation of realtime quantitative PCR data. Leukemia.

2007;21(4):604–611.

Fleiss JL, Levin B, Paik MC. Statistical Methods for Rates and Proportions.

Hoboken, NJ, USA. John Wiley & Sons, Inc.; 2003.

Uchiyama Y, Nakashima M, Watanabe S, et al. Ultra–sensitive droplet digital PCR for detecting a low–prevalence somatic GNAQ mutation in Sturge–Weber syndrome. Sci Rep. 2016;6(1):22985.

Hunter ZR, Xu L, Yang G, et al. The genomic landscape of Waldenstrom macroglobulinemia is characterized by highly recurring MYD88 and WHIM-like CXCR4 mutations, and small somatic deletions associated with B-cell lymphomagenesis. Blood. 2014;123 (11):1637– 1646.

Drandi D, Kubiczkova-Besse L, Ferrero S, et al. Minimal residual disease detection by droplet digital PCR in multiple myeloma, mantle cell lymphoma, and follicular lymphoma. J Mol Diagnostics. 2015;17 (6):652–660.

Petrikkos L, Kyrtsonis M-C, Roumelioti M, et al. Clonotypic analysis of immunoglobulin heavy chain sequences in patients with Waldenström’s macroglobulinemia:

haematologica | 2018; 103(6)

MYD88L265P mutation detection by ddPCR on BM, PB and ctDNA

33.

34.

35.

36.

37.

38.

correlation with MYD88 L265P somatic mutation status, clinical features, and outcome. Biomed Res Int. 2014;2014:809103.

Kang Q, Henry NL, Paoletti C, et al.

Comparative analysis of circulating tumor DNA stability In K3EDTA, Streck, and CellSave blood collection tubes. Clin Biochem. 2016;49(18):1354–1360.

El Messaoudi S, Rolet F, Mouliere F, Thierry AR. Circulating cell-free DNA: Preanalytical considerations. Clin Chim Acta. 2013;424: 222–230.

Devonshire AS, Whale AS, Gutteridge A, et al. Towards standardisation of cell-free DNA measurement in plasma: controls for extraction efficiency, fragment size bias and quantification. Anal Bioanal Chem.

2014;406(26):6499–6512.

Gustine J, Meid K, Xu L, Hunter ZR, Castillo JJ, Treon SP. To select or not to select? The role of B-cell selection in determining the MYD88 mutation status in Waldenström macroglobulinaemia. Br J Haematol. 2017;176(5):822–824.

Xu L, Hunter ZR, Yang G, et al. Detection of MYD88 L265P in peripheral blood of patients with Waldenström’s Macroglobulinemia and IgM monoclonal gammopathy of undetermined significance. Leukemia. 2014;28(8):1698–1704.

Treon SP, Tripsas CK, Meid K, et al.

Carfilzomib, rituximab, and dexamethasone (CaRD) treatment offers a neuropathy-sparing approach for treating

haematologica | 2018; 103(6)

39.

40.

41.

42.

43.

44.

Waldenström's macroglobulinemia. Blood.

2014;124(4):503-510.

Paiva B, Montes MC, García-Sanz R, et al.

Multiparameter flow cytometry for the identification of the Waldenström’s clone in IgM-MGUS and Waldenström’s Macroglobulinemia: new criteria for differential diagnosis and risk stratification.

Leukemia. 2014;28(1):166–173.

Kotrova M, van der Velden VHJ, van Dongen JJM, et al. Next-generation sequencing indicates false-positive MRD results and better predicts prognosis after SCT in patients with childhood ALL. Bone Marrow Transplant. 2017;52(7):962–968.

Kurtz DM, Green MR, Bratman SV, et al.

Noninvasive monitoring of diffuse large Bcell lymphoma by immunoglobulin highthroughput sequencing. Blood. 2015;125 (24):3679–3687.

Roschewski M, Dunleavy K, Pittaluga S, et al. Circulating tumour DNA and CT monitoring in patients with untreated diffuse large B-cell lymphoma: a correlative biomarker study. Lancet Oncol. 2015;16(5): 541–549.

Rossi D, Diop F, Spaccarotella E, et al.

Diffuse large B-cell lymphoma genotyping on the liquid biopsy. Blood. 2017;129(14): 1947–1957.

Rustad EH, Coward E, Skytøen ER, et al.

Monitoring multiple myeloma by quantification of recurrent mutations in serum.

Haematologica. 2017;102(7):1266–1272.

45. Sarkozy C, Huet S, Carlton VEH, et al. The prognostic value of clonal heterogeneity and quantitative assessment of plasma circulating clonal IG-VDJ sequences at diagnosis in patients with follicular lymphoma.

Oncotarget. 2017;8(5):8765–8774.

46. Bruscaggin A, Spina V, Di Trani M, et al.

Genotyping of classical Hodgkin lymphoma on the liquid biopsy. Hematol Oncol. 2017;3564–3565.

47. Camus V, Sarafan-Vasseur N, Bohers E, et al. Digital PCR for quantification of recurrent and potentially actionable somatic mutations in circulating free DNA from patients with diffuse large B-cell lymphoma. Leuk Lymphoma. 2016;57(9): 2171–2179.

48. Alcaide M, Yu S, Bushell K, et al. Multiplex doplet digital PCR quantification of recurrent somatic mutations in diffuse large Bcell and follicular lymphoma. Clin Chem.

2016;62(9):1238–1247.

49. Hattori K, Sakata-Yanagimoto M, Suehara Y, et al. Clinical significance of disease-specific MYD88 mutations in circulating DNA in primary central nervous system lymphoma. Cancer Sci. 2018;109(1):225–230.

50. Hiemcke-Jiwa LS, Minnema MC, Radersma-van Loon JH, et al. The use of droplet digital PCR in liquid biopsies: a highly sensitive technique for MYD88 p.(L265P) detection in cerebrospinal fluid.

Hematol Oncol. 6 Dec 2017. [Epub ahead of print].

1037

ARTICLE

Chronic Lymphocytic Leukemia

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):1038-1046

DNA polymerase ν gene expression influences fludarabine resistance in chronic lymphocytic leukemia independently of p53 status

Srdana Grgurevic,1* Patricia Montilla-Perez,1 Alice Bradbury,2 Julia Gilhodes,3 Sophie Queille,1 Sandrine Pelofy,4 Aurélien Bancaud,4 Thomas Filleron,3 Loïc Ysebaert,5 Christian Récher,5 Guy Laurent,5 Jean-Jacques Fournié,1 Christophe Cazaux,1,† Anne Quillet-Mary1 and Jean-Sébastien Hoffmann1*

CRCT, Université de Toulouse, Inserm, CNRS, UPS, France; Equipe Labellisée Ligue Contre le Cancer, Laboratoire d’Excellence Toulouse Cancer, France; 2University of Newcastle, UK; 3Clinical Trials Office - Biostatistics Unit, Institute Claudius Regaud, Institute Universitaire du Cancer Toulouse-Oncopole (IUCT-O), Toulouse, France; 4LAAS, Toulouse, France and 5Department of Hematology, Institut Universitaire du Cancer Toulouse-Oncopole, Toulouse, France 1

*SG and J-SH contributed equally to this work † In memoriam

ABSTRACT

A

Correspondence: jean-sebastien.hoffmann@inserm.fr or srdana.grgurevic@gmail.com Received: June 7, 2017.

Accepted: March 16, 2018.

Pre-published: March 22, 2018.

doi:10.3324/haematol.2017.174243 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1038 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1038

lteration in the DNA replication, repair or recombination processes is a highly relevant mechanism of genomic instability. Despite genomic aberrations manifested in hematologic malignancies, such a defect as a source of biomarkers has been underexplored. Here, we investigated the prognostic value of expression of 82 genes involved in DNA replication-repair-recombination in a series of 99 patients with chronic lymphocytic leukemia without detectable 17p deletion or TP53 mutation. We found that expression of the POLN gene, encoding the specialized DNA polymerase ν (Pol ν) correlates with time to relapse after first-line therapy with fludarabine. Moreover, we found that POLN was the only gene up-regulated in primary patients’ lymphocytes when exposed in vitro to proliferative and pro-survival stimuli. By using two cell lines that were sequentially established from the same patient during the course of the disease and Pol ν knockout mouse embryonic fibroblasts, we reveal that high relative POLN expression is important for DNA synthesis and cell survival upon fludarabine treatment. These findings suggest that Pol ν could influence therapeutic resistance in chronic lymphocytic leukemia. (Patients’ samples were obtained from the CLL 2007 FMP clinical trial registered at: clinicaltrials.gov identifer: 00564512).

Introduction Chronic lymphocytic leukemia (CLL) is the most common hematologic malignancy in the Western hemisphere.1 With a median age at diagnosis of 70 years, CLL is considered to be a disease of the elderly. The malignancy is characterized by the accumulation of mature B lymphocytes that carry B-cell markers, such as CD23, CD19 and CD20, along with a T-cell marker, CD5. Leukemic cells have a defective apoptosis pathway disabled by, among others, overexpression of Bcl-2 protein.2 However, a prolonged life-span of leukemic B cells is not the only cause of disease progression. Several studies have shown that proliferation also plays an important role in the development and clinical course of CLL.3 CLL cells proliferate in socalled proliferation centers (pseudofollicles) of the lymph nodes where they are provided anti-apoptotic and proliferative stimuli from other accessory cells, such as CD40 ligand from helper T lymphocytes or interleukin-2.4 In contrast to the lymph node compartment, the majority of leukemic lymphocytes in the peripheral blood are arrested in the quiescent G0 or G1 phase of the cell cycle.5 CLL is a highly heterogeneous disease in terms of its clinical course and outcome. haematologica | 2018; 103(6)

Contribution of Pol nu in CLL progression

Many patients diagnosed with CLL survive for several years without needing any treatment while others develop an aggressive form of the disease and require immediate therapeutic intervention. Apart from Binet and Rai staging, clinicians take into consideration several other biological characteristics, such as mutational status of the immunoglobulin heavy chain variable gene segment (IGHV) and genomic aberrations. Approximately 4 out of 5 CLL patients harbor at least one chromosomal aberration in their leukemic clone and these, likewise, have a clear clinical impact on patients’ survival.6 Recent studies have also revealed other genomic defects in CLL, including mutations in genes involved in RNA processing, such as SF3B17 and, in general, in B-cell development,8 and mutations in non-coding regions that potentially affect the clinico-biological course of CLL.9 Nevertheless, few of these markers are used in routine clinical practice and even fewer of them have been able to demonstrate any predictive value in terms of response to chemotherapy, such as 17p-deletion6 and mutational status of the TP53 gene.10 Genomic instability is a prominent source of genetic diversity within tumors, generating a different cell population that can be subject to selection in a given microenvironmental or therapeutic context. The past decade has seen a dramatic advance in our understanding of how genomic integrity is related to cancer and has revealed that replication stress in the S phase of the cell cycle is an important feature,11,12 placing studies on replication stress at the forefront of cancer research.13 The genomic instability in CLL may indicate the presence of underlying replication stress. In addition, the backbone of the standard, first-line treatment for CLL patients is fludarabine,14 which leads to excessive replication stress by inhibiting ribonucleotide reductase, an enzyme involved in the production of precursors essential for DNA synthesis. In contrast to solid cancer, DNA replication defects as a source of markers in hematologic malignancies have been poorly Table 1. Clinical characteristics of the chronic lymphocytic leukemia patients.

Characteristic Sex Female Male Age ≤ or 57 years old > 57 years Binet stage B C IGHV Mutated Unmutated Cytogenetics Deletion 13q Deletion 11q Trisomy 12

Category

explored. It was only recently that we revealed the importance of the DNA replication checkpoint CHEK1 as a predictor of survival in acute myeloid leukemia and resistance to therapy with cytarabine.15 Here, we hypothesized that mis-expression of DNA replication-repair-recombination (3R) genes could occur during CLL leukemogenesis and contribute to the evolution of the pathology. We reasoned that a modified DNA replication program might, likewise, have an impact on the response to fludarabine which targets several steps in chromosome duplication. We also speculated that specific 3R gene expression signatures would be relevant in the context of relapse, representing potential predictors of the patients’ outcome. Our data show that, among more than 80 3R genes analyzed in a series of primary samples from 99 patients with CLL prior to treatment, only expression of the POLN gene encoding the specialized A-family DNA polymerase nu (Pol ν), with yet unknown function, can forecast the time to relapse after first-line therapy. We also demonstrate that relatively high POLN expression protects CLL cells against replicative stress caused by fludarabine by allowing efficient DNA replication in the presence of an unbalanced nucleotide pool.

Methods Patients with chronic lymphocytic leukemia RNA from patients’ samples were obtained from the CLL 2007 FMP clinical trial (clinicaltrials.gov identifier: 00564512).16 In accordance with the Declaration of Helsinki, informed consent to participation in the trial was obtained from each patient. Patients recruited for the trial were medically fit, previously untreated, diagnosed with Binet stage B or C CLL, under 65 years of age and without detected 17p deletion or TP53 mutation. Patients were randomized into two treatment arms: fludarabine-cyclophosphamide-campath (FCCam) and fludarabine-cyclophosphamiderituximab (FCR). The primary outcome, namely, progression-free survival at 36 months, was the same for both treatment arms (P=0.21). Relevant data regarding the cohort of patients are summarized in Table 1.

N. (frequency, %)

Cell lines and culture 30/99 (30) 69/99 (70) 51/99 (52) 48/99 (48) 78/99 (79) 21/99 (21) 43/99 (43) 56/99 (57) Detected Not present Unknown Detected Not present Unknown Detected Not present Unknown

haematologica | 2018; 103(6)

52/99 (53) 42/99 (42) 5/99 (5) 21/99 (21) 78/99 (79) 13/99 (14) 80/99 (80) 6/99 (6)

MEF cells and culture Primary mouse embryonic fibroblasts (MEF) were derived from e13.5 embryos with genotypes Poln+/+, Poln+/- and Poln-/- (C57BL/6J strain). Exon 2 containing the initiation codon was deleted in the knockout allele. Primary MEF were cultured in medium containing a high concentration of glucose, glutamax-Dulbecco modified Eagle medium (Invitrogen), 15% Hyclone fetal bovine serum (Thermoscientific), non-essential amino acids, sodium pyruvate, modified Eagle medium vitamin solution, and penicillin/streptomycin (Invitrogen) and was maintained in air-tight containers filled with a gaseous mixture of 93% N2, 5% CO2 and 2% O2 (Praxair) at 37°C. Immortalized MEF were cultured in medium containing a high concentration of glucose, glutamax-DMEM (Invitrogen), 10% fetal bovine serum (Atlanta Biologics) and penicillin/streptomycin, and maintained in a humidified incubator in 5% CO2 at 37°C.

Cell culture MEC-1 and MEC-2 cell lines (DSMZ, Braunschweig, Germany) were cultured in Iscove modified Dulbecco medium with glutamax supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at a density of 5x105 and 3x105 cells/mL for 1039

S. Grgurevic et al.

MEC-1 and MEC-2, respectively. MEF cell lines were grown in Dulbecco modified Eagle medium with glutamax supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at a density of 1.5 x 105 cells/10 cm dish. Primary peripheral blood mononuclear cells from CLL patients were cultured in Roswell Park Memorial Institute medium supplemented with 1% glutamine, 10% fetal bovine serum and 1% penicillin/streptomycin at a density of 1 x 107 cells/mL allowing long-term viability.

DNA synthesis DNA synthesis was monitored by a Click-iT EdU imaging assay (Invitrogen, Saint-Aubin, France). 5-ethynyl-2’-deoxyuridine (EdU) was added at the concentration of 25 mM for 30 minutes (min) as indicated. During DNA staining, cells were incubated with propidium iodide at a concentration of 50 mg/mL, 0.1% RNase A and 0.1% Triton X-100 in a phosphate-buffered saline solution for 10 min away from light.

DNA combing in nanochannels Cells were incorporated into agarose plugs (200,000 cells/plug) and each experiment was carried out with half of an agarose plug deposited in a 1.5 mL tube filled with 800 mL of 0.5X TBE buffer supplemented with 2% poly-vinylpyrrolidone (40 kDa, below the overlapping concentration of 7%) and 5% dithiothreitol. Agarose was melted by heating the tube at 70°C for 15 min. The temperature was set at 42°C, and 2 mL of β-agarase (New England Biolabs) were added for overnight agarose digestion. Finally, DNA was stained with 2.5 mM fluorophore SYTOX-orange (Molecular Probes). Nanofluidic chips were fabricated using a two-step photolithography process to generate the array of nanochannels by projection lithography (ECI 1.2 mm photoresist, Stepper Canon 3000i4); these were then transferred into silicon by Reactive Ion Etching over a depth of 250 nm. Conventional photolithography was then performed to etch lateral microchannels of 20 mm in width and 7 mm in depth. For optical mapping, we used hydrodynamics to force the uptake of genomic DNA in nanochannels.

Imaging was performed with a wide-field inverted Zeiss microscope equipped with a 40X lens (NA=1.4). The light source was a LED engine (Lumencor) with 542/33 nm emission with the filter sets Cy3-4040C (Semrock) for SYTOX-orange visualization.

Images were collected with an Andor Zyla camera operating with a 2x2 binning (pixel size = 325 nm). The velocity of molecules was set at 200 ± 20 mm/s using a pressure source operating at 80-100 mbar. All images presented in the manuscript were filtered using the FFT bandpass filter implemented in ImageJ using minimum and maximum cut-offs of 3 and 40 pixels, then subtracting the background.

Statistical analysis The patients’ clinical data are summarized by frequency and percentage for categorical variables and by median and range for continuous variables. Time to progression was defined as the period from the first-line therapy to progression or last follow up (censored data) and estimated by the Kaplan-Meier method with 95% confidence intervals. The minimum P-value approach was used to dichotomize POLN expression,17 which selects the threshold that best discriminates patients’ outcomes. Selected values of the prognostic factor are examined as candidates for the threshold, after eliminating the top and bottom 10% of the extreme values. The value that best separates patients’ outcomes according to a minimum P-value obtained by the log-rank test is chosen. The P-value is adjusted using the Altman correction to account for the problem of multiple testing. Stability of the threshold was assessed using bootstrap internal validation. Backward selection was performed with a Cox proportional hazards model to identify clinical factors 1040

associated with time to progression. Multivariate analysis was also performed using a Cox proportional hazards model to study the influence of POLN on time to progression after adjusting for the clinical factors previously identified. Two-sided P-values of less than 0.05 were considered statistically significant. All statistical analyses were performed using STATA 12.0 software. Viability data obtained by the MTS assay were analyzed in GraphPad Prism using an ANOVA multiple test. The length of replicated domains assessed in the DNA combing in nanochannels experiments was analyzed in GraphPad Prism using the Mann-Whitney rank-sum t-test. FACS data were analyzed in GraphPad Prism using a paired or unpaired Student t-test depending on the experimental conditions.

Results POLN determines time to progression in chronic lymphocytic leukemia In order to investigate the implication of 3R genes in the clinical course of CLL, we first performed a large, highthroughput quantitative reverse transcriptase polymerase chain reaction-based gene expression analysis on 99 primary samples obtained from treatment-naïve patients included in the CLL 2007 FMP clinical trial (see above).

The patient’s baseline characteristics are summarized in Table 1. Unsupervised clustering on 3R gene expression values confirmed, as previously published,18 clear underlying differences between samples from healthy donors and CLL lymphocytes (Figure 1A) and showed a significantly higher expression in CLL of POLN, the gene encoding for the specialized DNA polymerase ν (Figure 1B), as compared to most of the 3R genes, including other specialized DNA polymerases or genes involved in double-stranded break repair (Online Supplementary Figure S1). Next, we performed a cross-analysis on the gene expression information obtained and the patients’ clinical data. The results of this analysis reveal that among all 3R genes, POLN was the only one linked to the therapeutic outcome of CLL. In univariate analysis, a high level of POLN expression was associated with a shorter time to relapse after first-line therapy, i.e. time to progression (threshold=11.9x10-2, stability=47.42%; P=0.0009, adjusted P=0.0227), as shown in (Figure 1C). Interestingly, we found that POLN expression maintained a significant correlation with time to progression in multivariate analysis after adjustments using previously identified prognostic factors (Table 2). The adjusted Hazard Ratio (HR) was 4.14 with a 95% Confidence Interval of 1.60-10.72 and a P-value of 0.003, defining POLN expression as the strongest prognostic marker of time to progression after fludarabine-based treatment, independently of Binet stage or IGHV mutational status.

Proliferating chronic lymphocytic leukemia lymphocytes over-express POLN Leukemic lymphocytes circulating in the CLL patients’ peripheral blood are arrested in the G0 and G1 phases of the cell cycle.5 To proliferate, CLL lymphocytes need to receive proliferative and pro-survival stimuli released by the accessory cells residing in the lymph node pseudofollicules.4 In order to investigate CLL replication parameters, we mimicked the proliferative lymph node microenvironment by stimulating primary CLL lymphocytes with interleukin-2 and DSP30. After obtaining proliferating leukemic cells, as confirmed by the CFSE dilution assay haematologica | 2018; 103(6)

Contribution of Pol nu in CLL progression

(Figure 2A and Online Supplementary Figure S2), we analyzed the full 3R gene expression profile of non-cycling and cycling purified CLL lymphocytes. Our data revealed that actively dividing primary CLL cells modified expression of only 11 out of a total of 82 3R genes (Figure 2B).

Ten of these 11 3R genes were down-regulated, including genes involved in DNA damage response, such as FANCD2, PRKDC, SIN3B, TIMELESS and LIG3, and several genes implicated in global DNA replication, e.g.

SHPRH, CDC25B, CHTF18, POLE and MCM9. In contrast, the genes encoding factors of replication origin licensing and firing, such as CDC6, CDT1, and GINS4, remained stable (Figure 2B). Interestingly, POLN was the only 3R gene found to be up-regulated in proliferating CLL cells. This finding led us to postulate that POLN might contribute, from the origin, to disease evolution, namely, in the proliferating leukemic lymphocyte.

POLN expression increases during disease progression according to a cellular model MEC-1 and MEC-2 are two CLL cell lines that were established from the same CLL patient sequentially during the course of the disease.19 More precisely, the MEC-1 cell line was established during an early stage of the disease, while the MEC-2 cell line was established later when the patient developed a more aggressive form of the disease.

Based on the clinical characteristics of the patient at the time of sampling, as well as the cellular immunophenotype, it is considered that these two cell lines reflect the clinical progression of CLL.20 In order to employ this cellular model in our further investigations, we first decided to analyze the 3R gene expression profiles of the two cell lines. This analysis revealed that the two cell models share a high level of similarity in their 3R gene profiles as they differ in the expression levels of only 6 3R genes (Figure 3).

A

B

C

Figure 1. POLN gene expression as an independent prognostic marker of time to progression in chronic lymphocytic leukemia (CLL). (A) Unsupervised clustering of CLL and healthy donor (HD) samples based on 3R gene expression data. (B) POLN gene expression in CLL primary samples in comparison to HD samples. (C) Kaplan-Meier graphical estimates of time to progression (TTP) according to POLN expression level. The cut-off between low and high expression is 11.9x10-2. The expression number ranges in under-expressed and over-expressed groups are 0.008-0.118 and 0.119-0.272, respectively.

haematologica | 2018; 103(6)

1041

S. Grgurevic et al.

These include the β isoform of the tumor suppressor TP53 β and the LIG4 gene coding for a protein required in double-stranded break repair through the non-homologous end joining pathway, which were up-regulated by the MEC-2 subclone, along with MCM10 and GINS4, coding for factors implicated in initiation of replication and POLH, a translesion synthesis DNA polymerase gene, which were slightly down-regulated (less than 2-fold) in the MEC-2 subclone. Strikingly, with more than a 5-fold higher expression level in the aggressive MEC-2 subclone, POLN was the gene we found to be most deregulated between the two CLL cell lines (Figure 3). These results suggest that POLN could be an advantageous trait for the CLL clone and could be selected during the evolution of the disease.

Relatively higher POLN expression contributes to fludarabine chemoresistance We next evaluated the sensitivity of the MEC-1 and MEC-2 cell lines in the presence of fludarabine. Cells were

treated with different doses of the nucleoside analog for 24 h and cellular viability was measured at the end of the treatment period. MTS data analysis showed that the MEC-2 cell line was more chemoresistant than the MEC-1 cell line, as indicated by higher cellular viability and fludarabine EC50 values (239.2±2.16 mM for MEC-1 and 433.3±2.38 mM for MEC-2) (Figure 4A).

In order to gain deeper insight into the molecular mechanism by which MEC-2 cells maintained higher survival than MEC-1 in the presence of fludarabine, we next evaluated DNA replication efficiency in these two cell lines upon 2 hours (h) of treatment with fludarabine at the doses of 180 mM and 600 mM. The nucleoside analog EdU was added at the end of the treatment to visualize DNA synthesis and cells were stained with propidium iodide. When normalized to untreated conditions, flow cytometry cell cycle analysis showed a higher percentage of EdU-positive cells among the chemoresistant MEC-2 cells (Figure 4B and Online Supplementary Figures S3 and S4), suggesting that

Table 2. Multivariate analysis of time to progression by Cox proportional hazard models.

POLN IGHV status Gender BINET stage

A

Parameter

Number

Hazard Ratio (HR)

95% HR Confidence limits

P

High expression Low expression Unmutated Mutated Male Female C B

90 9 56 43 69 30 21 78

4.14 1 (base) 2.04 1 (base) 2.11 1 (base) 1.61 1 (base)

1.60-10.72

0.003

1.02-4.12

0.045

0.92-4.83

0.078

0.76-3.42

0.214

B

Figure 2. Proliferation and 3R gene expression profiles of primary sample from patients with chronic lymphocytic leukemia (CLL). Peripheral mononuclear cells from CLL patients were cultured in vitro in the absence or presence of proliferation stimuli (interleukin-2/DSP30) for six days. (A) CFSE-labeled unstimulated/nonproliferating and stimulated/proliferating primary CLL lymphocytes. (B) Unsupervised clustering according to 3R gene expression in non-proliferating and proliferating purified CLL cells. Out of 82 3R genes, only one gene was up-regulated upon proliferation (POLN), 10 genes were down-regulated and the other 71 genes remained stably expressed.

1042

haematologica | 2018; 103(6)

Contribution of Pol nu in CLL progression

MEC-2 cells were able to sustain DNA synthesis more efficiently than the chemosensitive MEC-1 cells in the presence of the drug.

To test this hypothesis in more detail, we evaluated DNA replication parameters in both cell lines upon fludarabine treatment by using an innovative genome-wide DNA combing nanofluidics-technology developed recently in our laboratory.21 By applying differential pressure between the entry and exit microchannels, individual DNA molecules are directed into and stretched on their way through the nanochannels. The strategy allows us to monitor the length of the in cellulo-replicated domains at the level of individual replicating DNA molecules. The results of this analysis showed that treating chemosensitive MEC-1 cells with fludarabine reduced the length of the replicated domains. In contrast, the size of the replicated domains in MEC-2 cells was not decreased upon fludarabine treatment in comparison to the untreated condition. Track length data (scatter and bar plots), representative photographs of tracks and their quantified histograms are presented in Figure 4C. As we found that POLN expression is greater in the MEC-2 cell line than in the MEC-1 cell line, and since we observed relatively higher POLN expression to be a marker of negative patientsâ&#x20AC;&#x2122; outcome after fludarabine-based firstline therapy in our CLL cohort, we hypothesized that relatively higher POLN expression might contribute to enhanced DNA replication ability and cell survival upon fludarabine treatment. To test this, we used independent MEF expressing normal (MEF+/+) and relatively low (MEF+/-) levels of POLN, as well as knockout POLN MEF (MEF-/-) (Online Supplementary Figure S5A). These cell lines were treated with fludarabine for 4 h, then exposed to EdU for 30 min and stained with propidium iodide. DNA synthesis efficiency was then assessed by flow cytometry. Analysis of cell cycle and quantification of the proportion of EdU-positive cells confirmed that POLN-proficient cells (MEF+/+) could indeed sustain DNA synthesis more efficiently than MEF+/- or MEF-/- cells in the presence of fludarabine (Figure 5A and Online Supplementary Figures S4 and S5B). We did not observe any significant differences in fludarabine-mediated DNA replication inhibition between the MEF-/- and MEF+/- cell lines (Online Supplementary Figure S5B). This might be due to the poor expression of POLN in mouse fibroblasts, and consequently a non-significant differential expression between heterozygous and knock-out cells (Online Supplementary Figure S5A). Collectively, these results suggest that relatively high expression of POLN could contribute to higher replication efficiency and sustained cell survival upon fludarabine treatment.

could limit replication stress caused by fludarabine in the condition of dNTP starvation.

In order to confirm that fludarabine chemoresistance driven by POLN was specifically related to the condition of a reduced dNTP pool, we performed a rescue experiment by supplementing MEF cells with deoxynucleosides in the presence of fludarabine. MEF+/+ and MEF+/- cells were treated with a lower concentration of fludarabine (30 mM) for a longer period (8 h) in order to allow rescue of DNA synthesis with deoxynucleosides, which were added during the last 4 h of treatment. Finally, cells were pulsed with EdU for the last 30 min of the incubation period. Flow cytometry analysis showed no significant impact on the DNA synthesis rate in either cell line when the deoxynucleosides were added in the absence of fludarabine. As expected, in the absence of supplemental deoxynucleosides, inhibition of DNA synthesis by fludarabine was more pronounced in the MEF+/- cell line than in the MEF+/+ cell line (Figure 5C and Online Supplementary Figures S10 and S11). Importantly, we found that addition of deoxynucleosides could rescue DNA synthesis in the POLN+/- cell line abolishing the difference between two MEF cell lines (Figure 5C and Online Supplementary Figures S10 and S11). This result confirms that POLN reduces the impact of dNTP starvation on DNA synthesis. Collectively, these data suggest that a relatively high level of POLN might contribute to cell survival and provide advantages, not only upon endogenous replicative stress related to a pathology-associated dNTP pool imbalance, but also when CLL cells are treated with therapeutic agents that further affect dNTP metabolism, such as fludarabine.

