Cancer Therapy Volume 6 Issue A-2 by Cancer Therapy

CANCER THERAPY SPECIAL ISSUE

Presented at the Theilen Tribute Symposium, UC Davis 31st May - 1st June 2008.

Volume 6 Issue A-2 June 2008

CANCER THERAPY FREE ACCESS http://cancer-therapy.org

!!!!!!!!!!!!!!!!!!!!!!!! Editor

Teni Boulikas Ph. D., CEO Regulon Inc. 715 North Shoreline Blvd. Mountain View, California, 94043 USA Tel: 650-968-1129 Fax: 650-567-9082 E-mail: teni@regulon.org

Teni Boulikas Ph. D., CEO, Regulon AE. Gregoriou Afxentiou 7 Alimos, Athens, 17455 Greece Tel: +30-210-9853849 Fax: +30-210-9858453 E-mail: teni@regulon.org

!!!!!!!!!!!!!!!!!!!!!!!! Assistant to the Editor Maria Vougiouka B.Sc., Gregoriou Afxentiou 7 Alimos, Athens, 17455 Greece Tel: +30-210-9858454 Fax: +30-210-9858453 E-mail: maria@cancer-therapy.org

!!!!!!!!!!!!!!!!!!!!!!!! Editorial Board

Ablin, Richard J., Ph.D., Arizona Cancer Center, University of Arizona, USA Armand, Jean Pierre, M.D. Ph.D., European Organization for Research and Treatment of Cancer (EORTC), Belgium Aurelian, Laure, Ph.D., University of Maryland School of Medicine, USA Berdel, Wolfgang E, M.D., University Hospitals, Germany Bertino, Joseph R., M.D., Cancer Institute of New Jersey, USA Beyan Cengiz, M.D., Gulhane Military Medical Academy, Turkey Bottomley, Andrew, Ph.D., European Organization for Research and Treatment of Cancer Data Center (EORTC), Belgium Bouros, Demosthenes, M.D., University Hospital of Alexandroupolis. Greece Cabanillas, Fernando, M.D, The University of Texas M. D. Anderson Cancer Center, USA Castiglione, Monica, MHA, SIAK/IBCSG Coordinating Center, Switzerland Chou, Kuo-Chen, Ph.D., D.Sc., Pharmacia Upjohn, USA Chu, Kent-Man, M.D., University of Hong Kong Medical Center, Queen Mary Hospital, Hong Kong, China Chung, Leland W.K, Ph.D., Winship Cancer Institute, USA Coukos, George, M.D., Ph.D., Hospital of the University of Pennsylvania, USA Darzynkiewicz, Zbigniew, M.D., Ph.D., New York

Medical College, USA Der Channing, J. Ph.D, Lineberger Comprehensive Cancer Center, USA Devarajan, Prasad M.D., Cincinnati Children's Hospital, USA Dritschilo, Anatoly, M.D., Georgetown University Hospital, USA Duesberg, Peter H., Ph.D, University of California at Berkeley, USA El-Deiry, Wafik S. M.D., Ph.D., Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, USA Federico, Massimo, M.D. UniversitĂ di Modena e Reggio Emilia, Italy Fiebig, Heiner H, Albert-Ludwigs-UniversitĂ¤t, Germany Fine, Howard A., M.D., National Cancer Institute, USA Frustaci, Sergio, M.D., Centro di Riferimento Oncologico di Aviano, Italy Georgoulias, Vassilis, M.D., Ph.D., University General Hospital of Heraklion, Greece Giordano, Antonio, M.D., Ph.D., Sbarro Institute for Cancer Research and Molecular Medicine, Temple University, USA Greene, Frederick Leslie, M.D., Carolinas Medical Center, USA Gridelli, Cesare M.D., Azienda Ospedaliera, "S.G.Moscati", Italy Hengge, Ulrich, M.D., Heinrich-Heine-University Duesseldorf, Germany Huber, Christian M.D., Johannes-Gutenberg-

University, Germany Hunt, Kelly, M.D., The University of Texas M. D. Anderson Cancer Center, USA Kamen, Barton A., M.D. Ph.D, Cancer Institute of New Jersey, USA Kaptan, Kürsat, M.D., Gülhane Military Medicine Academy, Turkey Kazuma, Ohyashiki, M.D., Ph.D., Tokyo Medical University, Japan Kinsella, Timothy J. M.D., The research Institute of University Hospitals in Cleveland, USA Kmiec, Eric B, Ph.D., University of Delaware, USA Kosmidis Paris, M.D., "Hygeia" Hospital, Athens, Greece Koukourakis Michael, M.D., Democritus University of Thrace, Greece Kroemer, Guido, M.D. Ph.D., Institut Gustave Roussy, France Kurzrock, Razelle, M.D., F.A.C.P., M. D. Anderson Cancer Center, USA Leung, Thomas Wai-Tong M.D., Chinese University of Hong Kong, China Levin, Mark M.D., Sister Regina Lynch Regional Cancer Center, Holy Name Hospital, USA Lichtor, Terry M.D., Ph.D., Rush Medical College, USA Liebermann, Dan A., Ph.D., Temple Univ. School of Medicine, USA Lipps, Hans J, Ph.D., Universität Witten/Herdecke, Germany Lokeshwar, Balakrishna L., Ph.D., University of Miami School of Medicine, USA Mackiewicz, Andrzej, M.D., Ph.D., University School of Medical Sciences (USOMS) at Great Poland Cancer Center, Poland Marin, Jose J. G., Ph.D., University of Salamanca, Spain McMasters, Kelly M., M.D., Ph.D., University of Louisville, J. Graham Brown Cancer Center, USA Morishita, Ryuichi, M.D., Ph.D., Osaka University, Japan Mukhtar, Hasan Ph.D., University of Wisconsin, USA Ng, Eddie YK, Ph.D., Nanyang Technological University, Singapore Norris, James Scott, Ph.D., Medical University of South Carolina, USA Palu, Giorgio, M.D., University of Padova, Medical School, Italy Park, Jae-Gahb, M.D., Ph.D., Seoul National University College of Medicine, Korea Perez-Soler, Roman M.D., The Albert Einstein Cancer Center, USA Peters, Godefridus J., Ph.D., VU University Medical Center (VUMC), The Netherlands Poon, Ronnie Tung-Ping, M.D., Queen Mary Hospital, Hong Kong, China Possinger, Kurt-Werner, M.D., Humboldt University, Germany Rainov G Nikolai M.D., D.Sc., The University of Liverpool. UK Randall, E Harris, M.D., Ph.D., The Ohio State University, USA Ravaioli Alberto, M.D. Ospedale Infermi, Italy

Remick, Scot, C. M.D., University Hospitals of Cleveland, USA Rhim, Johng S M.D., Uniformed Services University of Health Sciences, USA Schadendorf, Dirk, M.D., Universitäts-Hautklinik Mannheim, Germany Schmitt, Manfred, Ph.D., Universität München, Klinikum rechts der Isar, Germany Schuller, Hildegard M., D.V.M., Ph.D., University of Tennessee, USA Slaga, Thomas J., Ph.D., AMC Cancer Research Center (UICC International Directory of Cancer Institutes and Organisations), USA Soloway, Mark S., M.D., University of Miami School of Medicine, USA Srivastava, Sudhir, Ph.D., MPH, MS, Division of Cancer Prevention, National Cancer Institute, USA Stefanadis, Christodoulos, M.D., University of Athens, Medical School, Greece, Stein, Gary S Ph.D., University Of Massachusetts, USA Tirelli, Umberto, National Cancer Institute, Italy Todo, Tomoki, M.D., Ph.D., The University of Tokyo, Japan van der Burg, Sjoerd H, Leiden University Medical Center, The Netherlands Wadhwa Renu, Ph. D., Nat. Inst. of Advan. Indust. Sci. and Technol. (AIST), Japan Waldman, Scott A. M.D., Ph.D., USA Walker, Todd Ph.D., Charles Sturt University, Australia Watson, Dennis K. Ph.D., Medical University of South Carolina, Hollings Cancer Center, USA Waxman, David J., Ph.D., Boston University, USA Weinstein, Bernard I., M.D., D.Sci (Hon.), Columbia University, USA Werner, Jochen Alfred M.D., Philipps-University of Marburg, Germany Wieand, Harry Samuel Ph.D., National Surgical Adjuvant Breast and Bowel Project (NSABP), USA Yamada, Akira Ph.D., Kurume University Research Center for Innovative, Japan Yu, Dihua M.D., Ph.D., The Univ. Texas M. D. Anderson Cancer Center, USA Zagon, Ian, Ph.D., The Pennsylvania State University, USA

!!!!!!!!!!!!!!!!!!!!!!!! Associate Board Members Chen, Jiguo, Ph.D, The University of Texas Health Science Center at San Antonio, USA Chen, Zhong, M.D, Ph.D, National Institute of Deafness and other Communication Disorders, National Institutes of Health, USA Dietrich Pierre Yves, Hopitaux Universitaires de GenFve Switzerland Jeschke Marc G, M.D., Ph.D. Universität Erlangen-Nürnberg. Germany Limacher Jean-Marc, MD Hôpitaux Universitaires de Strasbourg, France Los Marek J, M.D., Ph.D. University of Manitoba, USA Mazda Osam, M.D., Ph.D. Kyoto Prefectural University of Medicine, Japan Merlin Jean-Louis, Ph.D Centre Alexis Vautrin, National Cancer Institute University Henri Poincaré France Okada Takashi, M.D., Ph.D. Jichi Medical School Japan Pisa Pavel, M.D, Ph.D. Karolinska Hospital, Sweden Squiban Patrick, MD Transgene SA France Taupin, Philippe, Ph.D., National University of Singapore, Singapore Tsuchida Masanori, M.D, Ph.D Niigata University Graduate School of Medical and Dental Sciences Japan Ulutin, Cuneyt, M.D., Gulhane Military Medicine Academy, Turkey Xu Ruian, Ph.D., The University of Hong Kong, Hong Kong

!!!!!!!!!!!!!!!!!!!!!!!! For submission of manuscripts and inquiries: Editorial Office Teni Boulikas, Ph.D./ Maria Vougiouka, B.Sc. Gregoriou Afxentiou 7 Alimos, Athens 17455 Greece Tel: +30-210-985-8454 Fax: +30-210-985-8453 and electronically to maria@cancer-therapy.org

Instructions to authors: Cancer Therapy FREE ACCESS http://cancer-therapy.org

Scope This journal, bridging various fields is one of the most rapid with free access at www.cancertherapy.org. The scope of Cancer Therapy is to rapidly publish original and in-depth review articles on cancer embracing all fields from molecular mechanisms to results on clinical trials. Articles (both invited and submitted) review or report novel findings of importance to a general audience in cancer therapy, molecular medicine, gene discovery, and molecular biology with emphasis to molecular mechanisms and clinical applications. The journal will accept papers on all aspects of cancer, at the clinical, preclinical or cell culture stage on chemotherapy and new experimental drugs, gene discovery, cancer immunotherapy, DNA vaccines, use of DNA regulatory elements in gene transfer, cell therapy and drug discovery related to cancer therapy. The authors are encouraged to elaborate on the molecular mechanisms that govern a cancer therapy approach. To make the publication attractive authors are encouraged to include color figures.

Type of articles Both review articles and original research articles will be considered. Original research articles should contain a generous introduction in addition to experimental data. The articles contain information important to a general audience as the volume is addressed to researches outside the field. There is no limit on the length of the articles provided that the subject is interesting to a general audience and covers exhaustively a field. The typical length of each manuscript is 12-60 manuscript pages (approximately 4-20 printed pages) plus Figures and Tables. Free of Charge publication, Complimentary reprints & Subscriptions There are no charges for color figures or page numbers. Corresponding authors get a one-year free subscription (hard copy) plus 25 reprints free of charge. The free subscription can be renewed for additional years by having one paper per year accepted for publication. Sections of the manuscript Each manuscript should have a Title, Authors, Affiliation, Corresponding Author (with Tel, Fax, and E-mail), Summary, and Introduction; review articles are subdivided into headings I, II, III, etc. (starting with I. Introduction) and subdivided into A, B, C, etc. You can further subdivide into 1, 2, 3, etc. Research articles are divided into Summary; I. Introduction; II. Results; III Discussion; Acknowledgments IV. Materials and Methods and References. Please include in your text citations the name of authors and year in parenthesis; for three or more authors use: (name of first author et al, with year); for two authors please use both names. Please delete hidden text for references. In the reference list, please, type references with year and Journal in boldface and provide full title of the article such as: Buschle M, Schmidt W, Berger M, Schaffner G, Kurzbauer R, Killisch I, Tiedemann J-K, Trska B, Kirlappos H, Mechtler K, Schilcher F, Gabler C, and Birnstiel ML (1998) Chemically defined, cell-

free cancer vaccines: use of tumor antigen-derived peptides or polyepitope proteins for vaccination. Gene Ther Mol Biol 1, 309-321. Please use Microsoft Word, font “Times” (Mac users) or “Times New Roman” (PC users) and insert Greek or other characters using the “Insert/Symbol” function in the Microsoft Word rather than simple conversion to font “Symbol”. Please boldface Figure 1, 2, 3 etc. as well as Table 1, 2, etc. throughout the text. Please provide the highest quality of prints of your Figures; whenever possible, please provide in addition an electronic version of your figures (optional). Corresponding authors are kindly requested to provide a color (or black/white) head photo of themselves (preferably 4x5 cm or any size), as we shall include these in the publication.

Submission and reviewing Peer reviewing is by members of the Editorial Board and external referees. Please suggest 2-3 reviewers providing their electronic addresses, mailing addresses and telephone/fax numbers. Authors are being sent page proofs. Cancer Therapy (Volume 1, 2003) is published on high quality paper with excellent reproduction of color figures and electronically.

Reviewing is completed within 5-15 days from receiving the manuscript. Articles accepted without revisions (i.e., review articles) will be published online (www.cancertherapy.org) in approximately 1 month following submission. Please submit an electronic version of full text and figures preferably in jpeg format. The electronic version of the figures will be used for the rapid reviewing process. High quality prints or photograph of the figures and the original with one copy should be sent via express mail to the Editorial Office. Citation in MedLine Articles accepted for publication by GTMB or Cancer Therapy can be included in MedLine (PubMed) as full articles upon the request of authors provided that the authors have completed their published work under a government grant by NIH (or EU/Japan government grant). If this is you case, please consult the NIH Manuscript Submission System http://www.nihms.nih.gov/.

Editorial Office Teni Boulikas, Ph.D./ Maria Vougiouka, B.Sc. Gregoriou Afxentiou 7 Alimos, Athens 17455 Greece Tel: +30-210-985-8454 Fax: +30-210-985-8453 and electronically to maria@cancer-therapy.org The free electronic access to articles published in "Cancer Therapy" to a big general audience, the attractive journal title, the speed of the reviewing process, the no-charges for page numbers or color figure reproduction, the 25 complimentary reprints, the rapid electronic publication, the embracing of many fields in cancer, the anticipated high quality in depth reviews and first rate research articles and most important, the eminent members of the Editorial Board being assembled are prognostic factors of a big success for the newly established journal.

Gene Therapy and Molecular Biology (GTMB) is covered in the following Thomson Scientific services:

" Science Citation Index Expanded (also known as SciSearch# ) " Biotechnology Citation Index# " Journals Citation Reports/Science Edition

Table of Contents

Theilen Tribute Symposium, UC Davis 31st May - 1st June 2008

Pages *Vol 6 A

137-148

Type of Article

Article Title

Research Article

DNA methylation in cancer: techniques and preliminary evidence of hypermethylation in canine lymphoma

Review Article

Veterinary radiation oncology: technology, imaging, intervention and future applications

Review Article

Canine cancer genetics: transitional cell carcinoma in Scottish Terriers

Steven E. Crow

Review Article

Chemoimmunotherapy for canine lymphoma: tumor vaccines and monoclonal antibodies

Steven E. Crow

Research Article

Veterinary pathologists achieve 80% agreement in application of WHO diagnoses to canine lymphoma

Victor E. O Valli

Review Article

Shared pathogenesis of human and canine tumors - an inextricable link between cancer and evolution

Jaime F. Modiano, Matthew Breen

Review Article

Immunological concepts applied to pathologic diagnosis of proliferative diseases of the immune system

Peter F. Moore

Review Article

Search for oncogenic retroviruses in wild mice and man: Historical reflections

Murray Gardner

1-11 167-176 12-21 177-180

Authors *corresponding author is in boldface Jeffrey N. Bryan, Kristen H. Taylor, Carolyn J. Henry, Kimberly A. Selting, Farah Rahmatpanah, Michael R. Lewis, Charles W. Caldwell Ira K. Gordon, Michael S. Kent

22-25 181-186 26-31 221-226 32-37 239-246 38-45 263-270 46-53 285-302 54-81

Cancer Therapy Vol 6-A.2, page 1 Cancer Therapy Vol 6, 137-148, 2008

DNA methylation in cancer: techniques and preliminary evidence of hypermethylation in canine lymphoma Research Article

Jeffrey N. Bryan1, ,#, Kristen H. Taylor2, Carolyn J. Henry1,3, Kimberly A. Selting1, Farah Rahmatpanah2, Michael R. Lewis1,4, Charles W. Caldwell2,5 1

Department of Veterinary Medicine and Surgery, Department of Pathology and Anatomical Sciences, 3 Department of Internal Medicine, Division of Hematology/Oncology, University of Missouri-Columbia 4 Research Service, Harry S. Truman Memorial Veteransâ&#x20AC;&#x2122; Hospital, Columbia, MO 5 Ellis Fischel Cancer Center, University of Missouri-Columbia. 2

__________________________________________________________________________________ *Correspondence: Jeffrey N. Bryan, Washington State University, PO Box 646610, Pullman, WA 99164-6610, USA; Phone: (509) 335-0711; Fax: (509) 335-0880; E-mail: bryanjn@vetmed.wsu.edu # Dr. Bryanâ&#x20AC;&#x2122;s current address is: Department of Veterinary Clinical Sciences, Washington State University, PO Box 646610, Pullman, WA 99164-6610. Key words: DLC1, Dogs, Lymphoma, DNA hypermethylation, Epigenetic Abbreviations: basic local alignment search tool, (BLAST); chronic lymphocytic leukemia, (CLL); combined bisulfite restriction analysis, (COBRA); Differential methylation hybridization, (DMH); diffuse large B-cell lymphomas, (DLBCL); DNA methyltransferases, (DNMT); histone deacetylases, (HDAC); insulin-like growth factor-1, (IGF-1); interleukin-6, (IL-6); Methylation specific PCR, (MSP); immunoprecipitation, (IP); O6-methylguanine-DNA methyltransferase, (MGMT); polymerase chain reaction, (PCR); Quantitative MSP, (Q-MSP); restriction landmark genomic scanning technique, (RLGS); Rho-GTPase Activating Protein, (RhoGAP); sterile alpha motif homology 2, (SAM2); steroidogenic acute regulatory protein, (START); vascular endothelial growth factor, (VEGF) Received: 6 February 2008; electronically published: June 2008

This work was funded in part by the National Library of Medicine Biomedical and Health Informatics Research training grant LM07089. Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary Epigetic changes in cancer, including DNA methylation and histone modification, are now recognized to be integrally involved in generation and maintenance of the neoplastic phenotype. The addition of a methyl group to cytosines occurs normally in genomic, low-density CpG dinucleotides, and is normally blocked in high density accumulations of CpG dinucleotides in control regions of genes, termed CpG islands. Global hypomethylation of cytosines and local hypermethylation of CpG islands are hallmarks of the neoplastic epigenome. Hypomethylation can result in overexpression of growth factors or oncogenes, alterations in DNA repair enzymes, and loss of genomic stability. Hypermethylation of CpG islands can silence tumor suppressor genes, promoting cell cycling. In concert, alterations in methylation may be the primary players in pluripotent cell expansion that develops into cancer. In this paper we review the current molecular procedures to elucidate the methylation status of DNA on a genomewide basis and for specific gene determination for veterinary cancer researchers. We also report the results of a pilot study in dogs of the orthologue of a recognized human tumor-suppressor gene, DLC1. The functional canine sequence orthologous to human DLC1 was identified using web-based tools. Methylation specific PCR (MSP) and combined bisulfite restriction analysis (COBRA) were performed. MSP demonstrated methylated DNA to be present in six of 13 canine NHL samples and two of three canine chronic lymphocytic leukemia (CLL) samples. COBRA identified methylation in nine of 13 NHL and two of three CLL samples. This study provides compelling pilot data that hypermethylation occurs in canine NHL. This change in neoplastic cells warrants further detailed investigation as a marker for diagnosis and classification of disease or as a therapeutic target.

Bryan et al: DNA Methylation review using S-adenosyl methionine as the methyl donor. Maintenance methylation, occurring shortly after DNA replication, serves the purpose of conserving patterns of methylation that silence transposons and maintain monoallellic expression of genes imprinted during development (Rollins et al, 2006). De novo methylation allows the dynamic modulation of gene expression through methylation of CpG dinucleotides in control regions of genes, including the promoter, first exon, and first intron. Waves of methylation and demethylation are critical events in normal embryogenesis and in utero development (Siegfried et al, 1999; Herman and Baylin, 2003). Methylation patterns are mitotically heritable for generations (Jones and Baylin, 2007). Across the genome, CpG dinucleotides are statistically relatively scarce. This is likely because methylcytosine that undergoes spontaneous deamination becomes thymine, and the change is inefficiently corrected. As a result, most genomic CpG dinucleotides are low density and are methylated (Rollins et al, 2006). Regions with higher density of CpG dinucleotides are referred to as CpG islands. These are functional units defined by a mathematical construct rather than well-defined physical boundaries. CpG islands are regions of DNA spanning at least 500 bases with a G + C content > 55% and an observed/expected ratio of CpG > 0.6 (Takai and Jones, 2002). In most cases, CpG islands are associated with the promoter and first exon regions of genes, and are largely unmethylated (Takai and Jones, 2002; Rollins et al, 2006). When these regions are predominantly methylated, as in imprinted genes or parasitic DNA sequences, histone modifications occur, chromatin compaction occurs, transcription factors no longer bind, and transcription is silenced (Figure 2) (Herman and Baylin, 2003).

I. Introduction A. Epigenetic changes modifying gene expression In 1948, Rollin Hotchkiss reported the presence of a fifth base in a DNA chromatogram (Hotchkiss, 1949). Subsequently identified as 5-methylcytosine, a modification of the standard base, its role in eukaryotic cells was largely unexplored for decades. In 1993, Holliday postulated that DNA methylation contributes to the neoplastic phenotype (Holliday, 1993). Subsequent research is identifying the mechanisms of Dr. Hotchkissâ&#x20AC;&#x2122; contribution at an increasing rate. Epigenetic changes are covalent modifications of DNA bases or histones that alter the tertiary structure of chromatin, the binding of transcription factors, and, ultimately, the expression patterns of genes within a cell. In human cell lines and primary tumors, abnormal DNA methylation has now been demonstrated to contribute to cancer of the breast, colon, stomach, kidney, prostate, skin, hemolymphatic organs, and other tissues (Cui et al, 2003; Kim et al, 2004; Leu et al, 2004; Liu et al, 2004; Murai et al, 2005; Schulz and Hatina, 2006; Shamay et al, 2006). The presence of aberrant methylation patterns in neoplastic tissues offers the promise of novel markers for diagnosis and prognosis, untapped therapeutic targets, and may hold the key to stem cell carcinogenesis. In normal cells, methylcytosine mediates transcriptional silencing of transposons, imprinted silencing of somatic genes, inactivation of one copy of the X chromosome of females, and the discrimination of self from invading DNA of pathogens (Singal and Ginder, 1999; Rollins et al, 2006) Most commonly, cytosine in a C followed by G (CpG) dinucleotide is methylated by enzymes called DNA methyltransferases (DNMT) (Figure 1). These enzymes are responsible for either maintenance (DNMT1) or de novo methylation (DNMT3a, DNMT3b), Uracil

Thymidine

N O

CH3

N O

N Deoxyribose

Deamination

NH2

DNMT

N O

N Deoxyribose

S-Adenosyl Methionine

CH3

Homocysteine

N Deoxyribose

Cytosine

5-Methylcytosine

Figure 1. Cytosine becomes 5-methylcytosine by an electrophilic methyl substitution at the 5 position catalyzed by DNA methyltransferases. A deamination event results in transformation of cytosine to uracil and methylcytosine to thymine as illustrated.

Cancer Therapy Vol 6-A.2, page 3 Often this change is associated with modifications of histone tails as well. The order of events in silencing has not been entirely elucidated. Open, transcriptional chromatin typically contains histone H3 with acetylated tails, phosphorylation of Serine 10, and a methyl group added to lysine-4, as well as acetylation of lysine 8 of histone H4 (Herman and Baylin, 2003). Condensed, transcriptionally silenced chromatin, however, generally lacks acetylation and methylation of lysine-4 on histone H3, and demonstrates methylation of histone H3 lysine-9. The presence of DNA hypermethylation can attract methylcytosine binding proteins, which recruit histone deacetylases (HDAC) and histone methyltransferases (Herman and Baylin, 2003). Some studies have demonstrated that histone modification events may occur in the absence of promoter methylation as well (Bachman et al, 2003; Leu et al, 2004). Modifications of histone tails have been dubbed the â&#x20AC;&#x153;histone codeâ&#x20AC;? (Mellor, 2006).

to suggest that epigenetic changes, particularly global hypomethylation, are the earliest events in carcinogenesis. Recent literature advances the hypothesis that carcinogens injure many cells rather than mutate a few (Jaffe, 2003; Feinberg et al, 2006). A novel paradigm has been proposed that suggests that epigenetic disruption of progenitor cells precedes gatekeeper mutations, followed by a period of epigenetic and genetic plasticity that allows promotion and progression in the more classic sense (Feinberg et al, 2006). Epigenetic changes are clearly more than coincidental occurrences in cancer and represent targets of prevention and therapy. Methylation is a post-replicative event. As such, the standard polymerase chain reaction (PCR) adds cytosine, not methylcytosine to the growing DNA strand. To preserve information about methylcytosine bases, several techniques have been employed.

1. Restriction scanning

B. Laboratory techniques to evaluate methylation

landmark

genomic

Large-scale information about methylation of CpG dinucleotides can be determined by the use of methylation-sensitive restriction enzyme digestion with a restriction landmark genomic scanning technique (RLGS) (Park et al, 2005; Ando and Hayashizaki, 2006). In this procedure, genomic DNA is digested, but methylated targets will not be cut. Analysis of the pattern of fragments reveals the overall degree of methylation present at CpGcontaining restriction sites in the sample.

In human cancers, results of recent studies have begun to elucidate the contribution of DNA methylation to the neoplastic phenotype. Early studies identified a global loss of methylation in cancer cells. This was shown to lead to hypomethylation of Ras oncogenes, among others, that could result in overexpression of growth factors participating in neoplastic transformation (Feinberg and Vogelstein, 1983). Global genomic hypomethylation appears to contribute to genomic instability and the acceleration of the accumulation of genetic abnormalities that characterize cancer cells (Herman and Baylin, 2003). Concurrently, cancer cells exhibit hypermethylation of CpG islands in promoter regions, leading to the silencing of tumor suppressor genes (Herman and Baylin, 2003). The DNA methylation changes typical of cancer cells are illustrated in Figure 2. Accumulating evidence has begun

2. Bisulfite conversion To understand the precise pattern of cytosine methylation, DNA may be treated with bisulfite to convert unmethylated cytosine to uracil. This, in turn, becomes thymine in subsequent PCR amplification of the bisulfite converted DNA. When compared to the original sequence,

Normal gene

Hypomethylated , normally imprinted gene

Hypermethylated gene

Figure 2. Cytosine methylation patterns of DNA in normal cells, cells expressing normally imprinted alleles due to hypomethylation, and cells with silencing of tumor suppressor genes due to hypermethylation. Circles represent CpG dinucleotides along the DNA (double lines) with black representing the methylated condition and white the unmethylated condition. Boxes represent the promoter/first exon regions of the gene. The region containing more densely spaced CpG dinucleotides represents a CpG island. The grey arrows represent the activity of transcription.

Bryan et al: DNA Methylation review the presence of a thymine, instead of a cytosine, prior to a guanosine indicates the presence of unmethylated cytosine in the original DNA strand. Preservation of a cytosine is evidence that methylcytosine was present originally (Clark et al, 1994).

dinucleotides in each forward and reverse primer. Two sets of primers are designed: One, for methylated sequences, that retains the CpG complementarity; a second, for unmethylated sequences, that is complementary to a TpG sequence. The presence of a band using the methylated primers is evidence for methylation in the original sequence (Figure 3) (Cottrell and Laird, 2003). This technique is sensitive, but does not convey information about the density of methylation.

3. COBRA This assay uses bisulfite converted DNA as the template for PCR. A region of interest of a gene, usually in the promoter region, is amplified with primers designed to avoid CpG dinucleotides. Thus, the DNA will be amplified whether the CpG island is hypermethylated or not. Between these primers will have been identified restriction enzyme sites that contain CpG dinucleotides. When treated with methylation sensitive enzymes such as Hpy99I, HpyCH4IV, TaqaI, and BstUI, sequences that originally contained methylcytosine will be cut, whereas those with unmethylated cytosine will have been converted to thymidine and not be recognized (Figure 3). Some idea of the density of methylation can be gleaned by including multiple cut-sites within the sequence to be studied. Greater numbers of fragments positively correlate with a greater degree of methylation (Fraga and Esteller, 2002).

5. Quantitative MSP (Q-MSP) Similar to MSP, this is performed using a thermocycler with real-time detection capability. Using the TaqMan technology, a probe containing a fluorophore is designed complementary to the PCR amplicon. The relative intensity of fluorescence between the methylated and unmethylated primers allows quantification of the degree of methylation in the sample (Cottrell and Laird, 2003).

6. Bisulfite sequencing by cloning Fragments of bisulfite treated DNA amplified for COBRA or MSP analysis are cloned by phage into bacteria. The resulting colonies are harvested and submitted for sequencing with universal primers. The resulting sequence is compared to the original sequence to determine the degree of methylation in that region of the gene (Clark et al, 1994).

4. MSP Using bisulfite converted DNA, PCR is performed with primers designed to include at least two CpG

Methylation Specific PCR Normal Patient 1

Neoplastic U

2 3

Patient

Combined Bisulfite Restriction Analysis Normal

Neoplastic

Figure 3. The top panel represents the gel results for MSP. Unless imprinted, normal tissue will rarely yield a methylated band (column M) and will display an unmethylated band (column U). Neoplastic cells that contain a hypermethylated gene of interest will yield a methylated band, and possibly unmethylated bands of varying size as well. The combination may be due to heterogeneity of the tumor, heterozygosity within tumor cells, or a mixed tumor/normal cell population. This can occur with COBRA as well (lower panel). Normal samples will not demonstrate digestion with the restriction enzyme, whereas the hypermethylated gene will be digested, resulting in two smaller bands. An unmethylated band may be seen for the aforementioned reasons.

Cancer Therapy Vol 6-A.2, page 5

7. Differential methylation hybridization microarray (DMH)

in many other human and canine cancers. As a potential diagnostic screening test, the loss of imprinting of the IGF2 gene was examined in colon cancer. (Cui et al, 2003). The investigators identified an odds ratio for loss of imprinting in peripheral blood lymphocytes of 21.7 for patients with colorectal carcinoma. Such screening techniques could lead to much less invasive, and potentially more cancer-specific, tests for markers of risk for disease, of early disease, or the presence of residual disease. Promoter hypermethylation has been demonstrated to silence many tumor suppressor genes in human cancers and cancer cell lines (Herman and Baylin, 2003). The gene RASSF1A is frequently silenced in head and neck carcinoma (Dong et al, 2003). In one study the silencing of the gene was inversely associated with the presence of human papillomavirus infection, a common predisposing infection for these carcinomas (Dong et al, 2003). Such differentiation among etiologies could inform therapeutic decisions for clinical patients. In vitro studies have demonstrated that the gene could be demethylated and reexpressed by treatment with 5-aza-2â&#x20AC;&#x2122;-deoxycytidine, making methylated RASSF1A a potential therapy target (Dong et al, 2003). The presence of methylated RASSF1A has also been evaluated as a diagnostic test in urine samples of humans with urinary transitional cell carcinoma (Chan et al, 2003). The test was more sensitive than cytology for detecting low-grade and early cases. This demonstrates the potential utility of DNA methylation as a method of specific primary diagnostic screening. This would also yield a marker to be followed throughout therapy to determine efficacy and detect early relapse. Occasionally, aberrant promoter methylation may result in a beneficial change for cells. The gene coding O6methylguanine-DNA methyltransferase (MGMT) is an important DNA repair enzyme that can lead to chemotherapy resistance by removing potentially lethal DNA adducts that form after chemotherapy exposure. In human diffuse large B-cell lymphomas (DLBCL) lack of expression of MGMT due to DNA methylation has been shown to be positively prognostic for chemotherapy response and outcome (Al-Kuraya et al, 2006). The same has been shown to be true in glioblastoma, in which methylated MGMT can be used to predict a positive response to temozolamide chemotherapy (Hegi et al, 2005; Donson et al, 2007). Discovery of more such markers in cancer could be used to categorize patients into useful therapeutic subgroups. Genomic hypomethylation has been identified in canine lymphoproliferative disease (Pelham et al, 2003). Using RLGS, investigators demonstrated differences in methylation between normal and neoplastic lymphoid tissue (Pelham et al, 2003). Further, they demonstrated that the normal pattern of methylation was preserved in peripheral lymphocytes of dogs with lymphoma (Pelham et al, 2003). This was the first paper to demonstrate the similarity of methylation change between human and canine neoplastic disease. This underscores the potential for companion animal cancer to be instrumental as a clinical model for diagnosis and therapy of similar

Genomic DNA from a tumor is digested using restriction enzymes to 200 base-pair fragments, leaving CpG islands relatively intact. Methylation-sensitive restriction enzymes are then used to digest the fragments that are CpG rich. Cut sites that are methylated are ignored by the restriction enzyme, leaving those sequences intact. The resulting fragments, both methylated and unmethylated are tagged with a fluorophore different from that used to tag normal genomic DNA from a similar tissue. The two tagged samples are co-hybridized on a microarray chip containing sequences from regions of interest that span the cut sites of the restriction enzyme used. Those experimental sequences that remain intact will hybridize to the chip, and those that have been cut will not. The resulting fluorophore intensity correlates to the degree of methylation in the tumor sample relative to the normal tissue. This is a discovery tool for candidate methylated sequences. Validation is required by other described techniques. No canine chip exists (Yan et al, 2000).

8. CpG island and promoter tiling arrays Microarray platforms have been developed that give coverage at a resolution of 50 to 100bp covering all CpG islands or promoter regions of genes. Chromatin immunoprecipitation (IP) is performed using an antimethylcytosine antibody to concentrate fragmented genomic DNA that is rich in methylcytosine. This is then fluorescent labeled and competitively hybridized on the chip with the input, complete genomic DNA labeled with a different color. Islands can be viewed as a whole, looking for increased intensity of IP signal to genomic to suggest hypermethylation (Bock and Lengauer, 2008).

9. High-throughput bisulfite sequencing Current research is utilizing high-throughput sequencing methods to bisulfite sequence PCR product to evaluate the methylome. A limitation of this technique is the conversion of much of the genome to a nearly three base, rather than a four base, system. This can make alignment with the complete sequenced genome a bioinformatics challenge, as statistical uniqueness declines significantly (Taylor et al, 2007a).

C. Hypermethylation in cancer Loss of imprinting or normal methylation of a gene promoter can lead to overexpression of that geneâ&#x20AC;&#x2122;s product, contributing to the neoplastic phenotype. An example of this is the overexpression of the NOTCH ligand, JAG2, in malignant human plasma cells. Houde and others demonstrated that hypomethylation of the promoter region of this gene in malignant cells, compared to normal cells, resulted in higher levels of expression of the JAG2 protein (Houde et al, 2004). This induced the secretion of interleukin-6 (IL-6), vascular endothelial growth factor (VEGF), and insulin-like growth factor-1 (IGF-1). Secretion of IL-6 could be blocked by inhibition of the NOTCH pathway, confirming the mechanism (Houde et al, 2004). It is likely that such alterations exist

Bryan et al: DNA Methylation review diseases in humans. Recently, promoter hypermethylation was reported by Chuammitri and others in abstract form at the Genes, Dogs, and Cancer: Fourth Annual Canine Cancer Conference, 2006-Chicago, IL. This group identified promoter hypermethylation in the E-cadherin gene in canine mammary tumors, and in the TIMP3 and DAPK1 genes in a small number of canine lymphomas. No methylation was demonstrated in the RASSF1A gene in canine lymphomas. To date, no peer-reviewed literature exists that documents the presence of promoter hypermethylation in dogs. Expression of the tumor suppressor gene DLC1 has been shown to be silenced in multiple cancers, most recently in NHL. Originally identified as a deletion of chromosome 8p21.3-22 in human hepatocellular carcinoma, lack of expression was observed later in the absence such of a deletion (Yuan et al, 1998; Wong et al, 2003). Cells from non-small cell lung cancer, neuroectodermal cancer, breast, colon, prostate, and gastric cancer have shown lack of DLC1 mRNA when the gene promoter region was hypermethylated (Kim et al, 2003; Herman and Baylin, 2003; Wong et al, 2003; Yuan et al, 2003, 2004; Pang et al, 2005). Shi and colleagues examined human NHL cell lines and patient samples for hypermethylation of CpG islands using a CpG island microarray (Shi et al, 2006). The DLC1 gene was found to be hypermethylated in all six NHL cell lines examined, and, in every case, expression was silenced. In several lines, expression could be upregulated by treatment with a combination of a demethylating agent and a histone deacetylase inhibitor. Seventy-five NHL patient samples were examined for methylation of several candidate genes, including DLC1. Of these, 87% demonstrated hypermethylation of DLC1. Overall, expression of mRNA for this gene was significantly downregulated in tumor tissue compared to normal tissue.

suggested to be appropriate, naturally occurring clinical models of human NHL (Greenlee et al, 1990; Hansen and Khanna, 2004). The canine orthologue of the human gene DLC1 was identified and supported in silico using multiple prediction methods. This pilot study used DNA harvested from normal canine lymph nodes and lymph node aspirate samples from dogs with lymphoma stored in saline at -80oC to develop MSP and COBRA assays and evaluate the diseased samples for preliminary evidence of hypermethylation of a candidate gene. The overall hypothesis of this study was that hypermethylation occurs in canine lymphoma cells, as it does in human nonHodgkin’s lymphoma.