Discussion In our study, we report that peripheral blood leukemic lymphocytes from CLL patients express POLN at a relatively high level and that the relative expression of POLN is a

POLN-mediated resistance to fludarabine is related to dNTP pool starvation In order to investigate whether POLN could play a role in limiting replication stress specifically upon decrease of dNTP, we used the ribonucleotide-reductase inhibitor hydroxyurea, which mimics the effect of fludarabine on the dNTP pool. MEF+/+ and MEF+/- cells were treated with different concentrations of hydroxyurea for 4 h and pulsed with EdU for 30 min at the end of the treatment. DNA synthesis profiles were analyzed by flow cytometry, schematically represented in Online Supplementary Figure S7. We observed that the POLN-expressing MEF+/+ cells sustained DNA synthesis upon hydroxyurea treatment more efficiently than their POLN-deficient MEF+/- counterparts (Figure 5B and Online Supplementary Figures S8 and S9), implying that Pol ν haematologica | 2018; 103(6)

Figure 3. 3R gene expression profile change during clinical progression of chronic lymphocytic leukemia (CLL) based on a CLL cellular model. MEC-1 and MEC-2 cell lines were established sequentially from leukemic lymphocytes obtained from the same CLL patient. MEC-1 was isolated at the beginning of the disease while MEC-2 was isolated at the onset of active disease and represents a more aggressive CLL subclone. Expression fold change for each 3R gene was calculated using the comparative Ct method.

1043

S. Grgurevic et al.

A

B

C

Figure 4. POLN and fludarabine chemoresistance in vitro. (A) MTS viability assay of MEC-1 and MEC-2 cell lines in the absence or presence of fludarabine for a treatment period of 24 hours (h). The percentage of viable cells is expressed as the mean Âą Standard Deviation of three independent experiments. The fludarabine EC50 value for the MEC-1 cell line was 239 mM and that for the MEC-2 cell line was 433 mM. (B) Flow cytometry cell cycle analysis of MEC-1 and MEC-2 cell lines, control and treated with fludarabine for 2 h. Delta (D) is calculated as the difference of percent of cells present in the EdU+ gate by control condition (representative of three experiments). (C) DNA combing in nanochannels. The length of replicated domains is expressed in kilobase pairs (kbp). The number of tracks measured in all conditions was 60. The concentration of fludarabine used was 600 mM; the duration of treatment was 4 h. Medians in the dot plot are indicated by red lines.

1044

haematologica | 2018; 103(6)

Contribution of Pol nu in CLL progression

marker determining CLL patients’ outcome and fludarabine chemoresistance. By stimulating in vitro proliferation of primary CLL samples, we also revealed that POLN is the only up-regulated 3R gene among 82 genes analyzed, suggesting that higher expression of POLN could be important for CLL proliferation and progression, independently of external treatment. In an independent approach, using two cell lines established sequentially from the same CLL patient which reflected clinical progression of the disease, we found that the most striking difference between the two cell lines was, again, the relatively enhanced expression of POLN in the more aggressive CLL clone. Cell survival and DNA synthesis experiments with CLL cell lines and murine homozygous, heterozygous and knockout MEF for POLN demonstrated that relatively high expression of POLN conferred higher viability following nucleotide pool starvation by fludarabine, the backbone of therapeutic treatment in CLL. The link between fludarabine resistance and expression of POLN was further confirmed in the high POLNexpressing CLL cells by showing that 30% POLN deple-

tion by a small interfering RNA in these cells resulted in a mild but significant decrease in fludarabine sensitivity (data not shown). Collectively, these results suggest a possible role of POLN in the natural course of CLL and in cell survival upon external stress affecting the dNTP pool balance. This led us to reason that these two mechanisms might be mutually associated as endogenous replicative stress frequently results in dNTP pool starvation when replication origins are over-used. Upon nucleotide starvation by fludarabine, a high POLN level allowed globally higher DNA replication efficiency and, conversely, POLN depletion led to increased perturbation of replication dynamics.

Importantly, rescue experiments by supplementing POLNdepleted cells with deoxyribonucleosides in the presence of fludarabine showed a complete restoration of the rate of DNA synthesis.

POLN encodes a low fidelity A-family Pol ν, whose cellular role still remains elusive although some repair activity has been proposed.22,23 Taking into consideration the high sequence homology between Pol ν and the polymerase

A

B

Figure 5. POLN and DNA synthesis in the presence of replicative stress due to a limited dNTP pool. (A) Flow cytometry cell cycle analysis of MEF POLN+/+ and MEF POLN+/- cell lines in the control condition or treated with fludarabine for 4 hours (h). Delta (D) is calculated as the difference of percent of cells present in the EdU+ gate by control condition (representative of 3 experiments). (B) Flow cytometry DNA synthesis analysis of MEF POLN+/+ and MEF POLN+/- cell lines, control and treated with hydroxyurea for 4 h. Delta (D) is calculated as difference of percent of cells present in the EdU+ gate by control condition, (representative of 3 experiments). (C) Flow cytometry DNA synthesis analysis of MEF POLN+/+ and MEF POLN+/- cell lines, control and treated with fludarabine for 8 h at 30 mM concentration and supplemented or not with deoxynucleosides (dNs). Delta (D) is calculated as a fold change of EdU incorporation by control condition (representative of 3 experiments).

C

haematologica | 2018; 103(6)

1045

S. Grgurevic et al. domain Pol θ, whose function in DNA double-stranded break repair has been more thoroughly established (reviewed by Wood and Doubliè24), one could suppose that Pol ν might play partially similar roles as its paralog.

Enhanced repair of double-stranded breaks by an excess of Pol ν in CLL might contribute to maintaining cell survival in managing double-stranded breaks occurring at stalled replication forks upon endogenous replication stress and fludarabine treatment.

We recently reported that, besides its documented microhomology-mediated end joining repair activity, Pol θ is capable of regulating the activity of replication origins by interacting with replication origin licensing factors and regulating the timing of replication initiation.25 Whether Pol ν could also regulate origin activity will be another issue to explore in the future since it may also give an alternative mechanistic basis for adaptive response to replication stress.

During disease pathogenesis, CLL leukemic cells succumb to an underlying level of replication stress. This fact is evidenced by the presence of markers of genomic instability, e.g. recurrent chromosomal abnormalities and common somatic mutations. We could hypothesize that, for the CLL cell entering the cell cycle and starting division inside the lymph node pseudofollicule, enhanced expression of POLN could be an adaptive mechanism to limit replication stress caused by a suddenly elevated requirement for dNTP. Our study suggests that the MEC-2 cell line, which shows characteristics of a CLL subclone that has recently exited

References 1. Hallek M. Chronic lymphocytic leukemia: 2017 update on diagnosis, risk stratification, and treatment. Am J Hematol.

2017;92(9):946-965.

2. Hanada M, Delia D, Aiello A, Stadtmauer E, Reed JC. bcl-2 gene hypomethylation and high-level expression in B-cell chronic lymphocytic leukemia.

Blood.

1993;82(6):1820-1828.

3. Messmer BT, Messmer D, Allen SL, et al.

In vivo measurements document the dynamic cellular kinetics of chronic lymphocytic leukemia B cells. J Clin Invest.

2005;115(3):755-764.

4. Os A, Bürgler S, Ribes AP, et al. Chronic lymphocytic leukemia cells are activated and proliferate in response to specific T helper cells. Cell Rep. 2013;4(3):566-577.

5. Obermann EC, Went P, Tzankov A, et al.

Cell cycle phase distribution analysis in chronic lymphocytic leukaemia: a significant number of cells reside in early G1phase. J Clin Pathol. 2007;60(7):794-797.

6. Döhner H, Stilgenbauer S, Benner A, et al.

Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):1910-1916.

7. Rossi D, Bruscaggin A, Spina V, et al.

Mutations of the SF3B1 splicing factor in chronic lymphocytic leukemia: association with progression and fludarabinerefractoriness. Blood. 2011;118(26):69046908.

8. Landau DA, Tausch E, Taylor-Weiner AN, et al. Mutations driving CLL and their evolution in progression and relapse. Nature.

2015;526(7574):525-530.

1046

the lymph node, expresses a higher level of POLN, allowing leukemic cells to surpass replication stress imposed by treatment with fludarabine. In this context, we could hypothesize that enhanced expression of POLN could be a characteristic acquired in the lymph node as an adaptive mechanism to endogenous replication stress and an advantageous trait once the leukemic cell has experienced treatment with fludarabine.

In conclusion, the A-family DNA polymerases, whose enhanced expression is observed frequently in solid cancers and, as we reveal here for the first time, in hematologic neoplasia, could be considered as a response to replication stress, contributing to both cancer progression and therapeutic resistance, which makes these enzymes attractive targets for future anti-cancer therapies.

Acknowledgments The authors would like to thank Drs. Richard Wood and Keiichi Takata (Department of Epigenetics and Molecular Carcinogenesis, University of Texas M.D. Anderson Cancer Center, Smithville, TX, USA) for the kind gift of MEF POLN cell lines, and Dr. Romain Guieze for providing CLL FMP 2007 RNA samples. Work in the laboratory of JSH is supported by funding from INCa-PLBIO 2016, ANR PRC 2016, Laboratoire d’Excellence Toulouse-Cancer (TOU-CAN) La Ligue Contre le Cancer (Equipe Labellisée 2017) and ITMO Cancer Aviesan within the framework of the Cancer Plan. SG was funded by TOU-CAN and the “Association Action Leucémies and Société Française d’Hématologie”.

9. Puente XS, Beà S, Valdés-Mas R, et al. Noncoding recurrent mutations in chronic lymphocytic leukaemia. Nature. 2015;526 (7574):519-524.

10. Gonzalez D, Martinez P, Wade R, et al.

Mutational status of the TP53 gene as a predictor of response and survival in patients with chronic lymphocytic leukemia: results from the LRF CLL4 trial. J Clin Oncol. 2011;29(16):2223-2229.

11. Bartkova J, Horejsí Z, Koed K, et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis.

Nature. 2005;434(7035):864-870.

12. Gorgoulis VG, Vassiliou L-VF, Karakaidos P, et al. Activation of the DNA damage checkpoint and genomic instability in human precancerous lesions. Nature.

2005;434(7035):907-913.

13. Halazonetis TD, Gorgoulis VG, Bartek J. An oncogene-induced DNA damage model for cancer development. science. 2008;319 (5868):1352-1356.

14. Keating BMJ, Kantorjian H, Brien SO, et al.

Fludarabine: a new agent with marked cytoreductive activity in untreated chronic lymphocytic leukemia. J Clin Oncol.

1991;9(1):44-49.

15. David L, Fernandez-Vidal A, Bertoli S, et al.

CHK1 as a therapeutic target to bypass chemoresistance in AML. Science Signal.

2016;9(445):ra90.

16. Lepretre S, Aurran T, Mahé B, et al. Excess mortality after treatment with fludarabine and cyclophosphamide in combination with alemtuzumab in previously untreated patients with chronic lymphocytic leukemia in a randomized phase 3 trial.

Blood. 2012;119(22):5104-5110.

17. Mazumdar M, Glassman JR.Categorizing a

18.

19.

20.

21.

22.

23.

24.

25.

prognostic variable: review of methods, code for easy implementation and applications to decision-making about cancer treatments. Stat Med. 2000;19(1):113-32.

Grgurevic S, Berquet L, Quillet-Mary A, et al. 3R gene expression in chronic lymphocytic leukemia reveals insight into disease evolution. Blood Cancer J. 2016;6(6):e429.

Stacchini A, Aragno M, Vallario A, et al.

MEC1 and MEC2: Two new cell lines derived from B-chronic lymphocytic leukaemia in prolymphocytoid transformation. Leuk Res. 1999;23(2):127-136.

Rasul E, Salamon D, Nagy N, et al. The MEC1 and MEC2 lines represent two CLL subclones in different stages of progression towards prolymphocytic leukemia. PLoS One. 2014;9(8):e106008.

Lacroix J, Pélofy S, Blatché C, et al.

Analysis of DNA replication by optical mapping in nanochannels. Small. 2016; 12(43):1-8.

Moldovan G-L, Madhavan MV, Mirchandani KD, McCaffrey RM, Vinciguerra P, D'Andrea AD. DNA polymerase POLN participates in cross-link repair and homologous recombination.

Mol Cell Biol. 2010;30(4):1088-1096.

Zietlow L, Smith LA, Bessho M, Bessho T. Evidence for the involvement of human DNA polymerase N in the repair of DNA interstrand cross-links. Biochemistry.

2009;48(49):11817-11824.

Wood RD, Doublié S. DNA polymerase θ (POLQ), double-strand break repair, and cancer. DNA Repair (Amst). 2016;44:22-32.

Fernandez-Vidal A, Guitton-Sert L, Cadoret J-C, et al. A role for DNA polymerase θ in the timing of DNA replication.

Nat Commun. 2014;5:4285.

haematologica | 2018; 103(6)

ARTICLE

Plasma Cell Disorders

Treatment to suppression of focal lesions on positron emission tomography-computed tomography is a therapeutic goal in newly diagnosed multiple myeloma

Ferrata Storti Foundation

Faith E. Davies,1 Adam Rosenthal,2 Leo Rasche,1 Nathan M. Petty,1 James E. McDonald,3 James A. Ntambi,3 Doug M. Steward,1 Susan B. Panozzo,1 Frits van Rhee,1 Maurizio Zangari,1 Carolina D. Schinke,1 Sharmilan Thanendrarajan,1 Brian Walker,1 Niels Weinhold,1 Bart Barlogie,1 Antje Hoering,2 and Gareth J. Morgan1

1 Myeloma Institute, University of Arkansas for Medical Sciences, Little Rock, AR; 2Cancer Research and Biostatistics, Seattle, WA and 3Department of Radiology, University of Arkansas for Medical Sciences, Little Rock, AR, USA

Haematologica 2018 Volume 103(6):1047-1053

ABSTRACT

F

luorine-18 fluorodeoxyglucose positron emission tomography with computed tomography attenuation correction (PET-CT) in myeloma can detect and enumerate focal lesions by the quantitative characterization of metabolic activity. The aim of this study was to determine the prognostic significance of the suppression of PET-CT activity at a number of time points post therapy initiation: day 7, post induction, post transplant, and at maintenance therapy. As part of the TT4-6 trial series, 596 patients underwent baseline PET-CT and were evaluated serially during their disease course using peak standardized uptake values above background red marrow signal. We demonstrate that the presence of more than 3 focal lesions at presentation identifies a group of patients with an adverse progression-free survival and overall survival. At day 7 of therapy, patients with complete focal lesion signal suppression revert to the same prognosis as those with no lesions at diagnosis. At later time points, the continued suppression of signal remains prognostically important. We conclude that for newly diagnosed patients with focal lesions, treatment until these lesions are suppressed is an important therapeutic goal as the prognosis of these patients is the same as those without lesions at diagnosis. (clinicaltrials.gov identifiers: 00734877, 02128230, 00869232, 00871013).

Correspondence: fedavies@uams.edu

Received: July 21, 2017.

Accepted: March 8, 2018.

Pre-published: March 22, 2018.

Introduction

doi:10.3324/haematol.2017.177139

A key strategy to improve outcomes in myeloma is to customize the treatment used based on the response to therapy. Such an approach is becoming increasingly feasible as the range of treatment options with different mechanisms of action increases. The number of tools available to monitor response to therapy is also increasing, with minimal residual disease (MRD) assessment of the bone marrow (BM) using flow cytometry and next generation sequencing being the most widely used.1,2 Imaging techniques such as magnetic resonance imaging (MRI) and fluorine-18 fluorodeoxyglucose positron emission tomography with computed tomography (FDG PET-CT) have also been used as a method to assess the extent and distribution of disease at presentation and pre and post autologous transplant.3-10 These two imaging approaches rely on different biological features of the tumor and as such offer important complementary information. Both technologies identify focal lesions (FLs), which are anatomical lesions seen during myeloma progression from monoclonal gammopathy of uncertain significance (MGUS) to plasma cell leukemia (PCL). They are more characteristic of the later stages of disease and are associated with adverse prognosis. However, in contrast to PET-CT, where the imaging features respond rapidly to exposure to therapy, classic MRI features are slow to resolve and can remain positive long term. Therefore, PET-CT is a useful monitoring tool for disease response.

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1047

haematologica | 2018; 103(6)

Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1047

F.E. Davies et al.

Table 1. Positron emission tomography with computed tomography sample availability of cases entered into the TT4-6 studies at the different timepoints of assessment. Note that some patients did not have GEP data available and therefore could not be classified by GEP70 risk.

Time points

Baseline + day 7 + end of induction + post-first TX + maintenance Baseline + day 7 + end of induction + post-first TX only Baseline + day 7 + end of induction + maintenance only Baseline + day 7 + post-first TX + maintenance only Baseline + end of induction + post-first TX + maintenance only Baseline + day 7 + end of induction only Baseline + day 7 + post-first TX only Baseline + day 7 + maintenance only Baseline + end of induction + post-first TX only Baseline + end of induction + maintenance only Baseline + post-first TX + maintenance only Baseline + day 7 only Baseline + end of induction only Baseline + post-first TX only Baseline + maintenance only Baseline only Total

N. of patients

N. of GEP70 low-risk patients

N. of GEP70 high-risk patients

67 49 29 42 32 58 33 11 21 14 25 60 46 28 15 66 596

41 40 20 38 23 40 31 9 15 9 23 46 31 27 14 60 467

16 4 6 1 8 12 1 0 5 3 1 7 6 0 0 1 71

GEP: gene expression profile.

Table 2. Patients’ characteristics, overall and by protocol.

Previously we have evaluated the role of PET-CT at presentation and have demonstrated that it can refine the assessment of prognosis, with both the number and size of FLs giving clinically useful prognostic information.3-7 We have also shown that total lesion glycolysis (TLG), a calculation that takes into account total disease volume and glucose metabolism, can improve the assessment of disease burden and outcome prediction.11 In order to further determine the value of PET-CT for disease monitoring and prognosis, we have utilized data collected in the TT4-TT6 clinical trials of our Total Therapy program,12,13 where PETCT assessment was included both at presentation and during response as part of the clinical protocol. In a preliminary analysis, we also explored the potential for PET-CT analysis to enhance the value of conventional response assessment and MRD flow cytometry assessment.

Female

Methods

CRP ≥8 mg/L

Patients

Creatinine ≥2 mg/dL

Of the 606 patients entered into the TT4-6 studies, 596 patients had PET-CT analysis available and were included in this study.

Treatment included combination chemotherapy as induction with double autologous transplantation, post-transplant consolidation, and three years planned maintenance with lenalidomide, bortezomib, and dexamethasone.12,13 Protocols were approved by the Institutional Review Board of the University of Arkansas for Medical Sciences. All patients signed informed consent in keeping with institutional, federal, and international guidelines. Gene expression analysis and risk status (GEP70) were determined.14,15 The number of patients for analysis at each landmark is shown in Table 1. The most common reason for a missing PET-CT was lack of health insurance to cover the costs of the test. The 3-year survival estimates with corresponding 95% confidence intervals were 68% (65, 72) for progression-free survival (PFS) and 82% (78, 85) for overall survival (OS). Median follow up was 5.1 years (Table 2).

1048

Factor Age ≥ 65 years IgA isotype

White Albumin < 3.5 g/dL β2M ≥ 3.5 mg/L β2M > 5.5 mg/L

Hb <10 g/dL LDH ≥190 U/L Cytogenetic abnormalities ISS Stage 1 ISS Stage 2 ISS Stage 3

All patients

TT4

TT5

TT6

198/596 (33%) 109/588 (19%) 238/596 (40%) 500/596 (84%) 248/595 (42%) 321/593 (54%) 152/593 (26%) 160/594 (27%) 29/595 (5%) 239/595 (40%) 121/595 (20%) 258/590 (44%) 193/593 (33%) 248/593 (42%) 152/593 (26%)

110/376 (29%) 65/370 (18%) 143/376 (38%) 321/376 (85%) 166/376 (44%) 194/374 (52%) 98/374 (26%) 99/375 (26%) 17/376 (5%) 151/376 (40%) 53/376 (14%) 148/370 (40%) 121/374 (32%) 155/374 (41%) 98/374 (26%)

21/72 (29%) 20/72 (28%) 31/72 (43%) 61/72 (85%) 37/72 (51%) 56/72 (78%) 31/72 (43%) 28/72 (39%) 7/72 (10%) 46/72 (64%) 28/72 (39%) 48/72 (67%) 11/72 (15%) 30/72 (42%) 31/72 (43%)

67/148 (45%) 24/146 (16%) 64/148 (43%) 118/148 (80%) 45/147 (31%) 71/147 (48%) 23/147 (16%) 33/147 (22%) 5/147 (3%) 42/147 (29%) 40/147 (27%) 62/148 (42%) 61/147 (41%) 63/147 (43%) 23/147 (16%)

n.: number; β2M: beta-2-microglobulin; CRP: C-reactive protein; Hb: hemoglobin; LDH: lactate dehydrogenase; ISS: International Staging System; IgA: immunoglobulin A.

haematologica | 2018; 103(6)

Prognostic significance of myeloma PET-CT focal lesions

PET-CT Scans were performed using a standard clinical protocol following 6-8 hours of fasting and after intravenous administration of 1015mCi (370-555Mbq) of fluorodeoxyglucose (FDG). After 50-70 minutes of uptake, images were acquired on either a CTI-Reveal or a Biograph 6 PET/CT system (Siemens Medical Systems), both with full ring LSO crystal configurations. PET images were generated by 3D iterative reconstruction on a 168x168 matrix, with a zoom of 1.0 FWHM filter of either 5.0 or 6.0 mm, and 2 iterations with 8 subsets. CT data were used for localization and attenuation correction. Images underwent a 3D region of interest (ROI) analysis of the axial and appendicular skeleton using the US Food and Drug Administration approved “Mirada Medical PET-CT XD Oncology Review” software (Oxford, UK). Background red marrow was defined using a 1 cm3 ROI in the most inferior vertebral body that did not demonstrate focally increased uptake. FLs were defined as areas, measuring at least 1 cm, not otherwise demonstrated to be artefacts by comparison with co-registered CT and exhibiting a peak SUV greater than the peak SUV for the background red marrow. Radiologists used a standardized approach for reporting. All data for analysis were extracted from clinical reports.

nique (CD138/CD38/CD19/CD45/CD27/CD81/CD56/CD20). A minimum of 2 million cells were analyzed, giving a sensitivity of 1 in 105.

Statistical analysis The Kaplan-Meier method16 was used to estimate OS and PFS distributions. Cumulative incidences by GEP70 risk for complete response (CR), very good partial response (VGPR) and partial response (PR) were calculated.17 Group comparisons (overall and pairwise) for survival end points and cumulative incidence were performed using the log-rank test.18 Cox proportional hazards modeling was used to identify the association of risk factors with outcome. OS was defined as time from landmark to death from any cause. PFS was calculated as time from landmark to progression, relapse, or death from any cause. Patients experiencing none of these events were censored at the date of last contact. Fisher’s exact test was used to evaluate the association between categorical variables. P<0.05 was considered statistically significant. Cutoff points for FL parameters were applied as previously reported.6

Results

Response assessment

PET-CT at presentation and outcome

Clinical response assessment was performed using International Myeloma Working Group (IMWG) definitions.1 Minimal residual disease assessment was performed on BMs using an 8-color tech-

The presence of more than 3 FLs detected on PET-CT scan at baseline was associated with adverse PFS (P<0.0001) and OS (P<0.0001) (Figure 1). There was no

A

B

C

Figure 1. Survival data according to number of focal lesions (FLs). Progression-free survival (PFS) (upper panel) and overall survival (OS) (lower panel) for patients entered into TT4-6 trials by the number of FL detected at presentation: (A) all patients, (B) GEP70 low-risk patients, and (C) GEP70 high-risk patients. A significant difference was observed for patients with FLs at baseline compared to patients with no FL at baseline for both PFS (P<0.0001) and OS (P<0.0001). These differences were significant when considering separately GEP70 low-risk patients (P=0.0007 for PFS, P<0.0001 for OS) and GEP70 high-risk patients (P=0.04 for PFS, P=0.05 for OS).

haematologica | 2018; 103(6)

1049

F.E. Davies et al.

significant difference in either PFS (P=0.3022) or OS (P=0.7842) between the patient groups with 0 and with 13 FLs (Table 3 and Online Supplementary Figure S1).

Suppression of FL signal at serial time points and its relationship to outcome We show that the suppression of FL signal following treatment is prognostically important. Patients achieving 100% suppression of FL signal following treatment at each time point studied (day 7, end of induction, post transplantation, and maintenance) have PFS and OS values that are not significantly different from cases with no FL present at baseline. Importantly, at each time point, patients with no detectable FL signal at that time point have a significantly superior outcome compared to patients with at least one detectable FL at that time point, irrespective of whether they had a FL at baseline (Table 4, Figure 2 and Online Supplementary Figure S2). Conversely, failure to suppress the FL signal (i.e. continued positivity) was seen in 46.4% of patients at day 7, 23.6% at the end of induction, 11.4% post transplantation, and 7.3% at maintenance, and was associated with an impaired outcome.

Interaction of GEP70 risk status with PET-CT signal suppression and outcome At presentation, 33.6% of GEP70 low-risk (LR) patients had more than 3 FLs and were associated with an adverse outcome (P=0.007 for PFS and P<0.001 for OS). A higher percentage of patients with FLs was seen in the GEP70 high-risk (HR) group at presentation (50.7%), and these cases also had an adverse outcome (P=0.04 for PFS and P=0.05 for OS) (Figure 1, Online Supplementary Table S1 and Online Supplementary Figure S3).

Following treatment, the suppression of FL signal had a similar impact in both risk strata with total suppression of signal being associated with outcomes that are not significantly different from cases with no FLs at baseline. For LR patients, this was significant at all time points analyzed.

In contrast, the differences in outcome were not as obvious in HR patients due to the smaller number of cases and their adverse outcomes irrespective of FL status at baseline. Nonetheless, we observed a significant difference in OS and PFS between patients with no FL at baseline and day 7 compared to patients with at least one FL at day 7, and we observed a non-significant trend in OS and PFS

A

B

C

D

Figure 2. Paired day 1, 7, and end of induction positron emission tomography with computed tomography (PET-CT). (A) Progression-free survival (PFS) and (B) overall survival (OS) for patients entered into TT4-6 trials with paired day 1 and day 7 PET-CT studies. An overall difference in PFS and OS was noted. A significant difference was observed for patients with no focal lesion(s) (FL) at baseline and no FL at day 7 compared to those with lesions present at day 7 in PFS (P=0.0002) and OS (P<0.0001). A significant difference was observed for patients with resolution of FL at day 7 compared to those with lesions present at day 7 in PFS (P=0.0001) and OS (P=0.0015). (C) PFS and (D) OS for patients entered into TT4-6 trials with paired day 1 and end of induction PET-CT studies. A significant difference was observed in PFS for patients with no FL at baseline and no FL at the end of induction compared to those with FL (P=0.0069). A significant difference was observed in PFS for patients with resolution of FL at this time point compared to those still with lesions (P=0.0064).

1050

haematologica | 2018; 103(6)

Prognostic significance of myeloma PET-CT focal lesions

between patients with suppression of baseline FL by day 7 compared to patients with at least one FL at day 7 (Online Supplementary Table S1 and Online Supplementary Figures S4 and S5). Multivariate analysis and R2 values suggest that both GEP and persistent FL positivity contribute to clinical outcome both at presentation and at subsequent time points, with presence of FLs making a very significant contribution to outcome (Table 5).

of MRD assessment; of these, 8 were MRD positive and 5 were MRD negative. This distribution of MRD was not significantly different from the distribution in cases with 0 FL (55 positive and 37 negative) (Fisherâ&#x20AC;&#x2122;s exact test P=0.90). This observation highlights the importance of combining imaging with MRD assessment.

Discussion Relationship between imaging response and minimal residual disease To address how imaging response relates to BM MRD, we looked at cases who had achieved a standard CR (as defined by the IMWG criteria) and had MRD assessment at the level of 1 in 104 performed by flow cytometry analysis. We identified 13 cases with 1 or more FLs at the time

We demonstrate in a large statistically robust data set that the serial use of PET-CT assessment can contribute to risk assessment and the prediction of outcome. We show that 62% of patients have PET-CT detectable FLs at diagnosis with a greater percentage in HR compared to LR patients. We show that, following modern day therapy,

Table 3. Progression-free and overall survival estimates at each positron emission tomography with computed tomography time point according to the number of focal lesions.

N. of focal lesions

3-year estimated progression-free survival % (95% CI)

3-year estimated overall survival % (95% CI)

0 1-3 >3 0 1-3 >3 0 1-3 >3 0 1-3 >3 0 1-3 >3

74 (68, 80) 74 (67, 81) 59 (52, 65) 76 (67, 86) 72 (63, 81) 53 (45, 60) 72 (64, 80) 73 (67, 79) 54 (44, 64) 74 (65, 84) 72 (65, 79) 57 (40, 74) 76 (65, 86) 66 (58, 74) 52 (28, 76)

89 (84, 93) 85 (79, 91) 72 (66, 78) 89 (82, 96) 86 (80, 93) 72 (64, 78) 88 (82, 94) 82 (76, 87) 71 (63, 80) 87 (80, 94) 80 (73, 86) 76 (61, 90) 88 (79, 96) 80 (73, 86) 60 (37, 83)

Presentation

Day 7

End of induction

Post transplant

Maintenance

N.: number; CI: Confidence Interval.