II. Materials and Methods A. In silico methods The NCBI reference number for the human DLC1 isoform 1 gene, NM_182643, was used to search a canine genome database (http://genome.ucsc.edu/cgi-bin/hgGateway) for the canine ortholog. The 5‘ region of the gene was examined using the MethPrimer (http://www.urogene.org/methprimer/ index1/html) CpG island analysis tool to identify candidate CpG islands (Li and Dahiya, 2002). This region was also examined using the Promoterscan transcription factor binding analysis tool to identify a promoter region within the sequence (Prestridge, 1995). The putative promoter region of the gene was then confirmed to match the human gene promoter region using the UCSC Genome Browser (http://genome.ucsc.edu/cgibin/hgGateway). A predicted mRNA sequence was constructed by concatenating the canine orthologs of the human exons of DLC1 identified using a basic local alignment search tool (BLAST). The resulting sequence was examined using the InterProScan (www.ebi.ac.uk/InterProScan/index.html) web-tool to identify protein functional groups (Zdobnov and Apweiler, 2001). The same procedure was performed using the N-Scan predicted mRNA sequence.

B. In vivo methods 1. Sample collection and preparation

D. Pilot evaluation of methylation of DLC1 in canine NHL

Lymph node aspirates were performed on clinical patients with multicentric, node-based lymphoma using 22ga. needles and preserved in Hank’s balanced salt solution at -80ºC until analysis. Peripheral blood mononuclear cells were isolated over a ficoll-hypaque (Sigma-Aldrich, St. Louis, MO) gradient and preserved at -80ºC until analysis. DNA was extracted using the Qiagen DNeasy Tissue (Qiagen USA, Valencia, CA) kit, and then bisulfite treated using the Zymo Research EZ DNA Methylation Gold (Zymo Research Corporation, Orange, CA) kit. Bisulfite treatment converts unmethylated cytosine to uracil, which becomes thymidine in subsequent polymerase chain reactions. Methylated cytosine is protected from conversion. Bisulfite treated DNA from normal canine lymph nodes was treated with SssI (New England Biolabs, Ipswich, MA), a DNA methyltransferase, and SAMe to methylate all CpG dinucleotides in the sequence and serve as a positive control.

The DLC1 gene product functions as a tumor suppressor gene (Yuan et al, 1998). The protein is a RhoGTPase Activating Protein (RhoGAP) that counteracts the feed forward signaling of RhoA and Cdc42 among other RAS signaling proteins (Wong et al, 2003). Specifically, the RhoGAP protein causes catalysis of GTP to GDP when bound to the Rho proteins, causing them to become inactive (Wong et al, 2003). Loss of this function results in unconstrained growth signaling from the surface of the cell to the nucleus, changes in cell mobility, and signaling between the cell and its extracellular environment (Yuan et al, 1998; Sahai and Marshall, 2002; Wong et al, 2003). Such changes could confer significant growth advantages, contributing to the initiation, promotion, or progression of cancer, as well as metastasis. The loss of function of DLC1 has been demonstrated to be a significant contributor to many human cancers as described above. However, this gene and its protein have not yet been characterized in the dog. The purpose of the present study was to screen clinical lymphoma samples for the presence of hypermethylation. Companion dogs with NHL have been

2. Methylation specific PCR The MethPrimer website was used to construct primers to amplify a 183bp region in the proximal predicted first intron, located between bases 39,535,120-39,535,307 of chromosome 16 (Table 1). Using bisulfite treated normal canine DNA, the conditions for MSP were optimized. The methylated primer set was used to amplify the region with an annealing temperature of 62°C for 30s, an extension temperature of 72°C for 30s, and a melting temperature of 95°C for 15s, repeating for 32 cycles. The

Cancer Therapy Vol 6-A.2, page 7 unmethylated primer set was used to amplify the region with an annealing temperature of 62°C for 60s, an extension temperature of 72°C for 60s, and a melting temperature of 95°C for 15s, repeating for 32 cycles. The PCR products were run on a 1.5% agarose gel with ethidium bromide. The negative control was bisulfite converted DNA from normal lymph node, and the positive control was DNA from a normal lymph node methylated in vitro using SssI, then bisulfite converted.

III. Results Searching the canine genome using the reference number NM_182643 for DLC1 yielded a highly significant similarity with a canine sequence in chromosome 16 (score 1328, Expect Value 0.0). This region is part of a predicted RhoGAP protein. The sequence of the canine 5‘ region corresponding to the human DLC1 promoter was retrieved (Figure 1). This sequence yielded two CpG islands in the MethPrimer analysis (Figure 1). These islands are large, 1,187 and 557 bp, respectively, with a gap of 223bp between them. The PromoterScan analysis of the reverse strand yielded an extremely high promoter score of 260.20 with a minimum promoter cutoff of 53.00 from bases 39,535,766 to 39,536,015. This identified promoter region corresponds highly to the reported human gene promoter region (Yuan et al, 1998). The promoter analysis identified 16 Sp-1 binding sites in close proximity with proper orientation to gene transcription. The transcription factor binding sites identified by Promoterscan in dogs and humans are listed in Table 1. Analysis of the sequence downstream of the predicted promoter region using the InterProScan webtool, revealed the presence of code for all the major functional units of the human DLC1 protein. These include the Rho-GAP active site, a lipid binding steroidogenic acute regulatory protein (START) site, and a sterile alpha motif homology 2 (SAM2) unit. The algorithm N-Scan predicts the location of a gene in the region that this examination predicts the DLC1 gene of dogs with all functional subunits included (van Baren and Brent, 2006).

3. COBRA The MethPrimer website was used to construct primer pairs for COBRA to amplify a 284bp region in the proximal predicted first intron, located between bases 39,534,204-39,534,492 of chromosome 16 (Table 1). These primers do not contain CpG dinucleotides, and will amplify DNA whether or not methylation is present in the gene. The conditions for PCR were an annealing temperature of 58°C for 60s, an extension temperature of 72°C for 60s, and a melting temperature of 95°C for 15s, repeating for 32 cycles. The product size of these primers is 284 bp and contains two BstUI (New England Biolabs, Ipswich, MA) cutsites that recognize the sequence CGCG and yield fragments of 27 bp, 38 bp, and 219 bp. For BstUI, 10!L of PCR product was added to 2.5!L of Buffer 2, 1!L of BstUI, and 11.5!L of HyPure water, and incubated at 60ºC for 4h. Controls were identical to MSP. The PCR products were run on a 1.5% agarose gel with ethidium bromide.

4. Statistical analysis DNA sequences were considered similar in BLAST search if the score was > 900 and the Expect Value was < 0.0001. Promoter score cutoff was set at a minimum of 53 on the Promoterscan software to identify a promoter region (Prestridge, 1995). For the CpG island determination, parameters were set to identify a sequence with a minimum length of 100 bases, a GC percentage > 50%, and a CpG observed/expected ratio of 0.6 (Li and Dahiya, 2002).

Figure 4. This is a composite of gels showing the results from the first 13 lymph node aspirate samples and CLL samples evaluated. M denotes methylated primers; U denotes unmethylated primers; UCX is unmethylated control; MCX is the SssI-treated methylated control; lymphoma samples are numbered one through 13. CLL samples are one through three. Interpretation of each lane is presented in Table 2.

Bryan et al: DNA Methylation review

Table 1. Transcription factors identified by Promoterscan in orthologous dog and human promoter regions of DLC1. Dog Transcription Factor Similar Factors Sp1 JCV_repeated_seq T-Ag GCF Differing Factors AP-2 NGFI-C KROX24 UCE.2 MRE_CS6 JunB-US2

Number

Human Transcription Factor

Number

16 2 2 2

Sp1 JCV_repeated_seq T-Ag GCF

8 1 1 1

2 1 1 0 0 0

AP-2 NGFI-C KROX24 UCE.2 MRE_CS6 JunB-US2

0 0 0 1 1 1

Figure 5. This gel shows lymphoma samples one through 13. The symbol “–“ denotes the negative methylation control and “+” the SssI -treated methylation control. Digestion of the PCR product with BstUI could yield bands of 219bp, 38bp, and 27bp from the original 284bp product. Interpretation of each lane is presented in Table 2.

Table 2. Results of MSPCR for thirteen canine lymph node aspiration samples. The symbol “+” denotes amplification with primers for methylated or unmethylated DNA. – denotes failure to amplify for the specified primer. Three patients showed no amplification with either primer set. Immunophenotype is denoted by B for B-cell, T for T-cell of N for not determined. Dog Unmethylated Methylated No Amplification COBRA CpG I 2 Immunophenotype

Normal + -

1 + -

2 +

3 + +

4 + +

5 + -

6 + +

+ B

+ N

+ B

+ N

8 + +

Examination of the CpG island containing the promoter region by MSP demonstrated the presence of bands from methylated primers in six of 13 NHL samples and two of three CLL samples (Table 1, Figure 4). Nine

10 + +

11 + -

+ T

+ + N

13 + -

+ N

+ T

of 13 NHL samples and two of three CLL samples demonstrated bands from unmethylated primers. Three NHL samples did not amplify with either primer set. Examination of the second CpG island by COBRA 8

Cancer Therapy Vol 6-A.2, page 9 demonstrated methylation in nine of 13 samples with the same three samples failing to amplify (Table 1, Figure 5). Immunophenotype was determined for six of the dogs with four B-cell and two T-cell lymphomas. Of these, two of four B-cell samples showed methylation by MSP and one did not amplify. No T-cell sample showed methylation by MSP, but both showed methylation in the 3â&#x20AC;&#x2122; CpG island by COBRA. Immunophenotype was not determined for CLL. No normal sample was positive for methylation. The DNA of at least three individual normal dogs was used in the development of the assay.

COBRA assay, which examined the CpG island located further downstream. It is significant that this methylation was identified using samples collected by fine needle aspiration. This underscores the ease of analysis and potential diagnostic utility of epigenetic changes that can be assayed by aspiration. Three canine NHL samples did not amplify. All three had lower DNA concentration and appeared at least partially degraded on agarose gel analysis. The lack of amplification could also be a result of deletion. Loss of the region of canine chromosome 16 containing this gene location has been reported in one case of T-cell lymphoma in a female Cocker spaniel (Thomas et al, 2003). Further study will be necessary to elucidate the frequency of such an event. Five of the six NHL and both methylated CLL samples also had a positive unmethylated band. The significance of this ummethylated band is not entirely clear at this time. It is certain that the cell population from which DNA was extracted was heterogeneous in composition, with neoplastic and normal cells mixed. The samples were lymph node aspirates, so would have contained more than one cell type. It is also possible that the neoplastic populations were heterogeneous in nature. Finally, the neoplastic cells could be individually heterozygous for methylation. Simultaneous demonstration of methylated and unmethylated markers is seen in human tumors as well (Kovalchuk et al, 2004; Shi et al, 2006; Taylor et al, 2007b) Whatever the source of the unmethylated DNA, the presence of hypermethylation of an important tumor suppressor gene in naturally occurring canine cancer represents the first report of its kind and is a significant find. It is significant that the observed methylation was present only in neoplastic tissue samples. As for the human DLC1 gene, methylation of the canine gene at this location may serve as a useful marker of the neoplastic phenotype (Wang et al, 2007). A methylation assay may identify early or even pre-neoplastic lesions that could be used to monitor nodes for minimal residual disease, or detect early relapse. To have a cancer-specific test that is robust and could be performed on lymph node aspiration samples rapidly and inexpensively would facilitate efficient diagnosis of canine lymphoma. Such a diagnostic test could also serve as a pre-clinical evaluation of similar technology in humans. Finally, as the clinical use of demethylating therapy is better defined, such a test may serve as a prognostic marker of likely response to therapy, even if expression of the specific gene identified is not modified.

IV. Discussion Epigenetic mechanisms have been shown to play an important role in many human cancers, including NHL. Whether these changes are causative or the result of neoplastic transformation has not been clearly established (Herman and Baylin, 2003). Environmental influences have been identified which modify the methylation pattern of DNA in laboratory animals (Ho et al, 2006). Such changes may cause or promote human cancers, however the epidemiology remains controversial (Welshons et al, 2003). Companion animals tend to share the lives and environmental exposures of their human counterparts (Hansen and Khanna, 2004). This represents an untapped resource of information on the interplay of environment and disease. Developing evidence that demonstrates similar epigenetic alterations in human and canine disease is critical to the development of this naturally occurring model of cancer. The sequence for the canine RhoGAP gene identified using the NCBI site is clearly orthologous to the human DLC1 gene. The high score and low Expect value make it a statistical impossibility that this gene is anything but the DLC1 ortholog. This gene lies on the long arm of canine chromosome 16, a 64Mbp chromosome. Through eons of recombination and fragmentation, this chromosome in the dog contains portions of the genes of human chromosome 8, along with genes located on human chromosomes 4 and 7. The sequence of the predicted promoter region lies within a CpG island, similar to the arrangement of the human gene, with a second, smaller CpG island present that is not seen in the human sequence. The promoter score is extremely high for a dog sequence using the webbased tool. This is the highest score observed by the authors for a canine gene. It is much higher than the score for the human promoter region of DLC1, 99.09. The UCSC Genome Browser clearly shows good alignment of this predicted region with the promoter, first exon, and first intron of the human DLC1 gene. The presence of numerous Sp-1 transcription factor binding sites in the promoter region also suggests that methylation could play a role in control of the gene. This transcription factor has been demonstrated in humans to be inhibited from binding by the presence of methylation (Mancini et al, 1999). The similarity between the predicted canine promoter region and the human tumor suppressor gene makes the identified area the most likely 5â&#x20AC;&#x2122; region of this canine ortholog. Methylation of the DLC1 CpG islands was evident in the majority of the canine NHL and CLL when assayed by MSP and COBRA. The proportion was higher in the

Acknowledgements This work was presented in part at the 22nd Annual Meeting of the Veterinary Cancer Society.

References Al-Kuraya K, Narayanappa R, Siraj AK, Al-Dayel F, Ezzat A, El SH, Al-Jommah N, Sauter G, Simon R (2006) High frequency and strong prognostic relevance of O6methylguanine DNA methyltransferase silencing in diffuse

Bryan et al: DNA Methylation review large B-cell lymphomas from the Middle East. Hum Pathol 37, 742-748. Ando Y, Hayashizaki Y (2006) Restriction landmark genomic scanning. Nat Protoc 1, 2774-2783. Bachman KE, Park BH, Rhee I, Rajagopalan H, Herman JG, Baylin SB, Kinzler KW, Vogelstein B (2003) Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene. Cancer Cell 3, 89-95. Bock C, Lengauer T (2008) Computational epigenetics. Bioinformatics 24, 1-10. Chan MW, Chan LW, Tang NL, Lo KW, Tong JH, Chan AW, Cheung HY, Wong WS, Chan PS, Lai FM, To KF (2003) Frequent hypermethylation of promoter region of RASSF1A in tumor tissues and voided urine of urinary bladder cancer patients. Int J Cancer 104, 611-616. Clark SJ, Harrison J, Paul CL, Frommer M (1994) High sensitivity mapping of methylated cytosines. Nucleic Acids Res 22, 2990-2997. Cottrell SE, Laird PW (2003) Sensitive detection of DNA methylation. Ann N Y Acad Sci 983, 120-130. Cui H, Cruz-Correa M, Giardiello FM, Hutcheon DF, Kafonek DR, Brandenburg S, Wu Y, He X, Powe NR, Feinberg AP (2003) Loss of IGF2 imprinting: a potential marker of colorectal cancer risk. Science 299, 1753-1755. Dong SM, Sun DI, Benoit NE, Kuzmin I, Lerman MI, Sidransky D (2003) Epigenetic inactivation of RASSF1A in head and neck cancer. Clin Cancer Res 9, 3635-3640. Donson AM, Addo-Yobo SO, Handler MH, Gore L, Foreman NK (2007) MGMT promoter methylation correlates with survival benefit and sensitivity to temozolomide in pediatric glioblastoma. Pediatr Blood Cancer 48, 403-407. Feinberg AP, Vogelstein B (1983) Hypomethylation of ras oncogenes in primary human cancers. Biochem Biophys Res Commun 111, 47-54. Feinberg AP, Ohlsson R, Henikoff S (2006) The epigenetic progenitor origin of human cancer. Nat Rev Genet 7, 21-33. Fraga MF, Esteller M (2002) DNA methylation: a profile of methods and applications. Biotechniques 33, 632, 634, 636632, 634, 649. Greenlee PG, Filippa DA, Quimby FW, Patnaik AK, Calvano SE, Matus RE, Kimmel M, Hurvitz AI, Lieberman PH (1990) Lymphomas in dogs. A morphologic, immunologic, clinical study. Cancer 66, 480-490. Hansen K, Khanna C (2004) Spontaneous and genetically engineered animal models; use in preclinical cancer drug development. Eur J Cancer 40, 858-880. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de TN, Weller M, Kros JM, Hainfellner JA, Mason W, Mariani L, Bromberg JE, Hau P, Mirimanoff RO, Cairncross JG, Janzer RC, Stupp R (2005) MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 352, 9971003. Herman JG and Baylin SB (2003) Gene silencing in cancer in association with promoter hypermethylation. N Engl J Med 349, 2042-2054. Ho SM, Tang WY, Belmonte de FJ, Prins GS (2006) Developmental exposure to estradiol and bisphenol A increases susceptibility to prostate carcinogenesis and epigenetically regulates phosphodiesterase type 4 variant 4. Cancer Res 66, 5624-5632. Holliday R (1993) Epigenetic inheritance based on DNA methylation. EXS 64, 452-468. Hotchkiss RD (1949) Chemical studies on the transforming factor of pneumococci. Colloques intern centre natl recherche sci Unites biol douees contin genet (Paris) 8, 5765. Houde C, Li Y, Song L, Barton K, Zhang Q, Godwin J, Nand S, Toor A, Alkan S, Smadja NV, vet-Loiseau H, Lima CS,

Miele L, Coignet LJ (2004) Overexpression of the NOTCH ligand JAG2 in malignant plasma cells from multiple myeloma patients and cell lines. Blood 104, 3697-3704. Jaffe LF (2003) Epigenetic theories of cancer initiation. Adv Cancer Res 90:209-30., 209-230. Jones PA,d Baylin SB (2007) The epigenomics of cancer. Cell 128, 683-692. Kim TY, Jong HS, Jung Y, Kim TY, Kang GH, Bang YJ (2004) DNA hypermethylation in gastric cancer. Aliment Pharmacol Ther 20 Suppl 1, 131-142. Kim TY, Jong HS, Song SH, Dimtchev A, Jeong SJ, Lee JW, Kim TY, Kim NK, Jung M, Bang YJ (2003) Transcriptional silencing of the DLC-1 tumor suppressor gene by epigenetic mechanism in gastric cancer cells. Oncogene 22, 3943-3951. Kovalchuk O, Burke P, Besplug J, Slovack M, Filkowski J, Pogribny I (2004) Methylation changes in muscle and liver tissues of male and female mice exposed to acute and chronic low-dose X-ray-irradiation. Mutat Res 548, 75-84. Leu YW, Yan PS, Fan M, Jin VX, Liu JC, Curran EM, Welshons WV, Wei SH, Davuluri RV, Plass C, Nephew KP, Huang TH (2004) Loss of estrogen receptor signaling triggers epigenetic silencing of downstream targets in breast cancer. Cancer Res 64, 8184-8192. Li LC, Dahiya R (2002) MethPrimer: designing primers for methylation PCRs. Bioinformatics 18, 1427-1431. Liu H, Wang J, Epner EM (2004) Cyclin D1 activation in B-cell malignancy: association with changes in histone acetylation, DNA methylation, RNA polymerase II binding to both promoter and distal sequences. Blood 104, 2505-2513. Mancini DN, Singh SM, Archer TK, Rodenhiser DI (1999) Sitespecific DNA methylation in the neurofibromatosis (NF1) promoter interferes with binding of CREB and SP1 transcription factors. Oncogene 18, 4108-4119. Mellor J (2006) It takes a PHD to read the histone code. Cell 126, 22-24. Murai M, Toyota M, Satoh A, Suzuki H, Akino K, Mita H, Sasaki Y, Ishida T, Shen L, Garcia-Manero G, Issa JP, Hinoda Y, Tokino T, Imai K (2005) Aberrant DNA methylation associated with silencing BNIP3 gene expression in haematopoietic tumours. Br J Cancer 92, 1165-1172. Pang JC, Chang Q, Chung YF, Teo JG, Poon WS, Zhou LF, Kong X, Ng HK (2005) Epigenetic inactivation of DLC-1 in supratentorial primitive neuroectodermal tumor. Hum Pathol 36, 36-43. Park J, Brena RM, Gruidl M, Zhou J, Huang T, Plass C, Tockman MS (2005) CpG island hypermethylation profiling of lung cancer using restriction landmark genomic scanning (RLGS) analysis. Cancer Biomark 1, 193-200. Pelham JT, Irwin PJ, Kay PH (2003) Genomic hypomethylation in neoplastic cells from dogs with malignant lymphoproliferative disorders. Res Vet Sci 74, 101-104. Prestridge DS (1995) Predicting Pol II promoter sequences using transcription factor binding sites. J Mol Biol 249, 923-932. Rollins RA, Haghighi F, Edwards JR, Das R, Zhang MQ, Ju J, Bestor TH (2006) Large-scale structure of genomic methylation patterns. Genome Res 16, 157-163. Sahai E, Marshall CJ (2002) RHO-GTPases and cancer. Nat Rev Cancer 2, 133-142. Schulz WA, Hatina J (2006) Epigenetics of prostate cancer: beyond DNA methylation. J Cell Mol Med 10, 100-125. Shamay M, Krithivas A, Zhang J, Hayward SD (2006) Recruitment of the de novo DNA methyltransferase Dnmt3a by Kaposi's sarcoma-associated herpesvirus LANA. Proc Natl Acad Sci U S A 103, 14554-14559. Shi H, Guo J, Duff DJ, Rahmatpanah F, Chitima-Matsiga R, AlKuhlani M, Taylor KH, Sjahputera O, reski M, Wooldridge JE, Caldwell CW (2006) Discovery of novel epigenetic

Cancer Therapy Vol 6-A.2, page 11 markers in non-Hodgkin's lymphoma. Carcinogenesis 28, 60-70. Siegfried Z, Eden S, Mendelsohn M, Feng X, Tsuberi BZ, Cedar H (1999) DNA methylation represses transcription in vivo. Nat Genet 22, 203-206. Singal R, Ginder GD (1999) DNA methylation. Blood 93, 40594070. Takai D and Jones PA (2002) Comprehensive analysis of CpG islands in human chromosomes 21 and 22. Proc Natl Acad Sci U S A 99, 3740-3745. Taylor KH, Kramer RS, Davis JW, Guo J, Duff DJ, Xu D, Caldwell CW, Shi H (2007a) Ultradeep bisulfite sequencing analysis of DNA methylation patterns in multiple gene promoters by 454 sequencing. Cancer Res 67, 8511-8518. Taylor KH, Pena-Hernandez KE, Davis JW, Arthur GL, Duff DJ, Shi H, Rahmatpanah FB, Sjahputera O, Caldwell CW (2007b) Large-scale CpG methylation analysis identifies novel candidate genes and reveals methylation hotspots in acute lymphoblastic leukemia. Cancer Res 67, 2617-2625. Thomas R, Smith KC, Ostrander EA, Galibert F, Breen M (2003) Chromosome aberrations in canine multicentric lymphomas detected with comparative genomic hybridisation and a panel of single locus probes. Br J Cancer 89, 1530-1537. van Baren MJ, Brent MR (2006) Iterative gene prediction and pseudogene removal improves genome annotation. Genome Res 16, 678-685. Wang MX, Shi H.D., Zhao XH, Taylor KH, Duff D, Caldwell C.W. (2007) Molecular Detection of Residual Leukemic Cells by a Novel DNA Methylation Biomarker (Abstract #1206). Mod Pathol 20, 263A.

Welshons WV, Thayer KA, Judy BM, Taylor JA, Curran EM, vom Saal FS (2003) Large effects from small exposures. I. Mechanisms for endocrine-disrupting chemicals with estrogenic activity. Environ Health Perspect 111, 9941006. Wong CM, Lee JM, Ching YP, Jin DY, Ng IO (2003) Genetic and epigenetic alterations of DLC-1 gene in hepatocellular carcinoma. Cancer Res 63, 7646-7651. Yan PS, Perry MR, Laux DE, Asare AL, Caldwell CW, Huang TH (2000) CpG island arrays: an application toward deciphering epigenetic signatures of breast cancer. Clin Cancer Res 6, 1432-1438. Yuan BZ, Durkin ME, Popescu NC (2003) Promoter hypermethylation of DLC-1, a candidate tumor suppressor gene, in several common human cancers. Cancer Genet Cytogenet 140, 113-117. Yuan BZ, Jefferson AM, Baldwin KT, Thorgeirsson SS, Popescu NC, Reynolds SH (2004) DLC-1 operates as a tumor suppressor gene in human non-small cell lung carcinomas. Oncogene 23, 1405-1411. Yuan BZ, Miller MJ, Keck CL, Zimonjic DB, Thorgeirsson SS, Popescu NC (1998) Cloning, characterization, chromosomal localization of a gene frequently deleted in human liver cancer (DLC-1) homologous to rat RhoGAP. Cancer Res 58, 2196-2199. Zdobnov EM, Apweiler R (2001) InterProScan--an integration platform for the signature-recognition methods in InterPro. Bioinformatics 17, 847-848.

Bryan et al: DNA Methylation review

Cancer Therapy Vol 6-A.2, page 13 Cancer Therapy Vol 6, 167-176, 2008

Veterinary radiation oncology: technology, imaging, intervention and future applications Review Article

Ira K. Gordon, Michael S. Kent* Department of Surgical and Radiological Sciences, UC Davis School of Veterinary Medicine, Davis, CA 95616

__________________________________________________________________________________ *Correspondence: Michael S. Kent, Assistant Professor, Department of Surgical and Radiological Sciences, 2112 Tupper Hall, 1 Shields Ave, Davis, CA 95616, USA; Tel: 530-752-1393; Fax: 530-752-9620; E-mail: mskent@ucdavis.edu Key words: Radiation therapy, Oncology, Cancer, Tumor, Veterinary Abbreviations: image guided radiation therapy (IGRT); cone beam computed tomography (CBCT); clinical tumor volume (CTV); computed tomography (CT); functional magnetic resonance imaging (fMRI); Gray (Gy); gross tumor volume (GTV); intensity modulated radiation therapy (IMRT); kilovoltage (kV); magnetic resonance imaging (MRI); magnetic resonance spectroscopic imaging (MRSI); multileaf collimators (MLC); planning target volume (PTV); single-positron emission computed tomography (SPECT); stereotactic radiosurgery (SRS); ultrasound (U/S) Received: 1 February 2008; electronically published: June 2008

Presented at the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary Radiation therapy has become an important modality in treating the veterinary cancer patient. Recent advances in technology such as advanced imaging, electron therapy, custom blocking, computerized treatment planning and advanced treatment techniques such as 3D conformal therapy and intensity modulated radiotherapy have advanced the field. On the near horizon are particle therapy, stereotactic radiosurgery and functional imaging. All these advances allow better targeting and treatment of tumors with the possibility of decreased side effects. This paper reviews the current state of the art of veterinary radiation oncology and looks toward the near future to see where the specialty is going.

versus tumor tissues can be accentuated with dose fractionation. Keeping these principles in mind, most of the recent advances made in radiation therapy have been to spare normal tissue by better targeting of the tumor with external beam radiotherapy techniques and equipment. This allows greater doses of radiation to be delivered to the tumor without increasing the side effects of irradiation. As these are technology driven, they can be expensive to implement but the benefits to the patient can be great.

I. Introduction In simplest terms, the primary goal of the oncologist is to kill cancer cells while sparing normal cells and tissues. In medical oncology, this is usually accomplished by the use of agents that are given systemically but have preferential toxicity to neoplastic cells. Only occasionally is chemotherapy given by a route that targets the desired site (intratumoral, intrathecal, intracavitary). In surgical oncology, the goal is to surgically remove the tumor with the dose of surgery limited by the resulting functional or cosmetic defects. In contrast, the radiation oncologist primarily accomplishes this fundamental goal by maximizing radiation dose to a defined tumor target while minimizing the dose received by surrounding normal tissues. This can be accomplished through the use of various radiation techniques including brachytherapy, external beam radiotherapy, and systemic targeted radiotherapy. In addition, differences in radiation response of normal

II. History and progress in radiation therapy Radiation oncology is a medical science that has existed for just over 100 years. The discovery of X-ray photons is considered the birth of radiation oncology which occurred in Wilhelm Roentgenâ&#x20AC;&#x2122;s lab on November 8, 1895 (Smith et al, 2006). Within months, the first cancer patients were treated with radiotherapy. Shortly thereafter, radioactivity and radium were discovered by

Gordon and Kent: Veterinary radiation oncology Henri Becquerel and Pierre and Marie Curie (Hall and Giaccia, 2006). By 1902, over 100 different conditions were listed that could be treated with X-rays. Brachytherapy was first described and performed in 1904. Many of the early radiation researchers suffered serious injuries or death due to a lack of understanding of radiation biology and dose measurement. The first major advance in radiation technology was the invention of the Coolidge tube in 1912-13 which delivered a more reliable beam in terms of beam energy and penetration and was the precursor to orthovoltage radiotherapy machines (Bernier et al, 2004). In the 1920s and 1930s, in addition to continued improvement of radiation delivery technology, the major advances were the development of dose measurement techniques, central axis treatment planning, and understanding of the benefits of dose fractionation. In 1955, the first patient was treated using the first linear accelerator at Stanford (Hall and Giaccia, 2006). Over the next 30-40 years, significant improvements were made to radiation therapy equipment with higher energy and multiple energy machines being developed. Machines were also made that produce other types of radiation including electrons, neutrons, and protons. Additionally, advances in computer technology allowed for faster and more complex treatment planning software. Despite these advances, the basic process of radiation planning and delivery remained relatively unchanged for most of the 20th century. In the past 10-15 years, while equipment and technology have continued to improve, the most significant clinical improvements in radiation treatments have come from fundamental changes to the basic approach to radiation planning, setup, and treatment delivery. The development and advances of novel imaging techniques, intensity modulated radiation therapy (IMRT), adaptive and image guided radiation therapy (IGRT), and stereotactic radiosurgery (SRS) represent new frontiers for advancement toward the cure of diseases that were previously considered untreatable or uncurable.

deliver a variety of energies and many of the newer machines allow treatment with either photons or electrons (see below for a more detailed explanation). They also allow the use of smaller field sizes, allow for a more homogenous dose to be delivered to deep-seated tumors and can take advantage of newer treatment techniques. The use of advanced treatment planning software allows dosimetric calculations and three-dimensional treatment planning to deliver accurate doses to deep-seated tumors. Systemic radiotherapy using radionucleotides and brachytherapy are also used in treating veterinary patients but are not covered in this review.

IV. Electron therapy Linear accelerators deliver radiation through photons and in newer machines either photons or electrons. Photons provide better deep penetration and often a more homogenous dose distribution in tissue than electrons but there are times that this might pose a problem (Figure 1). If the tumor is located over critical normal structures that you do not want to irradiate such as the spinal cord, intestines or lungs, electrons may be more advantageous. Electrons deliver their dose through a certain depth and then fall off very quickly. By setting the energy of the electrons you can control the depth of penetration. A rule of thumb is penetration in cm is about 1/2 of the energy of electron selected, while the useful beam penetrates to a maximum of about 1/3 of the energy of the electron you choose in cm. For example, treating with 6 MeV electrons corresponds to about 2 cm of tissue effectively treated and using 20 MeV electrons results in approximately 6 cm of tissue being effectively treated. Many of the tumors that we treat in veterinary medicine are located in the skin or subcutaneous tissues allowing electrons to be used effectively.

V. How radiation is dosed The SI unit of absorbed dose used in radiation oncology is the Gray (Gy). The older unit is the rad. 1 Gy is equal to 100 cGy or 100 rads. When considering the total dose to be given, several things need to be taken into account; Specifically: The radiosensitivity of the tumor, the dose delivered in each fraction, the time between fractions, the total number of fractions, the goal of therapy (palliative vs. definitive) and often most importantly, the tolerance of the surrounding normal tissues.

III. Machines used to deliver radiation therapy There are three different types of machines used to deliver external beam radiation therapy in veterinary radiation oncology (McEntee, 2004). Orthovoltage machines, cobalt-60 machines and linear accelerators are used for teletherapy. Orthovoltage machines are no longer widely used and have the disadvantages of increased skin reactions, no computerized treatment planning systems, high bone absorption of dose and the lack of isocentric machines. Cobalt-60 machines are still in use in veterinary medicine but also have some disadvantages compared to linear accelerators, which include increased penumbra at the field edge (leading to larger treatment fields), lower energy (leading to increased skin dosing), lower dose rate, and less penetration of dose. Linear accelerators are becoming more widely available in veterinary radiation oncology and are the standard of care in treating humans. They are able to

VI. Target therapy

volumes

radiation

It is useful to define several â&#x20AC;&#x153;tumor volumesâ&#x20AC;? when prescribing a radiation dose to a given area. The simplest target to define is the gross tumor volume (GTV). This volume incorporates all palpably or visually abnormal tissue that is evident on physical examination or routine diagnostic imaging studies. In animals that present for radiotherapy after marginal surgical resection of the primary mass, there is no GTV.

Cancer Therapy Vol 6-A.2, page 15

Figure 1. CT image of a dog’s thoracic cavity with an incompletely resected soft tissue sarcoma showing the dose distribution for electrons.

The clinical tumor volume (CTV) represents the gross tumor volume plus all regions that are suspected to contain microscopic disease. In some cases, this may be several centimeters around the gross disease in all directions. In other instances, it may represent the area defined by surgical hemoclips with a several centimeter additional margin. When a tumor is contained by a bony structure or is not felt to contain significant microscopic extension, the CTV may not be much larger than the GTV. A major challenge with delineating the CTV exists as it is unknown how far microscopic disease extends. This decision is therefore based on the biologic behavior of the tumor and the clinical situation of the particular patient and has the risk of missing disease. More commonly, a radiation oncologist will try to overestimate the extent of disease to prevent “geographic miss”. The consequence of these overestimates is increasing the volume of normal tissue that full dose is prescribed to. With advances in physiologic and functional imaging such as PET/CT and fMRI, one goal is to better define the CTV rather than simply adding a margin to the GTV. The planning target volume (PTV) is the CTV plus an additional margin to account for uncertainties associated with radiation delivery (Figure 2) (Purdy, 2004). The PTV is the volume that you ultimately plan to treat with your treatment setup or treatment planning system. The extent of this additional margin is completely

dependant upon your ability to minimize setup uncertainty and patient positioning relative to the radiation field and commonly range from at least 3-5 millimeters up to 1-2 centimeters. These uncertainties include tolerances for the radiation equipment as well as setup inaccuracies and both interfraction and intrafraction patient motion (Purdy, 2004). Each of the following parameters can introduce at least 1-2 mm of uncertainty about the precision of the radiation field relative to the patient although not all are relevant for every patient: 1) Collimator position – The size the collimator is set for versus the actual size of the radiation field 2) Collimator drift- Movement in the collimator during treatment or when the gantry is at different positions 3) Isocenter uncertainty- Movement of the isocenter when the gantry is at different positions 4) Light field uncertainty- Deviation between the radiation field and the light field displayed on the patient 5) Portal imaging uncertainty- The edges and field center as defined by portal image (whether film or digital) 6) Diagnostic imaging uncertainty- This can be from limitations of spatial accuracy with MRI or constraints based on the slice thickness of your diagnostic CT scan 7) Image registration uncertainty- If you merge 2 diagnostic studies such as an MRI and CT scan, there is uncertainty in the precision of that match.

Gordon and Kent: Veterinary radiation oncology

Figure 2. CT image of a cat with an injection site sarcoma showing the GTV (yellow line) CTV (red line) and the PTV (green shading).

8) Positioning and movement uncertainties before and during each treatment 9) Movement of fiducial markers or skin marks relative to the tumor target 10) Penumbra of field- The penumbra of the field is the region at the field edge where dose falloff occurs. This depends on the type and energy of radiation you are using and can be greater than 10 millimeters for Co-60 ! rays or for electron beams. Penumbras for linear accelerators are much smaller. If this is taken into account with your treatment planning system, it may not need to be considered twice by adding a margin into your PTV. However, if not or if you are planning a treatment by hand, the uncertainty of the penumbra region should be incorporated into the PTV. Once again, the tradeoff between a small and large PTV margin is the risk of geographic miss versus increased normal tissue radiation exposure. In stereotactic radiosurgery, which is discussed later, the use of external fiduciary markers with precise positioning devices and lasers minimize setup uncertainty while finely collimated cones reduce penumbra and collimator uncertainties so that a tight PTV can be drawn around the CTV to even further minimize dose to normal tissue.

can be used. Custom blocks can be made for shaping both electron and photon beams (Figure 3). Blocking can be used with both hand and computer plans. A custom electron block can be made in less than 10 minutes. When the patient has finished their treatment course, the block can then be melted and the alloy reused.

VIII. Patient reproducibility

positioning

for

It is essential not to decrease the PTV beyond patient positioning and movement limitations or the tumor may not receive the planned dose. Using a 3-D conformal or IMRT plan that targets the tumor precisely on a patient that cannot be accurately positioned will not deliver the desired dose and does not make sense. Fortunately, imaging techniques to check patient positioning have also improved which allows checking the accuracy of patient positioning on each treatment if needed. A major question that must be answered is how precisely can a patient be setup each day. All of the precision and detail that goes into advanced treatment planning is useless if the patient is unable to be positioned precisely, reliably and reproducibly. The importance of this cannot be overstated, especially as technology allows for finer tumor targeting it can often be just millimeters separating a full dose from a region far below therapeutic doses. Many devices have been developed to help with this over the last few years. These include vacuum bags, bite blocks, headframes, masks and calibrated treatment couches (Figure 4) (Kippenes et al, 2000; Lester et al, 2001; Green et al, 2003). Port films, which are radiographs taken using the treatment machine to check patient positioning are key in determining the accuracy of positioning (Rohrer Bley et al, 2003). They are particularly important for deep seated tumors such as nasal tumors and brain tumors where you cannot directly see the

VII. Custom blocking While the standard collimator on a cobalt-60 machine or linear accelerator can make only squares or rectangles, planned radiation fields are often more complex in shape. There are several different ways that the beam can be shaped to conform to the planned field. The simplest but crudest way is to have pre-cast blocks which can be hand positioned onto plastic trays fitted into the machine. For increased accuracy, either a multileaf collimator, which most commonly have between 80 and 120 leaves, or a custom made block made from a low temperature melting lead alloy to shield part of the field

Cancer Therapy Vol 6-A.2, page 17

Figure 3. Images of a (a) photon block and a (b) electron block.

Figure 4. Image of a dog set-up for a radiotherapy treatment using both a vacuum bag and a head mask.

area you are treating. They should be done at least weekly and with particularly difficult fields should even be done on a daily basis for quality assurance (McEntee, 2008). Taking two orthogonal port films allows accurate positioning of a patient to within several millimeters, ensuring that the plan can be accurately carried out regardless of how many beams are used or at what angle they enter. Traditionally, port films have been made using specialized cassettes and film and processed like a regular

piece of radiograph film. Newer techniques such as digital on board portal imaging equipment allow imaging of the patient right before treatment and image registration to compare actual patient positioning with what was planned (Figure 5) (McEntee, 2006, 2008). One limitation of this technique is that imaging with the high energy megavoltage (MV) beam of a linear accelerator yields poor soft tissue contrast. One solution has been to equip a linear accelerator with an on-board kilovoltage (kV) imaging system for diagnostic quality radiographs. 17

Gordon and Kent: Veterinary radiation oncology

Figure 5. A digitally reconstructed radiograph created from the initial CT scan of a dog with a solitary plasma cell tumor of the maxilla used for treatment planning (a) and a port film taken using an electronic portal imaging device (b).

On board cone beam computed tomography (CBCT) imaging is a technique of producing volumetric CT images at the time of treatment and is now becoming available on newer linear accelerators for even more precise positioning. The use of real-time imaging to adjust patient positioning immediately before and even during treatment is called IGRT (Xing et al, 2006). The power of this technology may reach the point that as the patient or tumor shape changes during the course of treatment, adaptive treatment planning can adjust for positioning changes during a single treatment and shape changes within the patient and/or target over the course of radiotherapy as well. With the increased precision of radiation equipment, the importance of 4-D treatment planning and positioning is now being realized. The â&#x20AC;&#x153;4th dimensionâ&#x20AC;? this refers to is subtle movement over time during and between radiation treatments (intrafraction and interfraction motion). Respiratory motion can result in several centimeters of movement of a tumor target and critical normal tissues including heart and lung. Gastrointestinal peristalsis during radiation is now recognized as a major cause of both geographic miss and colonic overdosage in human prostate cancer patients undergoing radiation therapy. Through the use of fiducial markers or infrared sensors, there are now techniques that can actually adjust the radiation field to match patient motion. While organ

motion was once difficult to study, 4-dimensional CT can provide extensive information about the dynamic nature of tumors and internal organs (Webb, 2006). Linear accelerators are now available with integrated imaging devices to provide target identification, real time monitoring for motion and delivery modification, verification, dose reconstruction, and adaptive therapy. Imaging devices that allow for respiratory gating, breathing control, and adapt therapy to account for motion will allow further sparing of normal tissue and finer tumor targeting (Giraud et al, 2006). These techniques will become extremely important for techniques such as IMRT with extremely conformal dose distributions and gradients at the boundary of target volumes and organs at risk (Keall et al, 2006). Refining these techniques will likely allow for dose escalations resulting in higher rates of local tumor control and survival.

IX. Treatment planning A. Hand planning When treating an area surrounding a scar from an incompletely excised soft tissue sarcoma or mast cell tumor for example, a hand calculated plan is often sufficient and saves cost to the owner. This is done by placing an appropriate margin around the scar and then calculating an appropriate dose to the desired depth, taking into account the skin and exit doses. This becomes rapidly impractical if multiple beams, complex blocking or beam 18

Cancer Therapy Vol 6-A.2, page 19 modifying devices are used. The dose delivered also may not match the dose calculated across the radiation field, particularly if the patient’s body contour in the radiation field is not flat.

X. 3-D conformal radiation therapy and IMRT 3-D conformal radiation therapy is similar to conventional 3-D planning except that multiple beam angles and conformal blocks are used to shape the dose closely to the target volume and simultaneously allow sparing of normal tissues. These beam angles and blocks are selected using reconstructed imaging data and treatment planning software capable of performing thousands of calculations in a short amount of time. IMRT is a technique that takes advantage of exact patient positioning, multileaf collimators (MLC) and treatment planning software to shape and modify the intensity of the beam so that the delivered dose conforms to the tumor (Purdy, 2007). There are two fundamental differences between IMRT and 3-D conformal radiation therapy. The first fundamental difference is the use of the computer driven small mobile tungsten leaves of the MLC to various positions around the patient. The “step and shoot” technique of IMRT involves moving the leaves to multiple fixed positions around the patient. What is more common and useful is dynamic treatment in which the leaves of the MLC move while the linear accelerator beam is on, modulating the fluence of the beam and effectively treating thousands of “beamlets” per field. The calculation algorithms for dynamic IMRT are so complex that they cannot be planned by conventional methods (Bortfeld, 2006). The second fundamental difference between IMRT and conformal radiotherapy is the use of “inverse planning.” Recall that in conventional and conformal treatment planning, the radiation oncologist or dosimetrist creates the beam orientation, weight, shape, field size and modifiers. The treatment planning system then calculates the resulting dose. The parameters can then be adjusted by modifying beam weighting, adding or removing beams,

B. Computer planning With the advent of computer planning, radiation plans with multiple beams can be easily made using images from a CT or MRI. Most treatment planning systems use CT scans for treatment planning since it provides density information for dose calculation and because it provides better geometric accuracy than MRI. However many of these programs allow fusing an MRI to a CT scan, taking advantage of the superior soft tissue detail seen with MRI. This is particularly useful when planning brain tumors, which are much better visualized on MRI than CT. For most other tumor types usually a CT alone is sufficient. For treatment planning, the patient is scanned in the position that they will be treated in using any patient positioning devices that will be used. This allows for the most accurate dose calculation and delivery. The CT is then imported into the treatment planning software program. The area to be irradiated is then identified and marked and the CTV is drawn. The margin for the CTV is based on the tumor’s type, location, and biological behavior. The PTV is then drawn around the CTV. The radiation oncologist then adds beams onto the tumor and calculates the dose. 3-D treatment planning systems let you visualize dose on a 3-D image and also allow creation of dose volume histograms, which are graphs that show what volume of tissue is receiving what dose of radiation (Figure 6). This is useful in ensuring that the tumor is receiving the desired dose and that normal structures (such as lung, spinal cord or eyes) are getting the minimal possible dose. Individual plans can then be compared and the best one chosen.

Figure 6. (a) A plan from a dog with a pituitary tumor and (b) the dose volume histogram from the same plan showing dose to the CTV (purple), PTV (red) and to the brain (blue) and eyes (green).

Gordon and Kent: Veterinary radiation oncology and adjusting other dose modifiers in a trial and error process. In inverse planning, the process starts by prescribing a dose to the defined target volume and setting constraints for the tolerable doses to normal tissues. The treatment planning system is responsible for optimizing beam weights, field sizes, and beam fluence to meet those constraints (Bortfeld, 2006). When those constraints cannot be met, inverse planning becomes an interactive process whereby the radiation oncologist or dosimetrist must adjust or prioritize the dose constraints until the most ideal treatment plan is created. Because frequently the ratio of the dose delivered to normal tissues compared to the tumor is reduced to a minimum with IMRT, dose escalation to more effective radiation doses can safely be performed with fewer side effects compared with conventional radiotherapy techniques (Smith et al, 2006). There are several limitations to IMRT. First, not all treatment locations can be improved to a significant extent compared to more conventional techniques. It is much more time consuming to verify radiation dose with IMRT than conventional and conformal techniques. The problems of organ motion (see section on image guided radiation therapy) may also interfere with the ability to accurately carry out an IMRT plan (Purdy, 2004). Of unknown clinical importance at this time is the effect of adding more beams to treatment. In addition to daily treatment time increasing, the volume of normal tissue that receives some dose of radiation increases and the total body radiation dose increases as well. Because these techniques are still relatively new, it is not yet known whether or to what extent those increases may lead to increased rates of carcinogenesis or other adverse effects (Hall, 2006).

use these benefits without losing the benefit of the information provided by a CT scan by overlaying the data. Historically, U/S has found limited application for external beam radiation due to a lack of a well-defined method for using a three dimensional coordinate system. As image quality has improved, U/S is now able to verify patient setup and identify organs at risk in a radiation field (Smith et al, 2006). It also has wide applications in humans for image guided placement of seed implants. While radioactive implants may not be a viable option for many veterinary patients, the placement of fiducial markers into internal sites for daily treatment localization can be readily accomplished (Smith et al, 2006). Techniques such as fMRI are not yet widely available in veterinary medicine but can provide information about local oxygen concentrations, oxygen consumption, and perfusion which may give information about areas within a tumor that may be hypoxic or more resistant to radiation. MRSI has the ability to detect molecules that may reflect different disease states to differentiate normal and malignant tissues (Smith et al, 2006). For 50 years, radiation oncologists have tried to improve the homogeneity of dose throughout a tumor. It is well known that tumors are very heterogeneous. As advances in imaging permit detection of these tumor heterogeneities, it may be found that the best treatment is not one of homogeneous dose throughout the tumor but one of planned heterogeneous dosing with higher risk portions of the tumor boosted to a much increased dose. IMRT techniques can accomplish these dose profiles more readily than traditional methods of radiation therapy.

B. Particle radiotherapy

XI. Moving toward the future in veterinary radiotherapy

1. Neutron therapy The main potential advantage for neutron therapy is the radiobiologic advantage they confer. Due to the characteristics of their dose deposition, they cause less sublethal damage to cells that can be repaired. Additionally, the extent of cell killing by neutrons is not cell cycle dependant and is not decreased by the presence of hypoxia. As neutrons are not charged particles they do not cause direct ionization and are also likely to pass through tissue without interacting. One approach is the use of boronated compounds to capture neutrons in the tumor itself. The concept of â&#x20AC;&#x153;boron-neutron capture therapyâ&#x20AC;? is also an area where significant advances may still be made. There are very few facilities capable of treating veterinary patients with a neutron beam although there is a recent report of boron neutron capture therapy for feline nasal squamous cell carcinoma (Trivillin et al, 2007).

A. Advanced imaging The future of radiation therapy is intimately tied to technology driven and image guided techniques inside and outside of the linear accelerator vault. The ability to accurately delineate target volumes based on physiologic imaging tests represents an area where significant advancement can be made. The ability to register and fuse multiple imaging modalities in combination with immobilization devices reduces both the CTV and PTV by reducing geometric uncertainties. These improvements will incorporate computed tomography (CT), magnetic resonance imaging (MRI), ultrasound (U/S), positron emission tomography and PET/CT, single-positron emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI) and magnetic resonance spectroscopic imaging (MRSI). For 30 years, CT has been the most commonly used imaging technique for radiation treatment planning due to its spatial fidelity and the ability to reconstruct images into a three-dimensional model (Smith et al, 2006). Alternative imaging techniques including MRI and PET may provide superior visualization of tumor and normal structures. Fusion of CT and MRI images is common in veterinary medicine in treatment of brain, spinal and nasal tumors to

2. Proton therapy The main potential advantage for protons is that as a proton passes through tissue, the dose deposited increases slowly with depth until it reaches a sharp peak at its maximum depth of penetration, known as the Bragg peak. No dose is delivered beyond this peak. The depth of the peak can be adjusted by varying the energy of the beam or through the use of compensating material and specialized filters. This property makes it possible to create a dose 20

Cancer Therapy Vol 6-A.2, page 21 profile that is precisely confined to a tumor volume with extremely sharp dose fall off deep to the tumor in normal tissue. There are only scattered reports of proton therapy in veterinary medicine, including the treatment of brain tumors. The technique and dose distributions have also been evaluated for canine nasal tumors and may have some dosimetric advantages to photons (Kaser-Hotz et al, 2002; Bley et al, 2005).

tumors (Lester et al, 2001; Farese et al, 2004). Many other potential applications exist and there are anecdotal reports of effective SRS treatments for urinary bladder and urethral transitional cell carcinoma, pituitary dependant Cushingâ&#x20AC;&#x2122;s disease, and spinal tumors.

XII. Radiation training and expertise As radiation therapy becomes more complex and technology driven, the role of the medical physicist and board certified radiation oncologist has become far more important. Because the specialty of veterinary radiation oncology is still relatively new and small, there are many facilities that rely on radiologists, medical oncologists, general practitioners, and human radiotherapists to perform radiation. Implementation of more complex techniques with appropriate care requires more advanced training and better knowledge of radiation equipment, radiation safety, radiation physics, and radiobiology. The role of the medical physicist includes calibration of the machine, collection of dosing data, assuring radiation safety and assisting in treatment planning and assuring that dose delivered is the dose planned. The law only requires yearly testing of the equipment in California for veterinary facilities. In human facilities the requirements are much more strict, with a physicist on-site or on call at all times. As veterinary facilities begin to implement treatments that have similar complexity, regular quality assurance evaluations by a medical physicist are needed to prevent serious errors. Similarly, as target margins become smaller with the goals of decreasing toxicity and dose escalation, the risks associated with subtle errors in calculation and treatment become much greater. Having an on-site radiation oncologist to evaluate patients and treatment sites to develop an integrative plan for imaging and therapy is important. Even more critically, the radiation oncologist is responsible for verifying accurate positioning and port films before a treatment is administered. Because the consequences of even small errors are markedly increased with finely conformal therapy, the importance of such verification is greatly amplified.

C. Tomotherapy Tomotherapy represents a specific type of advanced IMRT with IGRT. It is a helical machine employing a linear accelerator mounted into a CT style gantry that rotates as a fan beam modulated by an MLC and rotating around the patient as the couch moves the patient into the gantry. Tomotherapy represents an integrated system for treatment planning, patient positioning, and treatment delivery utilizing an inverse planning system. Using this type of image registration for patient positioning provides more detail for not only the 3 translational degrees of freedom but also the 3 rotational degrees of freedom (pitch, roll, and yaw) (Forrest et al, 2004; Hong et al, 2007). The tomotherapy unit is also capable of verifying treatment delivery using a detector to compute the energy fluence delivered with any error outside of the specified tolerance range triggering treatment shutdown. This information can also be used for dose reconstruction to compare the planned dose delivery to the actual delivery for quality assurance and adaptive radiation therapy (Lawrence and Forrest, 2007; Tome et al, 2007).

D. Stereotactic radiosurgery SRS was originally designed by a neurosurgeon with the intent to treat functional disorders of the brain. It is now recognized as a treatment with many indications including tumors, vascular lesions, and pain syndromes, including but not exclusive to the brain and spine (Smith et al, 2006). There are three basic techniques currently available for stereotactic radiosurgery. The underlying principle is the use of many finely collimated beams focused on a specific (and usually small) target to deliver a very high dose to the small target with much lower doses to all surrounding tissue. Initially, all stereotactic radiosurgical treatments were performed with a single or very few fractions. Fractionated SRS is now preferred for many conditions (Smith et al, 2006). Linear accelerator based SRS uses stereotactic cones to modify and collimate the beam of a standard linear accelerator. The gantry rotates similar to 3-D conformal therapy or IMRT to multiple positions. Gamma knife based radiosurgery utilizes up to 201 small Cobalt-60 sources in a heavily shielded apparatus. The third technique is robotically controlled radiosurgery (Cyberknife) which employs a small linear accelerator mounted on a robotic arm to deliver multiple beams. Currently, the few veterinary facilities performing SRS are using primarily linear accelerator based treatments although there is currently one veterinary facility with a Cyberknife. Published veterinary applications for SRS include an alternative to surgery for appendicular osteosarcoma and treatment of intracranial

References Bernier J, Hall EJ, Giaccia A (2004) Radiation oncology: a century of achievements. Nat Rev Cancer 4, 737-747. Bley CR, Sumova A, Roos M, Kaser-Hotz B (2005) Irradiation of brain tumors in dogs with neurologic disease. J Vet Intern Med 19, 849-854. Bortfeld T (2006) IMRT: a review and preview. Phys Med Biol 51, R363-379. Farese JP, Milner R, Thompson MS, Lester N, Cooke K, Fox L, Hester J, Bova FJ (2004) Stereotactic radiosurgery for treatment of osteosarcomas involving the distal portions of the limbs in dogs. J Am Vet Med Assoc 225, 1567-1572, 1548. Forrest LJ, Mackie TR, Ruchala K, Turek M, Kapatoes J, Jaradat H, Hui S, Balog J, Vail DM, Mehta MP (2004) The utility of megavoltage computed tomography images from a helical tomotherapy system for setup verification purposes. Int J Radiat Oncol Biol Phys 60, 1639-1644. Giraud P, Yorke E, Jiang S, Simon L, Rosenzweig K, Mageras G (2006) Reduction of organ motion effects in IMRT and

Gordon and Kent: Veterinary radiation oncology conformal 3D radiation delivery by using gating and tracking techniques. Cancer Radiother 10, 269-282. Green EM, Forrest LJ, Adams WM (2003) A vacuum-formable mattress for veterinary radiotherapy positioning: comparison with conventional methods. Vet Radiol Ultrasound 44, 476479. Hall EJ (2006) Intensity-modulated radiation therapy, protons, and the risk of second cancers. Int J Radiat Oncol Biol Phys 65, 1-7. Hall EJ, Giaccia AJ, (2006) Radiobiology for the radiologist. Lippincott Williams & Wilkins, Philadelphia. Hong TS, Welsh JS, Ritter MA, Harari PM, Jaradat H, Mackie TR, Mehta MP (2007) Megavoltage computed tomography: an emerging tool for image-guided radiotherapy. Am J Clin Oncol 30, 617-623. Kaser-Hotz B, Sumova A, Lomax A, Schneider U, Klink B, Fidel J, Blattmann H (2002) A comparison of normal tissue complication probability of brain for proton and photon therapy of canine nasal tumors. Vet Radiol Ultrasound 43, 480-486. Keall P, Vedam S, George R, Bartee C, Siebers J, Lerma F, Weiss E, Chung T (2006) The clinical implementation of respiratory-gated intensity-modulated radiotherapy. Med Dosim 31, 152-162. Kippenes H, Gavin PR, Sande RD, Rogers D, Sweet V (2000) Comparison of the accuracy of positioning devices for radiation therapy of canine and feline head tumors. Vet Radiol Ultrasound 41, 371-376. Lawrence JA, Forrest LJ (2007) Intensity-modulated radiation therapy and helical tomotherapy: its origin, benefits, and potential applications in veterinary medicine. Vet Clin North Am Small Anim Pract 37, 1151-1165; vii-iii. Lester NV, Hopkins AL, Bova FJ, Friedman WA, Buatti JM, Meeks SL, Chrisman CL (2001) Radiosurgery using a stereotactic headframe system for irradiation of brain tumors in dogs. J Am Vet Med Assoc 219, 1562-1567, 1550. McEntee MC (2004) A survey of veterinary radiation facilities in the United States during 2001. Vet Radiol Ultrasound 45, 476-479. McEntee MC (2006) Veterinary radiation therapy: review and current state of the art. J Am Anim Hosp Assoc 42, 94-109. McEntee MC (2008) Portal Radiography in Veterinary Radiation Oncology: Options and Considerations. Vet Radiol Ultrasound 49, S57-S61.

Purdy JA (2004) Current ICRU definitions of volumes: limitations and future directions. Semin Radiat Oncol 14, 27-40. Purdy JA (2007) From new frontiers to new standards of practice: advances in radiotherapy planning and delivery. Front Radiat Ther Oncol 40, 18-39. Rohrer Bley C, Blattmann H, Roos M, Sumova A, Kaser-Hotz B (2003) Assessment of a radiotherapy patient immobilization device using single plane port radiographs and a remote computed tomography scanner. Vet Radiol Ultrasound 44, 470-475. Smith RP, Heron DE, Huq MS, Yue NJ (2006) Modern radiation treatment planning and delivery--from Rontgen to real time. Hematol Oncol Clin North Am 20, 45-62. Tome WA, Jaradat HA, Nelson IA, Ritter MA, Mehta MP (2007) Helical tomotherapy: image guidance and adaptive dose guidance. Front Radiat Ther Oncol 40, 162-178. Trivillin VA, Heber EM, Rao M, Cantarelli MA, Itoiz ME, Nigg DW, Calzetta O, Blaumann H, Longhino J, Schwint AE (2007) Boron neutron capture therapy (BNCT) for the treatment of spontaneous nasal planum squamous cell carcinoma in felines. Radiat Environ Biophys. Webb S (2006) Motion effects in (intensity modulated) radiation therapy: a review. Phys Med Biol 51, R403-425. Xing L, Thorndyke B, Schreibmann E, Yang Y, Li TF, Kim GY, Luxton G, Koong A (2006) Overview of image-guided radiation therapy. Med Dosim 31, 91-112.

From left to right: Ira K. Gordon, Michael S. Kent

Cancer Therapy Vol 6-A.2, page 23 Cancer Therapy Vol 6, 177-180, 2008

Canine cancer genetics: transitional cell carcinoma in Scottish Terriers Review Article

Steven E. Crow VCA Sacramento Veterinary Referral Center and VCA Highlands Animal Hospital, CA, USA

__________________________________________________________________________________ *Correspondence: Steven E. Crow, DVM. VCA Sacramento Veterinary Referral Center, 9801 Old Winery Place, Sacramento, CA 95827, USA; Tel: (916) 362-3111 VCA Highlands Animal Hospital, Cancer Treatment Center, 3451 Elkhorn Blvd, North Highlands, CA 95660, USA; Tel: (916) 3322845; e-mail: steve.crow@vcamail.com Key words: Canine cancer, Scottish Terriers, Biological and epidemiological factors Abbreviations: Cyclooxygenase, (COX); cyclooxygenase-1, (COX1); cyclooxygenase-2, (COX2); transitional cell carcinoma, (TCC) Received: 10 March 2008; electronically published: June 2008

Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary Scottish Terriers have a markedly increased risk of developing transitional cell carcinoma. Although environmental and occupational hazards may play a role in the pathogenesis of TCC in humans and dogs, the predominance of evidence supports a genetic basis for the predisposition in this breed. Careful study of families and specific lines of Scotties should help to identify specific genes or gene clusters that predispose to malignant transformation in urinary epithelial cells. Development of clinically useful tools to identify dogs at risk prior to breeding is a logical and desirable goal.

Treatment, usually consisting of cytoreductive surgery, cytotoxic chemotherapy, and cyclooxygenase inhibitors, is often helpful in palliating clinical signs and prolonging life (Knapp et al, 1992, 1994, 2000; Chun et al, 1996, 1997; Henry et al, 2003). Researchers at the Purdue Comparative Oncology Program have demonstrated measurable reduction in tumor volume in 12 of 18 pet dogs with naturally occurring invasive TCC of the urinary bladder treated with piroxicam. Tumor reduction was strongly associated with induction of apoptosis and reduction in urine basic fibroblast growth factor concentration (Mohammed et al, 2002). In an earlier study, the same group treated 34 dogs with histologicallyconfirmed, measurable, non-resectable TCC of the bladder with piroxicam. After 8 weeks, they reported two complete remissions, four partial remissions, 18 dogs with stable disease, and 10 dogs showed no response (overall response rate = 17%) (Knapp et al, 1994). Henry and colleagues reported in 2003 an overall response rate of 35% (one complete response and 16 partial responses) in 55 dogs with bladder TCC treated with mitoxantrone and piroxicam (Henry et al, 2003).

I. Introduction Transitional cell carcinoma (TCC) is the most common neoplasm of the canine urinary system, comprising approximately 1.5%-2% of all canine malignancies (Hayes 1976; Burnie and Weaver 1983; Norris et al, 1992). Greater than 90% of canine bladder cancer is intermediate to high-grade TCC (Burnie and Weaver 1983; Kahn et al, 2000). Mean age at diagnosis is 11 years (Burnie and Weaver 1983; Hayes 1976; Knapp 2001). TCC is usually an invasive, progressive and ultimately fatal cancer, resulting in death due to post-renal obstruction within 3-12 months of diagnosis (Norris et al, 1992; Knapp 2001). Gross metastatic disease is present in 15-20% of dogs at the time of diagnosis, but more than 50% have metastases in regional lymph nodes and/or lungs at the time of death (Osborne et al, 1968; Burnie and Weaver 1983; Walter et al, 1984). Clinical signs of TCC include incontinence, strangury, pollakiuria, and hematuria. In advanced cases, signs of renal failure (vomiting, anorexia, dehydration) may occur secondary to urethral or ureteral obstruction. Diagnosis is greatly aided by contrast cystography (Figure 1) and abdominal sonography (Figure 2).

Crow: Canine cancer genetics: transitional cell carcinoma in Scottish Terriers

Figure 1. Contrast cystogram demonstrating a mass in the dorsal bladder wall of a dog (courtesy of NCSU website).

Figure 2. Sonographic image of a large bladder wall TCC (courtesy of NCSU website).

II. Biological factors

phenoxy and possibly non-phenoxy herbicides, or herbicides in combination with insecticides, but not with insecticides alone (Raghavan et al, 2004). In human TCC, aneuploidy correlates with advanced grade and stage. Using flow cytometric DNA ploidy analysis, Clemo and colleagues found aneuploidy in 79% of DNA histograms from 43 canine TCC samples; however, no significant correlation was found between the presence of aneuploidy and age, sex, survival time, growth pattern, stage, grade, or morphology of the neoplasm. In contrast, all but one of 23 samples from normal or inflamed/hyperplastic bladder mucosa were diploid (Clemo et al, 1994). Cyclooxygenase (COX) activity, and response of TCC to cyclooxygenase-1 (COX1) and cyclooxygenase-2 (COX2) inhibitors, is one of the most studied characteristics of human and canine TCC (Kahn et al, 2000, Komhoff et al, 2000, Mohammed et al, 1999, Mohammed et al, 2002). Komhoff demonstrated elevated

and

epidemiological

The relative risk of different dog breeds is quite variable, but Scottish Terriers (Figure 3) appear to have the greatest risk of developing TCC (odds ratio =18/09; 95%CI = 7.3 - 44.86) compared to mixed breed dogs (Glickman et al, 2000). Other breeds with increased risk of TCC include West Highland White terriers, Wire Hair Fox terriers, Shetland Sheepdogs (Knapp, 2001), and Beagles (Nikula et al, 1989). Breed-associated risk may be due to differences in pathways that activate or detoxify carcinogens. In human beings, several environmental risk factors for TCC have been identified, including exposure to aniline dyes and smoking cigarettes (Hartge, 1987). Similarly, studies have been conducted to identify risk factors for canine bladder cancer. In 1981, Hayes and colleagues found a correlation between TCC and industrial activity in both humans and dogs, suggesting that environmental pollution may play a role in the pathogenesis of TCC (Hayes et al, 1981, Knapp et al, 2000). Macy described two dogs that developed TCC 6 and 13 weeks after starting treatment with cyclophosphamide (Macy et al, 1983). Acrolein, a metabolite of cyclophosphamide, is known to cause severe chemical cystitis in dogs and humans. Glickman and colleagues concluded in two casecontrol studies of several different dog breeds that risk of developing TCC was higher for dogs exposed to marshes that had been (1) sprayed for mosquito control or (2) exposed to topical insecticides for flea and tick control. In addition, obese dogs that received more than two flea and tick treatments per year had greatly increased odds of developing TCC (odds ratio = 24.5, 95%CI = 1.4 - 43.2) compared to non-obese dogs not treated with topical insecticides (Glickman et al, 1989, 2000). Raghavan and colleagues reported in a case control study in Scottish Terrier dogs that risk of TCC was significantly increased with exposure to lawn and garden

Figure 3. Photograph of a Scottish Terrier (courtesy of Scottish Terrier Club of America).

Cancer Therapy Vol 6-A.2, page 25

Figure 4. Partial Pedigree of a Scottish Terrier Transitional Cell Carcinoma Cluster.

expression of COX2 in 17 of 47 (38%) high-grade human bladder carcinomas, but COX2 immunoreactivity was absent in benign bladder mucosa and all 23 low-grade urothelial carcinomas tested (Komhoff et al, 2000). Mohammed and colleagues reported similar findings using immunohistochemistry and western blot analysis. COX2 was not expressed in benign bladder tissue but was detected in 25 of 29 human invasive bladder TCC (86%) and in 6 of 8 cases (75%) of carcinoma in situ (Mohammed et al, 1999). COX2 is overexpressed in the invasive portions of TCC of fishing cats (Prionailurus viverrinus) (Landolfi and Terio 2006). The chemopreventative effects of COX inhibitors, and in particular COX2 inhibitors, in chemically induced bladder tumors of rodents has been documented (Rao et al, 1996; Grubbs et al, 2000).

(Figure 4). The dam was diagnosed at ten years of age, her male offspring were diagnosed at six and eight years of age, and her female progeny were diagnosed at six, six, and eleven years of age. The sire of the first litter was diagnosed with bladder cancer (TCC) at eight years of age, and the second sire died of lymphoma at approximately six years of age. The breeder/owner reports that none of the dogs were exposed to herbicides or topical insecticides. None of the dogs have ever been overweight. Serum samples have been submitted to NIH for genetic screening. It will be interesting to see if a specific gene or gene mutation can be identified in this breed and, in particular, in this family. A fact is the point at which curiosity ends.â&#x20AC;? --- author unknown

References

III. Discussion

Burnie AG, Weaver AD (1983) Urinary bladder neoplasia in the dog: a review of seventy cases. J Small Anim Pract 24, 129-143. Chun R, Knapp DW, Widmer WR, DeNicola DB, Glickman MW, Kuczek T, DeGortari A, Han CM (1997) Phase II clinical trial of carboplatin in canine transitional cell carcinoma of the urinary bladder. J Vet Intern Med 11, 279283. Chun R, Knapp DW, Widmer WR, Glickman MW, DeNicola DB, Bonney PL (1996) Cisplatin treatment of transitional cell carcinoma of the urinary bladder in dogs: 18 cases (1983-1993). J Am Vet Med Assoc 209, 1588-1591. Clemo FA, DeNicola DB, Carlton WW, Morrison WB, Walker E (1994) Flow cytometric DNA ploidy analysis in canine transitional cell carcinoma of urinary bladders. Vet Pathol 31, 207-215. Glickman LT, Raghavan M, Knapp DW, Bonney PL, Dawson MH (2004) Herbicide exposure and the risk of transitional

In 2005, the Scottish Terrier Club of America conducted a health survey. Four hundred eighty-seven owners / breeders responded, reporting on approximately 1375 Scotties. Medical conditions or illnesses reported included elevated liver enzymes of unknown origin (n=102), lymphoma (n=30), and transitional cell carcinoma of the urethra and urinary bladder (n=65). Recent work by the Purdue group and NIH researchers have attempted to identify specific differences in the Scottish Terrier genome that may contribute to the breedâ&#x20AC;&#x2122;s increased risk for TCC (Ostrander and Giniger, 1997). The author has just this year become aware of a cluster household in central California, where a Scottish Terrier bitch and five of her puppies (all three siblings from one sire, and two other dogs from a separate sire) have been diagnosed with TCC within a two year period

Crow: Canine cancer genetics: transitional cell carcinoma in Scottish Terriers cell carcinoma of the urinary bladder in Scottish Terriers. J Am Vet Med Assoc 224, 1290-1297. Glickman LT, Schofer PS, McKee LI, Reif JS, Goldschmidt MH (1989) Epidemiology study of insecticide exposures, obesity, and risk of bladder cancer in household dogs. J Toxicol Environ Health 28, 407-414. Grubbs CJ, Lubet RA, Koki AT, Leahy KM, Masferrer JL, Steele VE, Kelloff GJ, Hill DL, Seibert K (2000) Celecoxib inhibits N-butyl-n-(4-hydroxybutyl)-nitrosamine-induced urinary bladder cancers in male B6D2F1 mice and female Fischer-344 rats. Cancer Res 60, 5599-5602. Hartge P, Silverman D, Hoover R, Schairer C, Altman R, Austin D, Cantor K, Child M, Key C, Marrett LD (1987) Changing cigarette habits and bladder cancer risk: a case-control study. J Natl Cancer Inst 78, 1119-1125. Hayes HM (1976) Canine bladder cancer: epidemiologic features. Am J Epidemiol 104, 673-677. Hayes HM, Hoover R, Tarone R (1981) Bladder cancer in pet dogs: a sentinel for environmental cancer? J Epidemiol 114, 229-233. Henry CL McCaw DL, Turnquist SE, Tyler JW, Bravo L, Sheafor S, Straw RC, Dernell WS, Madewell BR, Jorgensen L, Scott MA, Higginbotham ML, Chun R (2003) Clinical evaluation of mitoxantrone and piroxicam in a canine model of human invasive urinary bladder carcinoma. Clin Cancer Res 9, 906-911. Khan KNM, Knapp DW, DeNicola DB, Harris K (2000) Expression of cyclooxygenase-2 in transitional cell carcinoma of the urinary bladder in dogs. Am J Vet Res 61, 478-481. Knapp DW (2001) Tumors of the urinary system. In Withrow SJ, MacEwen EC (eds): Small Animal Clinical Oncology, 3rd ed, Philadelphia, WB Saunders, pp 490-499. Knapp DW, Glickman NW, DeNicola DB, Bonney PL, Lin TL, and Glickman LT (2000) Naturally-occurring canine transitional cell carcinoma of the urinary bladder. A relevant model of human invasive bladder cancer. Urol Oncol 5, 4759. Knapp DW, Glickman NW, Widmer WR, DeNicola DB, Adams LG, Kuczek T, Bonney PL, DeGortari A, Han CM, Glickman LT (2000) Cisplatin versus cisplatin combined with piroxicam in a canine model of human invasive urinary bladder cancer. Cancer Chemother Pharmacol 46, 221226. Knapp DW, Richardson RC, Bottoms GD, Teclaw R, Chan TC (1992) Phase I trial of piroxicam in 62 dogs bearing naturally occurring tumors. Cancer Chemother Pharmacol 29, 214218. Knapp DW, Richardson RC, Chan TC, Bottoms GD, Widmer WR, DeNicola DB, Teclaw R, Bonney PL, Kuczek T (1994) Piroxicam therapy in 34 dogs with transitional cell carcinoma of the urinary bladder. J Vet Intern Med 8, 273-278. Komhoff M, Guan Y, Shappell HW, Davis L, Jack G, Shyr Y, Koch MO, Shappell SB, Breyer MD (2000) Enhanced expression of cyclooxygenase-2 in high-grade human transitional cell bladder carcinomas. Am J Pathol 157, 2935. Landolfi JA, Terio KA (2006) Transitional cell carcinoma in fishing cats (Prionailurus viverrinus): pathology and expression of cyclooxygenase-1, -2, and p53. Vet Pathol 43, 674-681.

Macy DW, Withrow SJ, Hoopes J (1983) Transitional cell carcinoma of the bladder associated with cyclophosphamide administration. J Amer Anim Hosp Assoc 19, 965-969. Mohammed SI, Bennett PF, Craig BA, Glickman NT, Mutsaers AJ, Snyder PW, Widmer WR, DeGortari AE, Bonney PL, Knapp DW (2002) Effects of the cyclooxygenase inhibitor, piroxicam, on tumor response, apoptosis, and angiogenesis in a canine model of human invasive urinary bladder cancer. Cancer Res 62, 356-358. Mohammed SI, Knapp DW, Bostwick DG, Foster RS, Kahn KN, Masferrer JL, Woerner BM, Snyder PW, Koki AT (1999) Expression of cyclooxygenase-2 (cox-2) in human invasive transitional cell carcinoma of the urinary, bladder. Cancer Res 59, 5647-5650. Nikula KJ, Benjamin SA, Angleton GM, Lee AC (1989) Transitional cell carcinomas of the urinary tract in a colony of beagle dogs. Vet Pathol 26, 455-461. Norris AM, Laing EJ, Valli VE, Withrow SJ, Macy DW, Ogilvie GK, Tomlinson J, McCaw D, Pidgeon G, Jacobs RM (1992) Canine bladder and urethral tumors: A retrospective study of 115 cases (1980-1985). J Vet Intern Med 6, 145-153. Osborne CA, Low DG, Perman V, Barnes DM (1968) Neoplasms of the canine and feline urinary bladder: incidence, etiologic factors, occurrence and pathologic features. Am J Vet Res 29, 2041-2055. Ostrander EA, Giniger E (1997) Insights from model systems. Semper fidelis: What man's best friend can teach us about human biology and disease. Am J Hum Genet 61, 475-480: Raghavan M, Knapp DW, Dawson MH, Bonney PL, Glickman LT (2004) Topical flea and tick pesticides and the risk of transitional cell carcinoma of the urinary bladder in Scottish Terriers. J Am Vet Med Assoc 225, 389-394. Rao KV, Detrisac CJ, Steele VE, Hawk ET, Kelloff GJ, McCormick DL (1996) Differential activity of aspirin, ketoprofen and sulindac as cancer chemopreventive agents in the mouse urinary bladder. Carcinogenesis 17, 1435-1438. Walter PA, Haynes JS, Feeney DA, Johnston GR (1984) Radiographic appearance of pulmonary metastases from transitional cell carcinoma of the bladder and urethra of the dog. J Am Vet Med Assoc 185, 411-418.