Table 4. P-value for progression-free and overall survival estimates for patients with and without lesions at each positron emission tomography with computed tomography time point.

>0 FL at day 7 vs. no lesions at baseline vs. lesion(s) at baseline, resolved by day 7 >0 FL at post induction vs. no lesions at baseline vs. lesion(s) at baseline, resolved by day 7 >0 FL at post transplant vs. no lesions at baseline vs. lesion(s) at baseline, resolved by day 7 >0 FL at maintenance vs. no lesions at baseline vs. lesion(s) at baseline, resolved by day 7

Progression-free survival P

Overall survival P

0.0002 0.0001

0.0001 0.0015

0.0069 0.0064

NS NS

0.0035 0.0070

NS NS

NS NS

0.0020 0.0187

FL: focal lesion; P: P-value; NS: not significant.

haematologica | 2018; 103(6)

1051

F.E. Davies et al.

the signal from FLs can be suppressed and that this is associated with improved outcomes. Even at the very early time point of 7 days post chemotherapy, the continuing presence of PET positivity is associated with an adverse outcome. The prognostic significance of ongoing FL positivity is maintained post one cycle of chemotherapy, post induction therapy, post transplantation, and during maintenance. Importantly, in the context of induction, transplant and maintenance, the 28% of patients who suppress PET-CT FL activity by day 7 or by the end of induction (46%) have a similar outcome to patients who had no FLs at diagnosis. These novel findings are clinically informative because they shift the emphasis of PET-CT assessment of FLs from a one-time diagnostic scan to a scenario where follow-up scanning is important to interpret the true prognostic significance of these lesions for the individual patient in the context of the therapy used and the biology of their cancer cells.

The current results expand on previous data analyses which have shown the value of the presence of FLs on PET-CT at diagnosis in MGUS, smoldering myeloma, and

myeloma.3,5-7,11,19-21 In myeloma, the number of lesions, maximum standardized uptake values (SUVmax), TLG, and metabolic tumor volume have all been shown to correlate with PFS and OS.3-8,11 In the current study, based on the analysis of 596 patients entered into TT4-TT6 clinical studies, we confirm these findings and show convincingly that the presence of more than 3 focal lesions detected on PET-CT at baseline is associated with adverse PFS and OS. We also clarify how such scanning technology should be used following the initiation of therapy.3-5,8 The Italian group used SUVmax as the marker of PET-CT positivity after induction treatment with bortezomib, thalidomide, and dexamethasone followed by autologous tandem transplant, and showed that 63% of patients who were PET-CT positive at diagnosis were still PET-CT positive at the end of induction therapy, and that this was linked with adverse clinical outcome.8 At three months post transplantation, positivity was seen in 35%, and again was associated with an adverse outcome. The Intergroup Francophone du Myelome (IFM) group4 used a combination of FLs and/or diffuse marrow signal to define PET-CT

Table 5. Multivariate analyses of progression-free and overall survival.

Variable At Day 7 GEP 70 high risk 0 FL at baseline + day 7 (vs. >0 FL at day 7) >0 FL at baseline, resolved by day 7 (vs. >0 FL at day 7) At end of induction GEP 70 high risk 0 FL at baseline + end of induction (vs. >0 FL at end of induction) >0 FL at baseline, resolved by end of induction (vs. >0 FL at end of induction) At post-first transplant GEP 70 high risk 0 FL at baseline + post-first TX (vs. >0 FL at post-first TX) >0 FL at baseline, resolved by post-first TX (vs. >0 FL at post-first TX) At maintenance GEP 70 high risk 0 FL at baseline + maintenance (vs. >0 FL at maintenance) >0 FL at baseline, resolved by maintenance (vs. >0 FL at maintenance)

n/N (%)

Progression-free survival HR (95% CI)

P R2

HR (95% CI)

Overall survival P R2

50/336 (15%) 82/240 (34%) 96/254 (38%)

3.91 (2.70, 5.66) 0.41 (0.27, 0.63) 0.41 (0.28, 0.62)

P≤0.001, R2=20.7% P≤0.001, R2=34.2% P≤0.001, R2=34.2%

4.64 (2.99, 7.19) 0.31 (0.16, 0.58) 0.43 (0.26, 0.72)

P≤0.001, R2=3.5% P≤0.001, R2=47.5% P=0.001, R2=47.5%

62/300 (21%) 81/207 (39%) 126/219 (58%)

3.45 (2.43, 4.90) 0.72 (0.47, 1.11) 0.67 (0.45, 0.98)

P≤0.001, R2=26.1% P=0.141, R2=28.5% P=0.039, R2 =28.5%

4.46 (2.93, 6.78) 0.78 (0.45, 1.35) 0.80 (0.50, 1.29)

P≤0.001, R2=39.3% P=0.377, R2=40.1% P=0.354, R2=40.1%

37/287 (13%) 91/126 (72%) 161/196 (82%)

4.94 (3.15, 7.77) 0.36 (0.21, 0.62) 0.36 (0.22, 0.60)

P≤0.001, R2=25.3% P≤0.001, R2=34.7% P≤0.001, R2=34.7%

6.19 (3.68, 10.40) 0.43 (0.20, 0.91) 0.46 (0.23, 0.90)

P≤0.001, R2=40.8% P=0.026, R2=45.0% P=0.024, R2=45.0%

35/223 (16%) 64/81 (79%) 142/159 (89%)

4.71 (2.98, 7.46) 0.32 (0.16, 0.68) 0.47 (0.24, 0.90)

P≤0.001, R2=28.8% P=0.003, R2=34.4% P=0.022, R2=34.4%

6.20 (3.59, 10.70) 0.22 (0.09, 0.52) 0.31 (0.15, 0.65)

P≤0.001, R2= 43.8% P≤0.001, R2=52.9% P=0.002, R2=52.9%

HR: Hazard Ratio; 95%CI: 95% Confidence Interval; P-value from Score c2 test in Cox Regression. R2: R-squared using method by O’Quigley and Xu. Multivariate results not statistically significant at 0.05 level. All univariate P-values reported regardless of significance. Multivariate model uses stepwise selection with entry level 0.1 and variable remains if the 0.05 level is met. A multivariate P-value greater than 0.05 indicates variable forced into model with significant variables chosen using stepwise selection.

1052

haematologica | 2018; 103(6)

Prognostic significance of myeloma PET-CT focal lesions

positivity. In their study, 68% of patients remained positive at the end of lenalidomide, bortezomib, and dexamethasone (RVD) induction, 42% after RVD consolidation and 25% after transplantation, with positivity being associated with an adverse outcome. In the TT3 study,5 the number of PET-FLs both at diagnosis and pre-transplant were important independent variables associated with adverse outcome.5 On multivariate analysis, more than 3 FLs at day 7 was associated with inferior OS and PFS, even in patients with GEP70 defined high risk. However, in TT3, we did not report the outcome of patients who suppressed their FL activity. The finding that these patients have outcomes similar to patients without FLs at diagnosis is of crucial clinical importance and suggests that treatment should be continued until lesion resolution.

Previous studies have shown that patients with a conventionally defined complete response using IMWG criteria may have persistence of the FLs after therapy.8,22 Such findings have led to the refinement of the IMWG definitions of complete response with the addition of assessment of MRD using flow cytometry, next generation sequencing, and imaging.2 Using an effective therapeutic strategy combining immunomodulatory drugs, protea-

References 1. Rajkumar SV, Dimopoulos MA, Palumbo A, et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol.

2014;15(12):e538-548.

2. Kumar S, Paiva B, Anderson KC, et al.

International Myeloma Working Group consensus criteria for response and minimal residual disease assessment in multiple myeloma. Lancet Oncol. 2016;17(8):e328346.

3. Usmani SZ, Mitchell A, Waheed S, et al.

Prognostic implications of serial 18-fluorodeoxyglucose emission tomography in multiple myeloma treated with total therapy 3. Blood. 2013;121(10):1819-1823.

4. Moreau P, Attal M, Caillot D, et al.

Prospective evaluation of magnetic resonance imaging and [18F]fluorodeoxyglucose positron emission tomography-computed tomography at diagnosis and before maintenance therapy in symptomatic patients with multiple myeloma included in the IFM/DFCI 2009 trial: results of the IMAJEM study.

J Clin Oncol.

2017;35(25):2911-2918.

5. Bartel TB, Haessler J, Brown TL, et al. F18fluorodeoxyglucose positron emission tomography in the context of other imaging techniques and prognostic factors in multiple myeloma.

Blood.

2009; 114(10):2068-2076.

6. Waheed S, Mitchell A, Usmani S, et al.

Standard and novel imaging methods for multiple myeloma: correlates with prognostic laboratory variables including gene expression profiling data. Haematologica.

2013;98(1):71-78.

7. Rasche L, Angtuaco E, McDonald JE, et al.

haematologica | 2018; 103(6)

8.

9.

10.

11.

12.

13.

14.

some inhibitors, and transplant, we were able to demonstrate that imaging gives additional information to both the clinical assessment of response using the IMWG criteria and also to MRD detection using a flow cytometric approach sensitive to 1 in 10-5. The recent study by the IFM group4 showed similar findings with 14 of 86 patients being PET-CT positive at the same time as they were MRD negative, suggesting that both techniques are essential to truly define a stringent response.

In other tumor settings, a PET-CT scan during therapy is used to guide treatment decisions, including continuing therapy, changing therapy to a modality with a different mechanism of action, or stopping treatment altogether.

Initiating the individualization of therapy in myeloma based on a comprehensive disease assessment is one way to improve patient outcomes. This study suggests that a risk-adapted approach based on serial PET-CT analysis would be appropriate for myeloma patients as it can reliably identify a group of patients with poor prognosis at different stages of their therapy who may benefit from alternative therapy. On the basis of our results, serial PETCT should be integrated into follow-up algorithms and risk-adapted clinical trials should be implemented.

Low expression of hexokinase-2 is associated with false-negative FDG-positron emission tomography in multiple myeloma. Blood. 2017;130(1):30-34.

Zamagni E, Patriarca F, Nanni C, et al.

Prognostic relevance of 18-F FDG PET/CT in newly diagnosed multiple myeloma patients treated with up-front autologous transplantation. Blood. 2011;118(23):59895995.

Hillengass J, Fechtner K, Weber MA, et al.

Prognostic significance of focal lesions in whole-body magnetic resonance imaging in patients with asymptomatic multiple myeloma. J Clin Oncol. 2010;28(9):16061610.

Walker R, Barlogie B, Haessler J, et al.

Magnetic resonance imaging in multiple myeloma: diagnostic and clinical implications. J Clin Oncol. 2007;25(9):1121-1128.

McDonald JE, Kessler MM, Gardner MW, et al. Assessment of total lesion glycolysis by 18F FDG PET/CT significantly improves prognostic value of GEP and ISS in myeloma.

Clin Cancer Res. 2017;23(8):1981-1987.

Jethava Y, Mitchell A, Epstein J, et al.

Adverse metaphase cytogenetics can be overcome by adding bortezomib and thalidomide to fractionated melphalan transplants.

Clin Cancer Res. 2017;23(11):2665-2672.

Jethava Y, Mitchell A, Zangari M, et al.

Dose-dense and less dose-intense Total Therapy 5 for gene expression profilingdefined high-risk multiple myeloma. Blood Cancer J. 2016;6(7):e453.

Nair B, van Rhee F, Shaughnessy JD Jr, et al.

Superior results of Total Therapy 3 (200333) in gene expression profiling-defined low-risk multiple myeloma confirmed in subsequent trial 2006-66 with VRD main-

tenance. Blood. 2010; 115(21):4168-4173.

15. Shaughnessy JD, Qu P, Tian E, et al.

Outcome with total therapy 3 (TT3) compared to total therapy 2 (TT2): Role of GEP70-defined high-risk disease with trisomy of 1q21 and activation of the proteasome gene PSMD4. J Clin Oncol.

2010;28:15s (suppl; abstr 8027).

16. Kaplan EL, Meier P. Nonparametric-estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457-481.

17. Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18(6):695-706.

18. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966;50(3):163-170.

19. Zamagni E, Nanni C, Gay F, et al. 18F-FDG PET/CT focal, but not osteolytic, lesions predict the progression of smoldering myeloma to active disease. Leukemia.

2016;30(2):417-422.

20. Dhodapkar MV, Sexton R, Waheed S, et al.

Clinical, genomic, and imaging predictors of myeloma progression from asymptomatic monoclonal gammopathies (SWOG S0120). Blood. 2014;123(1):78-85.

21. Bhutani M, Turkbey B, Tan E, et al. Bone marrow abnormalities and early bone lesions in multiple myeloma and its precursor disease: a prospective study using functional and morphologic imaging. Leuk Lymphoma. 2016;57(5):1114-1121.

22. Zamagni E, Nanni C, Mancuso K, et al.

PET/CT improves the definition of complete response and allows to detect otherwise unidentifiable skeletal progression in multiple myeloma. Clin Cancer Res. 2015;21(19):4384-4390.

1053

ARTICLE

Stem Cell Transplantation

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):1054-1064

A second-generation 15-PGDH inhibitor promotes bone marrow transplant recovery independently of age, transplant dose and granulocyte colony-stimulating factor support

Amar Desai,1 Yongyou Zhang,1 Youngsoo Park,1,2 Dawn M. Dawson,3 Gretchen A. Larusch,1 Lakshmi Kasturi,1 David Wald,3,4,5 Joseph M. Ready,6,7 Stanton L. Gerson1,3,* and Sanford D. Markowitz1,3,4*

Department of Medicine, Case Western Reserve University, Cleveland, OH, USA; Department of Pathology, Ulsan University College of Medicine, Asan Medical Center, Seoul, Republic of Korea; 3Case Comprehensive Cancer Center, Case Western Reserve University, Cleveland, OH, USA; 4University Hospitals Seidman Cancer Center, Cleveland, OH, USA; 5Department of Pathology, Case Western Reserve University, Cleveland, OH, USA; 6 Department of Biochemistry, University of Texas Southwestern Medical Center, Dallas, TX, USA and 7Simmons Cancer Center, University of Texas Southwestern Medical Center, Dallas, TX, USA 1 2

*SLG and SDM contributed equally to this work.

ABSTRACT

Correspondence: sxm10@case.edu

Received: August 9, 2017.

Accepted: February 20, 2018.

Pre-published: February 22, 2018.

doi:10.3324/haematol.2017.178376 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1054 Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1054

H

ematopoietic stem cell transplantation following myeloablative chemotherapy is a curative treatment for many hematopoietic malignancies. However, profound granulocytopenia during the interval between transplantation and marrow recovery exposes recipients to risks of fatal infection, a significant source of transplant-associated morbidity and mortality. We have previously described the discovery of a small molecule, SW033291, that potently inhibits the prostaglandin degrading enzyme 15-PGDH, increases bone marrow prostaglandin E2, and accelerates hematopoietic recovery following murine transplant.

Here we describe the efficacy of (+)- SW209415, a second-generation 15PGDH inhibitor, in an expanded range of models relevant to human transplantation. (+)-SW209415 is 10,000-fold more soluble, providing the potential for intravenous delivery, while maintaining potency in inhibiting 15-PGDH, increasing in vivo prostaglandin E2, and accelerating hematopoietic regeneration following transplantation. In additional models, (+)-SW209415: (i) demonstrated synergy with granulocyte colonystimulating factor, the current standard of care; (ii) maintained efficacy as transplant cell dose was escalated; (iii) maintained efficacy when transplant donors and recipients were aged; and (iv) potentiated homing in xenotransplants using human hematopoietic stem cells. (+)-SW209415 showed no adverse effects, no potentiation of in vivo growth of human myeloma and leukemia xenografts, and, on chronic high-dose administration, no toxicity as assessed by weight, blood counts and serum chemistry. These studies provide independent chemical confirmation of the activity of 15-PGDH inhibitors in potentiating hematopoietic recovery, extend the range of models in which inhibiting 15-PGDH demonstrates activity, allay concerns regarding potential for adverse effects from increasing prostaglandin E2, and thereby, advance 15-PGDH as a therapeutic target for potentiating hematopoietic stem cell transplantation.

haematologica | 2018; 103(6)

15-PGDH inhibition in multiple models of murine BMT

Introduction Hematopoietic stem cells (HSC) are rare, primitive cells in the blood and bone marrow which give rise to all differentiated blood cells.1-6 HSC are transplanted therapeutically after myeloablative chemotherapy as part of potentially curative treatment regimes for a variety of malignant and non-malignant disorders.7-10 However, profound neutropenia during the period while awaiting hematopoietic recovery after hematopoietic stem cell transplantation (HSCT) results in a high risk of opportunistic infection, which is a major source of morbidity, mortality, and prolonged hospitalization associated with the procedure.11,12 Strategies to accelerate hematopoietic regeneration following HSCT are thus of therapeutic interest. The eicosanoid signaling molecule prostaglandin E2 (PGE2), synthesized by the cyclo-oxygenase isoenzymes, has been recognized by several investigators as of interest for supporting hematopoietic regeneration. Ex vivo incubation of whole bone marrow with PGE2 has been shown to enhance HSC homing and proliferation in murine and non-human primate transplant models, and non-randomized clinical studies suggested potential for similar benefit in humans.13-17 In addition, in vivo treatment with PGE2 shows protective effects on murine hematopoietic cell populations following sublethal radiation via upregulation of cellular survival pathways.18,19 These observations suggest that in vivo modulation of PGE2 signaling could also potentiate hematopoietic recovery following HSCT. We have previously identified 15-hydroxyprostaglandin dehydrogenase (15-PGDH), the enzyme that mediates the first and rate-limiting step in PGE2 degradation in vivo, as a therapeutic target that, by acting on the bone marrow HSC niche, can potentiate hematopoietic recovery after HSCT.20 In particular, we described the discovery of SW033291, a potent small molecule 15-PGDH inhibitor (Ki = 0.1 nM).20 In in vivo studies, we have shown that, by inhibiting 15-PGDH, SW033291 doubles bone marrow PGE2 levels, induces bone marrow stromal production of CXCL12 and stem cell factor, potentiates HSC homing to the bone marrow niche, potentiates bone marrow colonyforming capacity, and, in murine HSCT, accelerates recovery of neutrophil counts by 6 days, enhances survival, and shows no long-term effects on serial transplantation capacity.20 In subsequent medicinal chemistry studies we have now developed SW209415, a second-generation 15PGDH inhibitor, which has a 10,000-fold enhanced aqueous solubility of 4300 mg/mL (as an HCl salt) as compared to SW033921.21 This improved aqueous solubility of SW209415 would enable administration in a humanacceptable intravenous formulation.21 SW209415 modifies SW033291 by substituting a dimethyl-imidazole in place of a previous phenyl group and substituting a thiazole ring in place of a prior thiophene (Online Supplementary Figure S1). Moreover, we identified that all of the 15-PGDH inhibitory activity of both SW033291 and SW209415 lies in their respective (+)-enantiomers.21 We herein now describe the in vivo biological activities of (+)-SW209415 in modulating tissue PGE2, and in enhancing recovery from HSCT in an expanded range of biological models relevant to human HSCT, and furthermore evaluate concerns that 15-PGDH inhibitors, by increasing PGE2, could potentiate the in vivo growth of cancers. These studies provide posi-

haematologica | 2018; 103(6)

tive findings that advance 15-PGDH as a therapeutic target for potentiating HSCT.

Methods Bone marrow homing Bone marrow homing to the recipient niche was measured by labeling donor marrow with 5 mM CellTrace CFSE (Life Technologies) for 30 min at 37°C and transplanting 10x106 labeled cells into lethally irradiated recipient mice (of the same age, gender, and strain). Recipient mice were give three intraperitoneal injections of vehicle or 5 mg/kg (+)-SW209415.

Colony-forming unit â&#x20AC;&#x201C; spleen assay Eight-week old C57BL/6J mice were lethally irradiated with 11 Gy and transplanted with 200x103 total bone marrow cells.

Recipients were treated intraperitoneally with either vehicle or 2.5 mg/kg (+)-SW209415 twice daily for 12 days. On day 12 mice were sacrificed and spleens harvested and assessed for colony counts and SKL determination.

Bone marrow transplantation Eight-week old female C57BL/6J mice were lethally irradiated with 11 Gy and transplanted with 500x103 total bone marrow cells 16 h after irradiation. Recipients were treated intraperitoneally with either vehicle or 2.5 mg/kg (+)-SW209415 twice daily through the course of the experiment. In studies involving granulocyte colony-stimulating factor (G-CSF), cohorts of mice were additionally treated once daily subcutaneously with 250 mg/kg human G-CSF or the combination of human G-CSF and (+)SW209415.

Human bone marrow and umbilical cord blood studies Discarded, de-identified human umbilical cord blood (3 unique samples) and adult bone marrow aspirates [2 unique samples from a 28-year old male (28/M) and a 50-year old male (50/M)] were obtained from the Case Western Reserve University Hematopoietic Biorepository with permission from the Institutional Review Board. Buffy coat from umbilical cord samples and total marrow from aspirates were incubated with carboxyfluorescein succinimidyl ester (CFSE) and the homing assay performed as described above. Engraftment at 12 weeks was studied by transplanting 1x106 mononuclear cells into each recipient NSG mouse. This was performed using two adult bone marrow aspirates (37 M and 41 M). Mice were treated twice daily for 21 days after the transplant with vehicle or 2.5 mg/kg (+)-SW209415 and bled at serial time-points through to day 84. Peripheral blood multilineage differentiation was assessed via flow cytometry (using CD3, B220, and CD11b gated from human CD45 cells).

Xenograft studies In vivo growth of human acute myeloid leukemia (AML) and human multiple myeloma (MM) was established by transplanting 5x106 total human AML cells (cultured from a human AML patient at University Hospitals) or MM cells (ATCC line RPMI 8226), via the tail vein, into NSG mice receiving 2.5 Gy irradiation.

Treatment was initiated 2 weeks after transplantation, with mice then started on twice daily intraperitoneal injections with either vehicle or 2.5 mg/kg (+)-SW209415 until the animals began to demonstrate noticeable signs of hunching and lethargy, at which point they were sacrificed to assess human CD45+ cells in the bone marrow of those that had received AML cells and human CD38+ cells in the bone marrow of those that had received MM cells.

1055

A. Desai et al.

Results (+)-SW209415 inhibits 15-PGDH and induces prostaglandin E2 in vivo We first examined the potency of (+)-SW209415 in inhibiting enzymatic activity of recombinant 15-PGDH protein in vitro in a titration experiment in which (+)SW209415 was added to ~5 nM 15-PGDH protein. The half minimal inhibitory concentration (IC50) of (+)SW209415 was 1.1 nM, consistent with this compound acting as a tight-binding inhibitor of 15-PGDH (and suggesting approximately 50% activity of the recombinant protein) (Figure 1A). A Morrison design experiment estimated that the Kiapp for (+)-SW209415 is 0.06 nM (Figure 1B). (+)-SW209415 thus retains the tight-binding inhibitor characteristics of the parent SW033291 compound. To test the potency in inhibiting 15-PGDH in cells, (+)-SW209415 was added to cultures of A549 cells stimulated with interleukin-1β, and increases in levels of PGE2 secreted into the culture media were determined. In this cell-based assay, (+)-SW209415 showed a half maximal effective concentration (EC50) of approximately 10 nM, demonstrating improved potency compared to SW033291 which had an EC50 of 40 nM (Figure 1C, arrows denote approximate EC50 values). Furthermore, (+)-SW209415 retained high potency in inducing PGE2 in vivo in the bone marrow of treated mice: mice treated in vivo with a single intraperitoneal dose of 2.5 mg/kg showing a 2-fold induction of bone marrow PGE2 at 2 and 3 h following treatment, with levels falling at 6 h and returning to baseline at 12 h after

A

C

1056

treatment (Figure 1D).21 This PGE2 response matches the persistent 2-fold elevation of PGE2 observed in the 15PGDH knockout mouse. Last, in vivo administration of (+)SW209415 recapitulated the induction of two cytokines, stem cell factor and CXCL12, in the CD45– bone marrow stromal cell population, which was previously demonstrated by SW03329120 (Online Supplementary Figure S2).

(+)-SW209415 promotes hematopoietic homing, engraftment and recovery after hematopoietic stem cell transplantation We next investigated the effects of (+)-SW209415 on steps in hematopoietic recovery following murine HSCT. We studied four distinct biological steps in recovery from HSCT: (i) homing of transplanted HSC to recipient marrow; (ii) expansion of transplanted HSC as hematopoietic colonies in the spleen; (iii) expansion of the stem cellenriched SKL population in the bone marrow; and (iv) production of mature neutrophils in the peripheral blood. We first investigated the effects of (+)-SW209415 on enhancing efficiency of HSC homing to bone marrow following transplantation. CFSE-labeled cells that successfully homed to the recipient’s bone marrow were quantified 16 h following the transplant. Three doses of (+)-SW209415 induced a 2-fold increase in donor marrow cell homing to the marrow cavity of recipient mice tibiae (P=0.0002), with an activity essentially identical to that of three doses of (+)-SW033291 (Figure 2A). Homing of SKL cells to the bone marrow niche was increased by 1.8-fold in (+)SW209415-treated mice (P=0.0029) (Online Supplementary

B

D

Figure 1. (+)-SW209415 inhibits 15-PGDH and induces prostaglandin E2 (PGE2) in vivo. (A) Increasing concentrations of (+)-SW209415 were added to 5 nM of 15-PGDH protein and enzyme activity was measured in duplicate determinations, with percent inhibition at each concentration depicted in the graph. (B) Graph of the relative initial 15PGDH enzyme reaction velocities Vi/V0 versus concentration of (+)-SW209415. Vi values are the averages of triplicate determinations. (C) PGE2 levels in medium of A549 cells following stimulation with interleukin-1β and treatment with increasing concentrations of (+)-SW209415. The arrows indicate the approximate EC50 values. The graph shows the means of four independent determinations ± Standard Error of Mean. (D) Eight-week old C57/Bl6J female mice were administered a single intraperitoneal dose of 2.5 mg/kg (+)-SW209415.

Animals were sacrificed at the time points indicated after the injection and bone marrow flushed for analysis of PGE2 via enzyme-linked immunosorbent assay (3 mice per timepoint). hr: hour.

haematologica | 2018; 103(6)

15-PGDH inhibition in multiple models of murine BMT

Figure S3). We next examined the effects of (+)-SW209415 on splenic hematopoiesis following HSCT, where engraftment of individual HSC in the spleen can be assessed by their generation of macroscopic multilineage hematopoietic colonies termed CFU-S. On post-transplant day 12, (+)-SW209415-treated mice demonstrated a significant increase in total spleen colony counts (1.53-fold, P=0.003), an increased total spleen weight (Online Supplementary Figure S4), and a significantly higher number of splenic SKL cells (2.32-fold, P=0.02) (Figure 2B). We observed a similar 2.75-fold expansion of bone marrow SKL cells on day 18 in the same arm of (+)-SW209415-treated mice (P=0.0002). Thus (+)-SW209415 potentiates engraftment and expansion of donor hematopoietic cells in both bone marrow and spleen. Last, we examined the effect of (+)SW209415 on the recovery of blood counts. (+)SW209415-treated mice attained double the neutrophil

counts of vehicle-treated control animals at each of postHSCT days 8 (P<0.0001), 12 (P<0.0001), and 18 (P<0.0001) (Figure 2C). In addition, (+)-SW209415-treated mice demonstrated higher platelet counts on each of days 8 (P=0.005), 12 (P=0.06), and 18 (P=0.0005) and, in particular, showed a significantly higher day 8 nadir count (98x109/L in (+)-SW209415 treated mice versus 73x109/L in vehicle-treated controls; P=0.005) (Online Supplementary Figure S5).

To further distinguish differential effects of (+)SW209415 on early versus later steps in hematopoietic reconstitution, we compared the effects of giving (+)SW209415 for only the first 7 days following bone marrow transplantation (BMT) versus maintaining continuous dosing of (+)-SW209415 for 18 days. In this experiment mice that received continuous twice-daily doses of (+)SW209415 attained neutrophil counts almost 3-fold high-

A

B

C

Figure 2. (+)-SW209415 promotes early hematopoietic homing, engraftment and recovery after bone marrow transplantation (BMT). (A) Lethally irradiated (11 Gy) 8-week old female C57/BL6J mice were transplanted with 107 CFSE-labeled total bone marrow cells and treated intraperitoneally with either vehicle or 2.5 mg/kg (+)-SW209415 immediately after irradiation, after BMT, and 8 hours (h) after BMT. Mice were sacrificed 16 h after the BMT and CFSE+ cells were measured in two femora + two tibiae/mouse. The graph shows mean values and Standard Error of Mean (SEM) (6 mice/arm). *Student t-test; P<0.05. (B) Lethally irradiated (11 Gy) 8-week old female C57/BL6J mice were transplanted with 200x103 total bone marrow cells and treated as in (A). The mice were sacrificed on day 12 after BMT. The spleens were harvested, fixed in Bouin solution, and the colonies counted (6 mice/arm). A second cohort of mice were sacrificed on day 12 and a single cell suspension prepared for SKL analysis via FACS (3 mice/arm). Representative photographs of day 12 spleens are shown on the left, with means and SEM of CFU-S and SKL numbers represented graphically on the right. *Student t-test; with a P<0.05. (C) Lethally irradiated (11 Gy) 8-week old female C57/BL6J mice were transplanted with 500x103 total bone marrow cells and treated intraperitoneally with either vehicle or 2.5 mg/kg (+)-SW209415 (twice daily). Peripheral blood counts were measured on days 8, 12, and 18 after BMT. The graph shows the mean values and SEM (13 mice/arm). *Student t-test; P<0.05.

haematologica | 2018; 103(6)

1057

A. Desai et al.

er on days 12 and 18 after BMT than those achieved in the vehicle-treated control animals, while mice that received only 7 days of (+)-SW209415 showed no significant difference in neutrophil counts on day 12 and a moderate increase on day 18 (Online Supplementary Figure S6).