Steven E. Crow

Cancer Therapy Vol 6-A.2, page 27 Cancer Therapy Vol 6, 181-186, 2008

Chemoimmunotherapy for canine lymphoma: tumor vaccines and monoclonal antibodies Review Article

Steven E. Crow VCA Sacramento Veterinary Referral Center and VCA Highlands Animal Hospital, CA, USA

__________________________________________________________________________________ *Correspondence: Steven E. Crow, DVM. VCA Sacramento Veterinary Referral Center, 9801 Old Winery Place, Sacramento, CA 95827, USA; Tel: (916) 362-3111 VCA Highlands Animal Hospital, Cancer Treatment Center, 3451 Elkhorn Blvd, North Highlands, CA 95660, USA; Tel: (916) 3322845; e-mail: steve.crow@vcamail.com Key words: Chemoimmunotherapy, canine lymphoma tumor vaccines, canine lymphoma tumor vaccines, Canine Abbreviations: complete response, (CR); monoclonal antibodies, (MAb); non-Hodgkin's lymphoma, (NHL); overall response rate, (ORR); radioimmunoconjugates, (RICs); radioimmunotherapy, (RIT) Received: 10 March 2008; electronically published: June 2008

Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary Despite numerous clinical trials with drugs that reliably induce remission, cure for most cases of canine lymphoma has continued to elude veterinary oncologists over the last 40 years. Results of various chemotherapy protocols are remarkably similar, and no breakthrough drugs have been discovered. Although trials employing biological response modifiers have been few in number and small in size, results of those studies are reason for optimism. Favorable outcomes with tumor vaccines in canine lymphoma and monoclonal antibodies in human non-Hodgkin lymphoma support the need for additional studies of immunotherapeutic interventions in this common and devastating disease.

II. Review of results A. Canine

I. Introduction Lymphoma is the most common neoplasm of the canine hemolymphatic system. It represents approximately 4.5% of all canine neoplasms and 15% of all malignant neoplasms. Canine lymphoma (CL) is usually rapidly fatal, resulting in death within one to three months of diagnosis (Squire et al, 1973). Temporary remission of clinical signs without treatment is rare. Most CL cases are high or intermediate histologic grade; less aggressive, lowgrade lymphoma represents less than 5% of all CL cases reported (Squire et al, 1973; Schwartz 1988; Rosenberg 1991; Teske et al, 1994). Treatment of CL has been a topic of great interest for veterinary oncologists for almost 40 years. The principal mode of medical management has been chemotherapy (Rosenthal 1990; Jeglum and Steplewski, 1996) but various attempts at immunotherapy (biological response modification) have produced results that rival or surpass outcomes with drugs alone (Table 1) (Crow et al, 1977, 1996; Theilen et al, 1977; Weller et al, 1980; Jeglum et al, 1986, 1988; Jeglum and Steplewski, 1996).

In the early 1970s, Benjamini and others demonstrated significant delay in recurrence and progression of tumors in laboratory mice treated with surgery and chemically-modified tumor cell vaccine compared to mice treated with surgical excision only (Thompson et al, 1972; Benjamini, Scibienski 1974; Benjamini et al, 1976). Subsequently, Theilen and Worley conducted a preliminary clinical investigation in which 20 dogs given a similar vaccine had markedly improved mean survival times (341 days) versus 47 dogs treated with chemotherapy only (138 days) (Theilen et al, 1977). From 1974 through 1977, our research team, led by Dr. Gordon Theilen, completed two prospective, randomized clinical trials (Crow et al, 1977, Weller et al, 1980) in which we compared dogs treated for multicentric lymphoma with combination chemotherapy, followed by injections of placebo or autogenous tumor vaccine. In the first study, dogs with lymphoma had a single lymph node excised prior to receiving a nine-week combination chemotherapy protocol. Dogs that achieved and 27

Crow: Chemoimmunotherapy for canine lymphoma: tumor vaccines and monoclonal antibodies maintained complete remission throughout induction were then injected intramuscularly with either tumor vaccine, consisting of acetoacetylated tumor cell wall proteins suspended in complete Freund's complete adjuvant (Figure 1), or placebo. Individual vaccine was produced for each test dog from its own lymphoma cells. Tumor cell membrane protein alterations and immunoadjuvant were designed to abrogate blocking antibody formation and to induce cell-mediated immunity (Harris and Copeland, 1964; Eilber and Morton, 1970; Smith and Adler, 1970; Sjogren et al, 1971; Currie and Basham, 1972; Thompson

et al, 1972; Prager and Baechtel, 1973; Benjamini and Scibienski, 1974; Witney et al, 1974; Benjamini et al, 1976; Rosenthal and MacEwen, 1990). Dogs receiving vaccine achieved median first remission and overall survival times of 132 and 336 days, respectively, compared to 91 and 196 for dogs receiving placebo injections (Figures 2, 3) (Crow et al, 1977). Weller and colleagues later reported similar results in 32 dogs treated with chemotherapy and various components of the vaccine (median first remission = 136 days; median survival = 334 days) (Weller et al, 1980).

Table 1. Comparison of chemotherapy only to chemoimmunotherapy protocols for treatment of intermediate and high grade canine lymphoma. Reference/Study Theilen et al, 1977 Crow et al, 1977 Weller et al, 1980 Jeglum et al, 1988 Jeglum, Steplewski 1996 Crow et al, 1996

Chemotherapy only Median survival (days) 138 (n=47) 196 (n=9) NR 180 (n=30) NR NR

Chemoimmunotherapy Median survival (days) 341(n=20) 336 (n=12) 334 (n=32) 305 (n=56) 410 (n=215) 301 (n=85)

NR = not reported; historical control used for comparison. Using intralymphatic autochthonous tumor cell vaccine after remission induction with combination chemotherapy, Jeglum achieved a median survival time of 305 days in 56 dogs with lymphoma, compared to 180 days in a control group (n=30) that received only eight weeks of chemotherapy (Jeglum et al, 1988). In the mid-1980s, using the novel hybridoma technology of Kohler and Milstein (Figure 4), several investigators set out to produce monoclonal antibodies (MAb) that could be used for the detection, classification, and treatment of canine lymphoma. In contrast to the nonspecific nature of most chemotherapy, MAb bind with high specificity to cell-surface antigens, resulting in targeted killing of malignant cells, relative sparing of normal tissues, and low toxicity. Using an established canine lymphoma cell line, Jeglum and collaborators at the Wistar Institute produced a panel of murine monoclonal antibodies. They identified a specific monoclonal antibody (CL/MAb 231) that selectively bound to formalin-fixed, paraffin-embedded tumor tissue from 75% of dogs with lymphoma. In addition, this antibody demonstrated cytotoxicity against lymphoma cells in vitro and in vivo (Steplewski et al, 1987; Rosales et al, 1988; Jeglum and Steplewski al 1996). After a Phase I trial demonstrated no significant toxicity, a clinical trial of 215 previously untreated dogs with lymphoma was conducted. An immunoperoxidase assay for CL/MAb 231 binding was performed retrospectively on biopsy specimens from 129 of the dogs in that trial. Complete remission induction rate was 80.5%. Overall median survival was 410 days. Median survival times for non-responders (n=41) and dogs achieving remission and subsequently treated with CL/MAb 231 (n=174) were 113 days and 493 days, respectively (Jeglum and Steplewski, 1996).

Figure 1. Schematic representation of method used to produce autogenous canine lymphoma vaccine. Reproduced from Crow et al, 1977 with kind permission from Cancer.

Cancer Therapy Vol 6-A.2, page 29

Figure 2. First remission duration for dogs receiving chemotherapy and placebo compared to dogs treated with chemotherapy and autogenous tumor vaccine. Reproduced from Crow et al, 1977 with kind permission from Cancer.

Figure 3. Survival duration for dogs receiving chemotherapy and placebo compared to dogs treated with chemotherapy and autogenous tumor vaccine. Reproduced from Crow et al, 1977 with kind permission from Cancer.

A prospective multi-institutional study was conducted to evaluate the effectiveness of CL/MAb 231 with alternative chemotherapy protocols (Crow et al, 1996). Between June 1992 and January 1994, dogs with lymphoma were randomized to receive combination chemotherapy or single-agent doxorubicin, followed by immunotherapy (CL/MAb 231). Tissue was submitted for immunoperoxidase binding assay on 65 of the 87 dogs admitted to the trial. Dogs in the doxorubicin only group achieved complete remission much less often than the combination chemotherapy dogs, but if remission was accomplished, remission and survival durations were not significantly different. Overall median survival was 301 days (range = 8 -1022 days); median survival times for the doxorubicin and combination chemotherapy groups were 259 (range = 1 - 630 days) and 335 days (range = 8 1022), respectively. Written reports for immunoperoxidase

binding were provided for only 25 cases: 20 were strongly positive, 3 were weakly positive, and 2 were negative. Because of the small number of non-binders, statistical analysis was not done. Interestingly, one of the nonbinders was the longest survivor.

B. Human Therapeutic options for human beings with NHL have improved over the past 20 years, but almost all patients with low-grade lymphoma and approximately 50% of patients with high-grade lymphoma eventually die of their disease, regardless of the regimen used. Thus, there is a continuing need for novel therapeutic options. Two such strategies are unconjugated MAb and MAb conjugated to radionuclides, i.e., targeted radioimmunotherapy (RIT).

Crow: Chemoimmunotherapy for canine lymphoma: tumor vaccines and monoclonal antibodies

Figure 4. Schematic representation of hybridoma production.

In 1997, rituximab (Rituxan, Genentech Inc, South San Francisco, California, USA, and Biogen Idec Inc, Cambridge, Massachusetts, USA) became the first MAb approved by the US Food and Drug Administration for use in the treatment of cancer, specifically B-cell nonHodgkin's lymphoma. Rituximab has become a staple in the management of B-cell non-Hodgkin's lymphoma, but it has limited activity as a single agent, with responses in about half of recurrent follicular and low-grade lymphoma patients (Juweid, 2002; Forero and Lobuglio, 2003; Horning, 2003; Marcus, 2005; Witzig, 2006). In hopes of using rituximab in the treatment of canine lymphoma, Impellizeri et al, evaluated canine B-cell binding and depletion by rituximab using flow cytometry. Despite immunohistochemistry demonstration of CD20 expression, rituximab did not bind or deplete canine B cells ex vivo. They concluded that rituximab is unlikely to be effective in the treatment of canine lymphoma (Impellizeri et al, 2006). RIT is a particularly attractive approach for B-cell lymphoma because CD20 affords an outstanding target and lymphoma cells are inherently radiosensitive. Both efficacy and safety of RIT have been established in the treatment of relapsed or refractory indolent non-Hodgkin's lymphoma (NHL) (Juweid, 2002; Forero and Lobuglio 2003; Horning 2003; Marcus, 2005; Witzig, 2006). The two most commonly used MAb radioimmunoconjugates (RICs), ibritumomab tiuxetan (Zevalin, Biogen Idec Inc, San Diego, California, USA, and Schering AG, Berlin, Germany) and tositumomab (Bexxar, GlaxoSmithKline, Brentford, Middlesex, United Kingdom) target the CD20 antigen on B-cells. The former was the first radioimmunotherapy agent to be approved by the US Food and Drug Administration for the treatment of patients with

relapsed, low-grade B-cell NHL. It is comprised of the murine IgG1 anti-CD20 antibody ibritumomab covalently linked to the beta-emitter yttrium-90 (90Y) by a chelator, tiuxetan. Tositumomab is a murine IgG2a lambda monoclonal antibody covalently linked to iodine-131 (131I). Both agents have demonstrated high anti-tumor activity in patients who are refractory to rituximab (Juweid, 2002; Forero and Lobuglio 2003; Horning, 2003; Marcus, 2005; Witzig, 2006). Mechanisms of action appear to include apoptosis, complement-dependent cytotoxicity, and antibody-dependent cellular cytotoxicity. In addition, the attached radionuclide may kill tumor cells as well as adjacent normal cells from crossfire or â&#x20AC;&#x153;bystanderâ&#x20AC;? effect (Witzig, 2006). A prospective trial comparing 90Y-ibritumomab tiuxetan with single-agent rituximab showed an overall response rate (ORR) of 80% (34% complete response [CR]) for 90Y-ibritumomab tiuxetan compared with an ORR of 56% (20% CR) for rituximab (P = .002) (Marcus, 2005). Of patients achieving a CR, 32% were still in remission at 3 to 4 years of follow-up. Similar efficacy (83% ORR, 43% CR) has been reported with 90Yibritumomab tiuxetan in patients with relapsed or refractory low-grade NHL with mild thrombocytopenia and in patients with rituximab-refractory NHL (Marcus, 2005). RIT is very well tolerated and is delivered on an outpatient basis over 1 week. The only significant toxicity is reversible myelosuppression (Marcus, 2005, Witzig, 2006). Other RICs are being investigated for the treatment of NHL, as are several immunotoxins. The role of RIT in first-line therapy of indolent NHL and in diffuse large Bcell lymphoma is still to be determined (Juweid, 2002; Forero and Lobuglio, 2003; Horning, 2003; Witzig, 2006).

Cancer Therapy Vol 6-A.2, page 31 Brooks MB, Matus RE, Leifer CE, Patnaik AK. (1987) Use of splenectomy in the management of lymphoma in dogs: 16 cases (1975-1985). J Am Vet Med Assoc 191:1008-1010. Carter RF, Harris CK, Withrow SJ, Valli VE, Susaneck SJ (1987) Chemotherapy of canine lymphoma with histopathological correlation: doxorubicin alone compared to COP as first treatment regimen. J Am Anim Hosp Assoc 23, 587-596. Cotter SM (1983) Treatment of lymphoma and leukemia with cyclophosphamide, vincristine, and prednisone: I. treatment of dogs. J Am Anim Hosp Assoc 19, 159-165. Cotter SM, Goldstein MA (1987) Comparison of two protocols for maintenance of remission of dogs with lymphoma. J Am Anim Hosp Assoc 23, 495-99. Crow SE, Rogers KS, Barton CL, Knapp DM, Morrison WE, Susaneck SJ, Jeglum KA, Raskin RE, Fox LM (1996) Veterinary Cancer Society Collaborative Clinical Trial-CLMAb231. Unpublished data. Crow SE, Theilen GH, Benjamini E, Torten M, Henness AM, Buhles WC (1977) Chemoimmunotherapy for canine lymphosarcoma. Cancer 40, 2102-2108. Currie, GA, Basham C (1972) Serum-mediated inhibition of the immunological reaction of the patient to his own tumor-a possible role for circulating antigen. Br J Cancer 26, 427430. Dobson JM, Gorman NT (1994) Canine multicentric lymphoma 2: Comparison of response to two chemotherapeutic protocols. J Small Anim Pract 35, 9-15. Eilber FR, Morton DL (1970) Impaired immunological reactivity and recurrence following cancer surgery. Cancer 25, 362367. Forero A, Lobuglio AF (2003) History of antibody therapy for non-Hodgkin's lymphoma. Semin Oncol 30, 1-5. Greenlee PG, Filippa DA, Quimby FW, Patnaik AK, Calvano SE, Matus RE, Kimmel M, Hurvitz AI, Lieberman PH (1990) Lymphoma in dogs. A morphologic, immunologic, and clinical study. Cancer 66, 480-490. Hahn KA, Richardson RC, Teclaw RF, Cline JM, Carlton WW, DeNicola DB, Bonney PL. (1992) Is maintenance chemotherapy appropriate for the management of canine malignant lymphoma? J Vet Intern Med 6, 3-10. Harris J, Copeland D (1964) Impaired immunoresponsiveness in tumor patients. Ann NY Acad Sci 120, 56-75. Horning SJ (2003) Future directions in radioimmunotherapy for B-cell lymphoma. Semin Oncol 30, 29-34. Impellizeri JA, Howell K, McKeever KP, Crow SE (2006) The role of rituximab in the treatment of canine lymphoma: an ex vivo evaluation. Vet J 171, 556-8. Jeglum KA, Steplewski Z (1996). Chemoimmunotherapy of canine lymphoma with adjuvant canine monoclonal antibody 231. Vet Clin N Amer Sm Anim Pract 26, 73-85. Jeglum KA, Young KM, Barnsley K, Whereat A (1988) Chemotherapy versus chemotherapy with intralymphatic tumor cell vaccine in canine lymphoma. Cancer 61, 20422050. Jeglum KA, Young KM, Bransley K, Whereat A, McGrath D, Hutson C (1986) Intralymphatic autochthonous tumor cell vaccine in canine lymphoma. J Bio Resp Mod 5, 168-175. Juweid ME (2002) Radioimmunotherapy of B-cell nonHodgkin's lymphoma: from clinical trials to clinical practice. J Nucl Med 43, 1507-29. Keller ET, MacEwen EG, Rosenthal RC, Helfand SC, Fox LE (1993) Evaluation of prognostic factors and sequential combination chemotherapy with doxorubicin for canine lymphoma. J Vet Intern Med 7, 289-295. Klein MK (1991) The effect of previous corticosteroid exposure on response to doxorubicin in canine lymphoma patients. Vet Can Soc Newsletter 15, 22-23.

II. Discussion Many studies of combination chemotherapy protocols for canine lymphoma have been reported in the last four decades (Brick et al, 1968; Madewell, 1972; MacEwen et al, 1981; Cotter, 1983; Brooks et al, 1987; Carter et al, 1987; Cotter and Goldstein, 1987; MacEwen et al, 1987; Rosales et al, 1988; Postorino et al, 1989; Rogers, 1989; Greenlee et al, 1990; Rosenthal, 1990; Rosenthal and MacEwen, 1990; Klein, 1991; Ogilvie et al, 1991; Price et al, 1991; Stone et al, 1991; Hahn et al, 1992; MacEwen et al, 1992; Novotney et al, 1992; Keller et al, 1993; Vail, 1993; Dobson and Gorman, 1994; Matherne et al, 1994; Moore et al, 1994; Ruslander et al, 1994), but little progress has been made in the overall survival of dogs with this common malignancy. Novel approaches, including whole body hyperthermia, halfbody radiation therapy (Laing et al, 1989), continuous low dose chemotherapy (Rosenthal, 1990; Ogilvie et al, 1991), bone marrow transplantation (Rosenthal, 1990) and nutritional intervention (Williams, 1988) have been attempted, but case accessions have been quite small in most trials. In retrospect, it is disappointing that evaluation of CL/MAb 231 by independent investigators was not completed prior to its commercial release. Questions of binding sensitivity and specificity as well as retention of efficacy with continued passage remained unanswered when production was discontinued, presumably for financial considerations. Anecdotally, during the last 34 years I have personally treated more than 1900 dogs with intermediate or high grade multicentric lymphoma (stages III-V) using various combination chemotherapy protocols. Only 18 dogs have been “cured”, i.e., survived longer than three years and died free of any signs of lymphoma. All but two of those dogs never relapsed after the initial induction chemotherapy. Interestingly, seven of the 18 dogs were treated with CL/MAb 231 and five dogs received either Freund's complete adjuvant or autogenous tumor vaccine injections. Unfortunately, the search for magic potions (drugs) has been surprisingly unfruitful over the last three decades. The documented success of therapeutic monoclonal antibodies for human non-Hodgkin’s lymphoma, clearly highlights the opportunity that was missed by veterinary oncologists. It is my hope that the new generation of veterinary oncologists will look again at immune targeting and immunomodulation as possible paths toward cure of this devastating cancer.

References Benjamini E, Scibienski RJ (1974) Immunochemical approaches to immunotherapy. Proceedings XI International Cancer Congress 1, 327-332. Benjamini E, Theilen GH, Torten J, Fong S, Crow SE, Henness AM (1976) The use of tumor vaccines in immunotherapy of canine lymphosarcoma. Ann NY Acad Sci 277, 305-312. Brick J0, Roenigk WS, and Wilson GP (1968) Chemotherapy of malignant lymphoma in dogs and cats. J Am Vet Med Assoc 1153, 47-52.

Crow: Chemoimmunotherapy for canine lymphoma: tumor vaccines and monoclonal antibodies Laing EJ, Fitzpatrick PJ, Binnington AG, Norris AM, Mosseri A, Rider WD, Binnington AG, Baur A, Valli VE (1989) Halfbody radiotherapy in the treatment of canine lymphoma. J Vet Intern Med 3, 102-108. MacEwen EG, Brown NO, Patnaik AK, Hayes AA, Passe S (1981) Cyclic combination chemotherapy of canine lymphosarcoma. J Am Vet Med Assoc 178, 1178-1181. MacEwen EG, Hayes AA, Matus RE, Kurzman I (1987) Evaluation of some prognostic factors for advanced multicentric lymphosarcoma in the dog: 147 cases (19781981). J Am Vet Med Assoc 190, 564-568. MacEwen EG, Rosenthal RC, Fox LE, Loar AS, Kurzman ID (1992) Evaluation of L-asparaginase: polyethylene glycol conjugate versus native L-asparaginase combined with chemotherapy: a randomized double-blind study in canine lymphoma. J Vet Intern Med 6, 230-234. Madewell, BR (1972) Chemotherapy for canine lymphosarcoma. Am J Vet Res 36, 1525-1528. Marcus R (2005) Use of 90Y-ibritumomab tiuxetan in nonHodgkin's lymphoma. Semin Oncol 32 (suppl), 36-43. Matherne CM, Satterfield WC, Gasparini A, Tonetti M, Astroff AB, Schmidt RD, Rowe LD, DeLoach JR (1994) Clinical efficacy and toxicity of doxorubicin encapsulated in glutaraldehyde-treated erythrocytes administered to dogs with lymphosarcoma. Am J Vet Res 55, 847-853. Moore AS, Ogilvie GK, Ruslander D, Rand WS, Cotter SM, Getzy DM, Lâ&#x20AC;&#x2122;Heureux DA, Dennis RA (1994) Evaluation of mitoxantrone for the treatment of lymphoma in dogs. J Am Vet Med Assoc 205, 1903-1905. Novotney CA, Page RL, Macy DW, Dewhirst MA, Ogilvie GK, Withrow SJ, McEntee MC, Heidner GL, Allen SA, Thrall DE (1992) Phase I evaluation of doxorubicin and wholebody hyperthermia in dogs with lymphoma. J Vet Intern Med 6, 245-249. Ogilvie GK, Vail DM, Klein MK, Powers BE, Dickinson K (1991) Weekly administration of low-dose doxorubicin for treatment of malignant lymphoma in dogs. J Am Vet Med Assoc 198, 1762-1764. Postorino NC, Susaneck SJ, Withrow SJ, Macy DW, Harris C (1989) Single agent therapy with Adriamycin for canine lymphosarcoma. J Am Anim Hosp Assoc 25, 221-225. Prager MD, Baechtel FS (1973) Methods for modification of cancer cells to enhance their antigenicity. Methods Cancer Res 9, 339-400. Price GS, Page RL, Fischer BM, Levine JF, Gerig TM (1991) Efficacy and toxicity of doxorubicin/cyclophosphamide maintenance therapy in dogs with multicentric lymphosarcoma. J Vet Intern Med 5, 259-262. Rogers KS (1989) L-asparaginase for treatment of lymphoid neoplasia in dogs. J Am Vet Med Assoc 194, 1626-1630. Rosales C, Jeglum KA, Obrocka M, Steplewski Z (1988) Cytolytic activity of murine anti-dog lymphoma monoclonal antibodies with canine effector cells and complement. Cell Immunol 115, 420-28. Rosenberg MP, Matus RE, Patnaik AK (1991) Prognostic factors in dogs with lymphoma and associated hypercalcemia. J Vet Intern Med 5, 268-271. Rosenthal RC (1990) The treatment of multicentric canine lymphoma. Vet Clin North Am: Small Anim Pract 20, 1093-1104. Rosenthal RC, MacEwen EG (1990) Treatment of lymphoma in dogs. J Am Vet Med Assoc 196, 774-781. Ruslander D, Moore AS, Gliatto JM, Lâ&#x20AC;&#x2122;Heureux D, Cotter SM (1994) Cytosine arabinoside as a single agent for the induction of remission in canine lymphoma. J Vet Intern Med 8, 299-301.

Schwartz SN (1988) Spontaneous regression of lymphosarcoma in a dog. J Am Vet Med Assoc 192, 222-224. Sjogren HO, Hellstrom I, Bansal SC, Hellsrom KE (1971) Suggestive evidence that the blocking antibodies of tumor bearing individuals may be antigen-antibody complexes. Proc Natl Acad Sci 68, 1372-1375. Smith RT, Adler WH (1970) Humoral tumor immunity. N Engl J Med 282, 1320-2322. Squire RA, Bush M, Melby EC, Neeley LM, Yarbrough B (1973) Clinical and pathologic study of canine lymphoma: clinical staging, cell classification, and therapy. J Natl Cancer Inst 51, 565-574. Steplewski Z, Jeglum KA, Rosales C, Weintraub N (1987) Canine lymphoma-associated antigens defined by murine monoclonal antibodies. Cancer Immunol Immunother 24, 197-201. Stone MS, Goldstein MA, Cotter SM (1991) Comparison of two protocols for induction of remission in dogs with lymphoma. J Am Anim Hosp Assoc 27, 315-321. Teske E, van Heerde P, Rutteman GR, Kurzman ID, Moore PF, MacEwen EG (1994) Prognostic factors for treatment of malignant lymphoma in dogs. J Am Vet Med Assoc 205, 1722-1728. Theilen GH, Worley M, Benjamini, E (1977) Chemoimmunotherapy for canine lymphosarcoma. J Am Vet Med Assoc 170, 607-610. Thompson K, Harris M, Benjamini E, Mitchell G, Nobel M (1972) Cellular and humoral immunity-a distinction in antigenic recognition. Nature 238, 20-21. Vail DM (1993) Recent advances in chemotherapy for lymphoma of dogs and cats. Compend Contin Educ Pract Vet 15, 1031-1037. Weller RE, Theilen GH, Madewell BR, Crow SE, Benjamini E, Villalobos A (1980) Chemoimmunotherapy for canine lymphosarcoma: a prospective evaluation of specific and nonspecific immunomodulation. Am J Vet Res 41, 516-521. Williams JH (1988) The use of gamma linolenic acid, linoleic acid and natural vitamin E for the treatment of multicentric lymphoma in two dogs. J South African Vet Assoc 59, 141144. Witney, R, Levy J, Smith AG (1974) Influence of tumor size and surgical resection on cell mediated immunity in mice. J Natl Cancer Inst 53, 111-116. Witzig TE (2006) Radioimmunotherapy for B-cell non-Hodgkin lymphoma. Best Pract Res Clin Haematol 19, 655-68.

Steven E. Crow

Cancer Therapy Vol 6-A.2, page 33 Cancer Therapy Vol 6, 221-226, 2008

Veterinary pathologists achieve 80% agreement in application of WHO diagnoses to canine lymphoma Research Article

Victor E. O Valli Veterinary Lymphoma Study Group

__________________________________________________________________________________ *Correspondence: Victor E. O Valli, DVM Professor Department of Pathobiology College of Veterinary Medicine University of Illinois, MC 004, Rm 284 SAC, 1008 W Hazelwood Drive Urbana IL 61821, USA; Tel: 217 265 5415; Fax 217 333 5496; email vevalli@uiuc.edu Key words: canine lymphoma, WHO diagnoses, Case assembly, Histopathology, microscopic review, pathologists, statistical methods, Case derivation, Diagnostic pitfalls Abbreviations: American Kennel Club, (AKC); Anaplastic Large Cell Lymphoma, (ALCL); Burkitt-Like Lymphoma, (BKL); Diffuse Large cell B-cell Lymphoma, (DLBCL); Follicular Lymphoma, (FL); Histiocytic Sarcoma, (HS); not otherwise specified, (NOS); Peripheral T-cell Lymphoma, (PTCL); T-Lymphoblastic Lymphoma, (T-LBL) Received: 27 March 2008; electronically published: June 2008

Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary To assist the accuracy and impact of lymphoma diagnosis in animals a blinded study was carried out to test the accuracy and reproducibility of veterinary pathologists, not specialists in hematopathology in applying the World Health Organization (WHO) classification system for human lymphomas. Twenty pathologists reviewed 300 cases of canine lymphoma with a consensus diagnosis derived for each case by three pathologists specializing in hematopathology. The 17 Reviewing pathologists were given the signalment for each animal and required to render a diagnosis on biopsy tissue after reviewing an oversight stain of hematoxylin and eosin, followed by the same tissues stained by immunohistochemical methods for B and T-cell lineage. The diagnostic process required the reviewers to fill out a bubble sheet having 43 types of lymphoma as choices. The overall accuracy of the 17 reviewing pathologists on the 300 cases was 83.1%. In a review of accuracy on the 6 most common diagnoses that contained 79.5% of total cases their accuracy rose to 86.6%. In a test of reproducibility made by recutting 5% of the cases that were re-entered under a different number the overall agreement between the first and second diagnosis on the same tissue varied from 40-86.7%. Statistically there were 43,000 data points for each of the 17 reviewers.

(Harris et al, 1994, 2000; [No authors listed], 1997). Currently, lymphomas in dogs are treated as if they are all of the same type, but we now find that like those in humans, the canine lymphomas are of many types that also benefit from specific identification and treatment. The goal of this international standardization project is to demonstrate that veterinary diagnosticians can effectively apply the human criteria to the canine tumors and thus permit much more effective treatment by veterinary oncologists. The WHO system of disease classification is based on diagnosis of diseases rather than on cell types and requires complete patient data plus immunophenotyping 5,6. In an international review by human medical pathologists, their overall accuracy and reproducibility in testing the WHO system exceeded 85%,

I. Introduction Lymphoma is the most common canine cancer treated by chemotherapy and a most common neoplasm that afflicts dogs of all breeds and ages. The disease is recognized internationally and treated by veterinarians in both first opinion and referral practice. The completion of sequencing of the canine genome has shown the remarkable similarities between the genetic make-up of dogs and humans. Similarly, many of the malignancies that occur in dogs are also like their human counterparts especially the tumors of the lymphoid system. The World Health Organization has devised a new system of recognizing and categorizing the many subtypes of human lymphoid tumors with very different characteristics that must be considered in providing effective treatments 33

Valli: Veterinary pathologists in agreement of WHO diagnoses to canine lymphoma through the 300 cases using a dual viewing microscope and compared their interpretations with that derived by Dr Valli. About 10% of diagnoses were changed on the basis of their review with a full consensus required and reached for each case. That consensus diagnosis became the official interpretation against which all of the 17 additional reviewers were compared. The second group of 6 pathologists (Aug. 18-25), and the third group of 4 from Europe (Nov 18-24), all followed the same routine. Dr Scott Moroff was not able to attend for all of the first week and completed the half of the study by subsequent shipment of the slides. Dr. Tony Ross was the only participant who labored entirely alone on the study while on a working visit to Illinois. Each group began with a Monday morning review of cases in contention; the cases in question were projected and the diagnostic reasoning behind the interpretation was discussed. Throughout, reviewers would bring cases they were unsure of to one of the three principal pathologists for consultation, utilizing a 4 headed microscope. Lunch was provided on site and most days the pathologists worked till 7-8:00PM, depending on their rate of progress, and then all would break for dinner. The final review period was conducted during the Thanksgiving week for European pathologists while students were absent and facilities available. Plans to conduct the final review in Bristol University UK were changed to Illinois with the decline in the dollar against the Euro making the change essential.

rendering other systems obsolete ([No authors listed], 1997). The WHO system defines human lymphoid neoplasms as 16 disease subtypes of B- and of T-cell tumors that differ greatly in presentation, normal biology, rate of progression and response to therapy. Consequently, these neoplasms are very specifically identified and treated. Remarkably, canine, like human lymphomas, differ greatly in natural rate of progression and response to therapy. Canine lymphomas have been described in the WHO format and there is now very strong evidence that when specifically identified, they closely mimic the human counterparts in gaining remission and projected survival patterns (Valli et al, 2002, 2006; Jubb et al, 2006).

II. Materials and Methods A. Case assembly 1. Collection of cases Excellent and essential collaboration was requested from National Diagnostic Laboratories to provide histology slides to the ACVP study. Animal owners paid for the cost of biopsy and interpretation, including immunophenotyping. Veterinary oncologists were requested to submit cases for interpretation through cooperation of the Veterinary Cancer Society with the ACVP study. All cases processed at Illinois were sectioned cut at a thickness of 3 microns to facilitate recognition of cellular detail. Bubble sheet forms, derived from the medical pathologists Non-Hodgkin’s Lymphoma study were printed and used for each case reviewed by veterinary pathologists. Accession of cases began in January of 2006 with 670 received prior to beginning the review, and increased to 950 at the time of this writing. A target goal of 1000 cases is within reach and will allow determination of the value of specific diagnosis on response to treatment and survival.

C. The lymphoma review pathologists

2. Histopathology The slides were given a unique ACVP serial number in addition to the pathology number of the processing laboratory and the case reports were archived. A period of one week was projected for the pathology slide review. The number of cases to be evaluated by each pathologist was limited to 300 to permit completion of the entire slide set during that period. The signalment for each case was extracted from the case reports and assembled into an electronic file numbered 1-300. The 300 cases were divided into 8 separate groups of <38 cases in each group to allow the 8 pathologists to review cases simultaneously. An overview powerpoint program of diagnoses was prepared. CD’s, with over 250 images of the different diagnostic categories, were sent to pathologists potentially interested in participating in a blinded review. A shorter CD was prepared of major diagnostic types and teamed with a 4 page algorithm that provided a logical pathway for diagnostic decision points based on architecture (nodular or diffuse), cell size (small, intermediate or large), nuclear shape (round or indented), nuclear chromatin (dispersed or with parachromatin clearing), nucleoli number and placement (single central, immunoblastic or multiple peripheral, centroblastic), and mitotic rate (0-5 low, 6-10 medium, >10 high).

Name Barthel*A. Bienzle* D. Caswell* J. Colbatzky F.

Residence USA Canada Canada Germany

Training TAMU OVC OVC Hanover

Durham A.

USA

U Penn

Ehrhart* E. Jones* C. Kiupel* M. LaBelle* P. Lester*S. Miller*M. Moore*P. Moroff* S. Ramos-vara J. Roccabianca*** P.

USA USA USA USA Canada USA USA USA USA Italy

CSU Tufts Utrecht UCD UCD WSU UCD

Ross A.

Australia

Scase* T. Tvedten** H. Valli** V. Vernau* W

UK Sweden USA USA

Spain UCD & UNIMI U Sydney U FL MSU OVC OVC

Employment Antech University University BoehringerIngelheim Senior Resident University IDEXX University Antech Private Lab University University Antech University University Private lab University University University University

*Indicates diplomate ACVP, ** indicates diplomate ACVP AP and CP. ***indicates diplomate ECVP.

1. The problem

B. The microscopic review of cases

Lymphoid tumors in animals are often given a generic diagnosis of “lymphoma” by veterinary pathologists. Oncologists determining treatment thus receive a non-specific interpretation despite the marked variation in clinical behavior of animal lymphomas, as seen in human patients. Strategic solution: Demonstrate that veterinary pathologists who are not specialists in hematopathology can accurately and

In June/July of 2007, DVM pathologists were contacted largely by email to determine availability for a week in Illinois to participate in the review process. The Review program paid for air fare and hotel in Champaign IL and all meals. The first group of 8 pathologists (Aug 11-18) included Drs Peter Moore and William Vernau of UC Davis CA. They worked

Cancer Therapy Vol 6-A.2, page 35 consistently apply an upgraded system of lymphoma classification in a blinded study. If veterinary diagnostic efficacy is proven then pathologists can accept that more specific diagnosis of lymphoma is achievable.

Hodgkinâ&#x20AC;&#x2122;s Lymphoma Study Group, who found that immunophenotyping is essential for correct recognition of all human lymphomas except those with a follicular architecture. Since follicular lymphoma constituted less than 1% of total cases in this study, it is apparent that immunophenotyping is essential for all cases of animal lymphoma. Intraobserver reproducibility was tested by re-cutting 5% of the cases, including examples from the most common entities. These recuts were reentered into the study with a similar signalment but different pathology accession numbers and ACVP serial numbers. The mean overall agreement between the first and second diagnosis on the same tissue for the 20 test observers was 65.5% with a range of 40-86.7%. The answer to the final question (clinical presentation and survival by diagnosis) will be derived from analysis of the presenting signs recorded in the final group of 1,000 cases as part of the review of survival by diagnosis. Initial evaluation of clinical data has been tallied for the 300 case study. It is apparent that dogs with indolent lymphoma retain normal appetite and activity, two very key determinants of well being, with advanced stages of lymphoma. The specific lymphomas with these characteristics include: nodal and splenic Marginal Zone lymphoma, the follicular lymphomas, small cell lymphoma of B and T-cell types, T-Cell Rich Large BCell lymphoma and T-Zone lymphoma. An important conclusion to be derived from this study is that the current paradigm of a diagnosis of B-cell lymphoma is more favorable to the outcome than one of T-cell lymphoma is no longer valid. We are now in the era of understanding that there are very aggressive lymphomas like lymphoblastic lymphoma of B or T-cell type and very indolent lymphomas of both B and T-cell type. The latter include Mantle and Marginal Zone lymphoma of B-cell type and the most indolent of all T-Zone lymphoma.

2. The plan Duplicate the plan as carried out by medical pathologists for The Non-Hodgkinâ&#x20AC;&#x2122;s Lymphoma Classification Project (Blood 89: 3909-3918, 1997) ([No authors listed], 1997). Specific Study Goals: 1. Determine the ability of veterinary pathologists to apply the WHO system of lymphoma classification on a retrospective group of cases of canine lymphoma derived nationwide. 2. Determine the need for immunophenotyping and clinical history in the diagnostic process. 3. Determine the intraobserver and interobserver reproducibility in the diagnosis of the more common diagnostic entities. 4. Determine the presenting clinical signs for the various types of lymphoma and the survival by diagnosis.

3. The funding In association with case collections fund raising efforts resulted in donations of $35,000.00. Idexx Labs, Ft. Dodge Animal Health and the ACVP Council each provided $10,000.00, and an additional $5000.00 was received from Dr Sally Lester. Funding of $42,000.00 was received in 2007 from the American Kennel Club (AKC). Substantial support from Antech was provided by covering both travel and hotel expenses for participating Antech personnel, and on occasion, graciously paying for food for the whole group attending the first week of review. Antech also provided funding for related research to determine the clonality of cells in a subset of reviewed cases. Boehringer-Ingelheim paid the travel expenses of Dr. Florian Colbatzky from Germany to Illinois.

III. Results A. The statistical methods and data The data were subjected to a 2x2 tabular analysis that measured both sensitivity based on the number of times the consensus diagnosis was chosen and specificity that measured the rate of rejection of incorrect diagnoses. The first analysis compared all of the 17 reviewers with the consensus diagnoses for each case. The overall level of agreement was 83.1% which included all 43 diagnostic categories; the percentages ranged from 46.4 to 100%. Upon analyzing data from the 6 most common diagnostic entities, which constituted 79.5% of total cases, the overall agreement was 86.6%. These six subtypes of lymphoma included; marginal zone spleen, diffuse large B-cell, Tlymphoblastic, nodal T zone, peripheral T-cell not otherwise specified (NOS) and disease other than lymphoma. Remarkably, there are 43,000 data points for each reviewer, which indicates the complexity of the analysis. Clearly these are most gratifying results and the most clinically significant to be derived from this data. These results positively and emphatically answer the first two stated goals of this study. The first is that veterinary pathologists who are not specialists in hematopathology can achieve a high degree of accuracy in applying the WHO classification system for lymphomas. The second is that immunophenotyping is essential for diagnosis of canine lymphoma. This is in agreement with the Non-

B. Case derivation by state As of this writing 950 cases have been received with 911 derived from 34 American States and 3 other countries including Canada 14, Austria 13 and The Netherlands for 6 cases. The major States on the basis of cases submitted in alphabetical order are: Arizona California Colorado Connecticut Florida Illinois Massachusetts New Jersey New York

134 314 23 33 36 118 18 20 109

Other States submitting 16 or less cases include: Alaska, Alabama, Arkansas, Georgia, Iowa, Indiana, Louisiana, Maryland, Minnesota, Missouri, Montana, North Carolina, Nebraska, New Hampshire, Nevada, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Dakota, Texas, Washington State, Wisconsin. 35

Valli: Veterinary pathologists in agreement of WHO diagnoses to canine lymphoma

C. Case distribution by diagnosis: (950 cases) B-CELL LYMPHOMAS

T-CELL LYMPHOMAS

NOT LYMPHOMA

B-CLL

CUTANEOUS T-CELL

ANGIOSARCOMA

B-LBC

MYCOSIS FUNGOIDES

CANINE HISTIOCYTOMA

B-SLL

PERIPHERAL T-CELL

118

HISTIOCYTIC SARCOMA

8 1 0

BURKITT-LIKE

T ANGIOIMMUNOBLASTIC

MAST CELL TUMOR

CENTROCYTIC DIBCL

ENTEROPATHY TYPE T

MYCOTIC ADENITIS

DLBCL

370

T-ANAPLASTIC LARGE CELL

SOFT TISSUE SARCOMA

FL II&III

T-CLL

THYMOMA

LYMPHOPLASMACYTOID

T-LBC

MANTLE CELL

T-LBL

MARGINAL ZONE

T-NK

PLASMACYTOMA

T-SLL

PLASMABLASTOMA T-CELL RICH LARGE BCELL

T-ZONE

104

D. Diagnostic pitfalls

6. Marginal zone lymphoma can be difficult to distinguish from early DLBCL. Distinction involves assessment of mitotic rate. Additionally, although rare, the blastoid variant of mantle cell lymphoma is also difficult to recognize.

In the course of reaching consensus on all diagnostic entities and in going over cases with pathology reviewers it is apparent that there are a number of diagnoses that are “look-alikes” and must be carefully interpreted to reach a correct conclusion. 1. Diffuse Large cell B-cell Lymphoma (DLBCL), the most common diagnostic entity in the dog, and Peripheral T-cell Lymphoma (NOS) (PTCL) are indistinguishable without immunophenotyping. 2. PTCL with most cells of intermediate size can be mistaken for T-Lymphoblastic Lymphoma (T-LBL). The distinction is important since T-LBL is the most aggressive lymphoma and is characterized by the shortest survival time despite intensive therapy. 3. DLBCL may be confused with Burkitt-Like Lymphoma (BKL). The standard criteria are if there are large cells with nuclei 2.0 or more red cells in diameter in every field, then the diagnosis is Large Cell type despite 90% of the cells being of intermediate size (nuclei are 1.5 times the diameter of RBC). 4. Follicular Lymphoma and Benign or Atypical Follicular Hyperplasia are frequently confused with consequent overdiagnosis of follicular lymphoma. True Follicular Lymphoma (FL) has no mantle cell cuffs and few or no tingible body macrophages. In FL the same cell types are present in the same proportion in each neoplastic follicle as might be expected with a clonal neoplasm. 5. Anaplastic Large Cell Lymphoma (ALCL) of T cell and B cell lineage, and canine Histiocytic Sarcoma (HS) may be confused. It is essential to note that the large atypical neoplastic cells of T-ALCL have variable CD3 expression (often reduced). Both neoplasms are characterized by marked variation in nuclear size and both have unusually abundant cytoplasmic volume.

IV. Discussion The most important findings of this study are that veterinary pathologists can accurately apply the criteria of the WHO classification for the diagnosis of canine lymphomas. If the degree of diagnostic accuracy achieved in this study was generally applied to animal lymphomas, it would provide the basis for more specific therapy and result in longer survival of many animals. Equally important is the hope that this characterization will facilitate more effective research on those entities that are currently virtually untreatable, like lymphoblastic lymphoma. Confidence in the significance of these findings is based on feedback from oncologists who find they can now tailor their treatments to specific diagnostic entities such as TZL (indolent lymphoma). The first step in proving the application of the WHO classification is done. The second step is to prove the value of its application by determining survival as a function of specific diagnosis and stratified by therapeutic protocol. The selection of pathologists for the review was not random, but largely based on their availability to attend during the times requested. Input received from medical colleagues central to the operations of the Non-Hodgkin’s Lymphoma Study Group suggested that an international approach was vitally important, so that US pathologists did not dominate the overall review. With that in mind, 7 of the 20 pathologists currently reside outside of the US (Bienzle, Caswell, Colbatzky, Roccabianca, Ross, Scase and Tvedten). In addition Barthel, Kiupel, LaBelle, and

Cancer Therapy Vol 6-A.2, page 37 Ramos-Vara are originally from outside of the US, giving us more than a 50% international presence. In a previous smaller blinded microscopic review, it was found that a first year resident scored higher in applying the study criteria than a pathologist with many years of experience. On that basis, one participant - now a senior resident - was invited to participate. That confidence was rewarded with an overall mean accuracy in the upper half of the entire review group. The very relevant point to be gained from this participation is that a pathologist need not have a long period of experience in order to successfully apply the WHO criteria. In fact, it might be argued that a lack of background bias based in application of previous diagnostic systems may be a distinct advantage. The case material for the study brought up several important issues. First, at our current level of understanding, it is not possible to provide a specific diagnosis of many of the lymphoma subtypes based on cytologic assessment. There are fine points of recognition in interpretation of lymphoma on Wrights-stained cytologic preparations, but these need verification that can only be gained by having access to cytology and histology on the same tissues. Secondly, there is a minimum size of biopsy required for a firm histologic interpretation. Nodal excision is always preferable and a wide bore cutting needle biopsy may be preferable to a shallow incisional biopsy. It can be firmly stated that an 18 gauge tru-cut biopsy may be adequate for renal biopsy but not for lymphoma. In general a nodal biopsy should be near 2mm in width to provide sufficient architecture for diagnosis on lymphoid tissue. Additionally, specimen thickness is critical with optimal assessment requiring sections 4u or thinner. With respect to immunophenotyping, there appears to be significant variation in the quality of CD79 a immunohistochemical staining achieved between laboratories, with poor quality staining significantly complicating interpretation. Consequently, CD20 assessment can be an extremely useful adjunctive (but not replacement) stain.

To mount the study we needed access to new cases as retrospective studies could not routinely provide the immune stained slides felt to be essential. The Veterinary Cancer Society not only printed a request for case submissions in their newsletter but the membership responded and has continued to do so with the target of 1000 cases likely achieved at this presentation. In the formative stages of the study the members repeatedly gave input on the content of the clinical history format. The input of oncologists was essential and is gratefully recognized. An accurate assessment of hematopoietic tissue requires superb histological preparations with those processed at Illinois thinly cut and well stained both for routine and immune preparations. Further to mount the study is was essential to gain continuing access to the slides for the pathology review and for the ability to reexamine these further on the basis of survival data yet to be derived. All contributing laboratories have our thanks. Finally the next step is also dependent on our clinical colleagues as we determine the outcome of cases for impact of the nature of tumor subtype on response to therapy. We thank them in advance as well as their animal owning clients.

References [No authors listed] (1997) A clinical evaluation of the International Lymphoma Study Group classification of nonHodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood 89, 3909-3918. Harris NL, Jaffe ES, Diebold J, Flandrin G, Muller-Hermelink HK, Vardiman J, Lister TA, Bloomfield CD (2000) The World Health Organization classification of neoplastic diseases of the haematopoietic and lymphoid tissues: Report of the Clinical Advisory Committee Meeting, Airlie House, Virginia, November 1997. Histopathology 36, 69-89. Harris NL, Jaffe ES, Stein H, Banks PM, Chan JK, Cleary ML, Delsol G, De Wolf-Peeters C, Falini B, Gatter KC, et al (1994) A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 84, 1361-1392. Jubb KVF, Kennedy PC, Palmer N (2006) Pathology of Domestic Animals. Fifth Edition, Vol 3. Chapter 3, the Hematopoietic System, VE Valli. Valli VE, Jacobs RM, Parodi AL, Vernau W, Moore PF (2002) Histological Classification of Hematopoietic Tumors of Domestic Animals. WHO, Second Series, Vol VIII. Armed Forces Institute of Pathology Washington DC. Valli VE, Vernau W, de Lorimier LP, Graham PS, Moore PF (2006) Canine Indolent Nodular Lymphoma. Vet Pathol 43, 241-56.

Acknowledgements This document would not be complete without recognition of the assistance gained and essential to the study. The initial financial support gave the impetus to get the project underway and provided the credibility for the American Kennel Club proposal for which many pathologists and oncologists offered supporting statements. Thanks to all.

Valli: Veterinary pathologists in agreement of WHO diagnoses to canine lymphoma

Cancer Therapy Vol 6-A.2, page 39 Cancer Therapy Vol 6, 239-246, 2008

Shared pathogenesis of human and canine tumors an inextricable link between cancer and evolution Review Article

Jaime F. Modiano1,*, Matthew Breen2,* 1

Department of Veterinary Clinical Sciences, College of Veterinary Medicine and Masonic Cancer Center, University of Minnesota, Minneapolis/St. Paul, MN 2 Department of Molecular Biomedical Sciences, College of Veterinary Medicine, and Center for Comparative Medicine and Translational Research, North Carolina State University, Raleigh, NC

__________________________________________________________________________________ *Correspondence: Dr. Jaime F. Modiano, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, 1365 Gortner Ave., St Paul, MN 55108, USA; Tel: 612-625-7436; fax: 612-624-0751; e-mail: modiano@umn.edu Dr. Matthew Breen, Department of Molecular Biomedical Sciences, College of Veterinary Medicine, North Carolina State University, 4700 Hillsborough Street, Raleigh, NC 27606, USA; Tel: 919-513-1467; Fax: 919-513-7301; e-mail: Matthew_Breen@ncsu.edu Key Words: The “Raleigh Chromosome”, canine chronic myelogenous leukemia, c-Myc, sporadic Canine Burkitt Lymphoma, CFA 22, Canine CLL, pathogenesis Abbreviations: chronic lymphocytic leukemia (CLL); comparative genomic hybridization (CGH); dog chromosome 13 (CFA 13); fluorescence in situ hybridization (FISH) Received: 14 April 2008; electronically published: June 2008

Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary The etiological significance of pathognomonic molecular cancer signatures cannot be understated. Since the discovery that the Philadelphia Chromosome was present recurrently in karyotypes from human chronic myelogenous leukemia patients, the identification of other structural and/or numerical chromosomal abnormalities that are associated recurrently with specific tumors have implicated a long list of genes in the pathogenesis of hematological and solid tumors. Still, a gap remains in our understanding of the mechanisms driving these changes. We recently showed that morphologically equivalent cancers of humans and dogs share homologous molecular signatures. This not only confirms the causal relationship between defined genetic abnormalities and specific tumor types, but it also provides novel mechanistic insights for these events. Here, we review these findings and propose that at least some of these abnormalities occur due to promiscuous reassembly between fragile sites connecting chromosomal segments with preserved gene order. We also advance the notion that cancer susceptibility and tumor progression are the result of comparable gene-environment interactions in humans and dogs, thus allowing us to leverage the unique genetic structure of purebred dog populations to identify heritable factors that contribute to the pathogenesis of cancer in both species.

breed predisposition for certain canine cancers, including lymphoma and soft tissue sarcomas (Dorn et al, 1968; Priester and McKay, 1980; Onions, 1984). In the case of lymphoma, familial clustering has been reported in dogs, suggesting that genetic risk (or protective) factors for the disease have segregated with breed-specific traits. Finally, the status of both humans and dogs as “post-genomic” species (Lander et al, 2001; Kirkness et al, 2003; Lindblad-Toh et al, 2005) also opens a new window of opportunity for comparative cancer genomics. Despite the strong similarities in histological appearance, biological behavior and response to conventional therapies, a major obstacle that remained to fully validate the relevance of canine cancers to the human

I. Introduction Spontaneous tumors that are morphologically and clinically equivalent occur in humans and dogs. This allows for comparative studies in both species as a means to understand the pathogenesis of these conditions. Among the advantages that dogs provide for this approach are the greater prevalence of some types of cancer in dogs than in people, thus providing ready access to case materials, and the fact that cancer progression follows a compressed time course in dogs, allowing for the timely assessment of new interventions (MacEwen, 1990; Hansen and Khanna, 2004; Khanna et al, 2006; Paoloni and Khanna, 2008). In addition, there is a distinct, significant and reproducible

Modaino and Green: The Role of Evolution in Cancer Pathogenesis condition was the uncertainty as to whether naturally occurring canine tumors would recapitulate the molecular pathogenesis of their human counterparts. Hence, we decided to test the hypothesis that homologous cancers of humans and dogs would harbor evolutionarily conserved and pathognomonic genetic abnormalities. For this purpose, we used chronic myelogenous leukemia (CML), Burkitt lymphoma, and chronic lymphocytic leukemia (CLL), all of which occur spontaneously in humans and dogs, and each of which have well defined chromosomal aberrations in people; that is, the t(9;22) Philadelphia (Ph’) chromosome in CML, t(8;14) translocations superimposing c-Myc onto the immunoglobulin heavy chain enhancer in Burkitt lymphoma, and del13q14 deletion in CLL (Breen and Modiano, 2008).

III. Overexpression of c-Myc sporadic Canine Burkitt Lymphoma

Compelling evidence supports an essential role for deregulated c-Myc expression in BL (Hecht and Aster, 2000). A translocation of C-MYC from HSA 8q24 to HSA 14q32 places c-Myc under the control of the immunoglobulin heavy chain enhancer in virtually every case of BL or L3 subtype acute B-cell ALL, and in a small number of T-cell lymphomas and leukemias (Croce et al, 1984; McKeithan et al, 1986; Shima et al, 1986; MacEwen 1990; Navid et al, 1999). BL in humans occurs most frequently as the endemic form associated with Epstein Barr virus infection, although the molecular pathogenesis of the endemic and the sporadic forms are indistinguishable. In dogs, BL only occurs sporadically, thus this condition is somewhat less common than in humans (Fosmire et al, 2007; Modiano et al, 2007). As we did for BCR and ABL in canine CML, we established the locations for the MYC and IGH loci at CFA 13q12 and CFA 8q33, respectively, to examine if a translocation involving these genes was detectable in canine BL. As we predicted, using BAC clones for these loci to probe cells from canine BL patients, we demonstrated that they harbored a translocation that juxtaposed MYC and the IGH enhancer, leading to constitutive expression of c-Myc in canine BL (Breen and Modiano, 2008).

II. The “Raleigh Chromosome” in canine chronic myelogenous leukemia Various lines of evidence support the causal association between t(9;22) translocations that generate the Ph’ chromosome and the consequent BCR-ABL fusion proteins with CML. First, this group of aberrations is seen in as many as 95% of adult patients with CML (Kurzrock et al, 2003). Second, chimeric BCR-ABL proteins have constitutively elevated tyrosine kinase activity that is crucial to their oncogenic potential, as demonstrated by the ability to induce and maintain remission of CML patients with imatinib mesylate (Gleevec®), an antagonist of BCRABL kinase activity (Kurzrock et al, 2003). Third, adoptive transfer of hematopoietic stem cells carrying the BCR-ABL fusion product into lethally irradiated mice recapitulates the disease (Wertheim et al, 2002). Given the similarities between human and canine CML, we surmised that an homologous translocation might be detectable in dogs with naturally occurring CML. We established that the genomic locations for the canine ABL and BCR loci are at CFA 9q25dist-q26 and CFA 26q24-q25, respectively, and used canine BAC clones representing these loci to examine if a translocation involving these genes was detectable in canine CML. Indeed, we identified the canine equivalent of the Ph chromosome, which we refer to as the “Raleigh chromosome” recurrently associated with canine CML cells (Breen and Modiano, 2008). Not only was the translocation evident by fluorescence in situ hybridization (FISH) analysis of metaphase and interphase nuclei, but also, immunoprecipitation of Abl proteins followed by reciprocal immunoblotting with anti-BCR antibodies verified the presence of a BCR-Abl fusion protein that was comparable to the common ~190 kDa form seen in human CML. The Ph’ chromosome is also seen in a limited number of human acute lymphoblastic leukemias (ALL). Intriguingly, we also identified the Raleigh chromosome in a case of canine ALL, although leukemic cells from this dog did not harbor a BCR-Abl fusion protein, suggesting a distinct variant form of the translocation event in canine ALL (Breen and Modiano, 2008).

IV. Loss of CFA 22 as a Marker of Canine CLL The most common chromosome abnormality in adult CLL is hemizygous deletion of the q14 region of HSA 13 at the RB-1 locus (Dohner et al, 2000). A causal relationship was recently identified between CLL and the 13q14 deletion in humans, which was due to loss of two microRNAs (mir-15 and mir-16) with strong tumor suppressor activity (Cimmino et al, 2005; Calin and Croce, 2006). Both mir-15 and mir-16 are Bcl-2 antagonists, and reduced levels of mir-16 in NZB mice is responsible for the characteristic lymphoproliferative disease in these mice that recapitulates human CLL (Raveche et al, 2007). Two observations suggested this abnormality might be evolutionarily conserved in canine CLL. First, deletion of 13q14 is prognostically significant, defining the group with most favorable outcomes among human CLL patients. In dogs, CLL is an indolent disease that also shows favorable response to therapy with extended remissions. Second, the structure of the chromosomal region that includes RB-1 locus and miR15/16 at HSA13q14 is evolutionarily conserved back to zebrafish. In dogs, a deletion homologous to HSA 13q14 would manifest as a deletion within CFA 22q11.2. As we did for CML and BL, we examined if deletion of CFA 22q11.2 (based on use of a BAC clone containing RB-1) was a common occurrence in canine CLL. Perhaps not surprisingly, eight of nine cases of CLL we tested had hemizygous or homozygous deletions of CFA 22 that always included at least one copy of the RB-1 locus with reduced or absent expression of Rb, indicating that loss of

Cancer Therapy Vol 6-A.2, page 41 this region of CFA 22 was a functionally significant event (Breen and Modiano, 2008). One intriguing difference between CLL in humans and dogs is worth noting: in people, CLL is predominantly a disease of B-cells and T-CLL is a rare entity (Schlegelberger et al, 1994); whereas in dogs, CLL presents frequently as a disease of T-cells (Burnett et al, 2003). The observation that loss of syntenic DNA is a shared cytogenetic feature between the canine and humans diseases further supports the notion that this event facilitates proliferation of both B- and T-cells without overt malignant transformation, resulting in similar disease phenotypes and underscoring the essential role of mir-15 and mir-16 in lymphocyte cell cycle control.

sites of genomic reorganization are inherently fragile and may serve as evolutionary retained ‘hot spots’ for chromosomal reassembly associated with tumorigenesis (Pfeiffer et al, 1995; Murphy et al, 2005). There are recurrent cancer-associated translocations that do not involve such sites, such as the t(8;14) translocation observed in human (and canine) Burkitt lymphoma. Thus, other underlying mechanisms probably contribute to the pathogenesis of this disease in both species (Finger et al, 1986; Reiter et al, 2003).

VI. Canine cancers and heritable traits that control risk and progression Familial cancer syndromes associated with single gene defects have been characterized in humans, rats, and dogs (Fearon 1997; Hino, 2003; Lingaas et al, 2003), and myriad cancers have been modeled in laboratory mice using genetic engineering. It is curious that cancer syndromes with spontaneous mutation of the BHD gene have been identified in each of these species (Comstock et al, 2004; Okimoto et al, 2004). Nevertheless, most cancers occur sporadically as complex diseases where multiple genes with variable penetrance interact with environmental factors to progressively endow a single “tumor initiating cell” with malignant properties. Lymphoid cancers are among the most common cancers of dogs (Modiano et al, 2005a). Lymphoma and leukemia account for 20-25% of all canine tumors with an average lifetime risk of ~1 in 15 (Modiano et al, 2005a). Generally, lymphomas are “treatable,” but not “curable.” The median survival for dogs with lymphoma treated with standard-of-care is 10-14 months, and most will die from their disease. However, while lymphoma and leukemia can affect dogs of all breeds and all ages, there are significant breed-associated differences in the age of onset for these diseases (Figure 1) and in their incidence, and disease progression and response to therapy are different among distinct histologic and phenotypic subtypes of lymphoma and leukemia (Modiano et al, 2007). The breed predisposition for lymphoid tumors in dogs suggests that heritable risk factors for the disease have segregated with breed-specific traits. A number of dog breeds, including Boxers, Golden Retrievers, Labrador Retrievers, Cocker Spaniels, Bassett Hounds, and others are at increased risk for lymphoid cancer, while other breeds, such as Chihuahuas, Dachshunds and Pomeranians show a lower incidence of the disease (Priester and McKay 1980). To put this in proper context, the estimated lifetime risk for any dog to develop lymphoma or leukemia is approximately 1 in 15. The lifetime risk for Boxers is more than four times higher (Priester and McKay, 1980), while that for Golden Retrievers is approximately 1 in 8 (Glickman et al, 2000). The contribution of heritable factors to the risk for lymphoma and leukemia is further supported by the familial clustering observed in certain Rottweiler and Scottish Terrier lines (Teske et al, 1994), and by the fact that breed type can influence response to therapy (Garrett et al, 2002).

V. Mechanistic insights for the pathogenesis of chromosomal abnormalities in naturally occurring cancers of humans and dogs Our results show that comparable, recurrent chromosomal aberrations were identifiable in these naturally occurring tumors, supporting a strong causal association between each abnormality and the respective tumor phenotype. In addition, the data suggest that each of these pathognomonic lesions leads to similar consequences in both species; i.e., the translocations or the deletion seemed to generate a fusion protein, promote overexpression of a homologous protein, or silence a tumor suppressive activity. Curiously, the aberrations in CML and CLL involved chromosomal regions located near areas of genome reorganization (Breen et al, 1999; O'Brien et al, 1999; Yang et al, 1999), suggesting the existence of evolutionarily conserved, intrachromosomal fragile sites that are used for reassembly in different species, and which may be particularly susceptible to pathologic rearrangements. Our understanding of genome reorganization during speciation remains rudimentary. A comparative genomic map of 19 mammals representing various placental orders illustrates strong associations between large segments of chromosomal DNA, such as regions that comprise portions of chromosomes 12 and 22 in humans, but which are joined in a single chromosome in lower primates (lemur and tree shrew), carnivores, ungulates (including cetaceans), bats, and rodents (O'Brien et al, 1999). In contrast, there does not appear to have been strong selection to retain the full structural association of the chromosomes involved in the CML (HSA 9 and HSA 22) translocation and the CLL deletion (HSA 13). Instead, these regions are near putative sites of genomic reorganization and may be susceptible not only to translocation or deletion, but also to other abnormalities that are presumably acquired during DNA replication and mitosis (such as a peculiar inversion at CFA 22q11). Similar mechanisms appear to be operative for rearrangements involving AML1 and ETO (on chromosomes 8 and 21 in humans and chromosomes 13 and 31 in dogs) in acute myelogenous leukemia and for deletion of PTEN (on HSA 10 and CFA 26) in various types of tumors. Together, this supports a notion that such

Modaino and Green: The Role of Evolution in Cancer Pathogenesis

Figure 1. Age at onset for canine lymphoma differs according to breed. The proportion of affected dogs for five dog breeds from a large collaborative study (Modiano et al, 2005b) is shown as a function of the age at which dogs were diagnosed (in years). There are significant differences among breeds, with three groups becoming readily apparent: dogs that develop lymphoma at a significantly younger age than the mean (and median) for all dogs (exemplified by Rottweilers and Boxers), dogs that are diagnosed when they are younger than the mean (and median) for all dogs, but older than the first group (exemplified here by Golden Retrievers, but also including Labrador Retrievers), and dogs that are representative for the mean (and median) of all breeds (exemplified here by German Shepherd Dogs). When all purebred dogs were considered together, this group was not significantly different from mixed breed dogs. This suggests that mix breed dog populations indeed approximate the mean of all the heritable contributions of purebred dogs, and that these contributions may have variably penetrance with few, if any, dominant alleles. In this sample set, Cocker Spaniels seem to be diagnosed even later than the mean (and median) for all dogs, but the difference between these groups was not statistically significant.

Recently, we showed that this breed predisposition could be extended to specific phenotypes (“B-cell” and “T-cell”) of lymphoid tumors (Modiano et al, 2005b). We showed that Spitz breeds and Asian “lap dogs” belonging to the oldest domestic dog groups (Parker et al, 2004) almost exclusively develop T-cell tumors, and some European breeds like Cocker Spaniels and Bassett Hounds almost exclusively develop B-cell tumors. Boxers and Golden Retrievers also show more T-cell disease than what is seen as an average for the population, and as shown in Figure 1, lymphoproliferative diseases occur earlier in life (Modiano et al, 2007), but in these breeds, the frequency of B-cell and T-cell phenotypes is more balanced (Modiano et al, 2005b). The prevalence of B-cell and T-cell lymphoma and leukemia in most other breeds is not significantly different from the average of all dogs or from that seen in dogs of mixed breeding (i.e., ~2:1 B-cell to T-cell). Environmental factors cannot account for the observed breed predilections for lymphoid tumors. There is, therefore, a basis for the hypothesis that heritable risk factors predispose dogs to develop B-cell or T-cell malignancies. Such risk factors are presumably mutations or epigenetic changes in genes that regulate lymphocyte development, although the involvement of genes that regulate the fidelity of the genome cannot be ruled out, as familial non-Hodgkin lymphomas in people are largely associated with conditions such as the p53 mutation in LiFraumeni syndrome (Segel and Lichtman, 2004; Siddiqui et al, 2004). Whether these factors arose ancestrally, or

developed more recently during the process of breed derivation, they are now firmly embedded in the genome, and their identification will provide information that will be valuable for prevention and treatment. While this task may seem difficult, we already have documented the occurrence of breed-specific cytogenetic changes in lymphoid tumors from Golden Retrievers (Modiano et al, 2005b), and analyses of pedigrees from Golden Retriever families suggest there is shared susceptibility to specific forms of lymphoma or leukemia among family members. The use of genomic and expression arrays provide contemporary tools that will allow us to refine familial data by tumor type and clinical response, thus providing greater insight and statistical power to find specific genes that are important in disease susceptibility and progression. For example, preliminary analyses of gene expression signatures of lymphoma samples from Golden Retrievers and non-Golden Retrievers shows a ~1.2 to 1.4fold decrease in expression of genes that localize to CFA 14, and a corresponding increase in expression of genes that localize to CFA 15 in the Golden Retrievers samples, with p vales <0.05 (Table 1). The data suggest there is a significant difference in expression levels, but the extent of the reduction is small. Several non-mutually exclusive possibilities can account for the small magnitude of these changes, including low levels of expression for genes along these chromosomes, loss of a single copy (or portion thereof) of CFA 14 in tumors from Golden Retrievers, leading to haploinsufficiency rather than absolute deletion of genes encoded therein, or epigenetic modification.

Cancer Therapy Vol 6-A.2, page 43 Table 1. Gene Expression Profiles from Golden Retriever Lymphomas Reflect Genomic Aberrationsa CFA

GENBANK

PATHWAY

P-VALUE FOLD-CHANGE

CO600038

RNA processing

0.007

-1.2

PMRNA4890

Cell division

0.013

-1.2

PMRNA4163

IDDM

0.016

-1.2

PMRNA5677

DNA binding and repair

0.019

-1.1

CX008060

Fatty acid metabolism and apoptosis

0.041

-1.2

PMRNA3988

gamma-HCCH/Bisp A degradation

0.044

-1.2

CF406744

Transcriptional control

0.005

1.3

CO706972

Pantothenate and CoA biosynthesis

0.014

1.1

DN362831

Ischemia/ reperfusion

0.018

1.1

PMRNA4530

Glucose transport

0.020

1.1

PMRNA7126

MAPK/Insulin signaling

0.037

1.1

PMRNA9940

?Chromosome segregation

0.040

1.1

Significantly different genes show reduced expression along CFA 14 and increased expression along CFA 15 in tumors from Golden Retrievers as compared to tumors from non-Golden Retrievers. The Genbank identifier and the biochemical pathways in which the genes are known to participate are shown, as is the p value from Studentâ&#x20AC;&#x2122;s T-test after normalization and the fold-change. The small magnitude of change reflects consistent values with little variance, and may be due to relatively low levels of expression or epigenetic regulation.

Intriguingly, complementary experiments suggest that loss of CFA 14 may be a more widespread occurrence in multiple types of tumors from Golden Retrievers. This is the subject of additional investigation by our groups. New molecular and phenotypic classification of lymphoma and leukemia also may assist in predictions regarding the course of disease and outcomes in response to treatment. A series of genomic microarrays have been developed to analyze DNA copy number aberrations associated with canine cancers (Thomas et al, 2003a,b, 2005, 2007), and are being used now to investigate the cytogenetics of numerous canine cancers, including lymphoma and leukemia, brain tumors, osteosarcoma, and soft tissue sarcomas. The genome assembly-integrated nature of the most recent iterations of these genomic arrays facilitates rapid transition from the identification of a region within a chromosome associated with cancer presentation to the precise location with the canine genome assembly, and thus to the identification of specific genes within those regions (Thomas et al, 2007). Simultaneously, studies are underway to assess the predictive ability of gene expression profiles for lymphoma and leukemia. We are using chromosomespecific FISH to directly define and catalogue altered genome organization within individual cells comprising the malignant mass (Thomas et al, 2001, 2003c; Milne et al, 2004; Breen and Modiano 2008). We also are using canine comparative genomic hybridization (CGH) analysis to determine DNA copy number status across the entire genome (Dunn et al, 2000; Thomas et al, 2001, 2003c, 2005, 2007). FISH analysis also has enabled the precise definition of structural chromosome aberrations that show a convincing evolutionary history between homologous cancers of dogs and people (Breen and Modiano, 2008).

These analyses can be used to predict both tumor origin and response to therapy. For instance, gain of dog chromosome 13 (CFA 13), particularly in a region syntenic to human chromosome 4q (HSA 4q), occurs in ~70% of canine diffuse B-cell lymphoma (Thomas et al, 2003c). This suggests that this region of the genome contains heretofore-unidentified genes that are etiologically and prognostically significant for this disease. There are also indications that gain of CFA 13 also is predictive for chemotherapy response (Hahn et al, 1994; Thomas et al, 2003c), perhaps because amplification of the c-myc and c-kit oncogenes, both of which are encoded in this region of CFA 13, increases the proliferative rate of the malignant cells and consequently their susceptibility to anti-mitotic compounds. These approaches also can be used to define specific chromosomal regions that are associated with heritable risk by identifying unique tumor genomes that segregate with selected breeds or groups, and that can pinpoint new regions for gene discovery (Modiano et al, 2005b). Although they are at earlier stages of development, tools such as arrays that can be used to analyze global gene expression are making rapid progress toward clinical application. The ability to analyze thousands of genes at once maximizes the efficiency with which we can identify genetic alterations associated with tumor pathogenesis, as well as prognostic â&#x20AC;&#x153;gene signaturesâ&#x20AC;? (Rosenwald et al, 2002). Importantly, even simple tests that can be used as surrogates for cytogenetic changes such as the deletion of CFA11 in high-grade T-cell lymphomas have predictive value for this disease (Modiano et al, 2007). A contemporary area of emphasis is to define whether these properties are inherent to cancer cells or whether they arise by natural selection and clonal 43

Modaino and Green: The Role of Evolution in Cancer Pathogenesis by targeting BCL2. Proc Natl Acad Sci U S A 102, 1394413949. Comstock KE, Lingaas F, Kirkness EF, Hitte C, Thomas R, Breen M, Galibert F, Ostrander EA (2004) A high-resolution comparative map of canine Chromosome 5q14.3-q33 constructed utilizing the 1.5x canine genome sequence. Mamm Genome 15, 544-551. Croce CM, Tsujimoto Y, Erikson J, Nowell P (1984) Chromosome translocations and B cell neoplasia. Lab Invest 51, 258-267. Dohner H, Stilgenbauer S, Benner A, Leupolt E, Krober A, Bullinger L, Dohner K, Bentz M, Lichter P (2000) Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med 343, 1910-1916. Dorn CR, Taylor DO, Schneider R, Hibbard HH, Klauber MR (1968) Survey of animal neoplasms in Alameda and Contra Costa Counties, California. II. Cancer morbidity in dogs and cats from Alameda County. J Natl Cancer Inst 40, 307-318. Dunn KA, Thomas R, Binns MM, Breen M (2000) Comparative genomic hybridization (CGH) in dogs--application to the study of a canine glial tumour cell line. Vet J 160, 77-82. Fearon ER (1997) Human cancer syndromes: clues to the origin and nature of cancer. Science 278, 1043-1050. Finger LR, Harvey RC, Moore RC, Showe LC, Croce CM (1986) A common mechanism of chromosomal translocation in Tand B-cell neoplasia. Science 234, 982-985. Fosmire SP, Thomas R, Jubala CM, Wojcieszyn JW, Valli VE, Getzy DM, Smith TL, Gardner LA, Ritt MG, Bell JS, Freeman KP, Greenfield BE, Lana SE, Kisseberth WC, Helfand SC, Cutter GR, Breen M, Modiano JF (2007) Inactivation of the p16 cyclin-dependent kinase inhibitor in high-grade canine non-Hodgkin's T-cell lymphoma. Vet Pathol 44, 467-478. Garrett LD, Thamm DH, Chun R, Dudley R, Vail DM (2002) Evaluation of a 6-month chemotherapy protocol with no maintenance therapy for dogs with lymphoma. J Vet Intern Med 16, 704-709. Glickman L, Glickman N, Thorpe R (2000) The Golden Retriever Club of America National Health Survey, GRCA. 2000. Hahn KA, Richardson RC, Hahn EA, Chrisman CL (1994) Diagnostic and prognostic importance of chromosomal aberrations identified in 61 dogs with lymphosarcoma. Vet Pathol 31, 528-540. Hansen K, Khanna C (2004) Spontaneous and genetically engineered animal models; use in preclinical cancer drug development. Eur J Cancer 40, 858-880. Hecht JL, Aster JC (2000) Molecular biology of Burkitt's lymphoma. J Clin Oncol 18, 3707-3721. Hino O (2003) Hereditary renal carcinogenesis fitting Knudson's two-hit model: genotype, environment, and phenotype. Genes Chromosomes Cancer 38, 357-367. Khanna C, Lindblad-Toh K, Vail D, London C, Bergman P, Barber L, Breen M, Kitchell B, McNeil E, Modiano JF, Niemi S, Comstock KE, Ostrander E, Westmoreland S, Withrow S (2006) The dog as a cancer model. Nat Biotechnol 24, 1065-1066. Kirkness EF, Bafna V, Halpern AL, Levy S, Remington K, Rusch DB, Delcher AL, Pop M, Wang W, Fraser CM, Venter JC (2003) The dog genome: survey sequencing and comparative analysis. Science 301, 1898-1903. Kurzrock R, Kantarjian HM, Druker BJ, Talpaz M (2003) Philadelphia chromosome-positive leukemias: from basic mechanisms to molecular therapeutics. Ann Intern Med 138, 819-830. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, Funke R, Gage D, Harris K, Heaford A, Howland J, Kann L,

evolution. We can use molecular tools to predict risk, prognosis, and response to therapy in some cancers of companion animals, and we believe the availability and usefulness of such tools in clinical practice will expand rapidly. Therefore, as we improve our understanding of basic mechanisms that account for malignant transformation and tumor progression, we will be able to design even better strategies for cancer prevention and therapy.

VII. Conclusions In summary, we describe the identification of evolutionarily conserved chromosomal abnormalities associated with specific leukemia and lymphoma phenotypes in humans and dogs. We propose that a previously unidentified mechanism, which involves inappropriate reassembly of chromosomal regions that are involved in genome reorganization across species, may account for some of these abnormalities, offering a plausible explanation for the remarkable consistency and specificity of these cytogenetic signatures in human and canine tumors. Moreover, the predisposition of dog breeds to develop select types of cancer (Modiano et al, 2005b) suggests that, through the process of breed development, dogs have acquired genomes that render specific cells distinctly susceptible to malignant transformation. The presence of evolutionarily conserved chromosome aberrations in naturally occurring cancers will thus allow us to determine if the apparent structural â&#x20AC;&#x2DC;instabilityâ&#x20AC;&#x2122; in these regions is indeed a conserved feature of the ancestral genome. Furthermore, a thorough investigation of recurrent breakpoint regions in canine tumor genomes, especially those that have high breed specificity, offers the potential to define regions that contain genes that have so far not been associated with tumorigenesis in humans or other species (Modiano et al, 2005b). Evaluation of these genes may hence shed light on the observed disparity in response to treatment among different dog breeds (Garrett et al, 2002) and among people diagnosed with leukemia and lymphoma.