Moreover, while mice that received (+)-SW209415 continuously had a nearly 4-fold increase in bone marrow SKL cells (from 3707 to 14007, P=0.0026), mice that received only 7 days of (+)-SW209415 had only a 1.54-fold increase that did not reach statistical significance (Online Supplementary Figure S7). Collectively, these observations suggest that continuous in vivo dosing with the 15-PGDH inhibitor (+)-SW209415 optimally promotes improvement in hematopoietic recovery following murine BMT.

(+)-SW209415 accelerates neutrophil recovery after bone marrow transplantation even in mice treated with granulocyte colony-stimulating factor G-CSF (clinical names, Lenograstim and Filgrastim) is a Food and Drug Administration-approved glycoprotein that is the standard of care used to accelerate granulocyte recovery after human HSCT. As initial human studies of (+)-SW209415 would almost certainly be done in HSCT patients also receiving G-CSF, we used mice to model the effects of (+)-SW209415 in comparison to and in combination with G-CSF. On post-transplant day 8 neutrophil counts were significantly higher in (+)-SW209415-treated mice (339x103 cells/mL, P<0.0001), and also in G-CSFtreated mice (295x103 cells/mL, P<0.0001), than in mice

treated with vehicle control (150x103 cells/mL). However, mice treated with the combination of (+)-SW209415 and G-CSF developed 466x103 neutrophils/mL, which was significantly higher than the levels in mice treated with either (+)-SW209415 (P=0.0041) or G-CSF (P<0.0001) alone (Figure 3). Moreover, although, as compared to vehicle control, either (+)-SW209415 or G-CSF alone significantly increased neutrophil counts on post-HSCT days 8, 12 and 18, mice treated with the combination of (+)-SW209415 and G-CSF showed the highest neutrophil counts on each of these days (P=0.002 for the combination vs. (+)SW209415, P=0.01 for the combination vs. G-CSF) (Figure 3). Furthermore, on day 12, the increase in the number of neutrophils in mice treated with the combination of (+)SW209415 and G-CSF (an increase of 963x103 cells/mL over that in vehicle-treated mice) essentially equaled the additive increments of the two agents when given individually (412x103/mL for (+)-SW209415 and 555x103/mL for GCSF, or the predicted additive effect of 967x103 cells/mL).

Similarly, on day 8, the increase in number of neutrophils in mice treated with the combination of (+)-SW209415 and G-CSF (an increase of 316x103 cells/mL over that in the vehicle-treated mice) nearly equaled the additive increments produced by the two agents when given individually (189x103/mL for (+)-SW209415 and 145x103/mL for GCSF, or the additive effect of 334x103 cells/mL).

Furthermore, on post-HSCT day 18, mice treated with the combination of (+)-SW209415 plus G-CSF also showed the highest numbers of bone marrow SKL cells (Figure 3).

Figure 3. (+)-SW209415 accelerates neutrophil recovery after bone marrow transplantation (BMT) in mice treated with granulocyte colony-stimulating factor. Lethally irradiated (11 Gy) 8week old female C57/BL6J mice were transplanted with 500x103 total bone marrow cells and treated with either vehicle, 2.5 mg/kg (+)-SW209415 (twice daily, intraperitoneally), 250 mg/kg human G-CSF (once daily, subcutaneously), or the combination. The graphs display the means and Standard Error of Mean of peripheral blood neutrophil counts on days 8, 12, and 18 after BMT and bone marrow SKL cell numbers determined in mice sacrificed on day 18 (13 mice/arm). *Student t-test P<0.05.

1058

haematologica | 2018; 103(6)

15-PGDH inhibition in multiple models of murine BMT

In overview, treatment with (+)-SW209415 significantly accelerates neutrophil recovery after HSCT even in mice that are also receiving G-CSF, suggesting that this approach can provide an additive benefit over current standard therapy.

(+)-SW209415 is effective in promoting hematopoietic recovery after bone marrow transplantation in aged mice Our prior studies demonstrating the effects of the 15PGDH inhibitor SW033291 on accelerating hematopoietic reconstitution were performed with donor and recipient mice of 8-10 weeks of age, and thus roughly approximating humans of late teenage years.22 Humans undergoing HSCT are, however, primarily middle-aged and older, and the potency of human HSC in hematopoietic reconstitution is recognized to decline with age.23-25 To investigate potential age-related dependencies of the effects of 15PGDH inhibition, both in inducing tissue PGE2 and in promoting hematopoietic reconstitution, we also studied mice aged 52 weeks old, an age at which females are approaching reproductive senescence and exhibiting other biological changes of late middle-aged humans.

Reassuringly, in these older mice a single dose of (+)SW209415 induced a 2-fold increase in bone marrow PGE2 at 3 h after injection (Figure 4A), essentially recapitulating the same pharmacodynamic effect demonstrated in 8-week old mice (Figure 1D). We next investigated the activity of (+)-SW209415 in accelerating recovery from

HSCT in these older mice, performing murine HSCT in which both the bone marrow donor and lethally irradiated bone marrow recipient mice were 52 weeks old. Most notably, compared to control older mice, (+)-SW209415treated older mice had double the number of neutrophils on each of days 8, 12, and 18 after transplantation (P=0.01) (Figure 4B). Moreover, we also observed an ~2-fold upward trend in the total number of SKL cells in the (+)SW209415-treated cohort (Online Supplementary Figure S7).

In overview, these observations confirm that the approach of inhibiting 15-PGDH to accelerate recovery from HSCT remains efficacious even when both transplant donor and recipient animals are older, which may model older persons.

(+)-SW209415 accelerates hematopoietic recovery in hematopoietic stem cell transplant recipient mice that receive a â&#x20AC;&#x153;high-doseâ&#x20AC;? inoculum of donor cells (+)-SW209415 demonstrates activity in accelerating hematopoietic recovery in multiple different models of HSCT. However, hematopoietic recovery following HSCT can also be accelerated by providing a larger inoculum of transplanted donor cells. To investigate whether the benefits from inhibiting 15-PGDH are independent of the dose of transplanted donor cells, we re-examined our initial studies of murine HSCT using an escalated dose of 2x106 donor bone marrow cells. As expected, compared to mice transplanted with 500,000 donor cells, mice receiving a larger inoculum of 2x106 donor cells showed greater

Figure 4. (+)-SW209415 is effective in promoting hematopoietic recovery after bone marrow transplantation (BMT) in aged mice. (A) Fifty-two-week old aged C57/Bl6J female mice were administered a single intraperitoneal dose of 2.5 mg/kg (+)-SW209415. Animals were sacrificed at the time points indicated after injection and bone marrow flushed for analysis of PGE2 via enzyme-linked immunosorbent assay (3 mice/arm). (B) Fifty-two-week old female donor and lethally irradiated (11 Gy) recipient C57/BL6J aged mice were used in this study.

Mice were transplanted with 500x103 total bone marrow cells and treated with either vehicle or 2.5 mg/kg (+)-SW209415 (twice daily, intraperitoneally). The graphs display the means and Standard Error of Mean of peripheral blood neutrophil counts on day 8, 12, and 18 after BMT (10 mice/arm). *Student t-test P<0.05.

A

B

haematologica | 2018; 103(6)

1059

A. Desai et al.

neutrophil recovery on each of post-transplant days 8 (P=0.009), 12 (P<0.0001), and 18 (P=0.01), achieving an effect quite similar to that of accelerating a 500,000 donor cell transplant with (+)-SW209415 (Figure 5). Notably though, treatment with (+)-SW209415 significantly further accelerated hematopoietic recovery even in mice transplanted with the higher cell dose. Indeed, mice receiving 2x106 donor cells plus (+)-SW209415 showed greater neutrophil recovery than corresponding mice receiving vehicle control on each of post-HSCT days 8 (P=0.04), 12 (P=0.0019), and 18 (P=0.04) (Figure 5).

Furthermore, in mice transplanted with 2x106 donor bone marrow cells, added treatment with (+)-SW209415 increased post-HSCT day 18 numbers of marrow SKL cells by 55%, which is approximately the same proportion as the corresponding day 18 increase in neutrophil counts (Online Supplementary Figure S8). These data demonstrate that (+)-SW209415 continues to provide acceleration of hematopoietic recovery from HSCT even when the inoculum of transplanted donor cells is markedly increased.

(+)-SW209415 promotes bone marrow homing of human umbilical cord and human bone marrow hematopoietic stem cells To further model the potential efficacy of (+)-SW209415 in human HSCT, we examined the effects of (+)SW209415 in accelerating bone marrow homing of xenotransplanted human bone marrow and human umbilical

cord blood cells. In these studies, lethally irradiated immunodeficient NSG mice received inocula of CFSElabeled total human cells, and the frequency of the labeled cells was determined at 16 h after transplantation in recipient mouse femora and tibiae. We found that administering (+)-SW209415 increased homing of human bone marrow aspirate-derived cells by ~1.78-fold (P=0.0008) in a pooled analysis of two donors and increased homing of human umbilical cord blood-derived cells by ~1.71-fold (P=0.0001) in a pooled analysis of three donors (Figure 6A). A similar ~2-fold effect of (+)-SW209415 on increasing homing of purified human CD34+ cells was obtained in mice that were transplanted with CFSE-labeled isolated human CD34+ cells from two bone marrow aspirate donors (Online Supplementary Figure S9). Beneficial effects of (+)-SW209415 were maintained in follow-up studies assessing 12-week engraftment of human cells. In these experiments, NSG mice were transplanted with 1x106 mononuclear cells/mouse from total bone marrow aspirates of 2 donors and effects of (+)SW209415 human engraftment were assessed 84 days later. Treatment with (+)-SW209415 on days 1-21 days after BMT resulted in a 3.16-fold increase in day 84 human CD45+ bone marrow cells (P=0.024) (Online Supplementary Figure S10) and a similar 2- to 3-fold sustained increase in human peripheral blood chimerism from days 30-84 (P<0.03) (Figure 6B).

CD45+ human-derived cells were identified in each of the myeloid, T-cell, and B-cell compartments, at frequencies

Figure 5. (+)-SW209415 accelerates hematopoietic recovery in hematopoietic stem cell transplantation (HSCT) recipient mice that receive â&#x20AC;&#x153;high-doseâ&#x20AC;? donor cell inoculum. Lethally irradiated (11 Gy) 8-week old female C57/BL6J mice were transplanted with either 500 x103 (blue bars) or 2x106 (red bars) total bone marrow cells and treated with either vehicle (V), or 2.5 mg/kg (+)- SW209415 (S). Graphs display means and Standard Error of Mean of peripheral blood neutrophil counts as measured on days 8, 12, and 18 after BMT (12 mice/arm). *Student t-test P<0.05.

1060

haematologica | 2018; 103(6)

15-PGDH inhibition in multiple models of murine BMT

ranging from 3-10% of human cells (data not shown). These data support that effects of (+)-SW209415 in accelerating HSCT recovery will translate to humans as well.

(+)-SW209415 does not affect in vivo growth of human hematopoietic cancer cells Although hematopoietic growth factors such as G-CSF have been safely administered to human HSCT patients with hematologic malignancies, concerns have been raised regarding potential effects of increasing bone marrow PGE2 on growth of human cancer cells.26 To address this issue, we examined the effects of (+)-SW209415 on in vivo growth of xenografted cancer cells from a human AML and from a human MM, two cancers that are commonly treated by HSCT. We established cohorts of NSG mice xenotransplanted with primitive human AML blasts or with human MM cells, and treated these mice with (+)SW209415 2.5 mg/kg twice daily for 21 days, or with vehicle control. In both AML and MM xenotransplanted cohorts, administering (+)-SW209415 showed no effect on the rate of decline of animal weights and no effect on the number of human cancer cells that expanded in the mouse bone marrow (Figure 7A,B). While further studies with additional hematopoietic tumor models are warranted, these first experimental results do not support the hypothesis that a doubling of bone marrow PGE2 will potentiate the in vivo growth of human cancer cells.

Humans with homozygous inactivating mutations of 15-PGDH show chronically elevated PGE2 levels, but except for development of digital clubbing are otherwise phenotypically normal.27-32 However, a subset of individuals with mutations that inactivate SLCO2A1, the transporter that carries PGE2 to 15-PGDH, have been reported to develop myelofibrosis.33 As myelofibrosis has been observed in only some of these individuals, the question has arisen as to whether this finding is related to the increased PGE2 seen with loss of SLCO2A1, or whether it might be due to other genetic defects in these individuals who were all offspring of consanguineous marriages. To investigate this issue, we examined bone marrow from mice following 30 days of twice daily treatment with 2.5 mg/kg (+)-SW209415 or with vehicle control. The potential development of bone marrow fibrosis was assessed by reticulin staining. Essentially no myelofibrosis was detected in either control or (+)-SW209415-treated mice (Online Supplementary Figure S11). While longer term studies remain warranted, these initial results mitigate concerns of bone marrow fibrosis as a potential toxicity from treatment with (+)-SW209415.

(+)-SW209415 does not show off-target toxicity As an additional evaluation, we examined for potential off-target toxicities that might be induced by (+)SW209415. Mice were treated twice daily for 21 days

A

B

Figure 6. (+)-SW209415 promotes bone marrow homing of human umbilical cord and human bone marrow hematopoiesis stem cells. (A) Irradiated (2.5Gy) 8week old female NSG mice were transplanted with 20*106 CFSE-labeled total human umbilical cord blood buffy coat (3 samples transplanted into 4 mice/condition/sample) or total human marrow aspirate cells (2 samples transplanted into 3 mice/condition/sample). Mice were treated with either vehicle or 2.5 mg/kg (+)-SW209415 immediately after irradiation, after bone marrow transplantation (BMT), and 8 hours (h) after BMT. Graphs show mean and Standard Error of Mean for % of CFSE-labeled human donor cells homing to murine marrow. *Student t-test P<0.05. (B) Irradiated (2.5 Gy) 8-week old male NSG mice were transplanted with 1*106 white blood cells from 2 unique bone marrow aspirate donors and treated with either vehicle or 2.5mg/kg (+)-SW209415 twice daily for 21 days.

Peripheral blood was collected at multiple intervals after transplant and human CD45 measured via flow cytometry.

haematologica | 2018; 103(6)

1061

A. Desai et al.

with vehicle or with escalating doses of (+)-SW209415 over a range of doses from the therapeutically active dose of 2.5 mg/kg up to 10-fold above this, at 25 mg/kg. Parallel pharmacokinetic experiments confirmed dose-proportional increases in maximum and total plasma exposure (Cmax, area under the curve) over this dosing range.

Administering (+)-SW209415 did not produce any adverse effects on mouse weights (Online Supplementary Figure S12), activity, or grooming. Additionally, administering (+)-SW209415 did not have adverse effects on peripheral blood counts (Online Supplementary Figure S12) or on any of a battery of serum chemistry values (Online Supplementary Table S1). These data suggest that (+)SW209415 is without off-target toxicity at doses of up to 10-fold over the therapeutically active range.

Discussion In summary, (+)-SW209415, a second-generation 15PGDH inhibitor, provides independent chemical support that 15-PGDH is a therapeutic target whose inhibition accelerates hematopoietic recovery following HSCT. Moreover, these studies show that (+)-SW209415 advantages hematopoietic recovery across a broad range of models that newly interrogate key features of human

HSCT. Firstly, these studies demonstrate that (+)SW209415 accelerates neutrophil recovery after HSCT in mice that are also treated with G-CSF, which supports that (+)-SW209415 will provide added benefit in human HSCT. Secondly, these studies demonstrate that (+)SW209415 promotes hematopoietic recovery following HSCT in models in which both HSC donors and HSCT recipients are older. The current findings demonstrate that the pharmacodynamic activity of 15-PGDH and the downstream responses that potentiate bone marrow recovery both remain robust in older animals. Thirdly, these findings show that (+)-SW209415 advantages hematopoietic recovery, even when recovery is independently accelerated by administering a 4-fold increased number of donor HSC. This observation also suggests that (+)SW209415 would provide benefit at doses of HSC typically employed in human HSCT. Fourthly, and further supporting potential applicability of these findings to human HSCT, (+)-SW209415 potentiates homing of human HSC (that have multilineage differentiation capacity) when assayed by xenotransplantation into lethally irradiated NSG mice. Our further findings that the efficacy of (+)SW209415 is maximal when the drug is administered twice daily throughout the full period of bone marrow recovery informs the optimal method for administering 15-PGDH inhibitors in HSCT and also supports that the

A

B

Figure 7. (+)-SW209415 does not effect in vivo growth of human hematopoietic cancer cells. Irradiated (2.5Gy) 8-week old male NSG mice were transplanted with 5*106 human acute myeloid leukemia (AML) (A) or multiple myeloma (MM) cells (B). Two weeks after transplant mice began intraperitoneal treatment with either vehicle or 2.5 mg/kg (+)-SW209415 twice daily for 21 days. Upon signs of impending animal death (hunching, limited mobility) mice were sacrificed and the frequency of human CD45+ cells (for AML) or human CD38+ cells (for MM) in mouse marrow was measured via flow cytometry (4 mice/arm). Graphs show means and Standard Error of Mean for animal weights and for human cancer cell numbers in the mouse bone marrow.

1062

haematologica | 2018; 103(6)

15-PGDH inhibition in multiple models of murine BMT

mechanism of action of (+)-SW209415 is distinct from approaches that have examined acute ex vivo exposure of donor cells to PGE2 analogs prior to graft administration.1417 These findings, across multiple different models, are consistent with and extend our prior observations with SW033291, a first-generation 15-PGDH inhibitor, that demonstrated inhibiting 15-PGDH increased bone marrow stem cells as assessed by both cell surface markers and by marrow colony-forming assays.20 These demonstrations of efficacy are further buttressed by findings supporting the safety of (+)-SW209415, both when tested for evidence of hypothesized or potential ontarget toxicities and when tested for off-target effects. We have previously shown that administering a 15-PGDH inhibitor to mice receiving a bone marrow transplant had no adverse effects, as assayed by ability to serially transplant the regenerated bone marrow, and also did not show any evidence of tumor induction during 7 months following whole body irradiation.20 These new studies also produce no evidence that inhibiting tissue 15-PGDH has any effect on in vivo growth of either human MM cells or human AML cells. Although studies in additional human tumor models remain warranted, the absence of any effect of increased PGE2 on the in vivo growth of these cancer cells allays, at least, in part the hypothetical concern of this as a potential on-target toxicity of therapies employing 15PGDH inhibitors.26 This observed lack of effect of PGE2 modulation on growth of established human cancer cells is consistent with clinical observations of a lack of effects

References 1. Rossi L, Lin KK, Boles NC, et al. Less is more: unveiling the functional core of hematopoietic stem cells through knockout mice. Cell Stem Cell. 2012;11(3):302-317.

2. Calvi LM, Link DC. The hematopoietic stem cell niche in homeostasis and disease.

Blood. 2015;126(22):2443-2451.

3. Cheshier SH, Morrison SJ, Liao X, Weissman IL. In vivo proliferation and cell cycle kinetics of long-term self-renewing hematopoietic stem cells. Proc Natl Acad Sci USA. 1999;96(6):3120-3125.

4. Nygren JM, Bryder D. A novel assay to trace proliferation history in vivo reveals that enhanced divisional kinetics accompany loss of hematopoietic stem cell self-renewal.

PLoS One. 2008;3(11):e3710.

5. Morrison SJ, Scadden DT. The bone marrow niche for haematopoietic stem cells. Nature.

2014;505(7483):327-334.

6. Passegue E, Jamieson CH, Ailles LE, Weissman IL. Normal and leukemic hematopoiesis: are leukemias a stem cell disorder or a reacquisition of stem cell characteristics? Proc Natl Acad Sci USA. 2003;100 Suppl 1:11842-11849.

7. Rocha V, Cornish J, Sievers EL, et al.

Comparison of outcomes of unrelated bone marrow and umbilical cord blood transplants in children with acute leukemia.

Blood. 2001;97(10):2962-2971.

8. Bird JM, Russell NH, Samson D. Minimal residual disease after bone marrow transplantation for multiple myeloma: evidence for cure in long-term survivors. Bone

haematologica | 2018; 103(6)

of non-steroidal anti-inflammatory drugs, which lower tissue PGE2, in clinical trials testing these drugs as therapeutic agents for treating established cancers,34-37 and is consistent with prior hypotheses that oncogene activation may render tumor stem cells independent of PGE2 signaling.38 Furthermore, we found no evidence of off-target toxicity associated with (+)-SW209415 when chronically administered at doses 10-fold above the dose that is therapeutically effective in potentiating BMT recovery, thus demonstrating a substantial therapeutic index for this chemical scaffold.

In conclusion, our results identify (+)-SW209415 as a promising second-generation 15-PGDH inhibitor that validates the efficacy and safety of targeting 15-PGDH in multiple murine models and thereby advances 15-PGDH as a therapeutic target for potentiating HSCT. Funding This work was funded by NIH grants R35 CA197442, R01 216863, and U54 HL119810, by a Harrington Discovery Institute Scholar Innovator Award, by an award from the Case Western Reserve University Council to Advance Human Health, by the Welch Foundation (I-1612), and by a sponsored research agreement to Case Western Reserve University from Rodeo Therapeutics. This research was also supported by the Radiation Resources Core Facility (P30CA043703), the Hematopoietic Biorepository and Cellular Therapy Core Facility (P30CA043703), and the Cytometry & Imaging Microscopy Core Facility of Case Comprehensive Cancer Center (P30CA043703).

Marrow Transplant. 1993;12(6):651-654.

9. Gratwohl A, Baldomero H, Aljurf M, et al.

Hematopoietic stem cell transplantation: a global perspective. JAMA. 2010;303(16): 1617-1624.

10. Shenoy S. Hematopoietic stem-cell transplantation for sickle cell disease: current evidence and opinions. Ther Adv Hematol.

2013;4(5):335-344.

11. Gluckman E, Auerbach AD, Horowitz MM, et al. Bone marrow transplantation for Fanconi anemia. Blood. 1995;86(7):28562862.

12. Passweg JR, Baldomero H, Bader P, et al.

Hematopoietic SCT in Europe 2013: recent trends in the use of alternative donors showing more haploidentical donors but fewer cord blood transplants. Bone Marrow Transplant. 2015;50(4):476-482.

13. North TE, Goessling W, Walkley CR, et al.

Prostaglandin E2 regulates vertebrate haematopoietic stem cell homeostasis.

Nature. 2007;447(7147):1007-1011.

14. Hagedorn EJ, Durand EM, Fast EM, Zon LI. Getting more for your marrow: boosting hematopoietic stem cell numbers with PGE2. Exp Cell Res. 2014;329(2):220-226.

15. Hoggatt J, Singh P, Sampath J, Pelus LM. Prostaglandin E2 enhances hematopoietic stem cell homing, survival, and proliferation.

Blood. 2009;113(22):5444-5455.

16. Goessling W, Allen RS, Guan X, et al.

Prostaglandin E2 enhances human cord blood stem cell xenotransplants and shows long-term safety in preclinical nonhuman primate transplant models. Cell Stem Cell.

2011;8(4):445-458.

17. Cutler C, Multani P, Robbins D, et al.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Prostaglandin-modulated umbilical cord blood hematopoietic stem cell transplantation. Blood. 2013;122(17):3074-3081.

Porter RL, Georger MA, Bromberg O, et al.

Prostaglandin E2 increases hematopoietic stem cell survival and accelerates hematopoietic recovery after radiation injury. Stem Cells. 2013;31(2):372-383.

Hoggatt J, Singh P, Stilger KN, et al.

Recovery from hematopoietic injury by modulating prostaglandin E(2) signaling post-irradiation. Blood Cells Mol Dis. 2013;50(3):147-153.

Zhang Y, Desai A, Yang SY, et al. Tissue regeneration.

Inhibition of the prostaglandin-degrading enzyme 15-PGDH potentiates tissue regeneration. Science.

2015;348(6240):aaa2340.

Antczak MI, Zhang Y, Wang C, et al.

Inhibitors of 15-prostaglandin dehydrogenase to ptentiate tissue repair. J Med Chem.

2017;60(9):3979-4001.

Dutta S, Sengupta P. Men and mice: relating their ages. Life Sci. 2016;152:244-248.

Popplewell LL, Forman SJ. Is there an upper age limit for bone marrow transplantation?

Bone Marrow Transplant. 2002;29(4):277284.

Kim MJ, Kim MH, Kim SA, Chang JS. Agerelated deterioration of hematopoietic stem cells. Int J Stem Cells. 2008;1(1):55-63.

Geiger H, de Haan G, Florian MC. The ageing haematopoietic stem cell compartment.

Nat Rev Immunol. 2013;13(5):376-389.

FitzGerald GA. Biomedicine. Bringing PGE(2) in from the cold. Science. 2015;348 (6240):1208-1209.

Bergmann C, Wobser M, Morbach H, et al.

1063

A. Desai et al.

Primary hypertrophic osteoarthropathy with digital clubbing and palmoplantar hyperhidrosis caused by 15-PGHD/HPGD loss-of-function mutations. Exp Dermatol.

2011;20(6):531-533.

28. Diggle CP, Carr IM, Zitt E, et al. Common and recurrent HPGD mutations in Caucasian individuals with primary hypertrophic osteoarthropathy. Rheumatology (Oxford). 2010;49(6):1056-1062.

29. Erken E, Koroglu C, Yildiz F, Ozer HT, Gulek B, Tolun A. A novel recessive 15-hydroxyprostaglandin dehydrogenase mutation in a family with primary hypertrophic osteoarthropathy.

Mod Rheumatol.

2015;25(2):315-321.

30. Sinibaldi L, Harifi G, Bottillo I, et al. A novel homozygous splice site mutation in the HPGD gene causes mild primary hypertrophic osteoarthropathy. Clin Exp

1064

Rheumatol. 2010;28(2):153-157.

31. Uppal S, Diggle CP, Carr IM, et al. Mutations in 15-hydroxyprostaglandin dehydrogenase cause primary hypertrophic osteoarthropathy. Nat Genet. 2008;40(6):789-793.

32. Yuksel-Konuk B, Sirmaci A, Ayten GE, et al.

Homozygous mutations in the 15-hydroxyprostaglandin dehydrogenase gene in patients with primary hypertrophic osteoarthropathy. Rheumatol Int. 2009; 30(1):39-43.

33. Diggle CP, Parry DA, Logan CV, et al.

Prostaglandin transporter mutations cause pachydermoperiostosis with myelofibrosis.

Hum Mutat. 2012;33(8):1175-1181.

34. Lilenbaum R, Socinski MA, Altorki NK, et al.

Randomized phase II trial of docetaxel/ irinotecan and gemcitabine/irinotecan with or without celecoxib in the second-line treatment of non-small-cell lung cancer. J

Clin Oncol. 2006;24(30):4825-4832.

35. El-Rayes BF, Zalupski MM, Manza SG, et al. Phase-II study of dose attenuated schedule of irinotecan, capecitabine, and celecoxib in advanced colorectal cancer.

Cancer Chemother Pharmacol. 2008;61(2): 283-289.

36. Davies JM, Goldberg RM. First-line therapeutic strategies in metastatic colorectal cancer. Oncology (Williston Park). 2008;22(13): 1470-1479.

37. Maiello E, Giuliani F, Gebbia V, et al.

FOLFIRI with or without celecoxib in advanced colorectal cancer: a randomized phase II study of the Gruppo Oncologico dell'Italia Meridionale (GOIM). Ann Oncol.

2006;17(Suppl 7):vii55-59.

38. Markowitz SD. Aspirin and colon cancer-targeting prevention? N Engl J Med. 2007;356(21):2195-2198.

haematologica | 2018; 103(6)

ARTICLE

Cell Therapy & Immunotherapy

Targeting the anion exchanger 2 with specific peptides as a new therapeutic approach in B lymphoid neoplasms

Ferrata Storti Foundation

Jon Celay,1* Teresa Lozano,2* Axel R. Concepcion,3,4 Elena Beltrán,1,5 Francesc Rudilla,2 María José García-Barchino,1 Eloy F. Robles,1 Obdulia Rabal,6 Irene de Miguel,6 Carlos Panizo,7 Noelia Casares,2 Julen Oyarzabal,6 Jesús Prieto,2,3 Juan F. Medina,3 Juan José Lasarte2** and José Ángel Martínez-Climent1**

Division of Hematological-Oncology, Center for Applied Medical Research (CIMA), University of Navarra, CIBERONC, IDISNA, Pamplona, Spain; 2Program of Immunology and Immunotherapy, Center for Applied Medical Research (CIMA), University of Navarra, Pamplona, Spain; 3Division of Hepatology, Center for Applied Medical Research (CIMA), University of Navarra, Pamplona, Spain; 4Department of Pathology, New York University School of Medicine, New York, NY, USA; 5Department of Pharmacology, University of Navarra, Pamplona, Spain; 6Small Molecule Discovery Platform and Molecular Therapeutics Program, Center for Applied Medical Research (CIMA), University of Navarra, Pamplona, Spain and 7Department of Hematology, Clinica Universidad de Navarra, Pamplona, Spain 1

Haematologica 2018 Volume 103(6):1065-1072

*These authors contributed equally to the study and should both be considered first authors. **These authors contributed equally to the study and should both be considered senior and corresponding authors

ABSTRACT

R

egulatory T (Treg) cells can weaken antitumor immune responses, and inhibition of their function appears to be a promising therapeutic approach in cancer patients. Mice with targeted deletion of the gene encoding the Cl–/HCO3– anion exchanger AE2 (also termed SLC4A2), a membrane-bound carrier involved in intracellular pH regulation, showed a progressive decrease in the number of Treg cells. We therefore challenged AE2 as a potential target for tumor therapy, and generated linear peptides designed to bind the third extracellular loop of AE2, which is crucial for its exchange activity. Peptide p17AE2 exhibited optimal interaction ability and indeed promoted apoptosis in mouse and human Treg cells, while activating effector T-cell function. Interestingly, this linear peptide also induced apoptosis in different types of human leukemia, lymphoma and multiple myeloma cell lines and primary malignant samples, while it showed only moderate effects on normal B lymphocytes. Finally, a macrocyclic AE2 targeting peptide exhibiting increased stability in vivo was effective in mice xenografted with B-cell lymphoma. These data suggest that targeting the anion exchanger AE2 with specific peptides may represent an effective therapeutic approach in B-cell malignancies.