References Breen M, Modiano JF (2008) Evolutionarily conserved cytogenetic changes in hematological malignancies of dogs and humans - man and his best friend share more than companionship. Chromosome Res 16, 145-154. Breen M, Thomas R, Binns MM, Carter NP, Langford CF (1999) Reciprocal chromosome painting reveals detailed regions of conserved synteny between the karyotypes of the domestic dog (Canis familiaris) and human. Genomics 61, 145-155. Burnett RC, Vernau W, Modiano JF, Olver CS, Moore PF, Avery AC (2003) Diagnosis of canine lymphoid neoplasia using clonal rearrangements of antigen receptor genes. Vet Pathol 40, 32-41. Calin GA, Croce CM (2006) Genomics of chronic lymphocytic leukemia microRNAs as new players with clinical significance. Semin Oncol 33, 167-173. Cimmino A, Calin GA, Fabbri M, Iorio MV, Ferracin M, Shimizu M, Wojcik SE, Aqeilan RI, Zupo S, Dono M, Rassenti L, Alder H, Volinia S, Liu CG, Kipps TJ, Negrini M, Croce CM (2005) miR-15 and miR-16 induce apoptosis

Cancer Therapy Vol 6-A.2, page 45 Lehoczky J, LeVine R, McEwan P, McKernan K, Meldrim J, Mesirov JP, Miranda C, Morris W, Naylor J, Raymond C, Rosetti M, Santos R, Sheridan A, Sougnez C, StangeThomann N, Stojanovic N, Subramanian A, Wyman D, Rogers J, Sulston J, Ainscough R, Beck S, Bentley D, Burton J, Clee C, Carter N, Coulson A, Deadman R, Deloukas P, Dunham A, Dunham I, Durbin R, French L, Grafham D, Gregory S, Hubbard T, Humphray S, Hunt A, Jones M, Lloyd C, McMurray A, Matthews L, Mercer S, Milne S, Mullikin JC, Mungall A, Plumb R, Ross M, Shownkeen R, Sims S, Waterston RH, Wilson RK, Hillier LW, McPherson JD, Marra MA, Mardis ER, Fulton LA, Chinwalla AT, Pepin KH, Gish WR, Chissoe SL, Wendl MC, Delehaunty KD, Miner TL, Delehaunty A, Kramer JB, Cook LL, Fulton RS, Johnson DL, Minx PJ, Clifton SW, Hawkins T, Branscomb E, Predki P, Richardson P, Wenning S, Slezak T, Doggett N, Cheng JF, Olsen A, Lucas S, Elkin C, Uberbacher E, Frazier M, Gibbs RA, Muzny DM, Scherer SE, Bouck JB, Sodergren EJ, Worley KC, Rives CM, Gorrell JH, Metzker ML, Naylor SL, Kucherlapati RS, Nelson DL, Weinstock GM, Sakaki Y, Fujiyama A, Hattori M, Yada T, Toyoda A, Itoh T, Kawagoe C, Watanabe H, Totoki Y, Taylor T, Weissenbach J, Heilig R, Saurin W, Artiguenave F, Brottier P, Bruls T, Pelletier E, Robert C, Wincker P, Smith DR, Doucette-Stamm L, Rubenfield M, Weinstock K, Lee HM, Dubois J, Rosenthal A, Platzer M, Nyakatura G, Taudien S, Rump A, Yang H, Yu J, Wang J, Huang G, Gu J, Hood L, Rowen L, Madan A, Qin S, Davis RW, Federspiel NA, Abola AP, Proctor MJ, Myers RM, Schmutz J, Dickson M, Grimwood J, Cox DR, Olson MV, Kaul R, Shimizu N, Kawasaki K, Minoshima S, Evans GA, Athanasiou M, Schultz R, Roe BA, Chen F, Pan H, Ramser J, Lehrach H, Reinhardt R, McCombie WR, de la Bastide M, Dedhia N, Blocker H, Hornischer K, Nordsiek G, Agarwala R, Aravind L, Bailey JA, Bateman A, Batzoglou S, Birney E, Bork P, Brown DG, Burge CB, Cerutti L, Chen HC, Church D, Clamp M, Copley RR, Doerks T, Eddy SR, Eichler EE, Furey TS, Galagan J, Gilbert JG, Harmon C, Hayashizaki Y, Haussler D, Hermjakob H, Hokamp K, Jang W, Johnson LS, Jones TA, Kasif S, Kaspryzk A, Kennedy S, Kent WJ, Kitts P, Koonin EV, Korf I, Kulp D, Lancet D, Lowe TM, McLysaght A, Mikkelsen T, Moran JV, Mulder N, Pollara VJ, Ponting CP, Schuler G, Schultz J, Slater G, Smit AF, Stupka E, Szustakowski J, Thierry-Mieg D, Thierry-Mieg J, Wagner L, Wallis J, Wheeler R, Williams A, Wolf YI, Wolfe KH, Yang SP, Yeh RF, Collins F, Guyer MS, Peterson J, Felsenfeld A, Wetterstrand KA, Patrinos A, Morgan MJ, Szustakowki J, de Jong P, Catanese JJ, Osoegawa K, Shizuya H, Choi S, Chen YJ (2001) Initial sequencing and analysis of the human genome. Nature 409, 860-921. Lindblad-Toh K, Wade CM, Mikkelsen TS, Karlsson EK, Jaffe DB, Kamal M, Clamp M, Chang JL, Kulbokas EJ, 3rd, Zody MC, Mauceli E, Xie X, Breen M, Wayne RK, Ostrander EA, Ponting CP, Galibert F, Smith DR, DeJong PJ, Kirkness E, Alvarez P, Biagi T, Brockman W, Butler J, Chin CW, Cook A, Cuff J, Daly MJ, DeCaprio D, Gnerre S, Grabherr M, Kellis M, Kleber M, Bardeleben C, Goodstadt L, Heger A, Hitte C, Kim L, Koepfli KP, Parker HG, Pollinger JP, Searle SM, Sutter NB, Thomas R, Webber C, Baldwin J, Abebe A, Abouelleil A, Aftuck L, Ait-Zahra M, Aldredge T, Allen N, An P, Anderson S, Antoine C, Arachchi H, Aslam A, Ayotte L, Bachantsang P, Barry A, Bayul T, Benamara M, Berlin A, Bessette D, Blitshteyn B, Bloom T, Blye J, Boguslavskiy L, Bonnet C, Boukhgalter B, Brown A, Cahill P, Calixte N, Camarata J, Cheshatsang Y, Chu J, Citroen M, Collymore A, Cooke P, Dawoe T, Daza R, Decktor K, DeGray S, Dhargay N, Dooley K, Dooley K, Dorje P, Dorjee K, Dorris L, Duffey N, Dupes A, Egbiremolen O, Elong R, Falk J, Farina A, Faro

S, Ferguson D, Ferreira P, Fisher S, FitzGerald M, Foley K, Foley C, Franke A, Friedrich D, Gage D, Garber M, Gearin G, Giannoukos G, Goode T, Goyette A, Graham J, Grandbois E, Gyaltsen K, Hafez N, Hagopian D, Hagos B, Hall J, Healy C, Hegarty R, Honan T, Horn A, Houde N, Hughes L, Hunnicutt L, Husby M, Jester B, Jones C, Kamat A, Kanga B, Kells C, Khazanovich D, Kieu AC, Kisner P, Kumar M, Lance K, Landers T, Lara M, Lee W, Leger JP, Lennon N, Leuper L, LeVine S, Liu J, Liu X, Lokyitsang Y, Lokyitsang T, Lui A, Macdonald J, Major J, Marabella R, Maru K, Matthews C, McDonough S, Mehta T, Meldrim J, Melnikov A, Meneus L, Mihalev A, Mihova T, Miller K, Mittelman R, Mlenga V, Mulrain L, Munson G, Navidi A, Naylor J, Nguyen T, Nguyen N, Nguyen C, Nguyen T, Nicol R, Norbu N, Norbu C, Novod N, Nyima T, Olandt P, O'Neill B, O'Neill K, Osman S, Oyono L, Patti C, Perrin D, Phunkhang P, Pierre F, Priest M, Rachupka A, Raghuraman S, Rameau R, Ray V, Raymond C, Rege F, Rise C, Rogers J, Rogov P, Sahalie J, Settipalli S, Sharpe T, Shea T, Sheehan M, Sherpa N, Shi J, Shih D, Sloan J, Smith C, Sparrow T, Stalker J, Stange-Thomann N, Stavropoulos S, Stone C, Stone S, Sykes S, Tchuinga P, Tenzing P, Tesfaye S, Thoulutsang D, Thoulutsang Y, Topham K, Topping I, Tsamla T, Vassiliev H, Venkataraman V, Vo A, Wangchuk T, Wangdi T, Weiand M, Wilkinson J, Wilson A, Yadav S, Yang S, Yang X, Young G, Yu Q, Zainoun J, Zembek L, Zimmer A, Lander ES (2005) Genome sequence, comparative analysis and haplotype structure of the domestic dog. Nature 438, 803-819. Lingaas F, Comstock KE, Kirkness EF, Sorensen A, Aarskaug T, Hitte C, Nickerson ML, Moe L, Schmidt LS, Thomas R, Breen M, Galibert F, Zbar B, Ostrander EA (2003) A mutation in the canine BHD gene is associated with hereditary multifocal renal cystadenocarcinoma and nodular dermatofibrosis in the German Shepherd dog. Hum Mol Genet 12, 3043-3053. MacEwen EG (1990) Spontaneous tumors in dogs and cats: models for the study of cancer biology and treatment. Cancer Metastasis Rev 9, 125-136. McKeithan TW, Shima EA, Le Beau MM, Minowada J, Rowley JD, Diaz MO (1986) Molecular cloning of the breakpoint junction of a human chromosomal 8;14 translocation involving the T-cell receptor alpha-chain gene and sequences on the 3' side of MYC. Proc Natl Acad Sci U S A 83, 66366640. Milne BS, Hoather T, O'Brien PCM, Yang F, Ferguson-Smith MA, Dobson J, Sargan D (2004) Karyotype of canine soft tissue sarcomas: a multicolor, multi-species approach to canine chromosome painting. Chromosome Res 12, 825835. Modiano JF, Breen M, Avery AC, London CA (2005a). Breed Specific Canine Lymphoproliferative Diseases. The Dog and its Genome. Ostrander EA, Giger U, Lindblad-Toh K, Cold Spring Harbor, CSH Press. Modiano JF, Breen M, Burnett RC, Parker HG, Inusah S, Thomas R, Avery PR, Lindblad-Toh K, Ostrander EA, Cutter GC, Avery AC (2005b) Distinct B-cell and T-cell lymphoproliferative disease prevalence among dog breeds indicates heritable risk. Cancer Res 65, 5654-5661. Modiano JF, Breen M, Valli VE, Wojcieszyn JW, Cutter GR (2007) Predictive value of p16 or Rb inactivation in a model of naturally occurring canine non-Hodgkin's lymphoma. Leukemia 21, 184-187. Murphy WJ, Larkin DM, Everts-van der Wind A, Bourque G, Tesler G, Auvil L, Beever JE, Chowdhary BP, Galibert F, Gatzke L, Hitte C, Meyers SN, Milan D, Ostrander EA, Pape G, Parker HG, Raudsepp T, Rogatcheva MB, Schook LB, Skow LC, Welge M, Womack JE, O'Brien S J, Pevzner PA,

Modaino and Green: The Role of Evolution in Cancer Pathogenesis Lewin HA (2005) Dynamics of mammalian chromosome evolution inferred from multispecies comparative maps. Science 309, 613-617. Navid F, Mosijczuk AD, Head DR, Borowitz MJ, Carroll AJ, Brandt JM, Link MP, Rozans MK, Thomas GA, Schwenn MR, Shields DJ, Vietti TJ, Pullen DJ (1999) Acute lymphoblastic leukemia with the (8;14)(q24;q32) translocation and FAB L3 morphology associated with a Bprecursor immunophenotype: the Pediatric Oncology Group experience. Leukemia 13, 135-141. O'Brien SJ, Eisenberg JF, Miyamoto M, Hedges SB, Kumar S, Wilson DE, Menotti-Raymond M, Murphy WJ, Nash WG, Lyons LA, Menninger JC, Stanyon R, Wienberg J, Copeland NG, Jenkins NA, Gellin J, Yerle M, Andersson L, Womack J, Broad T, Postlethwait J, Serov O, Bailey E, James MR, Watanabe TK, Wakefield MJ, Marshall Graves JA (1999) Genome maps 10. Comparative genomics. Mammalian radiations. Wall chart. Science 286, 463-478. Okimoto K, Sakurai J, Kobayashi T, Mitani H, Hitrayama Y, Nickerson ML, Warren MB, Zbar B, Schmidt LS, Hino O (2004) A germ-line insertion in the Birt-Hogg-DubĂŠ (BHD) gene gives rise to the Nihon rat model of inherited renal cancer. Proc Natl Acad Sci U S A 101, 2023-2027. Onions DE (1984) A prospective survey of familial canine lymphosarcoma. J Natl Cancer Inst 72, 909-912. Paoloni M, Khanna C (2008) Translation of new cancer treatments from pet dogs to humans. Nat Rev Cancer 8, 147-156. Parker HG, Kim LV, Sutter NB, Carlson S, Lorentzen TD, Malek TB, Johnson GS, DeFrance HB, Ostrander EA, Kruglyak L (2004) Genetic structure of the purebred domestic dog. Science 304, 1160-1164. Pfeiffer C, Stein J, Southwood S, Ketelaar H, Sette A, Bottomly K (1995) Altered peptide ligands can control CD4 T lymphocyte differentiation in vivo. J Exp Med 181, 15691574. Priester WA, McKay FW (1980) The occurrence of tumors in domestic animals. Natl Cancer Inst Monogr 54, 1-210. Raveche ES, Salerno E, Scaglione BJ, Manohar V, Abbasi F, Lin YC, Fredrickson T, Landgraf P, Ramachandra S, Huppi K, Toro JR, Zenger VE, Metcalf RA, Marti GE (2007) Abnormal microRNA-16 locus with synteny to human 13q14 linked to CLL in NZB mice. Blood 109, 5979-5986 Reiter A, Saussele S, Grimwade D, Wiemels JL, Segal MR, Lafage-Pochitaloff M, Walz C, Weisser A, Hochhaus A, Willer A, Reichert A, Buchner T, Lengfelder E, Hehlmann R, Cross NC (2003) Genomic anatomy of the specific reciprocal translocation t(15;17) in acute promyelocytic leukemia. Genes Chromosomes Cancer 36, 175-188. Rosenwald A, Wright G, Chan WC, Connors JM, Campo E, Fisher RI, Gascoyne RD, Muller-Hermelink HK, Smeland EB, Giltnane JM, Hurt EM, Zhao H, Averett L, Yang L, Wilson WH, Jaffe ES, Simon R, Klausner RD, Powell J, Duffey PL, Longo DL, Greiner TC, Weisenburger DD, Sanger WG, Dave BJ, Lynch JC, Vose J, Armitage JO, Montserrat E, Lopez-Guillermo A, Grogan TM, Miller TP, LeBlanc M, Ott G, Kvaloy S, Delabie J, Holte H, Krajci P, Stokke T, Staudt LM (2002) The use of molecular profiling to predict survival after chemotherapy for diffuse large-Bcell lymphoma. N Engl J Med 346, 1937-1947.

Schlegelberger B, Himmler A, Godde E, Grote W, Feller AC, Lennert K (1994) Cytogenetic findings in peripheral T-cell lymphomas as a basis for distinguishing low-grade and highgrade lymphomas. Blood 83, 505-511. Segel GB, Lichtman MA (2004) Familial (inherited) leukemia, lymphoma, and myeloma: an overview. Blood Cells Mol Dis 32, 246-261. Shima EA, Le Beau MM, McKeithan TW, Minowada J, Showe LC, Mak TW, Minden MD, Rowley JD, Diaz MO (1986) Gene encoding the alpha chain of the T-cell receptor is moved immediately downstream of c-myc in a chromosomal 8;14 translocation in a cell line from a human T-cell leukemia. Proc Natl Acad Sci U S A 83, 3439-3443. Siddiqui R, Onel K, Facio F, Offit K (2004) The genetics of familial lymphomas. Curr Oncol Rep 6, 380-387. Teske E, de Vos JP, Egberink HF, Vos JH (1994) Clustering in canine malignant lymphoma. Vet Q 16, 134-136. Thomas R, Bridge W, Benke K, Breen M (2003a) Isolation and chromosomal assignment of canine genomic BAC clones representing 25 cancer-related genes. Cytogenet Genome Res 102, 249-253. Thomas R, Duke S, Bloom S, Breen T, Young AC, Feiste E, Seiser EL, C. TP, Langford CF, Ellis P, Karlsson EK, Lindblad-Toh K, Breen M (2007) A cytogenetically characterized, genome anchored 10-Mb BAC array for the domestic dog. J Hered 98, 474-484. Thomas R, Fiegler H, Ostrander EA, Galibert F, Carter NP, Breen M (2003b) A canine cancer-gene microarray for CGH analysis of tumors. Cytogenet Genome Res 102, 254-260. Thomas R, Scott A, Langford CF, Fosmire SP, Jubala CM, Lorentzen TD, Hitte C, Karlsson EK, Kirkness E, Ostrander EA, Galibert F, Lindblad-Toh K, Modiano JF, Breen M (2005) Construction of a 2-Mb resolution BAC microarray for CGH analysis of canine tumors. Genome Res 15, 18311837. Thomas R, Smith KC, Gould R, Gower SM, Binns MM, Breen M (2001) Molecular cytogenetic analysis of a novel highgrade canine T-lymphoblastic lymphoma demonstrating coexpression of CD3 and CD79a cell markers. Chromosome Res 9, 649-657. Thomas R, Smith KC, Ostrander EA, Galibert F, Breen M (2003c) Chromosome aberrations in canine multicentric lymphomas detected with comparative genomic hybridisation and a panel of single locus probes. Br J Cancer 89, 1530-1537. Thomas R, Duke SE, Karlsson E, Evans A, Ellis P, Lindblad-Toh K, Langford CF, Breen M (2007) Development of a 1Mb resolution genome integrated canine BAC resource application for cytogenetic analysis and array based comparative genomic hybridization. In press Wertheim JA, Miller JP, Xu L, He Y, Pear WS (2002) The biology of chronic myelogenous leukemia: mouse models and cell adhesion. Oncogene 21, 8612-8628. Yang F, O'Brien PC, Milne BS, Graphodatsky AS, Solanky N, Trifonov V, Rens W, Sargan D, Ferguson-Smith MA (1999) A complete comparative chromosome map for the dog, red fox, and human and its integration with canine genetic maps. Genomics 62, 189-202.

Cancer Therapy Vol 6-A.2, page 47 Cancer Therapy Vol 6, 263-270, 2008

Immunological concepts applied to pathologic diagnosis of proliferative diseases of the immune system Review Article

Peter F. Moore Department of Pathology, Microbiology and Immunology, School of Veterinary Medicine, University of California, Davis CA 95616, USA

__________________________________________________________________________________ *Correspondence:Peter F. Moore, Professor, Department of Pathology, Microbiology and Immunology 3315 Vet Med 3A, School of Veterinary Medicine, University of California, Davis, CA 95616, USA; Tel: 530 752 5204; Fax: 530 752 3349; e-mail: pfmoore@ucdavis.edu Key words: functional significance, immune system, Lymphocyte antigen receptor, lymphoma, diagnosis, Cell lineage Abbreviations: antigen presenting cells (APC); Canine cutaneous histiocytoma (CCH); chronic lymphocytic leukemia of T cell, (TCLL); Cluster of differentiation (CD); Cluster of differentiation (CD); complementarity determining region 3 (CDR3); dendritic cells (DC); heavy chain Ig genes (VHDJ); histiocytic sarcomas (HS); immunoglobulin (Ig); immunoglobulin heavy chain locus (IGH); intestinal epithelial compartment (IEC); intra-epithelial lymphocytes (IEL); large granular lymphocyte (LGL); light chain Ig genes (VLJ); major histocompatibility complex, (MHC); mucosal addressin cell adhesion molecule (MadCAM); natural killer (NK); polymerase chain reaction (PCR); T cell receptor molecules (TCR) Received: 21 February 2008; electronically published: June 2008

Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary Investigation of immunopathological diseases often entails precise identification of lineages of infiltrative cells and the nature of the disease process â&#x20AC;&#x201C; inflammatory or neoplastic. The development of the Cluster of differentiation (CD) antigen system and the associated monoclonal antibodies specific for CD molecules has facilitated immunostaining for the detection of the lineages of cells in hemopoietic neoplasia of animals (immunophenotyping). This has been important in the classification of lymphomas and proliferative histiocytic diseases in dogs and cats. The distinction of lymphocyte dominant inflammation from lymphoma has been facilitated by the development of methods to readily assess lymphocyte antigen receptor gene rearrangements. These assessments are almost exclusively performed by polymerase chain reaction (PCR) methodology with primers designed to amplify the T cell receptor ! locus (TCRG) for T cell clonality assessment, and the immunoglobulin heavy chain locus (IGH) for B cell clonality assessment. It is important to recognize that cell lineage is best determined by immunophenotypic analysis, and that molecular clonality assessment should be performed in the light of a full appreciation of the clinical and morphological features of the disease process.

rearrangement by polymerase chain reaction (PCR) provides valuable information pertaining to antigen receptor diversity in the lesion. Specifically, the detection of lymphocyte clones in lymphoproliferative disease, in the appropriate clinical and morphological context, is consistent with lymphoma. Detection of clonality in nonlymphoid proliferative diseases such as histiocytic proliferative disease is more challenging and has not been routinely applied in diagnosis in animals. The nomenclature and complexity pertaining to the biology of leukocyte surface molecules is intimidating.

I. Introduction The crux of investigation of immunopathological diseases involves identification of the cells involved and the process, inflammatory (reactive) or neoplastic. Cell lineage determination in the immune system has advanced since the inception of the â&#x20AC;&#x153;cluster of differentiationâ&#x20AC;? (CD) antigen system. Precise cell identification can be achieved by the application of monoclonal antibodies (mAb) specific for leukocyte antigens to isolated cells, cell smears and tissue sections. In the case of lymphoid proliferation, assessment of B or T cell antigen receptor 47

Moore: Immunological concepts applied to pathologic diagnosis There are currently 350 CDs assigned in the human immune system, as well as many other defined molecules that have not yet been assigned to clusters of differentiation (Zola et al 2005, 2007). Comprehensive reports of recent leukocyte antigen workshops have been published for ruminants (Naessens et al, 1997), swine (Haverson et al, 2001), and horses (Lunn et al, 1998). A single workshop was convened to identify canine leukocyte antigens (Cobbold and Metcalfe, 1994) and none has been conducted for cats. Despite the large number of CD molecules identified, not all are useful or available for use in immunodiagnostics. In most instances, CD molecules are detectable by mAb, which react only with native antigen in unfixed cells, which can include unfixed air dried cytological preparations, anti-coagulated blood or bone marrow, and fresh tissue which has been carefully snap frozen and sectioned. In some instances, antibodies that detect epitopes resistant to the deleterious effects of formalin fixation have been developed; these are valuable reagents for the study of archived tissue in paraffin blocks and routinely processed pathological material (Moore et al, 1998).

they assist B cell activation, affinity maturation and immunoglobulin class switching in response to antigen (Janeway et al, 2001). CD8 T cells interact with MHC class I peptide complexes on almost any cell including APC. Hence, CD8 cytotoxic T cells are well suited to the task of clearing virally infected cells and cancer cells via detection of virally encoded peptides or peptides derived from mutated self proteins presented in the peptide binding groove of MHC class I (Janeway et al, 2001). The identification of histiocytes in tissue sections also relies on the identification of functionally important molecules on these cells. Histiocytes include macrophages and dendritic cells (DC). Dendritic cells are the most potent APC. Nomenclature of DC populations is based on the differentiation pathway followed by the cell (the cell lineage), and by the location in tissue. There are 3 DC lineages: interstitial DC, epithelial localized DC (Langerhans cells or LC) and plasmacytoid DC. MHC class I and class II, together with CD1, are the molecules responsible for presentation of peptides, lipids and glycolipids to T cells. Hence, DC in humans and dogs are best defined by their abundant expression of molecules essential to their function as antigen presenting cells. Of these, the family of CD1 proteins is more or less restricted to DC and other APC such as subsets of B cells and monocytes; while MHC class I and II are more broadly expressed. Based on CD1 expression, DC of canine skin occur in 2 major locations: within the epidermis (LC), and within the dermis especially adjacent to postcapillary venules (dermal interstitial DC) (Moore and Mariassy, 1986; Moore et al, 1996). The expression of #2 integrins is differentially regulated in normal canine macrophages and DC; CD11c is frequently expressed by DC; while macrophages predominately express CD11b (or CD11d in the splenic red pulp and bone marrow (Danilenko et al, 1992, 1995). Langerhans cells and dermal DC are distinguishable by their Thy-1 (CD90) expression; epidermal LC lack Thy-1 and dermal DC express abundant Thy-1 (Moore et al, 1996). Epidermal LC also express E-cadherin, which assists in their localization in the epidermis via a homotypic adhesive interaction with Ecadherin expressed by keratinocytes (Borkowski et al, 1994; Blauvelt et al, 1995). Regardless, it is important to realize that DC arise in bone marrow and migrate through blood to a variety of epithelial sites (cutaneous and mucosal), where they take up residence either within epithelia or in dermis and lamina propria. In these sites they function as antigen processing and ultimately antigen presenting cells, which interact with T cells. Migration of cutaneous DC (as veiled cells) via lymphatics to the paracortex of lymph nodes occurs following contact with antigen. Dendritic cells alter their chemokine receptor expression profile (upregulate CCR7) and are attracted to chemokine ligands displayed on the surface of lymphatic endothelial cells (CCL21) and on stromal cells in the paracortex of lymph nodes(Cyster, 1999; Kellermann et al, 1999). Dendritic cells differentiate into potent APC during this migration and change their surface phenotype accordingly (see below). The interdigitating dendritic APC of lymph node paracortex are partially derived from such migration. Another major

II. Cell lineages revealed by markers of functional significance Very few leukocyte antigens are expressed only by a single lineage of cells. For example, T cell receptor molecules (TCR) are only expressed by T cells, but the associated signaling molecule, CD3 epsilon, is expressed on the surface of T cells, and in the cytoplasm of activated natural killer (NK) cells (Lanier et al, 1992). CD3 expression has also been observed in human NK lymphoma (Suzumiya et al, 1994). For many other CD molecules, the cellular expression patterns are promiscuous. Pathologists wishing to use utilize leukocyte antigen expression patterns in diseased tissues for cell lineage determination, must utilize a combinatorial logic based on the examination of a diagnostically relevant region in multiple sections stained with members of a panel of mAb specific for leukocyte antigens. This is distinct and opposite of the logic applied to the application of special histochemical stains to tissue sections. Simply stated, â&#x20AC;&#x153;the one stain - one cellâ&#x20AC;? paradigm must be abandoned in the former context. Cell lineage is best determined by evaluation of the expression pattern of leukocyte antigens that are functionally important for a particular cell. The use of arcane markers with unknown functional significance is usually less valuable. In this context, identification of T cells and important T cell subsets is best achieved by demonstration of components of the TCR/CD3 complex on the surface of the cell, in association with other functionally important TCR/CD3 co-receptor molecules such as CD4 (helper T cells) or CD8" (cytotoxic T cells) (Janeway et al, 2001). These co-receptor molecules are involved in interaction of T cells with antigen presenting cells (APC). CD4 T cells interact with major histocompatibility complex (MHC) class II peptide complexes on the surface of APC. CD4 helper T cells are critically involved in humoral immune responses in which

Cancer Therapy Vol 6-A.2, page 49 population of DC occurs in lymph nodes; these cells do not arrive after migration from tissues. Instead they enter lymph nodes from blood as immature DC precursors, which differ from other DC in their surface phenotype. These DC do not express many of the lineage determining molecules of tissue DC, rather, they are relatively undifferentiated upon arrival in lymph nodes. They express abundant CD4, CD45RA and CD123, the IL-3 receptor " subunit, indicating that these cells differentiate under a different cytokine/growth factor milieu than tissue DC (Olweus et al, 1997). Accordingly, these DC have been named plasmacytoid DC (Rothenfusser et al, 2002). Plasmacytoid DC have not been identified in the canine due to lack of available reagents specific for key identifying surface molecules (esp. CD123). Successful interaction of DC and T cells in response to antigenic challenge also involves the orderly appearance of co-stimulatory molecules (B7 family) on DC, and their ligands (CD28 and CTLA-4) on T cells (Sethna et al, 1994; Walunas et al, 1994; Tivol et al, 1995). Defective interaction of DC and T cells appears to contribute to the development of reactive histiocytic proliferative diseases (cutaneous and systemic histiocytosis), which are related diseases arising out of disordered immune regulation. Recognition of this has led to the development of effective treatments for reactive histiocytic proliferative diseases; these treatments rely on potent immunosuppression, which may need to be lifelong in severe cases. Much research is currently directed at the molecular events associated with maturation and migration of DC, and the co-stimulatory function of DC. To some extent this research should be of future benefit to understanding the pathogenesis of reactive proliferative disorders of DC in both human and dog. Aspects of the developmental and migratory program of cutaneous DC are recapitulated in the DC proliferative disorders of canine skin. Canine cutaneous histiocytoma (CCH) is a LC disorder, in which tumor histiocytes express CD1, CD11c, MHC II and E-cadherin. Migration (metastasis) of CCH to lymph nodes is observed, although rarely. Tumor histiocytes in these instances are likely following the migratory route of normal LC after contact with antigen (Moore et al, 1996, 1998). Accordingly, CCR7 transcripts have been demonstrated in CCH (Moore unpublished data). Tumor LC obliterate affected lymph nodes. In some instances, spontaneous regression occurs; this is the usual outcome in cutaneous tumors and the time course of regression in lymph nodes is similar. In other instances tumor LC persist in lymph nodes, but do not usually extend beyond them. Regression does not occur in these cases and many of these dogs are euthanized. Normal DC do not recirculate following migration to lymph nodes; efferent lymph lacks significant numbers of DC, and most migratory DC undergo apoptosis following interaction with paracortical T cells in the process of antigen presentation. The distinctive surface antigen phenotype of canine LC has also enabled recognition of more aggressive LC proliferative disorders in dogs similar to aggressive forms of Langerhans cell histiocytosis in humans. In canine LCH, widespread cutaneous lesions are present initially; this phase is followed by a phase of rapid

dissemination to lymph nodes and internal organs (especially lung). Canine LCH has a poor prognosis; all afflicted dogs have been euthanized (Moore and Affolter, unpublished data). Most canine histiocytic sarcomas (HS) involve infiltration and expansion of cells of DC lineage (Affolter and Moore, 2002). These include localized and disseminated HS; the latter is often equated with malignant hisiocytosis. However, proliferative disorders of macrophage lineage have also been recognized. The best example is the hemophagocytic HS (hemophagocytic malignant histiocytosis) of dogs (Moore et al, 2006). In most instances this disorder is associated with expansion of splenic red pulp macrophages, often coincident with expansion of bone marrow macrophages, both of which manifest prominent erythrophagocytosis. Application of #2 integrin markers to frozen or even formalin fixed tissue sections is of value in recognizing the lineage and origin of the histiocytic population (Danilenko et al, 1992). While macrophages in hemophagocytic HS do not express abundant CD1 as expected of DC, they do prominently express CD11d and CD18, which are diffusely expressed by normal resident macrophages of splenic red pulp and bone marrow (Danilenko et al, 1995; Moore et al, 1998). Immunostaining of tissue sections with CD11d can reveal early intravascular spread of hemophagocytic histocytic sarcoma to liver and lung, which has proven to be frequently overlooked by pathologists who examine only routine HE sections (Moore et al, 2006). The exquisite expression pattern of CD11d is also of value in identification of hepatosplenic T cell lymphoma. Peripheral lymphadenopathy is not a feature of this disease. Neoplastic T cells arise in splenic red pulp and invade liver and bone marrow sinuses. These T cells are often of !$ lineage, possess large granular lymphocyte (LGL) morphology and express CD11d (Farcet et al, 1990; Weidmann, 2000; Fry et al, 2003). An examination of normal splenic T cell populations in dogs revealed that !$ T cells are most abundant in splenic red pulp and express CD11d at a frequency 4 fold higher than !$ T cells of peripheral blood (McDonough and Moore, unpublished). To further underscore the usefulness of evaluation of CD11d in hematopoietic neoplasia, evaluation of canine chronic lymphocytic leukemia of T cell origin (T-CLL) revealed that more than 70% of the cases were of LGL type. Clinical evaluation of these cases revealed that TCLL of LGL type was usually an indolent disease in which splenic red pulp infiltration and splenomegaly were well developed prior to bone marrow infiltration (Vernau and Moore, 1999; McDonough and Moore, 2000). In fact, bone marrow infiltration was inconsistent, and was only seen late in the clinical course. Immunophenotypic analysis of the cases revealed that both "# T cells (69%) and !$ T cells (31%) were involved in this disease. CD11d and CD8 were expressed in 90% of cases of LGL T-CLL (McDonough and Moore, 2000); this is an unusual phenotype in peripheral blood and bone marrow, but not in the splenic red pulp. Accordingly, it is highly likely that LGL T-CLL originates as a proliferation of splenic red pulp T cells, which populate the blood and cause significant splenomegaly. The defining concept of

Moore: Immunological concepts applied to pathologic diagnosis leukemia as a primary bone marrow disorder is changing, and the modern definition of leukemia is best defined as hemopoietic neoplasia with predominant blood involvement. These few examples serve to highlight the multi-lineage expression of many leukocyte antigens. The dominant expression of CD11d by LGL T-CLL, coupled with the observation that T cell lymphomas and perhaps NK cell lymphomas in dogs also frequently express CD11d in diverse sites (skin, nervous system and others), clearly associates the granulated lymphocyte phenotype with CD11d expression. However, in cats we have documented a series of cases of LGL lymphocytosis, in which CD11d expression was not as prevalent (24%). These cats had intestinal LGL lymphomas most frequently of T cell lineage (90% CD3+). In many instances these LGL expressed the unusual phenotype CD3CD8" (60%), which is commonly observed in intra-epithelial lymphocytes (IEL) extracted from the small intestinal epithelial compartment (IEC). Normally these IEL contain about 30% LGL, and express the #-7 integrin CD103 ("E) (Roccabianca et al, 2000; Woo et al, 2002). We observed CD103 expression in about 60% of the LGL lymphocytosis cases tested, consistent with their intestinal origin (Roccabianca et al, 2006). CD103 expression by lymphocytes is rare in peripheral blood (about 1%) due to the nature of lymphocyte trafficking to the gut (Woo et al, 2002). Lymphocytes that traffic to small intestine express the integrin "4#7. These cells transmigrate across endothelial cells of the post-capillary venules which express the "4#7 ligand, mucosal addressin cell adhesion molecule (MadCAM) (Cheroutre, 2004; Cheroutre and Madakamutil, 2004). The chemokine receptor CCR9 and its ligand, thymus expressed chemokine (TECK or CCL25), which is produced by crypt epithelium of the small intestine are also involved in this process (Kunkel et al, 2000; Marsal et al, 2002; Mora et al, 2003). Once lymphocytes enter the lamina propria (LPL) they begin to down-regulate the "-4 subunit and up-regulate the "-E subunit and now express "E#7. The frequency of expression of "E#7 is greater in the IEL population than in the LPL population. The ligand for "E #7 in the IEC is the homotypic epithelial adhesion molecule E-cadherin (Cepek et al, 1994; Higgins et al, 1998). Expression of "E #-7 on mucosal lymphocytes is influenced by TGF-# secreted by the IEC. IEL are largely comprised by T cells, and lymphocyte trafficking to the small intestinal lamina propria and IEC has been mostly studied in this cell type. Evidence for broader application of the rules of intestinal lymphocyte trafficking to plasma cells has been published (Pabst et al, 2004). IgA producing plasma cells also use CCR9/CCL25 interaction to enter the mucosa of the small intestine and cluster around the epithelial crypts, which produce CCL25, and the polyIg receptor, which transports dimeric IgA to the intestinal lumen (Phalipon and Corthesy, 2003; Pabst et al, 2004). Morphologic evaluation of intestinal biopsies is impacted by lymphocyte trafficking patterns, which impart the normal segregation of lymphocytes and plasma cells within the lamina propria of the small intestine.

Understanding of these small intestinal lymphocyte trafficking principles, coupled with the development of PCR primer sets capable of detecting clonal expansion of B and T cells (see below), has led to the discovery of the high prevalence of mucosal-confined small T cell lymphoma in cats (Moore et al, 2005; Werner et al, 2005).