Correspondence: jamcliment@unav.es or jlasarte@unav.es

Received: July 10, 2017.

Accepted: November 24, 2017.

Pre-published: November 30, 2017. doi:10.3324/haematol.2017.175687 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1065

Introduction Despite the use of new diagnostic and therapeutic strategies that have improved the prognosis of mature B-cell malignancies, most patients cannot be cured with currently available therapies.1,2 To improve the clinical outcome of these patients, novel agents against specific cellular targets are being developed and tested.3,4 In addition, different types of therapy have become standard treatments for certain hematologic malignancies, while others undergo clinical testing.5,6 Among these, a promising experimental approach aims to inhibit the CD4+CD25+Foxp3+ T regulatory (Treg) cells, and prevent their suppressor activity against antitumoral T helper and cytotoxic T cells.7,8 The use of therapies combining a direct antitumoral effect with an enhancement of the T cell-mediated immune responses would represent a major advance in the treatment of B-cell malignancies.9 The regulation of intracellular pH (pHi) is critical for important cellular processes and functions in many cell types, including lymphocytes.10–12 To achieve acid–base haematologica | 2018; 103(6)

©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1065

J. Celay et al.

homeostasis, lymphoid cells are equipped with a coordinated network of ion channels and transporters in the plasma cell membrane that orchestrate the input and output of acid/base ions H+ and HCO3− to maintain the pHi within a narrow physiological range that is generally ~7.2.10–13 On the other hand, cancer cells with a high rate of metabolic activity have increased pHi while the extracellular space becomes acidified.14–17 Extracellular acidification of the tumor microenvironment suppresses the effector function of antitumor cytotoxic T cells and promotes tumor evasion.18,19 Moreover, early in vitro studies have shown that inhibition of the acid extruder Na+/H+ exchanger 1 (NHE1) in leukemic cells decreases their pHi leading to apoptosis.20,21 Accordingly, physiological pH sensors involved in the modulation of acid-loading and acid-extruding mechanisms hold promise as targets in cancer therapeutics.22–24 Among the SLC4 family of HCO3− transporters, the Na+-independent Cl−/HCO3− anion exchanger 2 (AE2, also referred to as solute carrier family 4 member 2, SLC4A2) is considered a master acid loader in many cell types.25,26 Under physiological conditions, AE2 favors the extrusion of intracellular HCO3− in exchange for extracellular Cl−, resulting in an acid load.27–29 Our group has shown that mice carrying targeted deletion of AE2 (Ae2a,b–/– mice) have lymphocytes with alterations in pHi, which eventually leads to a reduction in the number of Treg cells, among other alterations.30–33 These data prompted us to investigate the role of AE2 as a potential target for tumor immunotherapy.

Here we report the generation and characterization of specific peptides targeting AE2 exchanger function. Our results show that AE2 binding peptides induced opposite effects on different T-cell subsets, promoting apoptosis in Treg cells while activating effector T-cell function. Targeting peptides also promoted apoptosis in tumor cells from different types of leukemia, lymphoma and multiple myeloma,

while showing only moderate effects on non-tumoral B lymphocytes. These data suggest that targeting AE2 represents a novel therapeutic approach that may simultaneously promote apoptosis of tumor cells and enhance T cellmediated immune responses.

Methods Peptides A series of 24 linear peptides of 15 amino acids that potentially bind a short stretch of highly conserved amino acid sequences (NMTWAGARPT in human and NMTWATTI in mouse AE2), were designed. These conserved target sequences are within the third extracellular loop of the protein, which has been shown to play a key role in Cl-/HCO3- exchange function.25,34 The design of binding peptides followed a methodology that assigns potential interactions between peptides based on the hydrophilicity/hydrophobicity and the net charge of the amino acid side chains of the involved peptides, as described.35 AE2 binding peptides were synthesized by the solid phase method of Merrifield using the Fmoc alternative, as described.36 The purity of the peptides was analyzed by HPLC. To generate cycled peptides, based on the 3D structure of the p17AE2 peptide predicted using the de novo prediction server PEP-FOLD,37,38 two macrocyclic peptides with a head-to-tail cyclization (termed p17AE2-HT) and with a secondary amide as linker (termed p17AE2-Amide) were synthesized by solid phase synthesis (Wuxi AppTech, Shanghai, China).

Synthesis and measurement of peptide metabolic stability procedures are detailed in Online Supplementary Methods.

Surface plasmon resonance Peptide binding to the third extracellular loop of AE2 was analyzed by surface plasmon resonance using the ProteOn XPR36 (Bio-Rad) optical biosensor, as described.39 Briefly, N-terminally

A

B

1066

C

Figure 1. Generation of functional peptides interacting specifically with the third extracellular loop of AE2 protein. (A) Alignment of the third extracellular loop of murine and human AE2 and linear p17AE2, p19AE2, p20AE2 and truncated peptides. (B) p17AE2, p19AE2 and p20AE2 increased proliferation of effector T cells co-cultured with Treg cells in vitro. (C) Sensogram of surface plasmon resonance showing p17AE2 binding to AE2 in a dose dependent manner. ***, P<0.001.

haematologica | 2018; 103(6)

Peptides specifically targeting AE2 induce apoptosis in tumor biotinylated peptide biot-DSEAGSSSSSNMTWATTILVPDNSSASGQSGQEKPR, encompassing the AE2-specific third extracellular loop of AE2, was immobilized into a NLC-neutravidin chip (Biorad). Individual peptide solutions (from 0.15 to 2.5 ÎźM) were injected by triplicate in running buffer (phosphate buffered saline, 0.005% (v/v) Tween 20, pH 7.4) at a flow of 30 mL/min. The interspot signal (obtained in the chip surface not immobilized with protein) was used as reference.

Cell lines and patient samples The human lymphoma, leukemia and myeloma cell lines used in the study were cultured in RPMI or IMDM media supplemented with 10% fetal bovine serum and 1% Penicillin-Streptomycin (PS) as described.40,41 A series of ten fresh peripheral blood cells obtained from untreated patients with B-cell lymphoma with peripheral blood dissemination were included in the study.

Primary B cells were isolated from human peripheral blood using a CD19+ isolation kit (Miltenyi Biotec). To determine cell viability

and apoptosis, cells were cultured in IMDM medium with 20% fetal bovine serum and 1% P-S. The research protocol on human subjects was approved by the University of Navarra Institutional Review Board. Informed consent was obtained from patients and healthy donors, in accordance with the Helsinki Declaration.

Cell viability and apoptosis assays Both primary cells and established cell lines were cultured in flat bottom M96 wells, at a density of 50,000 cells/well, and incubated in 100 mL of cell medium supplemented with the corresponding isoform of the peptide at a final concentration of 50 mg/mL for 24 hours. To determine cell viability, 10 mL of MTS (Promega) were added to each well and plates were read at a wavelength of 450/690 nm. Cell viability was normalized to control group for each experiment. Apoptosis was measured using the FITC Annexin V Apoptosis Detection Kit I (BD), as reported.41,42 Briefly, cell were stained with Annexin V-FITC and ToPro3 and analyzed in a FACS Calibur cytometer (BD).

A

B

C

D

Figure 2. p17AE2 peptide induces proliferation of effector T cells and apoptosis of regulatory T lymphocytes. Effect of p17AE2 on cell viability, IL-2 secretion and apoptosis in murine (A) and human (B) effector T lymphocytes. (C) Effect of p17AE2 on cell proliferation, IL-10 secretion and apoptosis in murine regulatory T lymphocytes. (D) Cell viability of human Jurkat (T effector) and Karpas299 (Treg) cell lines upon incubation with p17AE2 peptide. (E) pHi of murine effector and regulatory T cells after p17AE2 treatment. *P<0.05; **P<0.01; ***P<0.001.

E

haematologica | 2018; 103(6)

1067

J. Celay et al.

In vitro assays for murine effector and regulatory T cells Regulatory (CD4+CD25+) and effector (CD4+CD25−) T cells were isolated from BALB/c mouse spleens using Treg isolation kits (Miltenyi Biotec). For the screening of the peptides targeting AE2, CD3-stimulated effector T cells were cultured in the presence or absence of purified Treg cells and peptides, as previously described.43 Then, cell proliferation was calculated by adding [methyl-3H] thymidine, and incorporated radioactivity was measured using a scintillation counter.44 The same procedure was carried out to determine the effect of p17AE2 in effector and regulatory T cell populations. IL-2 and IL-10 secreted to the supernatant of cell cultures were measured by ELISA (Pharmingen).

Intracellular pH measurement To asses pHi, cells were stained with the intracellular fluorescent pH indicator BCECF-AM (Biotium), as described.32,45 Briefly, cells were stained and washed in MACS buffer supplemented with the corresponding peptide at a final concentration of 50 mg/mL. Then, cells were excited at 488 nm and the ratio of emission wavelengths 530/661 nm was determined in a FACS Calibur cytometer. The

A

B

C

D

E

F

G

H

1068

nigericin clamp technique was used to estimate pHi values from calibration curves as described.31,46

Determination of anion exchange activity To determine the Cl−/HCO3− exchange activity, cells were stained with BCECF-AM in Krebs-Ringer bicarbonate buffer (KRB, which contains in mM: 115 NaCl, 4.7 KCl, 1.5 CaCl2, 25 NaHCO3, 1 Na+-pyruvate, 1.2 KH2PO4,1 MgSO4, 5 glucose, and carbogen − 95% CO2 + 5% O2) and incubated as described above. Then, cells were washed and resuspended in Cl–-free buffer (in mM: 115 Na+-isethionate, 4.7 K+-gluconate, 1.5 Ca2+-gluconate, 25 NaHCO3, 1 Na+-pyruvate, 1.2 KH2PO4, 1 MgSO4, 5 glucose, and carbogen) supplemented with 50 mg/mL of the corresponding peptide, and pHi was measured every 2 minutes by flow cytometry. Cl– was replaced by isothionate in order to maintain osmolality. Changes in anion exchange activity were represented as the variation of pHi along the time after extracellular Cl– removal. This removal forces the efflux of Cl– and the influx of HCO3–, and intracellular pH increases as a result of a reverse activity of the AE2 anion exchanger.29,32

Figure 3. Effect of p17AE2 on human normal B lymphocytes and tumor B-cell lines. (A) Cell viability of CD19+ cell isolated from human peripheral blood lymphocytes (PBLs) upon treatment with 5, 10 or 50 mg/mL of truncated or p17AE2 peptides. (B) Effect of p17AE2 on cell viability of human B-cell leukemia, lymphoma and multiple myeloma cell lines. Cell viability, apoptosis, and cell cycle abnormalities in tumor B-cell (C, D, E), AML (F) and T-ALL cell lines (G, H) after 48 hour incubation with p17AE2 peptide. *P<0.05; **P<0.01; ***P<0.001.

haematologica | 2018; 103(6)

Peptides specifically targeting AE2 induce apoptosis in tumor

Therapeutic in vivo assays in mouse xenografts OCI-Ly7, L363 and UPN1 tumor cells (5x106 cells per animal) were injected subcutaneously in 6-week-old male immunodeficient Rag2-/-IL2gc-/- mice. Eight animals per group were used on each experiment. Once tumors reached volumes of 100 mm3, treatment was started by intratumorally injecting 50, 200 or 400 mg of the peptide (a truncated peptide was used for control mice at the same dosage). Tumor size was measured every 2 days, as described previously.47 When tumors reached volumes of 2000 mm3, animals were euthanized.

Statistical Analysis Results are expressed as mean ± SEM. At least three independent experiments for cell viability, apoptosis and pHi assessment assays were performed (in duplicate each). The normal distribution of values was assessed by using the Shapiro-Wilks and Kolmogorof-Smirnov tests, and the statistical significance of differences was determined with the student’s-t test, taking two-tailed P<0.05 as the criterion for significance. For in vivo experiments, data were analyzed by two way ANOVA test and the Bonferroni post-tests to compare replicate means. Statistical analyses were performed using the Graph Pad Prism 5 program.

Results and Discussion Generation and characterization of functional peptides targeting AE2 A series of 24 linear peptides were designed to bind a short, highly conserved region in human and mouse AE2 (Figure 1A). Once synthesized, peptides were tested for their ability to inhibit the suppressor activity of natural CD4+CD25+Treg cells in vitro. Thus, CD4+CD25– effector T cells activated with anti-CD3 monoclonal antibody in the presence of Treg cells were used to analyze the capacity of each peptide to restore T-cell proliferation inhibited by Treg cells. Among the 24 peptides, three (p17AE2, p19AE2 and p20AE2) were able to restore and even enhance the proliferation of effector T cells (Figure 1B). Surface plasmon res-

A

onance showed a dose-dependent binding of the p17AE2 peptide to a 36 amino acids long peptide encompassing the third extracellular loop of human AE2 that is crucial for its exchange function,25,34 thus revealing physical binding (Figure 1C). The p17AE2 peptide was therefore selected for further experiments.

Functional targeting peptide p17AE2 shows opposite functions in different T-cell subsets Next, we tested the effect of p17AE2 on conventional effector T-cell proliferation in the absence of Treg cells.

In vitro experiments showed that the addition of p17AE2 to murine effector T cells slightly increased cell proliferation in response to anti-CD3 stimulation, and similarly promoted IL-2 secretion, while apoptosis was not affected (Figures 2A). This capacity to promote effector T-cell proliferation was also observed in human effector T cells from healthy donors stimulated with anti-CD3/CD28 beads (Figure 2B). Conversely, p17AE2 reduced cell proliferation of activated Treg cells in culture and induced cell apoptosis, while the production of IL-10 was not significantly altered, (Figures 2C). Likewise, p17AE2 decreased cell viability of the human-derived T-cell leukemia cell line Karpas299, which shows characteristics typical of natural Treg cells with a CD4+CD25+Foxp3+ phenotype.48 On the other hand, p17AE2 did not affect cell survival of Jurkat T-cell leukemia cells with a CD4+CD25– T-cell effector phenotype (Figure 2D).39 We then determined whether changes in cell survival were related to variations of the pHi. Upon incubation with p17AE2, the pHi in effector T lymphocytes remained similar to that in control cells or cells treated with the truncated peptide, while pHi values decreased in Treg cells over time (Figure 2E).

AE2 targeting promotes apoptosis of B-cell leukemia, lymphoma and multiple myeloma cells The effect of the p17AE2 peptide in vitro was also evaluated on peripheral blood B lymphocytes isolated from healthy donors as well as on cell lines derived from

B

D

C

E

Figure 4. Modulation of the exchanger function of AE2 driven by p17AE2 peptide. (A) Basal pHi in human peripheral blood B lymphocytes (PBLs) and tumor B-cell lines. (B) Changes in pHi values upon treatment with p17AE2. (C) Average basal pHi in sensitive and resistant tumor B-cell lines. (D-E) Effect of p17AE2 treatment on the AE2 activity in sensitive Jeko1 and resistant U266 cell lines. ***P<0.001.

haematologica | 2018; 103(6)

1069

J. Celay et al.

A

B

C

D

E

F

1070

G

Figure 5. Therapeutic effect of the linear and the macrocyclic p17AE2 peptides on B-cell lymphomas. (A) Modelled 3D structures of p17AE2, p17AE2-HT and p17AE2-Amide. Left, superposition of conformation 1 of p17AE2 (< 5 Å, orange) and p17AE2HT (green). Right, superposition of conformation 2 of p17AE2 (>10 Å, orange) and p17AE2-Amide (cyan). (B) Sensogram of surface plasmon resonance showing binding of p17AE2-Amide and p17AE2-HT to AE2 in a dose dependent manner. (C) Halflife and clearance of the different peptides in humans and mice. (D) Cell viability and apoptosis in UPN1, Jeko1, OCI-Ly1 cell lines after incubation with 50 µg/mL of truncated, p17AE2, p17AE2-HT and p17AE2-Amide peptides for 24 hours. (E) Similarly, the peptides reduced cell viability and promoted apoptosis of primary samples obtained from patients with Bcell lymphomas (n=10). (F, G) Representation of volumes of subcutaneous tumors in xenografted mice after intratumoral injection of 200 mg or 400 mg of truncated or p17AE2-HT peptides. *P<0.05; **P<0.01; ***P<0.001.

haematologica | 2018; 103(6)

Peptides specifically targeting AE2 induce apoptosis in tumor patients with a variety of B-cell malignancies. While p17AE2 showed a moderate effect on normal B lymphocytes, decreasing B-cell viability only at high dosage (50 µg/mL) (Figure 3A), this peptide markedly reduced cell growth in 44 of 69 (64%) malignant cell lines, as follows: 28 of 43 (65%) mature B-cell lymphomas (8 of 12 mantle cell lymphoma, 5 of 8 Burkitt lymphoma, 14 of 21 diffuse large B-cell lymphoma and 1 of 2 splenic marginal-zone B-cell lymphoma); 2 of 4 EBV transformed lymphoblastoid B-cell lines; 9 of 16 (56%) multiple myelomas; and 5 of 6 (83%) B-cell acute lymphoblastic leukemias (Figure 3B). The effects of p17AE2 on cell viability and apoptosis were dose-dependent in p17AE2 sensitive B-cell lymphoma cell lines, but not in resistant OCI-Ly1 B-cell lymphoma cells (Figure 3 C-D). Cell cycle, however, was not affected either in sensitive or in resistant cell lines (Figure 3E). Besides, p17AE2 also affected cell viability and apoptosis of several acute myeloid leukemia (AML) and T-cell acute lymphoblastic leukemia (T-ALL) (Figure 3 F-G). These results indicate that targeting AE2 with p17AE2 reduces cell survival in a variety of hematologic neoplasms, including B-cell tumors, while showing only moderate effects on nontumoral B lymphocytes.

Functional targeting peptide p17AE2 induces apoptosis in tumor cells by modulating intracellular pH and AE2 function To study how p17AE2 reduces cell viability of tumor B cells after targeting AE2 protein, basal pHi values were measured in normal B cells and in tumor B-cell lines.

Consistent with previous studies, human peripheral blood B lymphocytes from healthy donors showed basal pHi values ranging from 7.0 to 7.2, whereas cell lines derived from different B-cell malignancies exhibited higher pHi values ranging from 7.4 to 7.8 (Figure 4A). The alkaline pHi is permissive for tumor cell proliferation and evasion of apoptosis, while acidic pHi values promote cell apoptosis.10,14 Incubation of tumor B cells with p17AE2 led to dose-dependent pHi acidification in sensitive cells, which correlated with apoptosis rates (Figure 4B). In B lymphocytes and in tumor resistant cells, however, p17AE2 treatment did not affect pHi values. Nevertheless, there was not a correlation between basal pHi values in the different tumor B-cell lines and response to p17AE2 (Figure 4C). To evaluate whether the Cl−/HCO3− exchange activity of AE2 was modulated by p17AE2, cells were subjected to removal of extracellular Cl−, and the pHi changes with and without p17AE2 were measured at different time intervals; under these forced conditions, a reversed exchange activity of AE2 is reflected by an increase in pHi (see Methods for details). In the sensitive cell line Jeko1, p17AE2 induced an increase in pHi with respect to cells incubated in the absence of the peptide, while in the resistant cell line U266 the activity of AE2 remained unchanged (Figure 4 D-E). These results suggest that in sensitive cell lines, like Jeko1, p17AE2 is able to modulate AE2 activity, altering pHi and thus inducing cell apoptosis, while in resistant cell lines, such as U266, the peptide would not be able to alter AE2 activity and pHi would remain stable, not compromising cell viability. Hence, in sensitive cell lines, p17AE2 would activate the physiological Cl−/HCO3− exchange function of AE2, favoring HCO3− export in exchange with Cl−, thus reducing pHi and promoting apoptosis. haematologica | 2018; 103(6)

Therapeutic effect of AE2 targeting peptides in mouse xenograft models in vivo Finally, we assessed the effects of targeting AE2 in vivo.

For that purpose, 5x106 cells from three sensitive cell lines (Mantle cell lymphoma- UPN1, DLBCL-OCI-Ly7 and multiple myeloma-L363) were injected subcutaneously in Rag2-/-IL-2gc-/- immunodeficient mice. When tumors achieved volumes of 100 mm3, mice received daily intratumoral injection of p17AE2 or a truncated peptide (50 mg each). However, no changes in tumor size between control and treated mice were observed after 14 days of treatment, suggesting that p17AE2 effects could be limited by a poor bio-availability and susceptibility to degradation by proteases (data not shown). Peptide cyclization has been used as a strategy for stabilizing small peptides.49 Therefore, based on computational studies, we used this strategy to design and synthesize two macrocyclic peptides, one with a head-to-tail cyclization (termed p17AE2HT) and another one with a secondary amide as linker (termed p17AE2-Amide) (Figure 5A). Surface plasmon resonance showed a dose-dependent binding of both p17AE2-HT and p17AE2-Amide peptides to the third extracellular loop of human AE2 (Figure 5B). However, only the p17AE2-HT peptide, but not the p17AE2-Amide, showed longer half-life and reduced clearance with respect to the linear p17AE2 peptide (Figure 5C). In addition, p17AE2-HT decreased cell proliferation and promoted apoptosis in different B-cell lymphoma cell lines and in primary B-cell lymphoma samples, even more potently than the p17AE2 linear peptide (Figures 5D-E). These results indicated that p17AE2-HT was more efficient than linear peptides and hence a good candidate for therapeutic testing in vivo. To assess the therapeutic role of the macrocyclic peptides in vivo, we inoculated subcutaneously 5x106 cells from the UPN1 B-cell lymphoma-derived cell line into Rag2-/-IL-2gc-/- mice, due to its sensitivity in the in vitro model. When tumors reached a size of 100 mm3, 200 mg or 400 mg per day of p17AE2-HT or truncated peptides were injected intratumorally during 14 days. Mice treated with 400 mg of p17AE-HT exhibited reduction in tumor volumes with respect to those treated with 200 µg or with truncated peptide (Figures 5F-G). These results indicate that p17AE2-HT treatment moderately reduced tumor growth in vivo.

In summary, our results suggest that targeting AE2 might exert a dual therapeutic effect in B-cell malignancies. In our in vitro studies, the linear p17AE2 was able to induce apoptosis of tumor B cells, while potentially boosting anti-tumor immune responses by reducing the number of Treg cells. In in vivo models, however, the linear peptide had no effect, probably because of its low stability, while the head-to-tail cycled p17AE2-HT showed increased stability and managed to reduce tumor growth in vivo when given at a high dosage. Thus, AE2 seems to be a promising target in different B-cell malignancies, and modifications of AE2 targeting peptides may increase their potential therapeutic value in vivo.

Acknowledgments We thank Elena Ciordia and Eneko Elizalde (CIMA) for excellent animal care.

Funding The work was supported by grants from Fundación Ramón Areces, BAYER Pharma HealthCare (Grant4Targets 2015), 1071

J. Celay et al.

Fundación Arnal Planelles, and Roche Spain (Hemato-Oncology award 2014) (to JAM-C, JJL and JFM); Worldwide Cancer Research (WCR15-1322), ISCIII/FIS PI16/00581, and ISCIII/FIS-CIBERONC (to JAM-C); Ministerio de Educación y Ciencia SAF2013-42772-R and SAF2016-78568-R and

References 1. Swerdlow SH, Campo E, Pileri SA, et al.

The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016; 127(20):23752390.

2. Shaffer AL, Young RM, Staudt LM. Pathogenesis of human B cell lymphomas.

Annu Rev Immunol. 2012;30(1):565-610.

3. Roschewski M, Staudt LM, Wilson WH. Diffuse large B-cell lymphoma: treatment approaches in the molecular era. Nat Rev Clin Oncol. 2013;11(1):12-23.

4. Nowakowski GS, Blum KA, Kahl BS, et al.

Beyond RCHOP: A blueprint for diffuse large B cell lymphoma research. J Natl Cancer Inst. 2016;108(12):djw257.

5. Pavanello F, Zucca E, Ghielmini M. Rituximab: 13 open questions after 20 years of clinical use. Cancer Treat Rev. 2017;53:38-46.

6. Goodman A, Patel SP, Kurzrock R. PD-1– PD-L1 immune-checkpoint blockade in Bcell lymphomas. Nat Rev Clin Oncol.

2016;14(4):203-220.

7. Zou W. Regulatory T cells, tumour immunity and immunotherapy. Nat Rev Immunol. 2006;6(4):295-307.

8. Lozano T, Casares N, Lasarte JJ. Searching for the Achilles Heel of FOXP3. Front Oncol. 2013;3:294.

9. Gotwals P, Cameron S, Cipolletta D, et al.

Prospects for combining targeted and conventional cancer therapy with immunotherapy. Nat Rev Cancer. 2017; 17(5):286-301.

10. Feske S, Skolnik EY, Prakriya M. Ion channels and transporters in lymphocyte function and immunity. Nat Rev Immunol.

2012;12(7):532-547.

11. Cahalan MD, Chandy KG. The functional network of ion channels in T lymphocytes.

Immunol Rev. 2009;231(1):59-87.

12. King LB, Freedman BD. B-lymphocyte calcium inFlux. Immunol Rev. 2009; 231(1):265-277.

13. De Vito P. The sodium/hydrogen exchanger: A possible mediator of immunity. Cell Immunol. 2006;240(2):69–85.

14. Webb BA, Chimenti M, Jacobson MP, Barber DL. Dysregulated pH: a perfect storm for cancer progression. Nat Rev Cancer. 2011;11(9):671-677.

15. Parks SK, Chiche J, Pouyssegur J. pH control mechanisms of tumor survival and growth. J Cell Physiol. 2011;226(2):299308.

16. Cardone RA, Casavola V, Reshkin SJ. The role of disturbed pH dynamics and the Na+/H+ exchanger in metastasis. Nat Rev Cancer. 2005;5(10):786-795.

17. Lang F, Huber SM, Szabo I, Gulbins E. Plasma membrane ion channels in suicidal cell death. Arch Biochem Biophys.

2007;462(2):189-194.

18. Mendler AN, Hu B, Prinz PU, Kreutz M, Gottfried E, Noessner E. Tumor lactic acidosis suppresses CTL function by inhibition of p38 and JNK/c-Jun activation. Int J cancer. 2012;131(3):633–640.

1072

Fundación Bancaria La Caixa-Hepacare Project, Ministerio de Educación y Ciencia SAF2013-42772-R and SAF2016-78568-R (to JJL), and Ministerio de Economia y Competitividad by the Torres-Quevedo subprogram PTQ-14-07320 (to IdM). JC is a Juan de la Cierva post-doctoral fellow.

19. Nakagawa Y, Negishi Y, Shimizu M, Takahashi M, Ichikawa M, Takahashi H. Effects of extracellular pH and hypoxia on the function and development of antigenspecific cytotoxic T lymphocytes. Immunol Lett. 2015;167(2):72-86.

20. Rich IN, Worthington-White D, Garden O a, Musk P. Apoptosis of leukemic cells accompanies reduction in intracellular pH after targeted inhibition of the Na(+)/H(+) exchanger. Blood. 2000;95(4):1427-1434.

21. Marches R, Vitetta ES, Uhr JW. A role for intracellular pH in membrane IgM-mediated cell death of human B lymphomas. Proc Natl Acad Sci USA. 2001;98(6):3434-3439.

22. Yue Loo S, Ker Xing Chang M, Sui Huay Chua C, Prem Kumar A, Pervaiz S, Veronique Clement M. NHE-1: a promising target for novel anti-cancer therapeutics.

Curr Pharm Des. 2012;18(10):1372-1382.

23. Parks SK, Chiche J, Pouysségur J. Disrupting proton dynamics and energy metabolism for cancer therapy. Nat Rev Cancer. 2013;13(9):611-623.

24. Lin L, Yee SW, Kim RB, Giacomini KM. SLC transporters as therapeutic targets: emerging opportunities. Nat Rev Drug Discov. 2015;14(8):543-560.

25. Alper SL. Molecular physiology and genetics of Na+-independent SLC4 anion exchangers. J Exp Biol. 2009;212(11):16721683.

26. César-Razquin A, Snijder B, FrappierBrinton T, et al. A call for systematic research on solute carriers. Cell. 2015; 162(3):478-487.

27. Alper SL. Molecular physiology of SLC4 anion exchangers. Exp Physiol. 2006; 91(1):153-161.

28. Romero MF, Chen AP, Parker MD, Boron WF. The SLC4 family of bicarbonate (HCO3-) transporters. Mol Aspects Med. 2013;34(2–3):159-182.

29. Uriarte I, Banales JM, aez E, Arenas F, et al.

Bicarbonate secretion of mouse cholangiocytes involves Na-HCO3 cotransport in addition to Na-independent Cl/HCO3 exchange. Hepatology. 2010;51(3):891–902.

30. Medina JF, Recalde S, Prieto J, et al. Anion exchanger 2 is essential for spermiogenesis in mice. Proc Natl Acad Sci USA. 2003;100(26):15847-15852.

31. Salas JT, Banales JM, Sarvide S, et al.

Ae2a,b-deficient mice develop antimitochondrial antibodies and other features resembling primary biliary cirrhosis.