III. Lymphocyte antigen receptor gene rearrangement â&#x20AC;&#x201C; an aid to diagnosis of lymphoma During T cell development in the thymus, T cells rearrange their antigen receptor genes TCRA, TCRB, TCRG and TCRD, and in the process create 2 lineages of T cells, "# and !$ T cells. The majority of "# T cells rearrange TCRG prior to the rearrangement of TCRA and TCRB. Hence, TCRG gene rearrangement occurs in the majority of T cells regardless of surface the type of TCR expressed (Theodorou et al, 1994). The protein product of TCRG is TCR !, which contains a variable (V) domain and a constant (C) domain. The V domain is encoded by 2 segments of DNA, the variable and joining (J) segments. Although multiple V and J segments exist for TCRG, they are relatively limited in number and diversity by comparison with TCRA and TCRB. Gene rearrangement during T cell development in the thymus leads to random joining of a V segment to a J segment, leading to the formation of the complete V domain exon. The diversification of the TCRG repertoire is enhanced by the creation of P nucleotides, and the random insertion of N nucleotides by terminal transferase between the V and J segments. This creates the highly diverse third hypervariable region of the V domain, also known as the complementarity determining region 3 (CDR3). The CDR3 region is at the center of the antigen binding site and hence is the major contributor to antigen specificity (Janeway et al, 2001). In the rearrangement of TCRB and TCRD, a diversity (D) segment is interposed between V and J segments to create the CDR3 region, which is longer and more diverse due to the presence of 2 regions of N nucleotide addition (V-N-D-N-J). The TCRD locus is contained within the TCRA locus such that rearrangement of TCRA results in deletion of TCRD from the genome (Janeway et al, 2001). In consideration of the above, the preferred target for determination of clonality in T lymphocyte populations of both "# and !$ lineages is PCR amplification of the CDR3 region of TCRG. B cells rearrange their antigen receptor genes in the bone marrow during B cell development. B cells rearrange multiple V, J and D immunoglobulin (Ig) gene segments, initially in their heavy chain Ig genes (VHDJ), and later in either % or & light chain Ig genes (VLJ). This process of somatic recombination was first described for Ig genes and resembles the description of the process already outlined for T cells. A further property of Ig genes is their propensity to undergo V region somatic hypermutation, particularly during secondary antibody responses in germinal centers of follicles in peripheral lymphoid organs (Janeway et al, 2001). In addition, secondary V(D)J re-arrangement has also been described in germinal centers of mice through reactivation of 50

Cancer Therapy Vol 6-A.2, page 51 recombinase genes, which were previously thought to be active only in central lymphoid organs (Han et al, 1997). By these mechanisms, antibodies of high affinity are produced in secondary lymphoid responses (affinity maturation). The most commonly used target for determination of clonality in B lymphocyte populations is the IGH locus due to the extensive diversity of the CDR3 region and the conservation of IGH V and J segments, which facilitate PCR primer design. One pitfall of using IGH for molecular clonality determination is the extensive V segment gene mutation that occurs in post germinal center B cells can modify primer binding sites in the V segment. This can lead to false negative PCR results and reduced sensitivity of IGH as a molecular target for clonality determination in B cells (van Dongen et al, 2003). Lymphocyte antigen receptor clonality determination is a valuable adjunct to morphologic and immunophenotypic assessment of lymphoproliferative disorders in dogs and cats (Vernau and Moore, 1999; Burnett et al 2003; Moore et al 2005; Werner et al, 2005). It is not a primary diagnostic assay, and it cannot replace morphologic and immunophenotypic assessment. Lymphocye antigen receptor gene rearrangement can be promiscuous; both TCRG and IGH rearrangement can be observed in lymphomas of a single immunophenotype (B or T cell). Also, lymphocyte antigen receptor gene rearrangement has been observed in non-lymphoid leukocytic malignancy, such as acute myeloid leukemia (van Dongen et al, 2003). Molecular clonality determination is not needed to establish a diagnosis in most lymphoid proliferations in which architectural effacement of organized lymphoid tissue, cytological features of lymphocytes, and immunophenotyping are sufficient. Molecular clonality determination is indicated when morphological features of lymphocytes and immunophenotyping are inconclusive. These conditions are most often met in some lymphoid proliferations in gut and skin, or in organized lymphoid tissue when architecture is largely intact (for example: early marginal zone and T-zone lymphoid proliferations) (Moore et al, 2005; Valli et al, 2006). The relatively recent recognition of marginal zone lymphomas in dogs has complicated the interpretation of complex splenic lymphoid nodular lesions. Molecular clonality determination has become a decisive tool to unravel these complex lesions (Benak and Moore, unpublished data). Molecular clonality determination is also valuable in the assessment of the clonal relationship of lymphoid proliferations in separate sites. In this instance, it is possible to distinguish relapse from a second malignancy. The use of clonotypic primers (specific for a particular CDR3 sequence) can facilitate this investigation. However, development of clonotypic primers is expensive and not always possible based on the CDR3 sequence.

cells in diseased tissues. This coupled with the analysis of lymphocyte antigen receptor gene rearrangement in lymphoproliferative diseases, has greatly expanded our abilities to diagnose proliferative diseases of the immune system. However, the interrogative techniques of immunohistochemistry and molecular clonality determination, while powerful, should only be used as adjunctive aids coupled with careful clinical assessment of the patient and morphologic assessment of lesional tissue.

References Affolter VK, Moore PF (2002) Localized and disseminated histiocytic sarcoma of dendritic cell origin in dogs. Vet Pathol 39, 74-83. Blauvelt A, Katz SI, Udey MC (1995) Human Langerhans cells express E-cadherin. J Invest Dermatol 104, 293-6. Borkowski TA, Van Dyke BJ, Schwarzenberger K, McFarland VW, Farr AG, Udey MC (1994) Expression of E-cadherin by murine dendritic cells: E-cadherin as a dendritic cell differentiation antigen characteristic of epidermal Langerhans cells and related cells. Eur J Immunol 24, 2767-74. Burnett RC, Vernau W, Modiano JF, Olver CS, Moore PF, Avery AC (2003) Diagnosis of canine lymphoid neoplasia using clonal rearrangements of antigen receptor genes. Vet Pathol 40, 32-41. Cepek KL, Shaw SK, Parker CM, Russell GJ, Morrow JS, Rimm DL, Brenner MB (1994) Adhesion between epithelial cells and T lymphocytes mediated by E-cadherin and the aE b7 integrin. Nature 372, 190-3. Cheroutre H (2004) Starting at the beginning: new perspectives on the biology of mucosal T cells. Annu Rev Immunol 22, 217-46. Cheroutre H, Madakamutil L (2004) Acquired and natural memory T cells join forces at the mucosal front line. Nat Rev Immunol 4, 290-300. Cobbold S, Metcalfe S (1994) Monoclonal antibodies that define canine homologues of human CD antigens: summary of the First International Canine Leukocyte Antigen Workshop (CLAW). Tissue Antigens 43, 137-54. Cyster JG (1999) Chemokines and cell migration in secondary lymphoid organs. Science 286, 2098-102. Danilenko DM, Moore PF, Rossitto PV (1992) Canine leukocyte cell adhesion molecules (LeuCAMs): characterization of the CD11/CD18 family. Tissue Antigens 40(1), 13-21. Danilenko DM, Rossitto PV, Van der Vieren M, Le Trong H, McDonough SP, Affolter VK, Moore PF (1995) A novel canine leukointegrin, a d b2, is expressed by specific macrophage subpopulations in tissue and a minor CD8+ lymphocyte subpopulation in peripheral blood. J Immunol 155, 35-44. Farcet JP, Gaulard P, Marolleau JP, Le Couedic JP, Henni T, Gourdin MF, Divine M, Haioun C, Zafrani S, Goossens M, et al. (1990) Hepatosplenic T-cell lymphoma: sinusal/sinusoidal localization of malignant cells expressing the T-cell receptor gd. Blood 75, 2213-9. Fry MM, Vernau W, Pesavento PA, Bromel C, Moore PF (2003) Hepatosplenic lymphoma in a dog. Vet Pathol 40, 556-62. Han S, Zheng B, Takahashi Y, Kelsoe G (1997) Distinctive characteristics of germinal center B cells. Semin Immunol 9, 255-60. Haverson K, Saalmuller A, Alvarez B, Alonso F, Bailey M, Bianchi AT, Boersma WJ, Chen Z, Davis WC, Dominguez J, Engelhardt H, Ezquerra A, Grosmaire LS, Hamilton MJ, Hollemweguer E, Huang CA, Khanna KV, Kuebart G, Lackovic G, Ledbetter JA, Lee R, Llanes D, Lunney JK,

IV. Conclusion The discovery of CD molecules and their functions in diverse lineages of leukocytes, which has been aided by the development of mAb specific for them, have made it possible to precisely identify the lineages of infiltrative 51

Moore: Immunological concepts applied to pathologic diagnosis McCullough KC, Molitor T, Nielsen J, Niewold TA, Pescovitz MD, de la Lastra JM, Rehakova Z, Salmon H, Schnitzlein WM, Seebach J, Simon A, Sinkora J, Sinkora M, Stokes CR, Summerfield A, Sver L, Thacker E, Valpotic I, Yang H, Zuckermann FA, Zwart R (2001) Overview of the Third International Workshop on Swine Leukocyte Differentiation Antigens. Vet Immunol Immunopathol 80, 5-23. Higgins JM, Mandlebrot DA, Shaw SK, Russell GJ, Murphy EA, Chen YT, Nelson WJ, Parker CM, Brenner MB (1998) Direct and regulated interaction of integrin aEb7 with Ecadherin. J Cell Biol 140, 197-210. Janeway C, Travers P, Walport M, Shlomchik M (2001) Immunobiology: The immune system in health and disease. New York: Garland Publishing. Kellermann SA, Hudak S, Oldham ER, Liu YJ, McEvoy LM (1999) The CC chemokine receptor-7 ligands 6Ckine and macrophage inflammatory protein-3 b are potent chemoattractants for in vitro- and in vivo-derived dendritic cells. J Immunol 162, 3859-64. Kunkel EJ, Campbell JJ, Haraldsen G, Pan J, Boisvert J, Roberts AI, Ebert EC, Vierra MA, Goodman SB, Genovese MC, Wardlaw AJ, Greenberg HB, Parker CM, Butcher EC, Andrew DP, Agace WW (2000) Lymphocyte CC chemokine receptor 9 and epithelial thymus-expressed chemokine (TECK) expression distinguish the small intestinal immune compartment: Epithelial expression of tissue-specific chemokines as an organizing principle in regional immunity. J Exp Med 192, 761-8. Lanier LL, Chang C, Spits H, Phillips JH (1992) Expression of cytoplasmic CD3 epsilon proteins in activated human adult natural killer (NK) cells and CD3 g, d, epsilon complexes in fetal NK cells. Implications for the relationship of NK and T lymphocytes. J Immunol 149, 1876-80. Lunn DP, Holmes MA, Antczak DF, Agerwal N, Baker J, Bendali-Ahcene S, Blanchard-Channell M, Byrne KM, Cannizzo K, Davis W, Hamilton MJ, Hannant D, Kondo T, Kydd JH, Monier MC, Moore PF, O'Neil T, Schram BR, Sheoran A, Stott JL, Sugiura T, Vagnoni KE (1998) Report of the second equine Leucocyte Antigen Workshop, Squaw valley, California, July 1995 [In Process Citation]. Vet Immunol Immunopathol 62, 101-43. Marsal J, Svensson M, Ericsson A, Iranpour AH, Carramolino L, Marquez G, Agace WW (2002) Involvement of CCL25 (TECK) in the generation of the murine small-intestinal CD8a alpha+CD3+ intraepithelial lymphocyte compartment. Eur J Immunol 32, 3488-97. McDonough SP, Moore PF (2000) Clinical, hematologic, and immunophenotypic characterization of canine large granular lymphocytosis. Vet Pathol 37, 637-46. Moore P, Affolter V, Olivry T, Schrenzel M (1998) The use of immunological reagents in defining the pathogenesis of canine skin diseases involving proliferation of leukocytes. In: Kwotchka K, Willemse T, von Tscharner C, editors. Advances in Veterinary Dermatology. Oxford: Butterworth Heinmann. p 77-94. Moore P, Mariassy A (1986) Dendritic (Langerhans) cells in canine epidermis: ultrastructure and distribution. Anatomy Histology Embryology 15, 178-179. Moore PF, Affolter VK, Vernau W (2006) Canine hemophagocytic histiocytic sarcoma: a proliferative disorder of CD11d+ macrophages. Vet Pathol 43, 632-45. Moore PF, Schrenzel MD, Affolter VK, Olivry T, Naydan D (1996) Canine cutaneous histiocytoma is an epidermotropic Langerhans cell histiocytosis that expresses CD1 and specific b2-integrin molecules. Am J Pathol 148, 1699-708. Moore PF, Woo JC, Vernau W, Kosten S, Graham PS (2005) Characterization of feline T cell receptor gamma (TCRG)

variable region genes for the molecular diagnosis of feline intestinal T cell lymphoma. Vet Immunol Immunopathol 106, 167-78. Mora JR, Bono MR, Manjunath N, Weninger W, Cavanagh LL, Rosemblatt M, Von Andrian UH (2003) Selective imprinting of gut-homing T cells by Peyer's patch dendritic cells. Nature 424, 88-93. Naessens J, Howard CJ, Hopkins J (1997) Nomenclature and characterization of leukocyte differentiation antigens in ruminants. Immunol Today 18, 365-8. Olweus J, BitMansour A, Warnke R, Thompson PA, Carballido J, Picker LJ, Lund-Johansen F (1997) Dendritic cell ontogeny: a human dendritic cell lineage of myeloid origin. Proc Natl Acad Sci U S A 94, 12551-6. Pabst O, Ohl L, Wendland M, Wurbel MA, Kremmer E, Malissen B, Forster R (2004) Chemokine receptor CCR9 contributes to the localization of plasma cells to the small intestine. J Exp Med 199, 411-6. Phalipon A, Corthesy B (2003) Novel functions of the polymeric Ig receptor: well beyond transport of immunoglobulins. Trends Immunol 24, 55-8. Roccabianca P, Vernau W, Caniatti M, Moore PF (2006) Feline large granular lymphocyte (LGL) lymphoma with secondary leukemia: primary intestinal origin with predominance of a CD3/CD8(a)(a) phenotype. Vet Pathol 43, 15-28. Roccabianca P, Woo JC, Moore PF (2000) Characterization of the diffuse mucosal associated lymphoid tissue of feline small intestine. Vet Immunol Immunopathol 75, 27-42. Rothenfusser S, Tuma E, Endres S, Hartmann G (2002) Plasmacytoid dendritic cells: the key to CpG. Hum Immunol 63, 1111-9. Sethna MP, van Parijs L, Sharpe AH, Abbas AK, Freeman GJ (1994) A negative regulatory function of B7 revealed in B7-1 transgenic mice. Immunity 1, 415-21. Suzumiya J, Takeshita M, Kimura N, Kikuchi M, Uchida T, Hisano S, Eura Y, Kozuru M, Nomura Y, Tomita K, et al. (1994) Expression of adult and fetal natural killer cell markers in sinonasal lymphomas. Blood 83, 2255-60. Theodorou I, Raphael M, Bigorgne C, Fourcade C, Lahet C, Cochet G, Lefranc MP, Gaulard P, Farcet JP (1994) Recombination pattern of the TCR glocus in human peripheral T-cell lymphomas. J Pathol 174, 233-42. Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA, Sharpe AH (1995) Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3, 541-7. Valli VE, Vernau W, de Lorimier LP, Graham PS, Moore PF (2006) Canine indolent nodular lymphoma. Vet Pathol 43, 241-56. van Dongen JJ, Langerak AW, Bruggemann M, Evans PA, Hummel M, Lavender FL, Delabesse E, Davi F, Schuuring E, Garcia-Sanz R, van Krieken JH, Droese J, Gonzalez D, Bastard C, White HE, Spaargaren M, Gonzalez M, Parreira A, Smith JL, Morgan GJ, Kneba M, Macintyre EA (2003) Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia 17, 2257-317. Vernau W, Moore PF (1999) An immunophenotypic study of canine leukemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69, 145-64. Walunas TL, Lenschow DJ, Bakker CY, Linsley PS, Freeman GJ, Green JM, Thompson CB, Bluestone JA (1994) CTLA-4 can function as a negative regulator of T cell activation. Immunity 1, 405-13.

Cancer Therapy Vol 6-A.2, page 53 Weidmann E (2000) Hepatosplenic T cell lymphoma. A review on 45 cases since the first report describing the disease as a distinct lymphoma entity in 1990. Leukemia 14, 991-7. Werner JA, Woo JC, Vernau W, Graham PS, Grahn RA, Lyons LA, Moore PF (2005) Characterization of feline immunoglobulin heavy chain variable region genes for the molecular diagnosis of B-cell neoplasia. Vet Pathol 42, 596607. Woo JC, Roccabianca P, van Stijn A, Moore PF (2002) Characterization of a feline homologue of the aE integrin subunit (CD103) reveals high specificity for intra-epithelial lymphocytes. Vet Immunol Immunopathol 85, 9-22.

Zola H, Swart B, Banham A, Barry S, Beare A, Bensussan A, Boumsell L, C DB, Buhring HJ, Clark G, Engel P, Fox D, Jin BQ, Macardle PJ, Malavasi F, Mason D, Stockinger H, Yang X (2007) CD molecules 2006--human cell differentiation molecules. J Immunol Methods 319, 1-5. Zola H, Swart B, Nicholson I, Aasted B, Bensussan A, Boumsell L, Buckley C, Clark G, Drbal K, Engel P, Hart D, Horejsi V, Isacke C, Macardle P, Malavasi F, Mason D, Olive D, Saalmueller A, Schlossman SF, Schwartz-Albiez R, Simmons P, Tedder TF, Uguccioni M, Warren H (2005) CD molecules 2005: human cell differentiation molecules. Blood 106, 3123-6.

Moore: Immunological concepts applied to pathologic diagnosis

Cancer Therapy Vol 6-A.2, page 55 Cancer Therapy Vol 6, 285-302, 2008

Search for oncogenic retroviruses in wild mice and man: Historical reflections Review Article

Murray Gardner Center for Comparative Medicine, University of California at Davis

__________________________________________________________________________________ *Correspondence: Murray B. Gardner, MD, University of California School of Medicine, Department of Medical Pathology, One Shields Avenue, Davis, CA 95616, USA; Tel: 530 752 1245; fax: 530 752 7914; e-mail: mbgardner@ucdavis.edu Key words: viral oncogene, provirus hypotheses, Epidemiology, Virus properties, Exogenous transmission, Pathology, pathogenesis, Lymphoma, Paralysis, Control measures, Genetic resistance, gene therapy, oncogenic retrovirus antigens/antibodies Abbreviations: baboon endogenous virus, (BaEV); bovine leukemia virus, (BoLV); complement fixation, (CF); electron microscopy, (EM); feline leukemia virus, (FeLV); human t-cell leukemia virus, (HTLV); interleukin 2, (IL-2) murine leukemia virus, (MuLV); murine mammary tumor virus, (MMTV); National Cancer Institute, (NCI); polymerase chain reaction, (PCR); radioimmunoassay, (RIA); reverse transcriptase, (RT); Rous sarcoma virus (RSV); simian t-cell leukemia virus, (STLV); Virus Cancer Program, (VCP) Received: 14 March 2008; electronically published: June 2008

Presented in the Theilen Tribute Symposium at UC Davis 31st May- 1st June 2008.

Summary In 1968, the National Cancer Institute inaugurated the Virus Cancer Program to search for oncogenic retroviruses in humans and their domestic pets. I was recruited to help with this effort by building an interdisciplinary research team and, in particular, to search for such viruses in wild mice, the progenitor of laboratory mice which served as the major animal model for retroviral-induced cancer. Summarized here is the result of a 12 year effort of this team towards fulfilling these goals. We uncovered a new biology of oncogenic retroviruses in wild mice but failed to find any retrovirus in humans. Retroviruses in the wild mouse provided a more accurate model than retroviruses in the inbred laboratory mouse in predicting the natural history of the first human oncogenic retrovirus discovered elsewhere a decade later.

was predicated primarily on the belief that viruses of this type might cause cancer in humans as they were known to do naturally in chickens, lab mice, and domestic cats. By contrast, most of the known DNA tumor viruses, such as polyoma and adenovirus were considered to be lab artifacts and not tumorigenic in nature. When Ludwig Gross first isolated murine leukemia virus (MuLV) from AKR mouse embryos in 1951, he was convinced that the virus was transmitted vertically only as an infectious agent (Gross, 1951). However, in 1968, the emphasis was not on finding RNA tumor viruses in the conventional infectious disease sense; consensus opinion was that you did not “catch” cancer, and cancer in lab mice did not behave like a horizontally transmitted infectious disease. The favored expectation was that cancer was a genetic disease: i.e. “genes gone wild”. Therefore, the concept that potentially infectious RNA tumor viruses might be inherited as cellular, i.e. endogenous, virogenes in humans and other mammals, as had been found in chickens and inbred mice, had great intellectual appeal. Activation of such latent virogenes, under control by other host genes, by factors

I. Introduction Background: viral oncogene and provirus hypotheses Forty years ago, 1968, I was given the grand opportunity to embark on a new and exciting scientific adventure. I was recruited by Robert Huebner of the National Cancer Institute (NCI) to build an interdisciplinary research team at the USC School of Medicine in Los Angeles in order to investigate the possible role of oncogenic retroviruses in naturally occurring cancer in humans and their domestic house pets, including the common house mouse. From a medical background in general practice and academic pathology, but without any research training, I transformed overnight, at the age of 39, into a “virus hunter”. The chairman of my department, Hugh Edmondson, and his wife donated money for a building to house the research group and the NCI funded construction of the lab facilities. At that time, oncogenic retroviruses were called RNA tumor viruses and the NCI’s Virus Cancer Program (VCP), part of Nixon’s “War on Cancer” (Rettig, 1977), 55

Gardner: Search for oncogenic retroviruses in wild mice and man such as aging, x-irradiation, chemical carcinogens, sex hormones, DNA tumor viruses or other mutagens could theoretically produce infectious virus whose cell to cell spread would amplify the oncogenic signal and lead to cancer. This idea was analogous to the activation of integrated lysogenic bacteriophages described in the 1950s (Lwoff, 1960). In the late 1960s, the mammary tumor and leukemia RNA tumor viruses (MMTV and MuLV, respectively), derived from mammary tumor and lymphoma prone inbred lab mice, were, indeed, shown to be transmitted genetically as integrated viral genomes (Bentvelzen and Daams, 1969; Rowe, 1973). Of course, MMTV had earlier been shown to be spread also by milk but this was not considered typical of MuLV transmission. Endogenous inherited virogenes were considered the evolutionary relic of ancient RNA tumor virus infections and, surprisingly, several examples of cross-species spread with such viruses had apparently occurred in the distant past (Weiss, 2006). We discovered one such example with the isolation of the cat endogenous virus, RD114, which had apparently infected the ancestors of modern day cats from the baboon several million years ago (McAllister et al, 1972). The “rescue” of defective sarcoma viral genomes by “helper” leukemia viruses in chickens and mice, together with the knowledge of endogenous virogenes, led to the Viral Oncogene Hypothesis (Huebner and Todaro, 1969), which predicted that endogenous virogenes might include specific “oncogenes” that could trigger cancer, not only in animals, but possibly in man. Later on, the oncogenes were shown to be of cellular, not viral origin. Confirmation of the endogenous virogene concept in chickens and mice occurred in the 1970s, as summarized in Table 1. Howard Temin’s pioneering cell culture studies in the early 1960s suggested the presence of an inheritable DNA “provirus” determining the cell phenotype in the replication cycle of the Rous sarcoma RNA tumor virus of chickens (RSV). Temin’s Proviral Hypothesis (Temin, 1964) was not generally accepted at first. He also considered the possibility that an ongoing transfer of DNA to RNA to DNA, in normal cells, like what is now called retrotransposon activity, might generate RNA tumor viruses de novo and that secondary mutations in the cell genome might trigger cancer. The discovery of the

enzyme, reverse transcriptase (RT), in 1970, (Baltimore, 1970; Temin and Mizutani, 1970), whose existence had been inferred by the Viral Oncogene and Provirus Hypotheses, explained the ability of the RNA tumor virus genome to be completely copied as DNA provirus and inserted into chromosomal DNA. Temin’s proviral hypothesis was correct. The name of RNA tumor viruses was then changed to “retroviruses”. The RT enzyme became an indisputable reagent for detection of retroviruses and for modern day molecular biology. In 1976, the oncogene (src) present in RSV was demonstrated to be a cellular gene that had been transduced i.e. captured, by the helper avian leukemia virus (Stehelin et al, 1976). Subsequently, other sarcoma or acute leukemia viruses of chickens, lab mice, domestic cats, a turkey and a wooly monkey were shown to contain specific oncogenes transduced by the helper leukemia viruses. Such oncogene-containing viruses were defective because the envelope was deleted to accommodate the cellular oncogene. In 1981, it was first shown that the slowly acting avian leukosis virus induced cancer through activation of the cellular myc oncogene by nearby insertion of the proviral DNA (Hayward et al, 1981). Soon thereafter, myc and other retroviral transduced or activated oncogenes were shown to be involved in normal cell growth, each oncogene functioning at a certain level of cell signaling, either cell surface receptor, cytoplasmic second messenger, or regulatory gene in the nucleus (Bishop, 1987). Activation and mutation of cellular oncogenes by the control elements in the proviral long terminal repeats (LTR) enhanced the growth promoting properties of these normal cellular genes and is now recognized as the common mechanism by which oncogenic retroviruses trigger hematopoietic tumors in animals. Approximately 60 retroviruses containing oncogenes are now known along with over 100 known oncogenes targeted by provirus insertional mutagenesis (Rosenberg and Jolicoeur, 1997). In lab mice over 3000 retroviral integration sites have been cloned from retroviral induced hemotopoietic tumors (Davé et al, 2004). Many of these sites undoubtedly represent oncogenes yet to be defined. In humans, however, the only known oncogenic retrovirus (HTLV) does not contain or activate any known oncogenes.

Table 1. Evidence of endogenous virogenes in chickens and lab rodents (1968-1980)*. Detection of core antigen in uninfected embryos, under genetic control. Immunological tolerance to core antigen in adult animals. Spontaneous release of viral particles from aging or exogenously mutated cells in vitro. Activation of virus production from single cell clones of virus-free embryo cells. Rescue of endogenous virus by complementation of defective sarcoma virus. Homology between viral RNA and uninfected cellular DNA. Identification of specific genetic loci representing complete viral genomes. Transfection of infections virus from cellular DNA. Cloning and sequencing of endogenous virogenes. *For review see Weiss, 2006.

Cancer Therapy Vol 6-A.2, page 57 Transduction of cellular oncogenes by leukemia viruses is a rare event in animals that leads to the creation of highly oncogenic sarcoma or acute leukemia viruses. The sarcoma viruses are not infectious in nature and have not been found in humans. Nevertheless, the unique animal sarcoma viruses first pointed the way to cellular oncogenes in the 1960s and 1970s and opened the door to our understanding of their function in controlling cell growth. In human cancer, the cellular oncogenes are activated, not by any retrovirus, but by other mutagenic events, such as chromosomal translocation or point mutations (Croce, 2008). Several oncogenes first discovered after transduction by retroviruses in chickens, lab mice and cats and several other non-retroviral oncogenes that are targets of present day cancer therapy are shown in Table 2. Although the VCP emphasized endogenous virogenes, which may now seem an unrealistic expectation, it became readily apparent during the 1970s that the oncogenic Type C retroviruses found in other animals (chicken, turkey, mouse, cat, cow, gibbon ape) are transmitted only exogenously (for review see Gardner, 1980a). Horizontal transmission also accounts for spread of the recently discovered Type C leukemia virus of koala bears in Australia; interestingly this virus appears to be undergoing endogination (Tarlinton et al, 2006). We also learned in the 1970s that the exogenous Type C viruses of animals can cause non-oncogenic disorders such as wasting, immunosuppression, anemia, and neurological disease. In 1968, most of this story was yet to unfold. The major goal of the VCP was to look for activation of endogenous retrovirogenes and oncogenes in human cancer, as well as domestic pets. The possibility that

exogeneous retroviruses might be acquired horizontally, even perhaps transmitted from domestic animals to humans in close contact, could not be ignored. Because the Viral Oncogene Hypothesis was based primarily on inbred mice prone to mammary tumors and leukemia it was obviously critical to determine to what extent this concept applied to feral outbred mice, the progenitors of lab mice. This effort seemed all the more appropriate in so far as the domestic house mouse – Mus musculus domesticus - had been man’s inadvertent house pet since antiquity and had served as the major animal model for cancer research for well over a century. Therefore, our “marching orders” were to search for oncogenic retroviruses in wild mice, other domestic pets and in humans, and if found, to explore the natural history and ways to control these agents, including vaccination. We were starting, of course, with a rich background of retrovirus knowledge gained from lab mice (Sarma and Gazdar, 1974). Remember that this effort was undertaken about a decade before the onset of modern-day molecular biology. We relied on standard techniques such as complement fixation (CF) and later, radioimmunoassay (RIA), for detection of viral antigens and antibodies in blood or tissue extracts, tissue culture, with neutralization and interference tests, for viral isolation and classification, and electron microscopy (EM) for detection of Type C (MuLV) or Type B (MMTV) virus particles. Assays for reverse transcriptase activity as a virus marker were very useful (Roy-Burman et al, 1976). Viral proteins were analyzed by tryptic-peptide mapping. By the mid 1970s, oligonucleotide fingerprinting, restriction enzyme analysis and molecular hybridization for analysis of viral and cellular RNA and DNA had become available.

Table 2. Cancer therapies that target oncogenic proteins. Reproduced from Croce, 2008 with kind permission from The New England Journal of Medicine. Anticancer Drug Monoclonal antibodies Trastuzmab (Herceptin, Genentech) Cetuximab (Erbitux, ImClone) Bevacizumab (Avastin, Genentech) Small molecules

Target

Retrovirus Source

Disease

ERBB2

!ALV

Breast Cancer

EGFR

None

VEGF

None

Colorectal Cancer Colorectal Cancer, non-smallcell lung cancer

Imatinib (Gleevem Novartis)

ABL KIT

! MuLV ! FeLV

Gefitinib (Iressa, AstraZeneca) Erlontinib (Tarceva, Genentech)

EGFR EGFR

Sorafenib (Nexavar, Bayer/Onyx)

VEGFR, PDGFR, FLT3

Sunitinib (Sutent, Pfizer)

VEGFR, PDGFR, FLT3

None None None None ! FeLV None None ! FeLV

Chronic myelogenous leukemia, gastrointestinal stromal tumors, chordoma Non-small-cell lung cancer Non-small-cell lung cancer Renal-cell carcinoma Gastrointestinal stromal tumors, renal-cell carcinoma

EGF-epidermal growth receptor, VEGF-vascular endothelial growth factor, PDGFR-platelet-derived growth factor receptor, FLT3-FMSlike tyrosine kinase 3, ALV-avian leukemia virus.

Gardner: Search for oncogenic retroviruses in wild mice and man Immunofluorescense and immunoperoxidase staining techniques and monoclonal antibodies were now on hand. Enzyme linked immunoassays (ELISA), and cloning and sequencing of viral DNA came much later. Nevertheless, using the available biological and relatively primitive molecular techniques of that era, we made, in my admittedly biased opinion, considerable headway towards our goals, the results of a magnificent collaborative effort, which I will now summarize.

County (Figure 1) and an egg farm in La Puente were high MuLV expressors and lymphoma prone (Gardner et al, 1973a). Our first MuLV isolate was in 1971, after 3 years of searching, from two embryo cultures and their mother trapped at La Puente (Gardner et al, 1976a,d). Finally, we could begin to do some interesting virology. LC wild mice provided our most bountiful resource for studying the natural history of MuLV in a highly infected population (Gardner et al, 1976c). About 80% of the mice from here were laden with MuLV when trapped and were persistently infected throughout life. The animals were immune tolerant to their MuLV because of its acquisition at birth, primarily via milk. The 20% uninfected LC mice had no anti MuLV antibodies so, presumably, had not resisted infection because of an antiviral immune response. We realized later that these uninfected mice were the progeny of uninfected mothers and had escaped congenital infection. Followed over their life span, about 20% of the MuLV infected LC mice developed lymphoma but only after 1-2 years of observation. The lymphoma rate in LC mice was about 10-fold greater than in the low MuLV expressor wild mice. Also after 1-2 years, about 12% of MuLV-infected LC mice developed a fatal neurological disease featuring hind limb paralysis (Gardner et al, 1973b). This observation was totally unanticipated, and we were soon able to prove that the neurological disease was caused by the LC-MuLV. Some mice had both lymphoma and paralysis but the diseases could occur independently of each other. The LC mice were not immunosuppressed, and isoenzyme analysis showed that they were fully outbred (Rice et al, 1980). Polyoma virus infection was not present in the LC mice. Wild mice at another squab farm (Bouquet Canyon) were infected with polyoma virus which had no oncogenic effect (Gardner et al, 1974b). The squabs (baby pigeons) were not infected with MuLV or avian leukosis viruses. Incidentally, the squabs were raised for food in LA Chinatown. Because of human consumption, the LC squab farm was inspected regularly and had to be certified as “rodent free”. Therefore, our trapping operation was carried out in secret.

II. MuLV in wild mice A. Epidemiology How to find and trap wild mice was our initial challenge. The best locations were, not surprisingly, at places with ample grain such as an egg farm, race track, bird seed plant and several squab farms. At first we used single open ended traps but later found that we could capture the mice, several at a time, with gloved hand by stunning them with a bright light from a miner’s helmet while they were feeding at night. We eventually trapped mice from 15 different locations in LA County and environs and returned them to the lab where we set up separate “Leisure World” communities for each trapping area. The mice were allowed to live out their full life span housed individually in mason jars while we observed them daily. When moribund or recently dead, the mice were necropsied, studied histopathologically and assayed for MuLV. We also screened against a battery of other mouse pathogens. In later years, breeding colonies were established in the lab from several of the representative mouse populations. After several years of observing over 500 wild mice grow old, we concluded that these were hardy animals indeed. Most of them lived for 2-3 years in the lab after capture and fewer than 10% developed cancer only in old age (Gardner et al, 1973c). Lymphomas were the most common tumor. We could see Type C virus particles by EM and detect MuLV core antigen by CF test in a few lymphomas examined but not in any of the other tumors. Type C particles and antigen were also detected in a small percentage of sarcomas that were induced by 3methylcholanthrene in old wild mice (Gardner et al, 1971b). However, we could never isolate or transmit infectious MuLV from these animals or their tumors. At this point, about 1970, we could conclude that MuLV certainly was not prevalent or ubiquitous in wild mice, in marked contrast to the lymphoma prone inbred mouse stains (e.g. AKR). The MuLV that we detected in a few “old timers” seemed to be strongly repressed but we suspected that it was probably responsible for the few lymphomas observed. We surmised, based on our “mind set” at that time, that the virus we detected was activated endogenous virus. Several years later, after developing better techniques for virus isolation and characterization, we found how wrong we were. Had we stopped our study then we would have concluded that most wild mice are cancer resistant, low MuLV expressors and quite free of infectious MuLV. But, luckily, we discovered several populations of wild mice that were quite the opposite. Mice habitating a squab farm near Lake Casitas (LC) in southern Ventura

Figure 1. Bucolic smog-free location of the LC squab farm. Wild mice trapped here were laden with MuLV and prone to lymphoma and neurodegenerative disease.

Cancer Therapy Vol 6-A.2, page 59

B. Virus properties

lymphoma. Later, others showed that molecularly cloned LC ecotropic MuLV could induce both lymphoma and CNS disease in lab mice (Jolicoeur, et al 1983). Passage of amphotropic or ecotropic MuLV from wild mice into susceptible lab mice sometimes gave rise to recombinant viruses with enhanced virulence and altered cell tropism (Lai et al, 1982; Rasheed et al, 1982), just as happened in AKR mice when ecotropic MuLV recombined with endogenous virogenes to create highly virulent viral variants called mink cell focus (MCF), because they transformed mink cells (Hartley et al, 1977). But this type of recombination event was never observed in nature. We did not recover a sarcoma virus from the few wild mouse spontaneous or chemically induced sarcomas including several infected with MuLV. We concluded that recombination between exogenous MuLV and endogenous virogenes or oncogenes, a step required to generate highly leukemogenic or sarcomagenic viruses in lab mice, chickens and cats, does not occur in wild mice and therefore must represent, to some extent, an artifact of inbreeding and selection for cancer.

The wild mice MuLV is related to, but distinct from MuLV in lab mice. Together with other oncogenic type C retroviruses of animals (e.g. chicken, turkey, cat, gibbon ape, koala), MuLV is classified as a simple retrovirus in the newly designated gamma retrovirus genus. The viral genome structure is LTR-Gag-Pol-Env-LTR, without accessory genes. Gag is the group specific core antigen. Pol is the RNA dependent DNA polymerase (i.e. RT). Env is the envelope and LTR is the long terminal repeat. By end point dilution we identified two different neutralization or interference classes of MuLV that occurred separately or together in MuLV infected mice from LC and several other widely separated trapping areas (Rasheed et al, 1977; Gardner and Rasheed, 1982). The most prevalent virus was an entirely new class of MuLV, never found in lab mice. It was called “amphotropic” because of its wide in vitro host range including human cells, a surprising finding (Hartley and Rowe, 1976; Rasheed et al, 1976). Less common was “ecotropic” MuLV which, as in lab mice, grows only in rodent cells. “Xenotropic” virus which grows only in non-murine cells (Levy, 1976), and is isolated frequently from lab mice, was rarely detected in wild mice. The lack of xenotropic virus isolates was explained a decade later when it was found that many of our southern California wild mice were free of any xenotropic proviral DNA (Kozak and O’Neill, 1986). Molecular hybridization with lab mouse MuLV probes and, much later, cloning, sequencing and phylogentic analysis of the wild mouse MuLVs revealed that both amphotropic and ecotropic MuLV have a unique origin from the same ancestor, quite distinct from the MuLV of lab mice (Barbacid et al, 1979, Howard et al, 2006). These viruses have apparently been maintained as infectious agents among indigenous feral mouse populations of Southern California for a long time and they never changed their cell tropism, virulence or disease phenotype over the decade that we studied them. Although showing microheterogenity in envelope proteins by tryptic peptide mapping (Bryant et al, 1978), the retroviruses in wild mice were remarkably stable. Amphotropic and ecotropic MuLV did not recombine with each other or with endogenous virogenes in the wild mouse genome (Gardner and Rasheed, 1982). Whereas amphotropic and ecotropic MuLV were found in both healthy and lymphomatous wild mice, ecotropic MuLV seemed uniquely associated with the spontaneous paralytic disease of LC mice. In nature, the more prevalent amphotropic virus had rather low virulence because over 50% of infected mice lived a healthy long life without developing lymphoma. Experimental transmission of end point cloned wild mouse MuLV in virus-free newborn NIH Swiss lab mice showed that amphotropic virus induced lymphoma, but not paralysis, in about 20% of recipients after 1 year, whereas ecotropic MuLV induced both lymphoma and paralysis, with higher incidence and shorter incubation. By inoculating susceptible newborn lab mice with high titered LC ecotropic MuLV, the incubation period for inducing paralysis was shortened to just 1-2 months with 100% incidence; longer surviving recipients also developed

C. Exogenous transmission We found, contrary to our expectations, that MuLV in wild mice was transmitted exogenously, primarily by milk and to a lesser extent by the venereal route (Gardner et al, 1979). Exogenous transmission via milk apparently occurred in both low and high MuLV expressor populations of wild mice. But we only confirmed this by foster nursing experiments in the virus-laden LC mice. Numerous type C virus particles were visible in the breast tissue and milk of the infected LC mice (Figure 2a and b). By foster nursing newborn LC mice from MuLV infected mothers on uninfected lab mice, most of the infectious MuLV could be eliminated in the progeny. We were eventually able to derive by foster nursing a subpopulation of LC mice totally free of infectious MuLV. These mice remained uninfected over their entire lifespan and were as resistant to cancer of all types as were the low MuLV expressor populations of wild mice. Extensive attempts over a decade to activate endogenous MuLV in vitro from uninfected cells of wild mice, using techniques successful in activating endogenous MuLV from inbred mice, were universally unsuccessful (Gardner and Rasheed, 1982). We were able to transmit amphotropic and ecotropic MuLV to uninfected and highly susceptible C57L inbred mice by foster nursing on infected LC mothers. Virus could also be transmitted sexually from infected LC males to C57L females. Once infected the C57L mice transmitted the virus to their progeny through milk, generation after generation, just as occurred naturally in LC mice. The milk transmission of MuLV in wild mice resembled milk transmission of MMTV in high breast cancer strains of lab mice. In retrospect, some lab strains of MuLV (FMR) were previously found to be transmitted exclusively by mother’s milk (or doctor’s needles) (Law and Moloney, 1961). But these FMR viruses, carried for many years in transplant tumors (e.g. Erlich ascites), were considered lab artifacts or “passenger viruses”. Ironically, the exogenous transmission via milk of the FMR viruses in 59

Gardner: Search for oncogenic retroviruses in wild mice and man analog of Fli-1 and surrounding genes have been mapped to chromosome 11 in a conserved arrangement: this region of the human genome is associated with various leukemias. Other common integration sites, at EVI-1 and MYB loci, were also detected in different leukemia cell types induced in lab mice by LC ecotropic MuLV after recombination with endogenous virogenes. Tumor suppressor genes are not common targets of retroviral transduction or insertional mutagenesis but mutation of the p53 gene occurs in some LC ecotropic MuLV induced lymphomas in lab mice (Bergeron et al, 1993). This technology was not available in the 1970s, so we never examined proviral integration sites in the spontaneous LC lymphomas.

lab mice proved more like MuLV transmission in wild mice than the endogenous MuLV transmission in the AKR and related lab mouse models. The “passenger” lab mouse viruses were like the “real world”, whereas the AKR mouse virus was the artifact.