Gastroenterology. 2008;134(5):1482-1493.

32. Concepcion AR, Salas JT, Sarvide S, et al.

Anion exchanger 2 is critical for CD8+ T cells to maintain pHi homeostasis and modulate immune responses. Eur J Immunol. 2014;44(5):1341-1351.

33. Concepcion AR, Salas JT, Sáez E, et al.

CD8+ T cells undergo activation and programmed death-1 repression in the liver of aged Ae2a,b –/– mice favoring autoimmune cholangitis. Oncotarget. 2015; 6(30):28588v28606.

34. Romero MF, Fulton CM, Boron WF. The SLC4 family of HCO 3 - transporters.

Pflugers Arch. 2004;447(5):495-509.

35. Ezquerro IJ, Lasarte JJ, Dotor J, et al. A syn-

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

thetic peptide from transforming growth factor type III receptor inhibits liver fibrogenesis in rats with carbon tetrachloride liver injury. Cytokine. 2003;22(1–2):12–20.

Borrás-Cuesta F, Golvano J, Sarobe P, et al.

Insights on the amino acid side-chain interactions of a synthetic T-cell determinant.

Biologicals. 1991;19(3):187-190.

Maupetit J, Derreumaux P, Tufféry P. A fast method for large-scale de novo peptide and miniprotein structure prediction. J Comput Chem. 2010;31(4):72-738.

Thévenet P, Shen Y, Maupetit J, Guyon F, Derreumaux P, Tufféry P. PEP-FOLD: An updated de novo structure prediction server for both linear and disulfide bonded cyclic peptides. Nucleic Acids Res. 2012;40(W1):W288-293.

Lozano T, Villanueva L, Durántez M, et al.

Inhibition of FOXP3/NFAT interaction enhances T Cell function after TCR stimulation. J Immunol. 2015;195(7):3180-3189.

Rodríguez-Ortigosa CM, Celay J, Olivas I, et al. A GAPDH-mediated trans-nitrosylation pathway is required for feedback inhibition of bile salt synthesis in rat liver.

Gastroenterology. 2014;147(5):1084–1093.

Beltran E, Fresquet V, Martinez-Useros J, et al. A cyclin-D1 interaction with BAX underlies its oncogenic role and potential as a therapeutic target in mantle cell lymphoma. Proc Natl Acad Sci USA. 2011; 108(30):12461-12466.

Sagardoy A, Martinez-Ferrandis JI, Roa S, et al. Downregulation of FOXP1 is required during germinal center B-cell function.

Blood. 2013;121(21):4311-4320.

Lozano T, Soldevilla MM, Casares N, et al.

Targeting inhibition of Foxp3 by a CD28 2’-Fluro oligonucleotide aptamer conjugated to P60-peptide enhances active cancer immunotherapy. Biomaterials. 2016;91:7380.

Gil-Guerrero L, Dotor J, Huibregtse IL, et al. In vitro and in vivo down-regulation of regulatory T cell activity with a peptide inhibitor of TGF-beta1. J Immunol.

2008;181(1):126-135.

Thomas JA, Buchsbaum RN, Zimniak A, Racker E. Intracellular pH measurements in Ehrlich ascites tumor cells utilizing spectroscopic probes generated in situ.

Biochemistry. 1979;18(11):2210-2218.

Musgrove E, Rugg C, Hedley D. Flow cytometric measurement of cytoplasmic pH: A critical evaluation of available fluorochromes. Cytometry. 1986;7(4):347–355.

Robles EF, Mena-Varas M, Barrio L, et al.

Homeobox NKX2-3 promotes marginalzone lymphomagenesis by activating B-cell receptor signalling and shaping lymphocyte dynamics.

Nat Commun.

2016;7:11889.

Wolke C, Tadje J, Bukowska A, et al.

Assigning the phenotype of a natural regulatory T-cell to the human T-cell line, KARPAS-299.

Int J Mol Med. 2006;17(2):275-278.

Bhat A, Roberts LR, Dwyer JJ. Lead discovery and optimization strategies for peptide macrocycles. Eur J Med Chem. 2015; 94:471-479.

haematologica | 2018; 103(6)

ARTICLE

Coagulation & its Disorders

Structural and cellular mechanisms of peptidyl-prolyl isomerase Pin1-mediated enhancement of Tissue Factor gene expression, protein half-life, and pro-coagulant activity Kondababu Kurakula,1,2,* Duco S. Koenis,1,* Mark A. Herzik Jr,3,4 Yanyun Liu,3 John W. Craft Jr,3 Pieter B. van Loenen,1 Mariska Vos,1 M. Khang Tran,1 Henri H. Versteeg,5 Marie-José T.H. Goumans,2 Wolfram Ruf,6,7 Carlie J.M. de Vries1,# and Mehmet Şen3,#

Ferrata Storti Foundation

Haematologica 2018 Volume 103(6):1073-1082

Department of Medical Biochemistry, Amsterdam Cardiovascular Sciences, Academic Medical Center, University of Amsterdam, the Netherlands; 2Department of Cell and Chemical Biology, Leiden University Medical Center, the Netherlands; 3Department of Biology and Biochemistry, Chemical Biology Interdisciplinary Program, University of Houston, TX, USA; 4Department of Integrative Structural and Computational Biology, The Scripps Research Institute, La Jolla, CA, USA; 5The Einthoven Laboratory for Experimental Vascular Medicine, Leiden University Medical Center, the Netherlands; 6 Center for Thrombosis and Hemostasis, University Medical Center, Mainz, Germany and 7 Department of Immunology and Microbial Science, The Scripps Research Institute, La Jolla, CA, USA 1

*KK and DSK contributed equally to this work. #CJMdV and MS contributed equally to this work.

ABSTRACT

T

issue Factor is a cell-surface glycoprotein expressed in various cells of the vasculature and is the principal regulator of the blood coagulation cascade and hemostasis. Notably, aberrant expression of Tissue Factor is associated with cardiovascular pathologies such as atherosclerosis and thrombosis. Here, we sought to identify factors that regulate Tissue Factor gene expression and activity. Tissue Factor gene expression is regulated by various transcription factors, including activating protein-1 and nuclear factor-κ B. The peptidyl-prolyl isomerase Pin1 is known to modulate the activity of these two transcription factors, and we now show that Pin1 augments Tissue Factor gene expression in both vascular smooth muscle cells and activated endothelial cells via activating protein-1 and nuclear factor-κ B signaling. Furthermore, the cytoplasmic domain of Tissue Factor contains a well-conserved phospho-Ser258Pro259 amino-acid motif recognized by Pin1. Using co-immunoprecipitation and solution nuclear magnetic resonance spectroscopy, we show that the WW-domain of Pin1 directly binds the cytoplasmic domain of Tissue Factor. This interaction occurs via the phospho-Ser258-Pro259 sequence in the Tissue Factor cytoplasmic domain and results in increased protein half-life and pro-coagulant activity. Taken together, our results establish Pin1 as an upstream regulator of Tissue Factor-mediated coagulation, thereby opening up new avenues for research into the use of specific Pin1 inhibitors for the treatment of diseases characterized by pathological coagulation, such as thrombosis and atherosclerosis.

Introduction Tissue Factor (TF), an integral cell-surface glycoprotein, is the initiator of the blood coagulation cascade and a key regulator of hemostasis.1,2 Aberrant expression of TF plays a crucial role in several coagulation-driven pathologies, such as thrombosis, atherosclerosis, and acute coronary syndromes,2-4 but also in endotoxemia, angiogenesis, and cancer.5-8 Many vascular cells express TF constitutively, including smooth muscle cells (SMCs), pericytes, and adventitial fibroblasts, while TF expression is undetectable in vascular endothelial cells (ECs).1 However, in both ECs and SMCs, the expression and activity of TF can be enhanced by pro-inflammatory signaling molehaematologica | 2018; 103(6)

Correspondence: c.j.devries@amc.nl or msen@uh.edu

Received: October 23, 2017.

Accepted: March 1, 2018.

Pre-published: March 15, 2018. doi:10.3324/haematol.2017.183087 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1073 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1073

K. Kurakula et al.

cules such as tumor necrosis factor-α (TNF-α) and lipopolysaccharides (LPS), which induce TF expression via activating protein 1 (AP-1) and nuclear factor-kappa B (NFκB) signaling.1,9,10 Consistently, inflammation-induced coagulation is completely abrogated by inhibition of TF activity in vivo.11 The TF protein contains two fibronectin type-III ectodomains, a single type-I transmembrane domain, and a twenty-amino acid cytoplasmic domain (TFCD), which plays an important role in several of the above-mentioned pathologies that involve TF.6-8 The TFCD can undergo multiple post-translational modifications, including palmitoylation at residue Cys245, phosphorylation at residues Ser253 and Ser258 by PKC-α and p38α-kinase, respectively, and ubiquitination at residue Lys255.12-15 These modifications act in concert to attenuate the interaction of the TFCD with the membrane lipid bilayer, as well as make it accessible to other interacting proteins that regulate TF activity.13,14,16 Peptidyl-prolyl cis-trans isomerase, NIMA-Interacting 1 (Pin1) is an enzyme that catalyzes cis-trans isomerization of proline residues that are preceded by a phosphorylated serine or threonine (a pSer/pThr-Pro motif) within its target proteins. The C-terminal isomerase domain of Pin1 binds the motif and catalyzes proline cis-trans isomerization, while the N-terminal WW-domain is responsible for mediating protein-protein interactions and target specificity.17-19 Conformational changes induced by Pin1-catalyzed proline isomerization have been shown to alter the phosphorylation, localization, stability, protein-protein interactions, and transcriptional activity of its target proteins, which include c-Jun, NF-κB, AP-1, p53, β-catenin, and the nuclear receptors PPARg and Nur77.20-22 Here, we report that Pin1 enhances TF gene expression in activated vascular cells, and directly interacts with TF protein through a pSer258-Pro259 motif in the TFCD. We provide the solution structure of the TFCD in complex with the WW-domain of Pin1, which shows that this interaction requires both phosphorylation of Ser258 as well as transconfiguration of the pSer258-Pro259 peptide bond in the TFCD. Functionally, we demonstrate that Pin1 increases the protein half-life and pro-coagulant activity of TF in human vascular cells.

Methods Mice Mice in which the cytoplasmic domain of endogenous TF is deleted (TFDCD mice)23 and wild-type litter mates were maintained under pathogen-free conditions at the Scripps Research Institute Animal Facility with approved protocols of the institutional Animal Care and Use Committee (protocols #08-0008 and #080009).

Cell culture Human umbilical vein endothelial cells (HUVECs) and smooth muscle cells (SMCs) were isolated from umbilical cords anonymously collected at the Academic Medical Center, as previously described,22 in accordance with the institute’s ethical guidelines and with approval from the institute's Medical Ethical Committee.

Please see the Online Supplementary Methods for a detailed description of cell culture conditions.

Lentiviral transductions Recombinant lentiviral particles encoding Pin1, shPin1, or back1074

bone control constructs were produced as previously described.22 Cells were transduced at a multiplicity of infection of 100 for 24 hours (h), after which the medium was refreshed and cells were cultured for an additional 24 h before starting experiments.

Transfections and luciferase assays HEK293T cells and SMCs were transfected using the CalPhos Mammalian Transfection Kit (Clontech) or Lipofectamine 3000 (Invitrogen), respectively. Cells were transfected according to the manufacturer’s instructions with wild-type, AP-1 binding site mutated, or NF-kB binding site mutated TF promoter luciferase reporter constructs (a gift from Nigel Mackman;24 Addgene #15442-15444) together with Pin1 or Pin1 mutants (described by van Tiel et al.22). The pRL-TK Renilla-reporter was co-transfected as internal control. After 24 h, cells were stimulated with 100 ng/mL 12-O-Tetradecanoylphorbol-13-acetate (PMA; Sigma) for 6 h. For NF-κB or AP-1 signaling inhibition, cells were treated with vehicle (DMSO), 10 mM BAY-117085 (Calbiochem), or 10 mM SP600125 (SelleckChem). Assays were performed using the dual-luciferase reporter assay and a Glomax Multi detection system (Promega).

Nuclear magnetic resonance spectroscopy Nuclear magentic resonance (NMR) experiments were performed on Bruker DRX 600 MHz and 800 MHz spectrometers equipped with triple resonance 1H/13C/15N optimized for proton detection at 300K. Acquired data were converted from Bruker XWINNMR format to NMRVIEW v.5.3 format using NMRpipe software.25,26 All spectra were analyzed using NMRVIEW. All experiments were performed at the UH NMR Facility at the University of Houston. Please see the Online Supplementary Methods for a detailed description of NMR spectroscopy procedures.

TF protein half-life assays Smooth muscle cells (SMCs) or HEK293T cells were transfected with Pin1 or Pin1 mutants and either TF or TFDCD constructs. After 24 h, transfected cells were treated with 50 mg/mL cycloheximide (Sigma) for times indicated. TF protein levels were quantified by western blotting.

TF activity assays Tissue factor activity was determined in SMCs, HUVECs, and EC-RF24 cells as previously described.27 Briefly, transduced cells were serum-starved overnight followed by stimulation with 50 ng/mL TNF-α (Peprotech) for 3 h. Cells were washed with PBS and incubated with 1 nM human Factor VIIa and 100 nM human Factor X (Kordia) at 37°C. Supernatant samples were collected in 100 mM EDTA, 50 mM Tris after 10, 20, and 30 minutes (min), incubated with 0.4 mM of FXa chromogenic substrate S-2222, and absorbance was measured at 405 nm.

Statistical analysis Data are presented as mean±Standard Error of Mean.

Significance was determined by unpaired two-tailed Student’s ttest or one- or two-way ANOVA with Bonferroni post-hoc correction as indicated in the figure legends. P<0.05 was considered statistically significant.

Results Pin1 enhances TF gene expression via activation of NF-κB and AP-1 Pin1 modulates the activity of various transcription factors involved in TF gene expression.14,15 Therefore, we inihaematologica | 2018; 103(6)

Pin1 enhances TF expression and activity

tiated our study by assessing the effect of Pin1 on TF gene expression. TF gene expression was measured in cultured ECs and SMCs after Pin1 gain and loss-of-function. Pin1 overexpression significantly increased TF mRNA levels in human SMCs and TNF-α-activated ECs, while knockdown of Pin1 by siRNA or Pin1 isomerase activity inhibition with Juglone28 resulted in significantly decreased TF mRNA expression in ECs (Figure 1A and B). Additionally, pharmacological inhibition of Pin1 isomerase activity with Juglone significantly decreased PMA-induced activation of

A

D

G

B

E

the TF promoter in HEK293T cells (Figure 1C).

The promoter of the TF gene contains binding motifs for NF-κB, AP-1, Sp1, and Egr-1, which are important transcription factors for both basal expression and stimulusspecific induction of TF expression.14,15 Consistent with our gene expression data, Pin1 overexpression markedly enhanced the activity of the wild-type TF promoter reporter construct in both HEK293T cells and SMCs (Figure 1D-F). However, this activation of the TF promoter

by Pin1 was significantly diminished when treating

C

F

H

Figure 1. Pin1 enhances Tissue Factor (TF) mRNA expression and TF promoter activation by NF-κB and AP-1. (A and B) Expression of the TF gene (F3) in human umbilical vein endothelial cells (HUVEC) and smooth muscle cells (SMC) after Pin1 overexpression, Pin1 knockdown (siPin1), or treatment with the Pin1 inhibitor Juglone. HUVECs were treated with TNF-α for 6 hours to induce TF gene expression. (C) Activity of the wild-type TF gene promoter luciferase reporter construct in PMA-stimulated HEK293T treated with Pin1 inhibitor Juglone or vehicle (DMSO) control. Inset shows schematic representation of the TF gene promoter luciferase reporter construct. (D) Activity of the wild-type TF gene promoter luciferase reporter construct after Pin1 overexpression in PMA-stimulated HEK293T and either vehicle (DMSO), NF-κB signaling inhibitor (BAY-117085), or AP-1 signaling inhibitor (SP600125) treatment. (E and F) Activity of luciferase reporter constructs containing either the wild-type TF gene promoter or the TF gene promoter in which the response element for NF-κB (NF-κB RE) or AP-1 (AP-1 RE) was mutated in PMAstimulated HEK293T (E) or TNF-α-stimulated SMCs (F) transiently over-expressing Pin1. (G and H) Activity of the wild-type TF gene promoter luciferase reporter construct after transient overexpression of Pin1 or Pin1 mutants containing a disrupted WW-domain (Pin1:W34A) or lacking isomerase activity (Pin1:K63A) in HEK293T (G) or SMCs (H) left untreated or stimulated with PMA. Data are shown as mean±Standard Error of Mean. (A-C) P-values were calculated for comparisons between Pin1 overexpression, Pin1 knockdown (siPin1), or Juglone treatment versus control transduced (Ctrl) or vehicle-treated groups using the two-tailed Student’s t-test.

(D-H) P-values were calculated using one-way ANOVA as indicated. *P<0.05, **P<0.01, ***P<0.001. AU: arbitrary units; mut: mutated.

haematologica | 2018; 103(6)

1075

K. Kurakula et al.

HEK293T cells with inhibitors of NF-κB or AP-1 signaling (Figure 1D). Similarly, mutating either the NF-κB or AP-1 response element sequences to become non-functional, as previously described,24 greatly reduced the ability of Pin1 to activate the TF promoter in both HEK293T cells and SMCs (Figure 1E and F). Finally, neither a Pin1 mutant with a disrupted WW-domain (Pin1:W34A) nor a Pin1 mutant lacking isomerase activity (Pin1:K63A) could activate the TF promoter in HEK293T cells or SMCs (Figure 1G and H).

Taken together, these results show that Pin1 enhances TF gene expression and TF promoter activation via NF-κB

and AP-1, and that this activation of the TF promoter is dependent on both Pin1 isomerase activity and a functional WW-domain.

Pin1 interacts with the cytoplasmic domain of Tissue Factor The TF cytoplasmic domain (TFCD) contains the wellconserved Ser258-Pro259 amino acid motif, which may function as a recognition site for Pin1 when Ser258 becomes phosphorylated (Figure 2A). We performed coimmunoprecipitation experiments in human SMCs and PMA-activated HUVECs, which showed that a pull-down

A

B

C

D

E

1076

Figure 2. Pin1 interacts with Tissue Factor (TF) via the twenty-amino acid cytoplasmic domain (TFCD). (A) Amino acid sequence similarity of the TFCD in different species, showing strong conservation of the Pin1 recognition motif Ser/Thr-Pro (box). (B) Co-immunoprecipitation (CoIP) for TF and Pin1 in human PMA-stimulated HUVECs or smooth muscle cells (SMC) with control IgG or anti-Pin1 antibody (IP: IgG/Pin1) and analyzed by western blot with anti-TF antibody. Input TF and Pin1 levels are also shown. (C) CoIP for TF and Pin1 in HEK293T whole cell lysates over-expressing HA-tagged Pin1 and either full-length TF or TF∆CD using anti-HA antibody (IP: HA-Pin1) analyzed by western blot with anti-TF antibody.

Input TF/TF∆CD and HA-Pin1 levels are also shown. (D) Pull-down assay with biotinylated peptides encoding wild-type human TFCD (WT) or TFCD with an S258A mutation. Full-length Pin1, Pin1 mutants with a disrupted WWdomain (Pin1:W34A) or lacking isomerase activity (Pin1:K63A), or GFP (as a negative control) were detected by western blot. (E) Pull-down assay with cysteine-linked TFCDencoding peptides that were unphosphorylated or phosphorylated at either Ser253 or Ser258. Full-length Pin1 and GFP (as a negative control) were detected by western blot.

haematologica | 2018; 103(6)

Pin1 enhances TF expression and activity

with anti-Pin1 antibody resulted in co-precipitation of endogenous TF, while a pull-down with control IgG antibody did not, thereby indicating that TF and Pin1 interact in these two cell types (Figure 2B). Next, we performed co-immunoprecipitation experiments in HEK293T cells over-expressing HA-tagged Pin1 and either wild-type TF or truncated TF lacking the cytoplasmic domain (TFDCD).

These experiments showed that full-length TF interacts

with Pin1, while the TFDCD mutant does not, indicating that the TFCD is involved in the interaction with Pin1 (Figure 2C). Next, we performed pull-down assays with biotinylated peptides encoding the human TFCD, which confirmed our finding that Pin1 interacts with the TFCD, and showed that this interaction was completely disrupted by mutating the Ser258 residue in the TFCD (Figure 2D). This result is consistent with the observation that this

A

B

C

D

Figure 3. Nuclear magnetic resonance (NMR) spectroscopy shows interaction between twenty-amino acid cytoplasmic domain (TFCD) and Pin1 requires phosphorylation of Ser258 and trans-configuration of the pSer258-Pro259 peptide bond in the TFCD. (A) Superimposition of the assigned 1H/15N HSOC spectra of the Pin1 WW-domain with double phosphorylated TFCD (pSer253/pSer258) showing the chemical shift changes upon increasing amount of peptide to a 10x molar excess.

A non-linear regression fit of the Ser18 (peak shift shown in black dotted box) was used to calculate the binding constant of the complex. (B) Bundle of 20 NMR conformers sampling into two major conformers. Sidechains of Pro259 (TFCD), Arg21 and Trp34 (Pin1 WW-domain) are indicated with circles. (C) Representative models of the two lowest-energy conformations from (B) are shown in green and magenta. The polypeptide backbones are shown as ribbons. Pin1 WW-domain residues are underlined. (D) Contact of the TFCD pSer258-Pro259 motif with the Pin1 WW-domain loop1 and comparison of NMR and X-ray structures.

haematologica | 2018; 103(6)

1077

K. Kurakula et al.

residue is part of a Pin1 recognition motif. Furthermore, the interaction was not disturbed by mutations in either the WW-domain or isomerase domain of Pin1 Pin1:W34A and Pin1:K63A, respectively (Figure 2D). Finally, these interactions were specific, as control pull-downs without peptide did not precipitate Pin1, nor did TFCD peptides precipitate green fluorescent protein (GFP), a protein not predicted to bind the TFCD (Figure 2D).

We next investigated the effect of phosphorylation on the interaction between Pin1 and TF using cysteine-linked peptides encoding the TFCD with varying states of phosphorylation. These experiments suggest that Pin1 has the highest affinity for the TFCD peptide phosphorylated at Ser258 (Figure 2E). However, it should be noted that these co-immunoprecipitation experiments are not quantitative.

Furthermore, the observation that Pin1 seems to bind unphosphorylated and Ser253-phosphorylated TFCD peptides may also be due to TF-phosphorylating kinases present in the whole-cell lysates used for these pulldowns. The pull-down reactions were specific, as neither a pull-down without peptide or a pull-down for GFP resulted in precipitation of the target protein (Figure 2E).

Taken together, these results confirm that Pin1 interacts with TF and strongly suggest that this interaction involves the predicted pSer258-Pro259 Pin1 recognition motif in the TFCD.

Structural modalities mediating Pin1–TFCD complex formation To study the interaction between the TFCD and Pin1 in more detail, we performed isothermal titration calorimetry (ITC) with Ser253 or Ser258-phosphorylated TFCDencoding peptides and purified Pin1 WW-domain (Online Supplementary Figure S1). Consistent with our pull-down experiments, these assays showed that the Pin1 WWdomain interacts specifically with the TFCD peptide phosphorylated at Ser258 with a fitted Kd of approximately 137 mM (Online Supplementary Figure S1B and D).

To characterize the molecular basis of the WWdomain/TFCD interactions, we titrated the 15N-labeled WW-domain with an increasing concentration of pSer253/pSer258 TFCD peptide. Analysis of the 2D 1H/15N heteronuclear single-quantum correlation (HSQC) spectra showed that in all titrations, binding kinetics were in the fast-to-intermediate exchange regime, with at least 11 residues showing a large chemical shift (δg > 0.1 ppm) (Figure 3A). Affected residues in the Pin1 WW-domain were located at the C-terminus of the β1-strand (S16), the β1-β2 loop (R17, S18, and G20), the β2-strand (R21, Y23, and F25), and the C-terminus of the β3-strand (W34 and E35) (Figure 3A; residues in bold). During our NMR titrations, the 1H/15N HSQC spectra showed line broadening for the Arg17 peak, resulting in its disappearance (Figure 3A). We attribute the broadening of the Arg17 line to its close proximity with the TFCD, with interactions occuring in the intermediate exchange regime in the NMR time scale (Figure 3B). A Kd of 133 ± 13 µM was determined from the δg (ppm) for HN changes in Pin1 WW-domain residue Ser18 (Figure 3A; box around moving chemical shift and inset graph), which is similar to our calorimetric measurements showing a fitted Kd value of ~137 mM (Online Supplementary Figure S1).

We next prepared 13C/15N- isotopically-labeled Pin1 WW-domain in complex with double phosphorylated pSer253/pSer258 TFCD peptide for sequential backbone 1078

assignment and NMR structure determination. Analysis of the Cα and Cβ values based upon the weighted chemical shift index highlighted the three β-strands, which are characteristic of WW-domains (Online Supplementary Figure S2A). A total of 203 unique intra- and intermolecular distance constraints were derived from the analysis of the NOE spectroscopy (NOESY) experiments. We used these constraints together with the weighted chemical shift index (Online Supplementary Table S1 and Online Supplementary Figure S2A) to calculate the solution structure of the Pin1 WW-domain and TFCD complex (see Methods for more details). The 20 lowest energy conformers of the Pin1 WW-domain yielded a root-meansquare deviation (RMSD) of 1 Å (residues 6-39) (Figure 3B and Online Supplementary Figure S2B). The Pin1 WWdomain is a canonical WW-domain18 consisting of three twisted anti-parallel β-sheet strands with a conserved TrpTrp motif located at the N-terminus of the first β-strand and C-terminus of the third β-strand, respectively.

Analysis of the Pin1 WW-domain–TFCD complex reveals that TFCD residues Asn257, pSer258 and Pro259 are involved in the binding interface of the Pin1 WWdomain, consistent with our pull-down assays. While the Pin1 WW-domain as a whole shows a well-defined, single conformation, the β1-β2 loop1 binding region of the WWdomain (residues 17 to 21) and the pSer258-Pro259 motif of TFCD show two distinct ensembles of conformers (Figure 3B and C), consistent with the fact that less NOEs were detected for the β1-β2 loop1. The features that principally drive the complex formation are the charge-charge interaction and the hydrophobic interaction between Trp34 (the second invariant Trp of the WW motif) with invariant Pro259. The ionic interactions between the phosphate of pSer258 with the positively charged guanidinium groups of Arg14, Arg17, and Arg21 also appeared to stabilize the complex. However, as mentioned earlier, the resonances for Arg17, located at position 1 of loop1, disappeared during the NMR titration due to intermediate exchange in the NMR timescale (Figure 3A), suggesting high flexibility around loop1. Arg21 connects loop1 to the β2-strand into two conformations and involves multiple polar contacts with the double phosphorylated TFCD at pSer258 (Figure 3B and C). Similarly, Trp34-driven hydrophobic packing of Pro259 induces two ensembles of Trp34 side chain rotamers and Pro259 configurations, and both ensemble states still adopt a trans-configuration of the pSer258-Pro259 peptide bond (Online Supplementary Table S1). For comparison, the previously determined interactions of Pin1 WW-domain with Cell division cycle 25 (Cdc25) and the C-terminal domain of the RNA polymerase II largest subunit (RNAP II-CTD) are also shown (Figure 3D).29,30 In order to further confirm the reliability of our structures, step-wise energy minimization was performed on all Pin1 WW-domain-TFCD complexes. The RMSD of heavy atoms/all atoms before and after energy minimization was less than 0.45 Å, while the RMSD of the backbone was less than 0.3 Å (Online Supplementary Figure S2B). Therefore, the interactions between the Pin1 WWdomain and the TFCD, such as the electrostatic interactions with pSer258 and the hydrophobic interactions of Pro259 with the WW-domain are consistent with the NMR structures calculated. The trans conformation of the pSer258-Pro259 peptide bond in the TFCD was stable durhaematologica | 2018; 103(6)

Pin1 enhances TF expression and activity

ing 10,000 steps of energy minimization, further demonstrating that our structures satisfy energy-of-motion (EOM) criteria with reasonable convergence parameters.

Taken together, our structural analyses show the interaction between TFCD and Pin1 involves phosphorylation of Ser258 and trans-configuration of the pSer258-Pro259 peptide bond in the TFCD.

Pin1 extends TF protein half-life via the TFCD Given our findings that Pin1 and TF interact and that Pin1 is known to enhance the protein half-life of several of

its substrates,18 we sought to determine the effect of Pin1 on TF protein half-life using full-length TF and TF lacking the cytoplasmic domain in cycloheximide-treated HEK293T cells. Pin1 significantly enhanced the protein half-life of full-length TF (Figure 4A), but not that of TFDCD (Figure 4B). Furthermore, Pin1 mutants with an inactive WW-domain (Pin1:W34A) or lacking isomerase activity (Pin1:K63A) were both still able to enhance the protein half-life of full-length TF (Figure 4A). Similar effects of Pin1 and Pin1 mutants on TF protein half-life were observed for endogenous TF protein in PMA-stimulated

A

B

C

Figure 4. Pin1 increases Tissue Factor (TF) protein half-life via the twenty-amino acid cytoplasmic domain (TFCD). (A) HEK293T cells over-expressing full-length TF with or without Pin1 or Pin1 mutants were treated with cycloheximide (CHX) for times indicated in hours (hrs). TF protein levels were determined by western blot. The graph shows the amount of TF protein remaining after CHX treatment as a percentage of the starting TF protein level. (B) HEK293T cells over-expressing TFâ&#x2C6;&#x2020;CD with or without Pin1 were treated with CHX for times indicated. TF protein levels were determined by western blot. The graph shows the amount of TF protein remaining after CHX treatment as a percentage of the starting TF protein level. (C) PMA-stimulated smooth muscle cells (SMC) over-expressing Pin1 or Pin1 mutants were treated with CHX for times indicated. Endogenous TF protein levels were determined by western blot. The graph shows the amount of TF protein remaining after CHX treatment as a percentage of the starting TF protein level. Data are shown as meanÂąStandard Error of Mean. P-values were calculated using two-way ANOVA. *P<0.05, **P<0.01, ***P<0.001 versus Mock-transfected controls.

haematologica | 2018; 103(6)

1079

K. Kurakula et al.

SMCs (Figure 4C). These results indicate that Pin1 can stabilize TF, and that this stabilization is dependent on interaction of Pin1 with the TFCD, but is not necessarily dependent on Pin1 isomerase activity.

Pin1 enhances TF pro-coagulant activity via the TFCD Having established that Pin1 enhances TF gene expression and TF protein half-life, we next examined the impact of Pin1 on the pro-coagulant activity of TF. We measured the Factor Xa-generating potential of the ECRF24 cell line and primary human ECs and SMCs after Pin1 gain and loss-of-function. Pin1 overexpression markedly increased TF pro-coagulant activity in both ionomycin-treated SMCs and TNF-α treated EC-RF24 cells and ECs (Figure 5A-C). In contrast, knockdown of Pin1 by siRNA or inhibition of Pin1 isomerase activity by Juglone attenuated the TNF-α induced pro-coagulant activity of TF in these cells (Figure 5A-C). Furthermore, using SMCs derived from wild-type mice and mice expressing the TFDCD truncated protein from the endogenous TF gene locus,23 we found that Pin1 can increase the activity of fulllength TF, but not that of TF lacking the cytoplasmic domain (Figure 5D). Taken together, these results show that Pin1 interaction with TF via the TFCD strongly increases the pro-coagulation activity of TF.

complex showed apparent chemical shift perturbation (e.g. Tyr23) (Figure 3A), while these residues were stagnant in previous Pin1 NMR titrations.32 This discrepancy could potentially be the result of titration of the short pThr-Pro fragment of Cdc25 or tau in these experiments, rather than the full protein domain that was used here.

Furthermore, the β1-β2 loop1 conformer shows structural similarity to the isolated ligand-free Pin1 WW-domain X-

A

B

Discussion Tissue Factor plays a crucial role in initiating the extrinsic blood coagulation cascade. Therefore, detailed knowledge of the interacting proteins that modulate TF activity in vascular cells is essential to understand the regulation of thrombus formation under pathological conditions and may lead to novel intervention strategies for thrombotic disease. Here, we show that the peptidyl-prolyl isomerase Pin1 both enhances TF gene expression via activation of NF-κB and AP-1 signaling and directly interacts with TF through a well-conserved phosphorylated Ser258-Pro259 motif in its cytoplasmic domain. We elucidate the structural details of this interaction and show that Pin1 increases both the protein half-life and pro-coagulant activity of TF in vascular cells. Additional effects of Pin1 on TF activity may come from protease-activated receptor 2-induced release of TF on microvesicles, which was not studied in our experiments but were recently described.31 We demonstrate that Pin1 is a potent activator of TF gene expression in activated ECs and SMCs. The promoter of the human TF gene contains transcription factor binding sites for NF-κB, AP-1, Sp-1 and Egr-1, which are essential for inducing TF expression in many cell types.9,24 Interestingly, Pin1 has been shown to bind and regulate the activity of all four of these transcription factors in a cell type-dependent manner.15 The deletion of Egr-1 and Sp1 response elements in the TF promoter constructs revealed that there is a complete abrogation of TF promoter activity (data not shown). Therefore, the effect of Pin1 on activation of the TF gene promoter by Egr-1 and Sp1 cannot be determined.

Concerning our detailed structural analyses, we conclude that the TFCD-Pin1 WW-domain complex shows a larger contact area than what was observed in previous WW-domain/phosphopeptide NMR structures, but similar to the RNAP II-CTD:Pin1 crystal structure.32 Residues in the anchoring zone of the Pin1 WW-domain-TFCD 1080

C

D

E

Figure 5. Pin1 enhances Tissue Factor (TF) pro-coagulant activity via the twenty-amino acid cytoplasmic domain (TFCD). (A-C) Factor Xa generation as a measure of TF activity in human human umbilical vein endothelial cells (HUVECs) (A), EC-RF24 cells (B), or smooth muscle cells (SMC) (C) after either overexpression or knockdown of Pin1 or treatment with the Pin1 inhibitor Juglone for 16 hours (hrs) followed by serum-starvation and treatment with ionomycin for 3 hrs (SMCs) or TNF-α for 16 hrs [EC-RF24 and endothelial cells (ECs)]. (D) Factor Xa generation as a measure of TF activity in mouse SMCs derived from wild-type mice (WT) or mice expressing TF∆CD from the endogenous TF gene locus after overexpression of Pin1 and stimulation with ionomycin for 3 hrs. (E) Summary of Pin1 effects on TF: Pin1 enhances the protein half-life and pro-coagulant activity of TF through interaction with the conserved pSer258Pro259 motif in the TFCD and enhances the activity of the transcription factors AP-1 and NF-κB to increase TF gene expression. Data are shown as mean±Standard Error of Mean. P-values were calculated using two-tailed Student’s t-test (A-C) or two-way ANOVA (D). *P<0.05, **P<0.01, ***P<0.001.

AU: arbitrary units.

haematologica | 2018; 103(6)

Pin1 enhances TF expression and activity

ray structures, as well as to the Cdc25 pThr peptide complex NMR structure, suggesting that loop1 undergoes a thermodynamic switch between ligand-bound and ligandfree states (Figure 3D).

Our calorimetric experiments revealed a weak interaction between the unphosphorylated TFCD and the Pin1 WW-domain (data not shown), with unphosphorylated TFCD having a lower affinity for Pin1 (Kd of 1.53 mM) than double phosphorylated TFCD (Kd of 137 mM).

Similarly, our pull-down assays showed that unphosphorylated TF peptides can still pull-down Pin1 protein (Figure 2E). These two results suggest that Pin1 can interact with TF in both phosphorylation-specific and phosphorylationindependent manners. It has previously been shown that pre-attachment of Pin1 to an unphosphorylated “dynamic anchoring” region of c-Myc that is distant from its pSerPro motif is functionally important for the interaction between these two proteins.33 It could, therefore, be speculated that a similar distal, phosphorylation-independent interaction occurs between Pin1 and TF. Further studies are needed to confirm whether or not this is indeed the case.

It is also important to emphasize that intermolecular stacking of TFCD trans-Pro259 and Pin1 WW-domain Trp34 residues is highly orthologous to the intramolecular stacking between Trp8 and Pro37 of the Pin1 WWdomain. Given that mutational studies have shown that Trp8-Pro37 stacking can stabilize the WW-domain fold,32 we think that a similar, but intermolecular trans-Pro259 to Trp34 interaction-mechanism is utilized to promote the observed affinity between Pin1 and TF even when TFCD is unphosphorylated at Ser258. Indeed, complex formation due to the hydrophobic Pro-Trp stacking with unphosphorylated Ser-Pro or Ser-Thr motifs is consistent with previous studies in mammalian cell lines34 and in vitro studies of the WW-domain binding to unphosphorylated Ser-Pro motifs such as the cytoplasmic polyadenylation element binding (CPEB) protein.35,36 In our peptide pull-down and protein half-life experiments, we found that full-length Pin1 as well as Pin1 mutants with a disrupted WW-domain (Pin1:W34A) or lacking isomerase activity (Pin1:K63A) could bind the TFCD and stabilize TF protein. The discrepancy between these findings and our NMR data showing the importance of the Pin1 Trp34 residue in binding the TFCD can be

References 1. Versteeg HH, Heemskerk JW, Levi M, Reitsma PH. New fundamentals in hemostasis. Physiol Rev. 2013;93(1):327-358.

2. Furie B, Furie BC. Mechanisms of thrombus formation. N Engl J Med. 2008; 359(9):938-949.

3. Taubman MB, Fallon JT, Schecter AD, et al.

Tissue factor in the pathogenesis of atherosclerosis. Thromb Haemost. 1997; 78(1):200-204.

4. Moons AH, Levi M, Peters RJ. Tissue factor and coronary artery disease. Cardiovasc Res. 2002;53(2):313-325.

5. Aras O, Shet A, Bach RR, et al. Induction of microparticle- and cell-associated intravascular tissue factor in human endotoxemia.

haematologica | 2018; 103(6)

explained by the fact that purified Pin1 WW-domain was used in our NMR studies, and that both domains of Pin1 can recognize and bind pSer-Pro motifs independently of each other.37 Therefore, it is possible that the Pin1 WWdomain mutant (Pin1:W34A) interacts with and stabilizes TF via its isomerase domain. However, whether the interaction between TF and Pin1 involves sequential and synergistic action of the Pin1 WW-domain and isomerase domain remains to be determined.

Several human and animal studies have shown that TF plays a crucial role in thrombosis development in vivo. We demonstrated that Pin1 has a regulatory role in FXa generation in both activated ECs and SMCs in vitro. Additional studies with Pin1-deficient mice may further delineate the role of Pin1 in TF gene expression, TF activity, and coagulation in vivo. However, Pin1-deficient mice already suffer from decreased body weight, testicular and retinal atrophies, and deficiencies in breast proliferative changes during pregnancy, which may interfere with in vivo coagulation experiments.38 In summary, our results demonstrate that Pin1 enhances TF gene expression, interacts with TF via the TFCD, and positively modulates TF half-life and pro-coagulant activity in vascular cells. Our findings suggest that Pin1 contributes to a hypercoagulable state in the local and systemic circulation. Specific Pin1 inhibitors could, therefore, be considered as potential novel therapeutic agents for the prevention of coagulation-driven pathologies, such as thrombosis and atherosclerosis.

Funding This work was supported by the research program of the BioMedical Materials institute, co-funded by the Dutch Ministry of Economic Affairs as a part of Project P1.02 NEXTREAM. This work was also supported by the Rembrandt Institute for Cardiovascular Research (RICS-grant 2013), the Netherlands CardioVascular Research Initiative (CVON: 2012-08), and the National Institute of Health (grants HL-60742 and P01HL16411).

Acknowledgments We would like to thank Dr. Youlin Xia and the UH NMR facility-DOR for NMR data acquisition, CACDS for the structure calculations, and Amr Elnashai, Amy Sater, and Glen Legge for their support of this research.

Blood. 2004;103(12):4545-4553.

6. Ahamed J, Niessen F, Kurokawa T, et al.

Regulation of macrophage procoagulant responses by the tissue factor cytoplasmic domain in endotoxemia.

Blood.

2007;109(12):5251-5259.

7. Belting M, Dorrell M, Sandgren S, et al.

Regulation of angiogenesis by tissue factor cytoplasmic domain signaling. Nat Med. 2004;10(5):502-509.

8. Schaffner F, Versteeg HH, Schillert A, et al.

Cooperation of tissue factor cytoplasmic domain and PAR2 signaling in breast cancer development. Blood. 2010; 23;116(26): 6106-6113.

9. Mackman N. Regulation of the tissue factor gene. FASEB J. 1995;9(10):883-889.

10. Schönbeck U, Mach F, Sukhova GK, et al.

CD40 ligation induces tissue factor expres-

sion in human vascular smooth muscle cells. Am J Pathol. 2000;156(1):7-14.

11. Levi M, ten Cate H, Bauer KA, et al.

Inhibition of endotoxin-induced activation of coagulation and fibrinolysis by pentoxifylline or by a monoclonal anti-tissue factor antibody in chimpanzees. J Clin Invest.

1994;93(1):114-120.

12. Ettelaie C, Collier ME, Featherby S, Greenman J, Maraveyas A. Oligoubiquitination of tissue factor on Lys255 promotes Ser253-dephosphorylation and terminates TF release. Biochim Biophys Acta. 2016;1863(11):2846-2857.

13. Ettelaie C, Elkeeb AM, Maraveyas A, Collier ME. p38alpha phosphorylates serine 258 within the cytoplasmic domain of tissue factor and prevents its incorporation into cell-derived microparticles. Biochim

1081

K. Kurakula et al.

Biophys Acta. 2013;1833(3):613-621.

14. Sen M, Herzik M, Craft JW, et al.

Spectroscopic Characterization of Successive Phosphorylation of the Tissue Factor Cytoplasmic Region. Open Spectrosc J. 2009;3:58-64.

15. Dorfleutner A, Ruf W. Regulation of tissue factor cytoplasmic domain phosphorylation by palmitoylation. Blood. 2003; 102(12):3998-4005.

16. Rothmeier AS, Liu E, Chakrabarty S, et al.

Identification of the integrin-binding site on coagulation factor VIIa required for proangiogenic PAR2 signaling. Blood. 2018; 131(6):674-685.

17. Ranganathan R, Lu KP, Hunter T, Noel JP. Structural and functional analysis of the mitotic rotamase Pin1 suggests substrate recognition is phosphorylation dependent.

Cell. 1997;89(6):875-886.

18. Joseph JD, Yeh ES, Swenson KI, Means AR. The peptidyl-prolyl isomerase Pin1. Prog Cell Cycle Res. 2003;5:477-487.

19. Wulf G, Finn G, Suizu F, Lu KP. Phosphorylation-specific prolyl isomerization: is there an underlying theme? Nat Cell Biol. 2005;7(5):435-441.

20. Lu KP, Zhou XZ. The prolyl isomerase PIN1: a pivotal new twist in phosphorylation signalling and disease. Nat Rev Mol Cell Biol. 2007;8(11):904-916.

21. Fujimoto Y, Shiraki T, Horiuchi Y, et al.

Proline cis/trans-isomerase Pin1 regulates peroxisome proliferator-activated receptor gamma activity through the direct binding to the activation function-1 domain. J Biol Chem. 2010;285(5):3126-3132.

22. van Tiel CM, Kurakula K, Koenis DS, van der Wal E, de Vries CJ. Dual function of Pin1 in NR4A nuclear receptor activation:

1082

23.

24.

25.

26.

27.

28.

29.

30.

enhanced activity of NR4As and increased Nur77 protein stability. Biochim Biophys Acta. 2012;1823(10):1894-1904.

Melis E, Moons L, De Mol M, et al.

Targeted deletion of the cytosolic domain of tissue factor in mice does not affect development. Biochem Biophys Res Commun. 2001;286(3):580-586.

Oeth P, Parry GC, Mackman N. Regulation of the tissue factor gene in human monocytic cells. Role of AP-1, NF-kappa B/Rel, and Sp1 proteins in uninduced and lipopolysaccharide-induced expression.

Arterioscler Thromb Vasc Biol. 1997;17(2): 365-374.

Johnson BA. Using NMRView to visualize and analyze the NMR spectra of macromolecules. Methods Mol Biol. 2004; 278:313-352.

Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A. NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR. 1995;6(3):277293. van den Hengel LG, Osanto S, Reitsma PH, Versteeg HH. Murine tissue factor coagulant activity is critically dependent on the presence of an intact allosteric disulfide.

Haematologica. 2013;98(1):153-158.

Hennig L, Christner C, Kipping M, et al.

Selective inactivation of parvulin-like peptidyl-prolyl cis/trans isomerases by juglone.

Biochemistry. 1998;37(17):5953-5960.

Zhou XZ, Kops O, Werner A, et al. Pin1dependent prolyl isomerization regulates dephosphorylation of Cdc25C and tau proteins. Mol Cell. 2000;6(4):873-883.

Verdecia MA, Bowman ME, Lu KP, Hunter T, Noel JP. Structural basis for phosphoserine-proline recognition by group IV WW-

31.

32.

33.

34.

35.

36.

37.

38.

domains. Nat Struct Biol. 2000;7(8):639643.

Ettelaie C, Collier MEW, Featherby S, Greenman J, Maraveyas A. Peptidyl-prolyl isomerase 1 (Pin1) preserves the phosphorylation state of tissue factor and prolongs its release within microvesicles. Biochim Biophys Acta. 2018;1865(1):12-24.

Wintjens R, Wieruszeski JM, Drobecq H, et al. 1H NMR study on the binding of Pin1 Trp-Trp domain with phosphothreonine peptides. J Biol Chem. 2001;276(27):2515025156.

Helander S, Montecchio M, PilstĂĽl R, et al.

Pre-Anchoring of Pin1 to Unphosphorylated c-Myc in a Fuzzy Complex Regulates c-Myc Activity.

Structure. 2015;23(12):2267-2279.

Macias MJ, Gervais V, Civera C, Oschkinat H. Structural analysis of WW-domains and design of a WW prototype. Nat Struct Biol.

2000;7(5):375-379.

Nechama M, Lin CL, Richter JD. An unusual two-step control of CPEB destruction by Pin1. Mol Cell Biol. 2013;33(1):48-58.

Schelhorn C, Martin-Malpartida P, Sunol D, Macias MJ. Structural Analysis of the Pin1-CPEB1 interaction and its potential role in CPEB1 degradation. Sci Rep. 2015;5:14990.

Innes BT, Bailey ML, Brandl CJ, Shilton BH, Litchfield DW. Non-catalytic participation of the Pin1 peptidyl-prolyl isomerase domain in target binding. Front Physiol.

2013;4:18.

Liou YC, Ryo A, Huang HK, et al. Loss of Pin1 function in the mouse causes phenotypes resembling cyclin D1-null phenotypes. Proc Natl Acad Sci USA. 2002; 99(3):1335-1340.

haematologica | 2018; 103(6)

ARTICLE

Coagulation & its Disorders

Mass spectrometry-assisted identification of ADAMTS13-derived peptides presented on HLA-DR and HLA-DQ

Ferrata Storti Foundation

Johana HrdinovĂĄ,1* Fabian C. Verbij,1* Paul H.P. Kaijen,1 Robin B. Hartholt,1 Floris van Alphen,2 Neubury Lardy,3 Anja ten Brinke,4 Karen Vanhoorelbeke,5 Pooja J. Hindocha,6 Anne S. De Groot,6,7 Alexander B. Meijer,1,2,8 Jan Voorberg1,9 and Ivan Peyron1

Department of Plasma Proteins, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, University of Amsterdam, the Netherlands; 2Department of Research Facilities, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, University of Amsterdam, the Netherlands; 3Department of Immunogenetics, Sanquin, Amsterdam, the Netherlands; 4Department of Immunopathology, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, University of Amsterdam, the Netherlands; 5Laboratory for Thrombosis Research, IRF Life Sciences, KU Leuven Campus Kulak Kortrijk, Belgium; 6EpiVax Inc., Providence, RI, USA; 7Institute for Immunology and Informatics, University of Rhode Island, Providence, RI, USA; 8 Pharmaceutics, Utrecht Institute for Pharmaceutical Sciences, Utrecht University, the Netherlands and 9Department of Experimental Vascular Medicine, Academic Medical Center, Amsterdam, the Netherlands 1

Haematologica 2018 Volume 103(6):1083-1092

*JH and FCV contributed equally.

ABSTRACT

F

ormation of microthrombi is a hallmark of acquired thrombotic thrombocytopenic purpura. These microthrombi originate from insufficient processing of ultra large von Willebrand factor multimers

by ADAMTS13 due to the development of anti-ADAMTS13 autoantibodies. Several studies have identified the major histocompatibility complex class II alleles HLA-DRB1*11, HLA-DQB1*03 and HLA-DQB1*02:02 as risk factors for acquired thrombotic thrombocytopenic purpura development. Previous research in our department indicated that ADAMTS13 CUB2 domain-derived peptides FINVAPHAR and LIRDTHSLR are presented on HLA-DRB1*11 and HLA-DRB1*03, respectively. Here, we describe the repertoire of ADAMTS13 peptides presented on HLA-DQ. In parallel, the repertoire of ADAMTS13-derived peptides presented on HLA-DR was monitored. Using HLA-DR- and HLA-DQ-specific antibodies, we purified HLA/peptide complexes from ADAMTS13-pulsed monocyte-derived dendritic cells. Using this approach, we identified ADAMTS13-derived peptides presented on HLA-DR for all 9 samples analyzed; ADAMTS13-derived peptides presented on HLA-DQ were identified in 4 out of 9 samples. We were able to confirm the presentation of the CUB2 domain-derived peptides FINVAPHAR and LIRDTHSLR on HLA-DR. In total, 12 different core-peptide sequences were identified on HLA-DR and 8 on HLA-DQ. For HLA-DR11, several potential new corepeptides were found; 4 novel core-peptides were exclusively identified on HLA-DQ. Furthermore, an in silico analysis was performed using the EpiMatrix and JanusMatrix tools to evaluate the eluted peptides, in the context of HLA-DR, for putative effector or regulatory T-cell responses at the population level. The results from this study provide a basis for the identification of immuno-dominant epitopes on ADAMTS13 involved in the onset of acquired thrombotic thrombocytopenic purpura.

Introduction Thrombotic thrombocytopenic purpura (TTP) is a severe life-threatening disorder caused by decreased levels of functional ADAMTS13 (a disintegrin and metalloproteinase with thrombospondin type 1 motifs, member 13). In healthy individuals, ADAMTS13 regulates the size of von Willebrand Factor (VWF) multihaematologica | 2018; 103(6)

Correspondence: j.voorberg@sanquin.nl

Received: August 21, 2017.

Accepted: March 14, 2018.

Pre-published: March 22, 2018. doi:10.3324/haematol.2017.179119 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/6/1083 Š2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1083

J. Hrdinová et al.

mers through cleavage of a Tyr1605-Met1606 peptide bond in the A2 domain of VWF.1,2 Functional or quantitative defects in ADAMTS13 levels in the circulation lead to the accumulation of high molecular weight VWF multimers and the formation of platelet- and VWF-rich thrombi. Within the microvasculature, these thrombi cause mechanical fragmentation of erythrocytes inducing hemolytic anemia.1,2 In addition, the presence of hyperadhesive VWF multimers results in platelet consumption.

As a consequence, patients with TTP often present with skin petechiae due to thrombocytopenia-induced blood loss from small vessels in the skin.1,3 Additional clinical symptoms may include fever, renal failure or neurological abnormalities.1,2 In the majority of patients with TTP, the decrease in ADAMTS13 levels is due to the development of autoantibodies directed towards ADAMTS13. Most of these autoantibodies are composed of IgG1 and IgG4 subclasses;4-6 these antibodies either inhibit the proteolytic function of ADAMTS13 or enhance its clearance from the circulation.6-9 While the mechanisms responsible for the development of anti-ADAMTS13 antibodies are currently unknown, several reports have suggested that infections, pregnancy or transplantation may be considered to be risk factors for the onset of acquired TTP.10-12 The generation of high affinity antibodies against ADAMTS13 is dependent on the help of specific CD4+ T cells. Priming of antigenspecific CD4+ T cells requires presentation of ADAMTS13-derived peptides on major histocompatibility complex class II (MHC-II) on professional antigen presenting cells.13 The MHC-II genes are highly polymorphic allowing for the selection of a broad repertoire of CD4+ T cells that is needed to combat infections. Specific MHC-II alleles have been linked to autoimmune disorders such as rheumatoid arthritis and celiac disease.14 Similarly, association studies from three different cohorts of patients with acquired TTP have identified HLA-DRB1*11 as a risk factor.15-17 Conversely, the frequency of HLA-DRB1*04 was significantly lower in patients with acquired TTP, suggesting a protective effect of this allele.15-17 In addition to HLA-DRB1*11, higher frequencies of alleles HLA-DQB1*0315,16 and HLA-DQB1*02:0217 were found in patients with acquired TTP when compared to healthy controls. A recent study of 190 Italian TTP patients and 1255 healthy controls suggested that HLA-DQB1*05:03 was less prevalent in patients with acquired TTP.18 This study also proposed that the common single nucleotide polymorphism rs6903608, which is located between the genes encoding the alpha and beta5 chains of the HLA-DR complex, combined with HLA-DQB1*05:03 explains most of the observed association between the HLA locus and acquired TTP.18 As yet, the molecular mechanism underlying the observed association between polymorphic sites within the MHC II locus and acquired TTP has not been identified.

Previous observation from our laboratory has shown that monocyte-derived dendritic cells (mo-DCs) from healthy donors preferentially presented two peptides derived from the CUB2 domain of ADAMTS13.19 Both of these peptides were found to activate CD4+ T cells of patients with acquired TTP.20 In addition, CUB2 domainderived peptide ADAMTS131239-1253 was identified as an immunodominant T-cell epitope in an HLA-DRB1 transgenic mouse model.21 The same study revealed that ADAMTS131239-1253 reactive CD4+ T cells were present in 1084

patients with acquired TTP as well as in peripheral blood of healthy individuals.21 As yet, the presentation of ADAMTS13-derived peptides on HLA-DQ has not been investigated. In the present work, we aimed to define the repertoire of ADAMTS13-derived peptides presented on HLA-DQ and prospectively identify putative effector and tolerated/tolerogenic T-cell epitopes using computational tools (EpiMatrix and JanusMatrix).

Methods Materials Recombinant full length ADAMTS13 was produced in stable transfected HEK293 cells and purified as described previously.9 Concentration of purified ADAMTS13 was determined using the Bradford assay. Lipopolysaccharide (LPS) was obtained from Sigma-Aldrich (St. Louis, USA). The hybridoma producing the HLA-DQ-specific antibody (SPV-L3)22 was a kind gift from Prof. dr. H. Spits (Academic Medical Center, Amsterdam, the Netherlands). The hybridoma producing the HLA-DR-specific monoclonal antibody (L243) was purchased from ATCC (Wesel, Germany). Antibodies were purified from culture supernatant via protein A Sepharose (GE Healthcare) and coupled to CNBr Sepharose 4B at a final concentration of 2 mg/mL (Amersham Biosciences, Buckinghamshire, UK).

Endocytosis of ADAMTS13 and affinity purification of HLA-DR and HLA-DQ Monocytes were obtained from healthy volunteers in accordance with Dutch regulations after approval from the Sanquin Ethical Advisory Board, in accordance with the Declaration of Helsinki. After five days of differentiation, 5x106 immature moDCs were incubated with 100 nM of recombinant ADAMTS13 in Cellgro medium supplemented with 800 U/ml IL-4 and 1000 U/mL GM-CSF. After five hours of incubation, the medium was supplemented without washing with 1 mg/mL of LPS and 1% fetal calf serum to allow for mo-DCs maturation overnight. After maturation, the cells were detached from the plate using trisodium citrate and gentle pipetting. HLA-DR and HLA-DQ peptide complexes were purified using a modification of a previously described protocol.23 Briefly, cells were resuspended in 500 mL lysis buffer containing 10 mM Tris-HCl (pH 8.0), 0.25% octyl-b-D-glucopyranoside, 1% sodium deoxycholate and Halt Protease and Phosphatase Inhibitor Cocktail (Thermo Scientific), and incubated at 4°C for 30 minutes (mins). The lysed cells were centrifuged at 20,000xg for 15 mins at 4°C. HLA-DR and HLA-DQ peptide complexes were purified from the supernatant by adding L243 or SPVL3 coupled CNBr Sepharose, respectively, and incubating the tubes at 4°C overnight ‘end-over-end’, using a laboratory rotator.

L243 or SPV-L3 Sepharose beads were then washed 2 times with lysis buffer and 5 times with 10 mM Tris-HCl (pH 8.0). HLA-peptide complexes were eluted using 500 mL 10% acetic acid for 10 mins at room temperature. Then the samples were centrifuged at 400xg for five mins at room temperature and the supernatant transferred to low-binding 1.5 mL Eppendorf tubes and heated for 15 mins at 70°C. Samples were desalted using C18 STAGE Tips (Dr. Maisch Gmbh, Amersfoort, the Netherlands).24 STAGE Tips were eluted with 60 mL of 1% formic acid/30% acetonitrile and eluates were concentrated using a SpeedVac (Savant, SPP1110, Thermo Scientific) to a final volume of 5 mL.

Mass spectrometry data analysis Raw data files from the Orbitrap Fusion Tribid were scored against the uniprot-organism_9606_AND_keyword_kw_0181 haematologica | 2018; 103(6)

Presentation of ADAMTS13 peptides on HLA-DR and HLA-DQ Table 1. MHC-II genotype of healthy donors included in the study. The donors included in this study were typed for HLA-DRB1/DQB1 using PCRSBT and HLA-DQA1 using next generation sequencing workflow.

Donor 1 2 3 4 5 6 7 8 9

HLA-DRB1 07:01 04:04 01:01 01:01 01:01 01:01 01:01 04:01 03:01

14:54 15:01 15:01 08:01 07:01 11:01 11:01 11:01 13:01

HLA-DQA1 01:04 01:02 01:01 01:01 01:01 01:01 01:01 03:01 01:01

database using Proteome Discover 1.4 (Thermo Scientific) with a 20 ppm tolerance for precursor mass and 10 ppm tolerance for fragment mass. Oxidation (+15.995 Da) on methionine was selected as dynamic modification. A decoy database comprising the reverse protein sequences from the same database was used to obtain a false discovery rate (FDR). Only peptides with a high or medium confidence (FDR threshold 0.05%) were considered for protein scoring.

Results Presentation of ADAMTS13-derived peptides on HLA-DQ To assess the contribution of HLA-DQ to the presentation of ADAMTS13-derived peptides, we pulsed mo-DCs from a panel of 9 HLA-typed healthy donors with 100 nM ADAMTS13. The HLA type of the donors included in this study is shown in Table 1. Mo-DCs from the same panel of donors were recently analyzed for the presentation of FVIII-derived peptides on HLA-DQ.25 Mo-DCs were incubated with 1 mg/mL LPS and 1% fetal calf serum to allow for their maturation. HLA-DR and HLA-DQ molecules were purified employing monoclonal antibodies directed against HLA-DR (L243) and HLA-DQ (SPV-L3),22 and the eluates were analyzed by mass-spectrometry. In agreement with our previous reports,19,23,25 peptides derived from endogenously expressed as well as internalized proteins were presented on HLA-DR and HLA-DQ (Online Supplementary Table S1A and B, respectively). A subset of peptides was derived from proteins that are found within the endolysosomal compartment; these include HLA-DR itself, several proteases (such as cathepsin B) and endocytic receptors (such as the macrophage mannose receptor and prolow-density lipoprotein receptor-related protein 1).

The total number of unique peptides that was identified in the pulldowns of HLA-DR and HLA-DQ is shown in Figure 1A and Online Supplementary Figure S1. The total number of peptides presented on HLA-DQ was approximately 3-fold lower when compared to HLA-DR (median of 566 unique peptides on HLA-DQ compared to 1521 on HLA-DR). Similarly, although not statistically significant, the number of ADAMTS13-derived peptides presented on HLA-DQ was 3-fold lower when compared to HLA-DR (Figure 1B). HLA-DR-presented ADAMTS13-derived peptides were identified for all donors whereas ADAMTS13derived peptides presented on HLA-DQ were only found haematologica | 2018; 103(6)

HLA-DQB1 02:01 03:01 04:01 04:01 02:01 05:05 05:05 05:05 05:01

03:03 03:02 05:01 04:02 02:02 03:01 03:01 03:01 02:01

05:03 06:02 06:02 05:01 05:01 05:01 05:01 03:02 06:03

in 4 out of 9 donors (Table 2A and B, and Online Supplementary Figure S1B).

Repertoire of ADAMTS13-derived peptides presented on HLA-DR In order to compare the repertoire of ADAMTS13derived peptides presented on HLA-DR and HLA-DQ, the binding cores of the ADAMTS13-derived unique peptides were predicted using the NetMHCIIpan 3.1 software (Table 2A and B).26 ADAMTS13-derived peptides that shared the same core sequence were grouped (Figure 2).

Previously, we have shown that ADAMTS13-derived peptides containing the sequence FINVAPHAR were presented on HLA-DRB1*11.19 Peptides sharing the same sequence were also presented on non-HLA-DRB1*11 positive donors when mo-DCs were pulsed with 500 nM of ADAMTS13.19 In the current study, peptides containing the FINVAPHAR sequence were presented on mo-DCs of HLA-DRB1*11 positive donors 7 and 8. Peptides with core-sequence FINVAPHAR were also presented by moDCs derived of HLA-DRB1*11 negative donors 1 and 9 (Table 2A). However, no peptides containing the FINVAPHAR sequence were identified in HLA-DRB1*11 positive donor 6 (Table 2A). Peptides derived from the sequence IHALATNMG which is located adjacent to the FINVAPHAR peptide were found in DRB1*11 negative donor 2 (Table 2A). Peptides derived from sequence LIRDTHSLR were identified in DRB1*03 positive donor 9.

In a previous study, LIRDTHSLR-derived peptides were also presented by mo-DCs derived of DRB1*03 positive donors.19 It is noteworthy that donors 6 and 7 with the same HLA-DRB1 haplotype (DRB1*01:01/DRB1*11:01) both presented peptides derived of core sequence LKTLPPARC originating from the TSP3 domain of ADAMTS13 (Table 2A). The majority of the HLA-DR-presented peptides identified in this study were derived from the CUB domains (Table 2A and Figure 2).

HLA-DQ contributes to presentation of ADAMTS13-derived peptides We found ADAMTS13-derived peptides associated with HLA-DQ in 4 out of 9 donors (Table 2B and Figure 2). We identified 4 sets of peptides that were exclusively presented by HLA-DQ that originated from the cysteinerich, TSP2, and TSP2-linker 1 of ADAMTS13 (Figure 2).

This suggests that the repertoire of peptides presented on HLA-DQ differs from that presented on HLA-DR. We also identified peptides that were presented both on HLA-DR 1085

J. HrdinovĂĄ et al.

and HLA-DQ. This is exemplified by the peptide derived from the CUB1 domain of ADAMTS13, spanning amino acid sequence 1206-1222 â&#x20AC;&#x201C; RGPGQADCAVAIGRPLG, that was identified on both HLA-DR and HLA-DQ of donors 2 and 3 (Figure 2). However, in the case of donor 9 this peptide was found on HLA-DQ but not on HLA-DR. Similarly, two peptides derived from the metalloprotease and CUB2 domains of ADAMTS13 were identified both on HLA-DR and HLA-DQ for donors 8 and 9, respectively.

Our data show that, while HLA-DQ is involved in ADAMTS13, HLA-DR is more likely to be presenting peptides from ADAMTS13, and that there is considerable overlap between the repertoires of ADAMTS13-derived peptides presented on HLA-DR and HLA-DQ.

Evaluation of ADAMTS13 peptide immunogenicity using bioinformatic approaches The above findings document which ADAMTS13 pep-

A

B

Figure 1. Relative presentation of total peptides and ADAMTS13derived peptides on HLA-DR and HLA-DQ. (A) Total number of peptides identified after elution from either HLA-DR or HLA-DQ. (B) The ADAMTS13-derived peptides found on either HLA-DR or HLADQ. Statistical differences in peptide presentation were determined using the non-parametric Mann-Whitney U-test. **P<0.01. ns: not significant.

Figure 2. ADAMTS13-derived peptides presented on HLA-DR and HLA-DQ. Peptides identified to be presented by DCs of 9 studied donors on HLA-DR and HLA-DQ. Longest amino acid sequence of overlapping peptides is shown in the third column. Amino acids that are predicted to be a part of the peptide-MHC-II binding core are shown in bold. Blue: peptides identified only in HLA-DR condition. Red: peptides identified exclusively in HLA-DQ condition. Green: peptides identified in both HLA-DR and HLA-DQ conditions.

1086

haematologica | 2018; 103(6)

Presentation of ADAMTS13 peptides on HLA-DR and HLA-DQ Table 2. HLA-DRB1 and HLA-DQA1/DQB1 peptide presentation and core-peptide binding affinity prediction. The ADAMTS13-derived peptides identified for each donor were analyzed for their binding affinities to the donor specific HLA-DRB1 (A) or HLA-DQA1/HLA-DQB1 alleles (B) using the NetMHCIIpan 3.1 software. The affinity value is presented in nM. For each peptide, a binding core was predicted. Binding cores with the highest affinity to the particular MHC-II are shown in yellow. Right panel represents the domain origin of each peptide.

A

haematologica | 2018; 103(6)

1087

J. Hrdinová et al.

B

tides are displayed on MHC class II following pulsing of dendritic cells with ADAMTS13. To complement our analysis of peptides eluted from MHC, we employed a bio-informatic approach to identify whether the ADAMTS13-derived peptides identified in this and a previous study (Table 3)19 might be immunogenic. Firstly, we used the EpiMatrix algorithm, which assesses the potential for individual peptides to bind to HLA-DR based on amino acid motifs preferred by nine “supertype” HLA-DR allele families. These HLA-DR motif families cover 95% of the human population worldwide.27 EpiMatrix assesses peptides for binding potential against a matrix of amino acid preferences for each HLA allele. This score is adjusted to a normalized (Z) score. The EpiMatrix Score of a peptide represents a sum of significant assessments for all of the nine-mers in a sequence, adjusted for length. An EpiMatrix Score of 40.66 was obtained for metallo-pro1088

tease-derived peptide SRRQLLSLLSAGRAR (residues 266280); this value indicates that this peptide has a significant potential to bind to multiple HLA-DR molecules (Table 3).

Also, the CUB2 domain-derived peptide FSEGFLKAQASLRGQYW (residues 1390-1406) was predicted to bind to multiple HLA-DR alleles (Table 3). A number of other CUB1-2 domain-derived peptides were found to be potentially immunogenic, as revealed by an EpiMatrix Score higher than 10 (Table 3).

A second tool, JanusMatrix,28 was used to evaluate whether the HLA-DR-presented ADAMTS13 peptides displayed homology to proteins from the human proteome.

JanusMatrix differentiates HLA-binding peptides that are more cross-conserved at the TCR face and thus more likely tolerated (or actively tolerogenic) from those less cross-conserved that are more likely to be immunogenic.29,30 This analysis revealed that six of the peptides (Table 3 and haematologica | 2018; 103(6)

Presentation of ADAMTS13 peptides on HLA-DR and HLA-DQ

Figure 3. Schematic representation of eluted peptide immunogenicity. EpiMatrix and JanusMatrix Homology Scores for each eluted peptide is depicted in the graph.

In the context of nine population-spanning â&#x20AC;&#x153;supertypeâ&#x20AC;? HLA-DR alleles, peptides with high (>10, orange) EpiMatrix Score are predicted to be promiscuous epitopes, presented by multiple HLA-DR alleles. Additionally, peptides with elevated to high (>3) JanusMatrix Homology Scores display high homology with human proteome at the TCR face and are predicted to be tolerated or actively tolerogenic (green squares). Peptides with high EpiMatrix Score (>10) in combination with low JanusMatrix Homology Score (<3) are predicted to be immunogenic (red squares). All other peptides, with an EpiMatrix Score <10, have limited or HLA-restricted immunogenic potential (black squares).

Figure 3) showed promiscuity in binding to supertype HLADR alleles in combination with low cross-conservation with human proteins. These peptides could potentially initiate an effector T-cell response. In contrast, four of the peptides (Table 3 and Figure 3) were found to be cross-conserved within the human protein repertoire and could thus be potentially tolerated or actively tolerogenic in patients with acquired TTP. The remaining 10 ADAMTS13 peptides did not register high EpiMatrix Scores for binding to HLADR. However, they did have significant assessments for individual HLA-DR alleles and therefore could be presented on a subset of MHC II molecules.

Discussion In this study, we explored the repertoire of HLA-DQpresented peptides on ADAMTS13 pulsed monocytederived dendritic cells. In parallel, we also assessed the HLA-DR-presented peptide repertoire. This approach allows for a direct comparison of the peptide repertoires presented on HLA-DR and HLA-DQ. For this, we used two different monoclonal antibodies, L243 and SPV-L3, that had been previously used for peptide presentation profiling on either HLA-DR or HLA-DQ.19,23,31-33 Our data revealed that the number of peptides presented on HLADQ is 2- to 3-fold lower when compared to the repertoire presented on HLA-DR. This is most likely explained by the higher expression of HLA-DR on dendritic cells when compared to HLA-DQ.25,34,35 Furthermore, this could also result from an overall lower binding affinity of peptides for HLA-DQ when compared to HLA-DR. Indeed, the NetMHCIIpan3.1 prediction tool suggests a higher binding affinity of ADAMTS13-derived peptides to HLA-DR when compared to HLA-DQ. In a recent study, we also haematologica | 2018; 103(6)

observed preferential presentation of blood coagulation FVIII-derived peptides on HLA-DR when compared to HLA-DQ.25 Interestingly, using higher-energy collisional dissociation fragmentation, the total number of peptides identified in HLA-DR eluates ranged from 1115 to 2247. This represents a 5- to 6-fold increase when compared to a previous study from our group that employed collisioninduced dissociation.19 We anticipate that the use of a highly sensitive mass spectrometer (Orbitrap Fusion Tribrid) has allowed for an increased number of peptides identified on HLA-DR. Surprisingly, despite the use of a highly sensitive mass-spectrometry strategy, the number of ADAMTS13-derived peptides remained relatively low in the case of HLA-DR and even lower for HLA-DQ (an average of 7 vs. 3 ADAMTS13-derived peptides, respectively). In the current study, we pulsed mo-DCs with 100 nM of recombinant ADAMTS13, whereas circulating levels of ADAMTS13 in healthy individuals ranges from 3.57 nM (740-1420 ng/mL).36 Previous work from our group has identified the mannose receptor as an endocytic receptor for ADAMTS13 by mo-DCs.37 More recently, we described the involvement of CD163 in the internalization of ADAMTS13 by macrophages in vitro.38 In the latter study, we observed that the uptake efficiency of CD163positive macrophages was approximately 10 times higher when compared to that of mo-DCs (that are devoid of CD163 expression). Based on this finding, it is tempting to speculate that the limited presentation of ADAMTS13derived peptides on mo-DCs is due to a relatively low efficiency to internalize ADAMTS13, thereby limiting the number of ADAMTS13-derived peptides that are presented. Whether the limited presentation of ADAMTS13 by dendritic cells explains the low incidence of TTP in the general population despite the high prevalence of the risk 1089

J. Hrdinovรก et al.

Table 3. Evaluation of ADAMTS13 peptide immunogenicity in silico.

Binding of peptides to nine common HLA-DR allele families was predicted using the EpiMatrix algorithm developed by EpiVax. An EpiMatrix Score that reflects peptide binding potential to each of the 9 HLA-DR alleles was assigned to each peptide. Peptides with EpiMatrix scores > 5 are predicted to have elevated immunogenic potential; peptides with scores > 10 are predicted to have significant immunogenic potential. In addition, JanusMatrix was used to predict potential cross-reactivity between the peptides and the human proteome, based on conservation of TCR-facing residues. A JanusMatrix Homology Score was assigned to each peptide. Higher scores indicate greater conservation with the human proteome. At the population level, given elevated to high EpiMatrix scores, peptides with low cross-conservation with human proteome (scores < 3) are considered likely to be immunogenic, while peptides with scores > 3 have elevated to high cross-conservation with human proteome and are thus potentially tolerated or tolerogenic.

alleles HLA-DRB1*11 and DQB1*03 in the general population represents an interesting question that needs further study.

The overall low number of ADAMTS13-derived peptides reported may also be due to post-translational modifications of HLA-DR/DQ-presented peptides that interfere with their identification. For instance, the presence of a glycan results in a shift in the net mass of the glycanbearing peptide, impairing its identification by mass spectrometry. Ten N-glycosylation sites, 8 O-fucosylation sites and 3 C-mannosylation sites have been described for plasma-derived ADAMTS13.39,40 As yet, the impact of posttranslational modifications such as N-linked glycosylation on MHC-II peptide presentation has not been extensively studied. To our surprise, we also identified a peptide derived from the TSP2 domain of ADAMTS13 (NYSCLDQAR) that has been reported to contain an Nlinked glycosylation site.39 A recent publication from our group showed that the asparagine present within this peptide contains a bi-antennary complex-glycan.40 As mentioned earlier, the presence of a glycan interferes with the identification of the glycan-bearing peptide by mass spectrometry. However, it was previously described that the third position in peptides with consensus sequence Asn1090

X-Ser/Thr affects glycosylation efficiency, Thr being associated with a higher degree of glycosylation compared to Ser.41 It is, therefore, likely that the identified peptide NYSCLDQAR is partially glycosylated, allowing for the identification of the non-glycosylated version in our study.

Whether the glycosylated counterpart of this peptide is also presented on HLA molecules remains to be determined. How the presence or absence of an N-linked glycan at this position modulates peptide/MHC-II/T-cell receptor (TCR) binding represents an interesting question.

On one hand, the presence of an N-linked glycan at this position could interfere with the binding of the peptides to HLA-DQ or with the recognition of the HLA-DQ/peptide complex by a complementary T-cell receptor (TCR).

Interestingly, absence of an O-linked glycan on type II collagen has been previously suggested to create a novel Tcell epitope that has been implicated in the development of autoimmune arthritis.42-44 Similarly, CD4+ T cells recognizing non-glycosylated forms of ADAMTS13-derived peptides that normally contain a glycan may contribute to the onset of autoimmune TTP. On the other hand, the presence of a glycan can also lead to the formation of a neo-epitope.45 In view of the large number of N-and Olinked glycans on ADAMTS13, it cannot be excluded that haematologica | 2018; 103(6)

Presentation of ADAMTS13 peptides on HLA-DR and HLA-DQ

glycan-containing peptides derived of ADAMTS13 are presented on MHC-II.40 Identification of the immunogenic determinants of ADAMTS13 represents a major challenge for the understanding of TTP pathogenesis. Our study shows that there is an overlap, but also differences, in the repertoire of ADAMTS13-derived peptides that are presented on HLADQ and HLA-DR. This is illustrated by the peptide with core sequence FINVAPHAR derived from the CUB2 domain that was identified exclusively in the case of HLA-DR. In contrast with the study of Sorvillo et al., only 2 out of the 3 HLA-DRB1*11 donors presented the FINVAPHAR peptide. In addition, this peptide was also identified in 2 non-HLA-DRB1*11 donors, confirming that the presentation of FINVAPHAR is not restricted to HLADRB1*11.19 Interestingly, several peptides were identified associated with HLA-DQ only. Especially, two peptides derived from the cysteine-rich domain were identified exclusively on HLA-DQ. Lastly, we identified sets of peptides that were presented on both HLA-DR and HLA-DQ, as exemplified by the CUB1-derived peptide with coresequence CAVAIGRPL. Whether the novel peptides identified in this study play a role in the onset of acquired TTP remains to be determined. A previous study of Verbij et al. showed that CUB2 domain-derived peptides FINVAPHAR and ASYILIRD are able to activate CD4+ T cells from an HLA-DRB1*11 and an HLA-DRB1*03-positive acquired TTP patient, respectively.20 A recent study by Gilardin et al. did not identify CD4+ T cells responding to peptide ADAMTS131239-1253 containing the FINVAPHAR core-sequence in HLA-DRB1*11positive patients.21 In their hands, another CUB2 domainderived peptide ADAMTS131239-1253 was identified as an immunodominant T-cell epitope for both DRB1*01 and DRB1*11-positive acquired TTP patients.21 Surprisingly, the ADAMTS131239-1253 peptide was not identified to be presented on HLA-DR or DQ in our current study, despite the fact that several HLA-DRB1*01 and HLA-DRB1*11 positive donors were included in our cohort. Two peptides containing the same core-sequence were, however,

References 1. Zheng X, Majerus EM, Sadler JE. ADAMTS13 and TTP. Curr Opin Hematol.

2002;9(5):389-394.

2. Kremer Hovinga JA, Coppo P, Lämmle B, Moake JL, Miyata T, Vanhoorelbeke K. Thrombotic thrombocytopenic purpura.

Nat Rev Dis Prim. 2017;3:17020.

3. George JN. Clinical practice. Thrombotic thrombocytopenic purpura. N Engl J Med. 2006;354(18):1927-1935.

4. Ferrari S, Palavra K, Gruber B, et al.

Persistence of circulating ADAMTS13-specific immune complexes in patients with acquired thrombotic thrombocytopenic purpura. Haematologica. 2014;99(4):779787.

5. Ferrari S, Mudde GC, Rieger M, Veyradier A, Kremer Hovinga JA, Scheiflinger F. IgG subclass distribution of anti-ADAMTS13 antibodies in patients with acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2009;7(10):1703-1710.

6. Pos W, Luken BM, Sorvillo N, Kremer Hovinga JA, Voorberg J. Humoral immune

haematologica | 2018; 103(6)

7.

8.

9.

10.

identified in a DRB1*0401/DRB1*1301 positive donor in a previous study from our group.19 In silico approaches (EpiMatrix) identified a majority of ADAMTS13 peptides (70%) with significant EpiMatrix scores in the CUB1/2 domains. Six of the eluted ADAMTS13 peptides with low cross-conservation with the human proteome (based on JanusMatrix analysis) were considered to be more likely to elicit an effector Tcell response. These predicted immunogenic peptide sequences included ADAMTS131322-1341 (FINVAPHAR) and ADAMTS131355-1373 (ASYILIRD); CD4+ T cells targeting these epitopes have indeed been found in patients with acquired TTP.20 In contrast, four of the peptides had TCR facing residues that were highly conserved with multiple peptides found in the human proteome and putatively restricted by the same HLA, and, therefore, could potentially be tolerated or actively tolerogenic in patients with acquired TTP. These peptides include peptide ADAMTS131238-1252, containing a minimal core sequence recognized by the ADAMTS13-specific DRB1*01:01restricted T-cell hybridomas as described by Gilardin et al.21 CD4+ T cells targeting different epitopes may develop in patients with acquired TTP. Phenotypic profiling of CD4+ T-cell responses in large numbers of patients with acquired TTP will be needed to define whether multiple immune-dominant and potential immunoregulatory CD4+ T-cell epitopes contribute to the etiology of acquired TTP. Acknowledgments The authors would like to thank Erik Mul and Mark Hoogeboezem for elutriation of the buffy coats. We are grateful to William Martin and Frances Terry (EpiVax) for their help with the in silico data analysis.

Funding This study was supported by grants from the Landsteiner Foundation of Blood Transfusion Research and Horizon 2020 Framework Programme for Research and Innovation of the European Union (Grant Agreement No. 675746).

response to ADAMTS13 in acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2011;9(7):1285-1291.

Long Zheng X, Wu HM, Shang D, et al.

Multiple domains of ADAMTS13 are targeted by autoantibodies against ADAMTS13 in patients with acquired idiopathic thrombotic thrombocytopenic purpura. Haematologica. 2010;95(9):15551562.

Luken BM, Kaijen PH, Turenhout EA, et al.

Multiple B-cell clones producing antibodies directed to the spacer and disintegrin/ thrombospondin type-1 repeat 1 (TSP1) of ADAMTS13 in a patient with acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2006;4(11):2355-2364.

Pos W, Crawley JT, Fijnheer R, Voorberg J, Lane DA, Luken BM. An autoantibody epitope comprising residues R660, Y661, and Y665 in the ADAMTS13 spacer domain identifies a binding site for the A2 domain of VWF. Blood. 2010;115(8):1640-1649.

Gerth J, Schleussner E, Kentouche K, Busch M, Seifert M, Wolf G. Pregnancy-associated thrombotic thrombocytopenic purpura.

Thromb Haemost. 2009;101(2):248-251.

11. Kosugi N, Tsurutani Y, Isonishi A, Hori Y, Matsumoto M, Fujimura Y. Influenza A infection triggers thrombotic thrombocytopenic purpura by producing the antiADAMTS13 IgG inhibitor. Intern Med. 2010;49(7):689-693.

12. Erdem F, Kiki I, Gundo du M, Kaya H. Thrombotic thrombocytopenic purpura in a patient with Brucella infection is highly responsive to combined plasma infusion and antimicrobial therapy. Med Princ Pract.

2007;16(4):324-326.

13. Schlienger K, Craighead N, Lee KP, Levine BL, June CH. Efficient priming of protein antigen-specific human CD4(+) T cells by monocyte-derived dendritic cells. Blood.

2000;96(10):3490-3498.

14. Trowsdale J, Knight JC. Major histocompatibility complex genomics and human disease. Annu Rev Genomics Hum Genet.

2013;14:301-323.

15. Coppo P, Busson M, Veyradier A, et al.

HLA-DRB1*11: a strong risk factor for acquired severe ADAMTS13 deficiencyrelated idiopathic thrombotic thrombocy-

1091

J. Hrdinová et al.

16.

17.

18.

19.

20.

21.

22.

23.

24.

1092

topenic purpura in Caucasians. J Thromb Haemost. 2010;8(4):856-859.

Scully M, Brown J, Patel R, Mcdonald V, Brown CJ, Machin S. Human leukocyte antigen association in idiopathic thrombotic thrombocytopenic purpura: Evidence for an immunogenetic link. J Thromb Haemost. 2010;8(2):257-262.

John M-LL, Hitzler W, Scharrer I. The role of human leukocyte antigens as predisposing and/or protective factors in patients with idiopathic thrombotic thrombocytopenic purpura. Ann Hematol. 2012;91(4): 507-510.

Mancini I, Ricaño-Ponce I, Pappalardo E, et al. Immunochip analysis identifies novel susceptibility loci in the human leukocyte antigen region for acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2016;14(12):2356-2367.

Sorvillo N, van Haren SD, Kaijen PHP, et al.

Preferential HLA-DRB1*11-dependent presentation of CUB2-derived peptides by ADAMTS13-pulsed dendritic cells. Blood.

2013;121(17):3502-3510.

Verbij FC, Turksma AW, de Heij F, et al.

CD4+ T cells from patients with acquired thrombotic thrombocytopenic purpura recognize CUB-2 domain derived peptides.

Blood. 2016;127(12):1606-1609.

Gilardin L, Delignat S, Peyron I, et al. The ADAMTS13(1239-1253) peptide is a dominant HLA-DR1-restricted CD4+ T-cell epitope. Haematologica. 2017;102(11):18331841.

Spits H, Keizer G, Borst J, Terhorst C, Hekman A, de Vries JE. Characterization of monoclonal antibodies against cell surface molecules associated with cytotoxic activity of natural and activated killer cells and cloned CTL lines. Hybridoma. 1983; 2(4):423-437. van Haren SD, Herczenik E, ten Brinke A, Mertens K, Voorberg J, Meijer AB. HLADR-presented Peptide Repertoires Derived From Human Monocyte-derived Dendritic Cells Pulsed With Blood Coagulation Factor VIII. Mol Cell Proteomics 2011; 10(6):M110.002246.

Rappsilber J, Ishihama Y, Mann M. Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics. Anal Chem. 2003;75(3):663-670.

25. Peyron I, Hartholt RB, Pedró-Cos L, et al.

Comparative profiling of HLA-DR and HLA-DQ associated factor VIII peptides presented by monocyte-derived dendritic cells. Haematologica. 2018;103(1):172-178.

26. Karosiene E, Rasmussen M, Blicher T, Lund O, Buus S, Nielsen M. NetMHCIIpan-3.0, a common pan-specific MHC class II prediction method including all three human MHC class II isotypes, HLA-DR, HLA-DP and HLA-DQ. Immunogenetics. 2013; 65(10):711-724.

27. De Groot AS, Martin W. Reducing risk, improving outcomes: bioengineering less immunogenic protein therapeutics. Clin Immunol. 2009;131(2):189-201.

28. Moise L, Gutierrez AH, Bailey-Kellogg C, et al. The two-faced T cell epitope. Hum Vaccin Immunother. 2013;9(7):1577-1586.

29. Liu R, Moise L, Tassone R, et al. H7N9 Tcell epitopes that mimic human sequences are less immunogenic and may induce Treg-mediated tolerance. Hum Vaccin Immunother. 2015;11(9):2241-2252.

30. Losikoff PT, Mishra S, Terry F, et al. HCV epitope, homologous to multiple human protein sequences, induces a regulatory T cell response in infected patients. J Hepatol.

2015;62(1):48-55.

31. de Verteuil D, Muratore-Schroeder TL, Granados DP, et al. Deletion of immunoproteasome subunits imprints on the transcriptome and has a broad impact on peptides presented by major histocompatibility complex I molecules. Mol Cell Proteomics. 2010;9(9):2034-2047.

32. Wahlström J, Dengjel J, Persson B, et al.

Identification of HLA-DR-bound peptides presented by human bronchoalveolar lavage cells in sarcoidosis. J Clin Invest.

2007;117(11):3576-3582.

33. Fissolo N, Haag S, de Graaf KL, et al.

Naturally presented peptides on major histocompatibility complex I and II molecules eluted from central nervous system of multiple sclerosis patients. Mol Cell Proteomics. 2009;8(9):2090-2101.

34. Brooks CF, Moore M. Differential MHC class II expression on human peripheral blood monocytes and dendritic cells.

Immunology. 1988;63(2):303-311.

35. Fernandez S, Wassmuth R, Knerr I, Frank C, Haas JP. Relative quantification of HLADRA1 and -DQA1 expression by real-time

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

reverse transcriptase-polymerase chain reaction (RT-PCR). Eur J Immunogenet.

2003;30(2):141-148.

Rieger M, Ferrari S, Kremer Hovinga JA, et al. Relation between ADAMTS13 activity and ADAMTS13 antigen levels in healthy donors and patients with thrombotic microangiopathies (TMA).

Thromb Haemost. 2006;95(2):212-220.

Sorvillo N, Pos W, van den Berg LM, et al.

The macrophage mannose receptor promotes uptake of ADAMTS13 by dendritic cells. Blood. 2012;119(16):3828-3835.

Verbij FC, Sorvillo N, Kaijen PHP, et al. The class I scavenger receptor CD163 promotes internalization of ADAMTS13 by macrophages. Blood Adv. 2017;1(5):293305.

Sorvillo N, Kaijen PHP, Matsumoto M, et al.

Identification of N-linked glycosylation and putative O-fucosylation, C-mannosylation sites in plasma derived ADAMTS13. J Thromb Haemost. 2014;12(5):670-679.

Verbij FC, Stokhuijzen E, Kaijen PHP, van Alphen F, Meijer AB, Voorberg J. Identification of glycans on plasma-derived ADAMTS13. Blood. 2016;128(21):e51-e58.

Nicolaes GAF, Villoutreix BO, Dahlbäck B. Partial glycosylation of Asn2181 in human factor V as a cause of molecular and functional heterogeneity. Modulation of glycosylation efficiency by mutagenesis of the consensus sequence for N-linked glycosylation. Biochemistry. 1999;38(41):1358413591.

Enouz S, Carrié L, Merkler D, Bevan MJ, Zehn D. Autoreactive T cells bypass negative selection and respond to self-antigen stimulation during infection. J Exp Med. 2012;209(10):1769-1779.

Dzhambazov B, Holmdahl M, Yamada H, et al. The major T cell epitope on type II collagen is glycosylated in normal cartilage but modified by arthritis in both rats and humans. Eur J Immunol. 2005;35(2):357-366.

Yamada H, Dzhambazov B, Bockermann R, Blom T, Holmdahl R. A transient posttranslationally modified form of cartilage type II collagen is ignored by self-reactive T cells. J Immunol. 2004;173(7):4729-4735.

Purcell AW, van Driel IR, Gleeson PA. Impact of glycans on T-cell tolerance to glycosylated self-antigens. Immunol Cell Biol.

2008;86(7):574-579.

haematologica | 2018; 103(6)