D. Pathology and pathogenesis 1. Lymphoma Most of the spontaneous lymphomas in old, low MuLV expressor wild mice were of histiocytic cell origin, then called reticulum cell sarcoma. By contrast, the LC lymphomas, occurring naturally or experimentally induced by amphotropic or ecotropic MuLV, were of lymphocytic origin and had remarkably similar gross, microscopic (Figure 3a-e) and functional characteristics. In contrast to the AKR mouse, the tumors spared the thymus, arose primarily in the splenic red pulp, became leukemic and were comprised of stem or null cells lacking T and B cell markers (Bryant et al, 1981). Numerous Type C virus particles were seen budding from and lying between the tumor cells. Several lymphoma clonal cell lines were of diploid karyotype and positive for surface antigens of lymphoid stem cells or pre-B-cells. The MuLV lymphomas in LC mice thus provided a useful model for childhood leukemia and for study of the early steps of B cell differentiation. As mentioned, on passage through lab mice, the LC MuLV sometimes recombined with endogenous virogenes which changed its cell tropism, virulence and type of leukemia induced. Interestingly, the CBL (“Casitas B. Lymphoma”) oncogene was derived from an ecotropic LC MuLV after several passages through lab mice (Langdon et al, 1989). As eventually found with the other slowly acting animal leukemia retroviruses, the mechanism for leukemogenesis by the LC MuLV is that of proviral insertional mutagenesis. In 1990, a common site of proviral integration was found in many of the clonal null cell lymphomas induced by the LC ecotropic MuLV in NIH Swiss mice (Bergeron et al, 1992). This site, called Fli-1, is localized on chromosome 9, near several other previously described proto-oncogenes and other common retroviral integration sites, including CBL. The human

2. Paralysis This disease, caused by LC ecotropic MuLV (Gardner et al, 1973b), has been studied in many labs over the past 30 years (Wiley and Gardner, 1993). It is the first example of a naturally occurring degenerative neurological or motor neuron disease caused by an oncogenic retrovirus (Officer et al, 1973). At first, it was considered a model for amyotrophic lateral sclerosis (ALS) in humans, but we were never able to find a retrovirus in ALS patients (Gardner and Henderson, 1974a). Subsequently, several lab mouse MuLV strains have induced a similar CNS disease (Zachary et al, 1986; Portis, 1990). The natural and experimental CNS disease is characterized histopathologically by a non-inflammatory spongiform change with reactive gliosis and loss of anterior horn motor neurons in the lower spinal cord (Figure 4a-g) (Gardner, 1985). Numerous Type C particles are present in the extra-cellular space. In some of the experimental models, the same changes occur also in the brain stem and cerebellum. Because of the loss of anterior horn neurons and motor nerve denervation, the lower limbs suffer severe muscle atrophy with tremors and eventual total paralysis. Pathogenesis clearly involves an early viremic spread of virus to the CNS, a productive infection of CNS endothelial cells and a non-immune mediated pathology related to virus level in the CNS.

Figure 2. (A) (left) MuLV in milk from a pregnant LC mouse: sucrose density gradient. (B) MuLV Type-C particles located extracellularly in breast tissue of a pregnant LC mouse.

Cancer Therapy Vol 6-A.2, page 61 The pathogenesis of the spongiform change and neuronal loss is still not understood. Viral recombination or prions are not involved. Disagreements among the different models center on whether glial cells and motor neurons become infected and whether direct or indirect effects of the virus cause the spongiform degeneration and loss of motor neurons. Infection of motor neurons was our initial suspicion based on EM detection of aberrant budding of virus particles into cytoplasmic vacuoles or the endoplasmic reticulum of anterior horn neurons of naturally diseased LC mice (Andrews and Gardner, 1974). However, we could not distinguish whether these aberrant particles were the result of aborted exogenous infection or activated endogenous virogenes. Infection of microglial

cells and not neurons has been a prominent feature of some versions of the disease in lab mice. Neurotoxicity of the viral envelope glycoprotein is suspected in some models, including a mild form of the disease induced in transgenic mice expressing only the Env gene of the LC ecotropic MuLV under control of its own LTR (Kay et al, 1993). Equally plausible is the neurotoxic effect of cytokines released by infected glial or endothelial cells. Such cytokines could damage cell membranes and lead to the spongiform change. Better understanding of the molecular pathogenesis of this retrovirus induced neurologic disease, discovered in the early 1970s, could reveal mechanisms shared in common with AIDS encephalopathy and prion diseases.

Figure 3. (A) Enlarged spleen and cervical lymph nodes of a LC wild mouse with spontaneous lymphoma. (B) Poorly differentiated lymphoid tumor cells invading wall of bronchiole. (C) Poorly differentiated lymphoid tumor cells invading wall of bronchiole. (D) Lymphoid tumor cells circulating in blood. (E) Extracellular Type-C virus particles lie between lymphoid cells of LC wild mouse lymphoma. Insert shows typical Type-C particle budding from lymphoid cell X84000.

Gardner: Search for oncogenic retroviruses in wild mice and man

Figure 4. (A) Paralyzed LC wild mouse: Note severe hind limb atrophy. (B) Spongiform degeneration of lower spinal cord. (C) (left) Degenerating motor neurons (arrows) in anterior horn of lower spinal cord. (right) Gliosis (arrows) in same location. (D) Spongiform degeneration of anterior horn motor neuron in lower spinal cord. One micron section of tissue fixed before death to avoid post-mortem shrinkage artifact. (E) Neurogenic muscle atrophy of hind limb. (F) Aberrant virus particles budding into cisterns of the endoplasmic reticulum of anterior horn motor neuron. (G) Extracellular Type-C virus particles in lower spinal cord.

Cancer Therapy Vol 6-A.2, page 63

E. Control measures

population of LC mice accounts for why the ecotropic MuLV is less prevalent than amphotropic MuLV. Only about 20% of LC mice fail to inherit one or both Fv-4 resistance alleles and therefore are susceptible to congenital infection with the LC ecotopic virus and subsequent disease. The other 80% of LC mice carry one or two Fv-4 alleles and are therefore resistant to LC ecotropic virus. Accordingly, the CNS disease occurs in only a few aging, ecotropic MuLV infected LC mice that survive long enough in nature. Because amphotropic virus is not restricted, it is much more prevalent in LC mice. The Fv-4 gene is not found in North America lab mice because of the narrow genetic base from which these mice were derived. Based on the signature of endogenous virogene DNA, including xenotropic, MCF and Fv-4 sequences, it was later shown that wild mice from southeast Asia (Mus castaneous) were the source of the Fv-4 resistance gene (Gardner et al, 1991). The Asian wild mouse had passed this gene to both Japanese wild mice and LC wild mice. LC mice are thus a hybrid between feral mice (M.m. domesticus), which crossed the Atlantic Ocean from western Europe to the eastern USA and feral mice (M.m. castaneous) which crossed the Pacific Ocean from Asia to Southern California. Fv-4 represents a fascinating example of a useful endogenous defective, but functional, provirus that is still segregating in a freeranging population of wild mice to regulate a serious infectious disease. Similar defective endogenous proviral genes that protect against exogenous retrovirus infection were also found in chickens, (Robinson et al, 1981) and, later in Asian wild mice (Mus castaneus) (Wu et al, 2005), and in DBA/2 lab mice (Jung et al, 2002). Since HTLV-1 or HIV-1 were not endogenous in humans, this type of genetic resistance probably does not apply to these retroviruses. However, genetic resistance to HIV-1 infection does similarly occur at the cell surface level; this is attributed to mutations in the CCR chemokine coreceptors.

Contrary to expectations based on the AKR mouse model, control of MuLV and prevention of lymphoma and paralytic disease in the high MuLV expressor LC mice was dramatically accomplished, as mentioned above, by foster nursing on virus-free NIH Swiss mothers (Gardner et al, 1979). Surgical removal of the spleen, the â&#x20AC;&#x153;virus factoryâ&#x20AC;? early in life, lowered the virus burden sufficiently to prevent the CNS disease completely and markedly reduced the incidence of lymphoma. Passive immunization of newborn LC mice with goat immunoglobulin, having a high neutralization titer to ecotropic virus and a low titer to amphotropic virus, completely prevented the paralytic diseases but only slightly lowered the incidence of lymphoma (Gardner et al, 1980b). Active immunization with inactivated LC MuLV had no effect, of course, in already infected immune tolerant LC mice, although they responded well to a heterologous MuLV vaccine and other foreign antigens (Klement, 1976). A decade later, the beneficial effect of transplacental anti-retroviral therapy with azidothymidine was shown in lab mice that were protected against the CNS disease after inoculation with LC ecotropic MuLV during mid-gestation or at birth (Sharpe et al, 1987). This is the first example of successful antiviral therapy for congenitally transmitted retrovirus. Finally, selective breeding could strongly suppress LCMuLV. Because the MuLV of LC and other wild mice is N-tropic, i.e. grows preferentially in NIH Swiss mouse cells, and not in BALB-C mouse cells (B-tropic), introduction of the FV-1B virus resistance gene from C57 BL-10 inbred mice completely blocked LC MuLV expression in the F-1 hybrids (Gardner et al, 1976b). More dramatic, however, was the discovery of a natural virus resistance gene in LC mice, described next.

F. Genetic resistance The exciting discovery of a naturally segregating MuLV-resistance gene in LC wild mice was serendipitous. I was curious as to what would happen to MuLV expression in the progeny of genetic crosses between LC and AKR mice. Analysis of AKR ecotopic virus in the FI progeny revealed a pattern of virus expression and leukemia consistent with a dominant AKR MuLV resistance gene in the LC mouse population (Gardner et al, 1980c). Back crosses of virus-free FI progeny to AKR proved that a single dominant gene was responsible for the virus restriction. A similar gene called Fv-4, because it restricted the Friend strain of MuLV, had been described in Japanese wild mice (Mus molossinus) (Odaka, et al, 1981). In crosses with Japanese wild mice, this restriction gene, initially called Akvr-1, was found allelic on chromosome 12 with the Fv-4 restriction gene (Oâ&#x20AC;&#x2122;Brien et al, 1983). The Akvr-1 and Fv-4 resistance genes proved to be sequence identical, including flanking DNA, and to represent a defective, truncated endogenous provirus encoding an envelope glycoprotein (gp70) closely related to ecotropic MuLV of both AKR and LC mice (Dandekar et al, 1987; Ikeda and Sugimura, 1989). Expression of this envelope protein on the cell surface blocks the receptor for the ecotropic class of MuLV. Segregation of this dominant polymorphous gene, now called Fv-4, in the natural

G. Reagents for gene therapy Because of their wide cell tropism, including human cells, the amphotropic MuLV from wild mice and the cat endogenous virus RD114 have provided the envelope glycoprotein for recombinant retroviruses employed in experimental gene therapy (Cosset et al, 1995). One of the tumor cell lines, HT-1080 (Rasheed et al, 1974), which we derived, is a favorite packaging cell because of its ability to release high-titered viruses that are resistant to inactivation by human serum. Such vectors are quite capable of gene transfer to human cord blood CD34+ cells, which can then be transplanted into an immunosuppresed host (Kelly et al, 2000, Relander et al, 2005). Progress in this direction depended on first defining, in the 1980s, the cell surface receptors for different retroviruses. The ecotropic MuLV receptor is a cationic amino acid transporter, whose gene is mapped to chromosome 5 in the mouse and a highly related, but non permissive gene, mapped to chromosome 13 in humans (Albritton et al, 1989). The receptor for amphotropic MuLV and for FeLV-B and GALV are phosphate transporters that map to mouse chromosome 8 and human 63

Gardner: Search for oncogenic retroviruses in wild mice and man chromosome 8 (Miller et al, 1994). The receptor for cat endogenous RD114, BaEV and avian reticuloendotheliosis virus is a neutral amino acid transporter mapped to human chromosome 19 (Tailor et al, 1999). All of these receptors have many membrane spanning domains and an external envelope binding domain. In the 1970s, we had no idea that our newly discovered viruses and cell lines would have a practical purpose for therapy. However, their clinical application is still “on hold” because of safety concerns i.e. insertional activation of oncogenes, a disasterous event which has occurred on several occasions (Davé et al, 2004).

EM in milk from about 50% of our wild mice (Rongey et al, 1973), (Figure 2c and d) and MMTV antigen was detected by immunoperoxidase staining of lactating mammary gland in about 20% of all wild mice (Fine et al, 1978); about 50% of the spontaneous breast tumors contained MMTV antigen (Gardner et al, 1980c), yet, less then 1% of these animals developed breast tumors and only after 1 year of age. Therefore, the virus may not be essential for breast tumors to develop in wild mice. Under natural conditions, then, the highly prevalent MMTV is only weakly carcinogenic, and it is also only weakly tumorigenic when transmitted by foster nursing to susceptible lab mice. Lab mouse derived MMTV was also of low pathogenicity in uninfected wild mice, probably because of strong genetic resistance. We could conclude that MMTV infected breast cancer prone strains of lab mice, like MuLV infected lymphoma prone lab mice, are an artifact of “inbreeding” and selection for this cancer. Interestingly, unlike lab mice, we found some LC wild mice totally lacking any endogenous MMTV proviral DNA. A wild mouse line was derived that had no MMTV DNA (Cohen et al, 1982), yet had normal mammary gland development and was susceptible to mammary hyperplasia and neoplasia (Faulkin et al, 1984). Thus, endogenous MMTV has no role in normal mammary gland development or function, as had been suggested, and is not required for breast tumorigenesis. When present in lab

H. Conclusion In its natural history features, the wild mice MuLV model was more accurate than the laboratory mouse MuLV model for predicting the natural history of HTLV1, which was discovered a decade later. The similarities and differences between the wild mouse MuLV model and HTLV-1 are listed in Table 3A and B. The inbreeding of mice and selection for cancer clearly created an artificial biology not accurately reflective of the real world, but, nevertheless, useful for experimental biology.

III. MMTV in wild mice A. Natural history In contrast to the regional, familial clustering of MuLV in wild mice, MMTV was equally prevalent in all of the populations examined. MMTV was detectable by Table 3 A. Similarities between MuLV in wild mice and HTLV1 in humans Viruses are not ubiquitous; they occur in regional and familial clusters associated with specific types of leukemia. Viruses are of low pathogenicity and long latency. Viruses are exogenous and transmitted mainly by milk. Sexual transmission can also occur. Viruses are stable, replication competent and do not undergo recombination with endogenous virogenes or oncogenes. Viruses integrate monoclonally in leukemia cells.

B. Differences between MuLV in wild mice and HTLV-1 humans HTLV-1 is more T-cell restricted and induces T-cell leukemia. Wild mouse MuLV is more B-cell tropic and induces a null or pre-B-cell leukemia. HTLV-1 is more latent and cell associated without viremia. HTLV-1 induces an immune response whereas wild mice are immunologically tolerant to their congenitally acquired virus. HTLV-1 can transform T-cells in vitro; MuLV does not transform T or B-cells. HTLV-1 does not induce leukemia by insertional activation of cellular oncogenes as does wild mouse MuLV. HTLV-1 has a more complex genetic composition than MuLV. It contains accessory genes not present in MuLV. One of these accessory genes called TAT transactivates several transcription factors which indirectly promote T-cell immortalization. HTLV-1, STLV-1 and BoLV have a similar pathogenesis (and genetic structure) in humans, monkeys and cows, respectively; these viruses are now classified in the deltaretrovirus genus.

Viruses are neuropathogenic as well as lymphomagenic.

Viruses and disease can be controlled by foster nursing or avoidance of breast feeding (Tsuji et al, 1990).

Cancer Therapy Vol 6-A.2, page 65

Figure 2. (C) MMTV in milk from a pregnant LC mouse: sucrose density gradient. (B) MMTV Type-B particles in breast tissue of pregnant LC mouse.

mice bred for susceptibility to breast tumors, MMTV behaves as a strong carcinogen, initiating early mammary hyperplasia (Cardiff, 1984). Interestingly, the oncogenes that are activated by insertion of MMTV proviral DNA (Peters et al, 1983) are not activated in human breast cancer. Probably the most important contribution of our MMTV work was to give credibility to our negative results in searching for a similar virus in human milk (Roy-Burman et al, 1973). Reports were appearing at that time of MMTV like virus particles and reverse transcriptase activity in human milk and of MMTV related antigens in human breast tumors. None of these reports could be confirmed. We could easily detect MuLV and MMTV in a drop of wild mouse milk, but could detect no virus in â&#x20AC;&#x153;gallonsâ&#x20AC;? of human milk. No convincing evidence exist today of MMTV or related retroviruses in human milk or breast cancer despite a few reports to the contrary.

Figure 5. A smoggy day in LA-Mid 1960s.

IV. Search for oncogenic retroviruses in humans

were performed on fresh samples which were also frozen or shipped to collaborators. Cell cultures were established when possible and tumor and control samples were shipped weekly to the Naval Biological Lab in Alameda, CA, where they were cultured and karyotyped. Animal tumors and other tissues were obtained from local veterinarians and pet owners. We received the complete cooperation of the LA Health Department and LA County Veterinarians. Nonhuman primate and other animal tissues were provided by veterinarians at the LA and San Diego Zoos and the Alamogordo, NM chimpanzee colony. Thankfully, the 1970s were not as litigious or encumbered by paper-work and legal permissions as nowadays and, there was no proprietary interest involved. All resources were available to all VCP investigators and new discoveries, such as viruses and tumor cell lines, were distributed freely after initial publication. I will now summarize our negative efforts to find an oncogenic human retrovirus.

A. Methods used and negative results Despite our success in finding oncogenic retroviruses in wild mice, wild rats (Rasheed et al, 1978a) and domestic cats (Gardner et al, 1971c) we could not find any such viruses in humans, dogs, parakeets (Gardner et al, 1981) and a few captive non-human primates (Rasheed and Gardner, 1979). The effort made in humans was truly exhaustive, many man-years of time over a 12 year period (1968-1980) (Gardner et al, 1977). For tracking the ongoing occurrence of specific types of human cancer by geographic location and smog exposure in LA (Figure 5), we developed a Tumor Registry of all 264 hospitals in LA County and pinpointed the cases on detailed maps of LA County and its freeways. Smog particulate matter was collected on giant filters in mobile trailers, analyzed for chemical content and tested for activation of endogenous virogenes in rodent tissue cultures and for mutagenicity by AMES assays. We collected fresh human tumors and normal tissues, blood, bone marrow, milk, aborted embryos and placentas from 16 major hospitals in LA county. Virologic, immunologic and biochemical assays

1. Epidemiology No geographical or familial cluster of lymphoma, sarcoma or other cancers were spotted in LA County that

Gardner: Search for oncogenic retroviruses in wild mice and man might suggest the presence of an activated endogenous or exogenous retrovirus. In particular, no cases of adult T cell leukemia, the eventual source of HTLV-1, were recognized in our tumor surveys. No evidence for an association of cancer in humans and household pets was found (Hanes et al, 1970; Gardner, 1971a). Specifically, no seroepidemiologic or virologic evidence was found to suggest the cross-species spread of FeLV or MuLV to humans in close contact. No evidence was obtained to indicate that the concentration, content, or in vitro cell transforming activity of LA smog particulate matter (Rhim et al, 1972) was related to the distribution of human or household pet cancers. Five years of prior research had found no lung tumorigenic effect in lab mice after lifelong exposure to ambient LA smog (Gardner et al, 1970), and expression levels of MuLV in wild mice were not correlated with levels of smog exposure.

including wild mice, rats and domestic cats. It is important to note that we had eight known occasions on which, despite good lab conditions, contamination of human or dog cell cultures inadvertently occurred. Efforts to isolate viruses from cultured nonhuman primate tissues were also negative, with the exception of two baboon embryo cell cultures which spontaneously released the endogenous baboon virus (BaEV). Although infection of virus-free rat cells in vitro by helper rat leukemia virus could rescue an endogenous SRC gene (Rasheed et al, 1978b), this result could never be duplicated in human or other mammalian cells. Interestingly, skin fibroblasts from individuals with a familial high risk to colorectal cancer, who we know now have mutations in tumor suppressor genes, were supersusceptible to transformation by animal sarcoma viruses (Rasheed and Gardner, 1981).

4. Attempts to identify oncogenic retrovirus antigens or antibodies in human materials

2. EM search for Type C or Type B virus particles

By using radioimmunoassay (RIA) for speciesspecific or interspecies viral core determinants, we sought evidence in humans for known mammalian oncogenic retrovirus antigens from cats, gibbon apes and mice or antibodies to these viruses. As a positive control we were able to detect MuLV core and envelope antibodies in terminal cancer patients experimentally immunized with formalin inactivated Rauscher MuLV (Hersh et al, 1974). No evidence of reactivity against such viruses was found in 100 human cancer and control sera. We were also unable to detect virus-specific antigens and antibodies by complement fixation or, later immunofluoresence tests. No reactivity by RIA was found in individuals with amyotropic lateral sclerosis (31), cat scratch disease (11) or systemic lupus erythematosis (26). Household contacts (71) of FeLV infected cats were all negative by RIA. No antibody to FeLV reverse transcriptase was found in 15 humans exposed to FeLV-positive cats, although antibody to this FeLV enzyme was detected in most cats naturally or experimentally infected with FeLV (Roy-Burman et al, 1972). Nor was any evidence found for infection in humans with known household or laboratory exposure to oncogenic retroviruses from mice and gibbon apes or in veterinarians and zoo primate handlers. The absence of antibodies to these viral antigens in many hundreds of human sera is strong evidence against the horizontal infection of humans with any of the known mammalian Type C oncogenic retroviruses, including those that could grow well in human cells in vitro. These exhaustive and fruitless efforts to find an oncogenic retrovirus in humans indicated that humans (like most wild mice) do not harbor such viruses. Most importantly, the activation of replication competent or infectious endogenous virogenes does not seem involved in human (or wild mouse) cancer.

More than 200 primary tumors including 102 sarcomas, 20 lymphomas and 86 carcinomas were examined. Tissues from 52 primary embryos at 10-20 weeks gestation and milk from 27 lactating women, one with breast cancer, were examined. No virus particles were seen in any fresh or cultured human tissue except when purposely or accidentally introduced. The closest resemblance was the budding Type C virus-like structures seen in 14 of 50 placentas examined. These structures, initially reported in baboon placentas from which the baboon endogenous type C virus was isolated (Benveniste et al, 1974), were located solely along the base of the syncytiotrophoblast layer. No mature or extracellular Type C virus particles were seen in the human placentas, and no virus was ever isolated. In recent years, these structures have been associated with the expression of a human endogenous retrovirus whose envelope protein may be involved in formation of this placental layer (Blond et al, 1999). EM search for virus particles in bone marrow, peripheral blood leukocytes or tumor cell lines from about 20 children with leukemia or sarcomas was also negative. No virus particles were seen, in human milk despite our ability, mentioned above, to readily detect MMTV and MuLV particles in wild mouse milk.

3. Attempt to isolate infectious oncogenic retrovirus from human tissues No virus was isolated from hundreds of fresh tumors and short-term tumor cell cultures and fewer established cell lines. No virus was isolated from blood cells, lymphoid tissues or cell lines from more than 30 children with acute leukemia or lymphoma. We grew many B-cell lines immortalized by the latent Ebstein-Barr herpes virus (EBV). We could not grow T-cells because T-cell growth factor (IL-2) had not yet been discovered. Placental and embryo tissues were also free of detectable infectious virus. Assays of different tumor cell cultures included cocultivation with numerous different human and nonhuman indicator cells and treatment of the cultures with various chemical and physical agents capable of inducing Type C viruses in our animal model systems,

V. Summary A. Importance of real world biology and animal models My 40 year adventure in oncogenic retrovirus research taught me the importance of keeping an eye on

Cancer Therapy Vol 6-A.2, page 67 natural history and the value of comparative or “one medicine” using animal models. Natural biology often seems eclipsed nowadays by the ascent of molecular biology and genomics. This is certainly true of genetically engineered mice (GEM) which, although genetic artifacts, have rightfully taken the place of conventional inbred mice in cancer research. Oncogenic retroviruses have become museum exhibits, no longer in fashion. However, integration sites of endogenous retrovirogenes continue to have an impact upon the phenotype of GEM (Barthold, 2002). Worth reemphasizing is the not surprising realization that the proviral insertional activation of cellular oncogenes discovered in naturally occurring oncogenic retrovirus animal models also potentially occurs in humans following gene therapy with vectors using these viruses.

Positive results win in scientific research. Negative results seldom get published, except in the context of positive results. Yet, looking back 40 years later, I would say that the significance of the negative results (Table 4) of our research outweighed, by far, the positive results. Our inability to find an oncogenic retrovirus in humans or in most wild mice, seems far more important now than the finding of such viruses in a small minority of wild mice, no matter how interesting it was back then. The absence of human infection with animal oncogenic retroviruses is a significant negative result. There was much public health concern in the 1970s that FeLV might pose a risk to humans in contact with infected cats. Finding no retroviruses in household dogs or parakeets is reassuring. Our negative results in humans have stood the test of time. Over the years numerous “sightings” and even isolations of putative human oncogenic retrovirus have been reported in high-profile scientific journals. All of the isolated viruses were contaminations with extant animal retroviruses. Inability to confirm false claims usually remained unpublished.

retrotransposons that may represent either the relics or predecessors of retroviruses (Griffiths, 2001). In contrast, the lab mouse genome contains 37% endogenous retroviral DNA and retrotransposon activity is far greater in mice than humans. Many families of human endogenous retroviruses (HERVs) have been mapped to chromosomal loci and used to create primate phylogenies (Johnson and Coffin, 1999). Most of these HERVs contain sequences closely related to animal oncogenic retroviruses, but not to HTLV or HIV. As mentioned above, during evolution some animal and human endogenous viruses were apparently acquired by cross species infection. Importantly, all of the HERVs are defective in that they contain mutations that prevent the production and release of complete virus particles (Löwer et al, 1995). Most of the HERVs have undergone recombinational deletion (Belshaw et al, 2007). Therefore, all that effort made during the 1970s to discover infectious or replication competent HERV was destined to fail; we now know almost for certain that there exists no infectious endogenous human oncogenic retrovirus. However, we also have to recognize that, since the human genome sequence is based on a single individual, it remains a remote possibility that genomes of other individuals might reveal “islands” of replication competent HERV. The absence of replication competent virus does not preclude the possibility that expression of HERV genes might have some biologic function. HERV-RNA and proteins, including envelope and incomplete budding particles, have been detected in some germ cell tumors, and as mentioned above, in the placental syncytial trophoblast layer (Blond et al, 1999, Frendo et al, 2003). HERV gene expression has been incriminated in several non-infectious diseases of humans such as insulin-dependent diabetes mellitus (Medstrand and Mager, 1998) but, as yet, without convincing evidence of a pathogenic role. Rarely, an endogenous retroviral insertion site has caused a recessive gene disorder in humans (Hughes and Coffin, 2001; Villesen et al, 2004).

C. Significance retroviruses in humans

endogenous

D. New claims of oncogenic retroviruses in humans

With the human genome now sequenced, we see that about 8% consists of endogenous retroviral DNA sequences and even a greater percentage consists of

New claims of putative oncogenic retroviruses in human tissues continue to appear. These reports rely mainly on PCR-based amplification of retroviral-related DNA sequences and are not accompanied by EM detection of typical virus particles, immunological or biochemical markers, or of virus isolated in tissue cultures. Report of MMTV related ENV-gene DNA in human breast cancer (Wang et al, 1995) has not been confirmed (Mant et al, 2004). A recent report of mouse xenotropic MuLV-related DNA in human prostate cancer (Urisman et al, 2006) remains a mystery. In the 1970s, a putative oncogenic retrovirus had to be isolated in vitro, its ultrastructure visualized, its antigenicity determined, and it had to survive shipment to other labs for confirmation. Clearly, retrovirus hunters of the past have transformed into modern day molecular geneticists.

B. Importance of negative results

Table 4. Important negative results. Lack of any replication competent endogenous or exogenous retrovirus in humans. Lack of any replication competent retroviruses in dogs and parakeets. Lack of any evidence of human infection with animal ongenic retroviruses. Lack of replication competent endogenous retrovirus in wild mice. Lack of any detectable oncogenic effect of LA smog on wild mice or humans Lack of rescueable oncogene by helper leukemia virus in vitro in any species other then rat.

Gardner: Search for oncogenic retroviruses in wild mice and man multiple membrane-spanning protein and confers susceptibility to virus infection. Cell 57, 659-666. Andrews JM, Gardner MB, (1974) Lower motor neuron degeneration associated with type C RNA virus infection mice: Neuropathological features. J Neuropathol Exp Neurol 33, 285-307. Baltimore D (1970) RNA-dependent DNA polymerase in virons of RNA tumor Viruses. Nature 226, 1209-1211. Barbacid M, Robbins KC, Aaronson SA (1979) Wild mouse RNA tumor viruses. A nongenetically transmitted virus group closely related to exogenous leukemia viruses of laboratory mouse strains. J Exp Med 149, 254. Barthold SW (2002) “Muromics”: genomics from perspective of the laboratory mouse. Comp Med 52, 206-223 Belshaw R, Watson J, Katzourakis, Howe A, Wooven-Allen J, Burt A, Tristem M (2007) Rate of recombinational deletion among human endogenous retroviruses. J Virol 81, 94379442. Bentvelzen P, Daams JH (1969) Hereditary infections with mammary tumor viruses in mice. J Natl Cancer Inst 43, 1025-1035. Benveniste RE, Lieber MM, Livingston DM, Sherr CJ, Todaro GJ, Kalter SS (1974) Infectious C Type virus isolated from a baboon placenta. Nature 248, 17-20. Bergeron D, Poliquin L, Houde J, Barbeau B, Rassart E (1992) Analysis of provirus integrated in Fli-1 and Evi-1 regions in Cas-Br-E MuLV-induced Non-T-Non-B-cell leukemias. Virology 191, 272-282. Bergeron D, Houde J, Poliquin L, Barbeau B, Rassart E (1993) Expression and DNA rearrangement of proto-oncogenes in Cas-Br-E-induced non-T, non-B leukemias. Leukemia 7, 954-962. Bishop JM (1987) The molecular genetics of cancer. Science 235, 305-311. Blond JL, Besème F, Duret L, Bouton O, Bedin F, Perron H, Mandrand O, Mallet F (1999) Molecular characterization and placental expression of HERV-W, a new human endogenous retrovirus family. J Virol 73, 1175-1185. Bryant ML, Pal BK, Gardner MB, Elder JH, Jensen FC, Lerner RA (1978) Structural analysis of the major envelope glycoprotein (gp70) of the amphotropic virus and ecotropic type C viruses of wild mice. Virology 84, 348-358. Bryant ML, Scott JL, Pal BK, Estes JD, Gardner MB (1981) Immunopathology of Natural and Experimental Lymphomas Induced by Wild Mouse Leukemia Virus. Am J Pathol 104, 272-282. Cardiff RD (1984) Protoneoplasia: the molecular biology of murine mammary hyperplasia. Adv Cancer Res 42, 167190. Cohen JC, Traina VL, Breznik T, Gardner MB (1982) Development of a mouse mammary tumor virus-negative mouse strain: a new system for the study of mammary carcinogeneis. J Virol 44, 882-885. Cosset FL, Takeuchi Y, Battini JL, Weiss RA, Collins MKL (1995) High-Titer packaging cells producing recombinant retroviruses resistant to human serum. J Virol 69, 74307436. Croce CM (2008) Molecular Origins of Cancer Oncogenes and Cancer. N Engl J Med 358, 502-511. Dandekar S, Rossitto P, Picket S, Mockli G, Bradshaw H, Cardiff R, Gardner MB (1987) Molecular characterization of Akvr-1 restriction gene: A defective endogenous retrovirus identical to Fv-4R. J Virol 61, 308-314. Davé UP, Jenkins NA, Copeland NG (2004) Gene Therapy Insertional Mutagenesis Insight. Science 303, 333. Fine DL, Arthur LO, Gardner MB (1978) Prevalence of murine mammary tumor virus antibody and antigens in normal and tumor-bearing feral mice. J Natl Cancer Inst 61, 485-491.

E. Other infectious causes of human cancer Hepatitis B and C viruses (HBV/HCV), Human Papilloma Virus (HPV), Ebstein-Barr Virus (EBV), Kaposi’s Herpes Sarcoma Virus (KSHV), Heliocobacter pylori and Schistosomiasis mansoni, have all been established, mostly after 1980, as important infectious causes of cancer. It is estimated that infections may be responsible for over 15% of all malignancies worldwide (Kuper et al, 2000). HBV vaccination has markedly reduced the incidence of liver cancer, and the new HPV vaccine promises to protect against cervical cancer. Thus, the 1968 idea of a possible human Type C virus vaccine never came to fruition, but vaccination works against other cancer viruses. Tumor suppressor genes, important “players” in cell behavior, were identified in the mid 1980s by their inactivation by DNA tumor viruses (e.g. polyoma, SV40, adenovirus), which had been less emphasized by the VCP. The initial recognition of AIDS in humans in 1981 was totally unexpected as was the discovery that HIV-1, the causative retrovirus, was in the lentivirus genus, members of which were not known to induce cancer in animals and were, therefore, rather neglected during the 1970s. Needless to say, the VCP did prepare investigators like myself and others to suspect that a retrovirus might be responsible for AIDS in humans and for a similar disease that was recognized soon thereafter in captive macaques. After coming to UC Davis Medical School in 1981, I was able to study new animal lentivirus models of AIDS in macaques (Gardner et al, 2004) and cats (Gardner, 1992), but that is another story). In the late 1980s our efforts to find a lentivirus in wild mice from the Lake Casitas squab farm and elsewhere around the world was another important negative and unpublished study. . This long scientific journey was only possible because of the opportunity given me 40 years ago to enter the world of retroviruses and learn on the job and because of the support provided to me by colleagues at the USC School of Medicine and at the UC Davis Veterinary School, Primate Center and Center for Comparative Medicine. Foremost among these colleagues at UC Davis was Gordon Theilen who helped to recruit me there and whose pioneering research on animal retroviruses was my inspiration.

Acknowledgements The following investigators at the USC School of Medicine participated in the research summarized in this review: Robert McAllister, Earle Officer, Suraiya Rasheed, Vaclov Klement, Marty Bryant, Howard Charman, P Roy Burman, B.K. Pal, Michael Lai, Bob Rongey, Robert Bryan, John Estes, Brian Henderson, Malcolm Pike, John Cassagrande, Ron Ross, and John Hisserick. Support was primarily from NCI contracts. For their critical review of this manuscript I thank Suraiya Rasheed, Peter Barry, Paul Luciw and Steve Barthold.

References Albritton ML, Tseng M, Scadden D, Cunningham JM (1989) A punative murine ecotropic retrovirus receptor gene encodes a

Cancer Therapy Vol 6-A.2, page 69 Frendo JL, Olivier D, Cheynet V, Blond JL, Bouton O, Vidaud M, Rabreau M, Evain-Brion D, Mallet F (2003) Direct Involvement of HERV-W env glycoprotein in human trophoblast cell fusion and differentiation. Molecular and Cellular Biology 23, 3566-3574. Faulkin LJ, Mitchell DJ, Young LJT, Morris DW, Malone RW, Cardiff RD, Gardner MB (1983) Hyperplastic and neoplastic charges in the mammary glands of feral mice free of endogenous MMTV provirus. J Natl Cancer Inst 73, 971982. Gardner MB, Loosli CG, Hanes B, Blackmore W, Teebken D (1970) Pulmonary changes in 7000 mice following prolonged exposure to ambient and filtered Los Angeles air. Arch Environ.Health 20, 310-317. Gardner MB (1971a) Current information on feline and canine cancers and relationship or lack of relationship to human cancer. J Natl Cancer Inst 46, 281-290. Gardner MB, Officer JE, Rongey RW, Estes JD, Turner H, Huebner RJ (1971b) C-type RNA tumor virus genome expression in wild house mice. Nat New Biol 232, 617-620. Gardner MB, Rogney RW, Johnson EY, DeJournett R, Huebner RJ (1971c) C-type tumor virus particles in salivary tissue of domestic house cats. J Natl Cancer Inst 47, 561-568. Gardner MB, Henderson BE, Estes JD, Menck H, Parker JC, Huebner RJ (1973a) Unusually high incidence of spontaneous lymphomas in wild house mice. J Natl Cancer Inst 50, 1571-1579. Gardner MB, Henderson BE, Officer JE, Rongey RW, Parker JC, Olwin C, Estes JD, Huebner RJ (1973b) A spontaneous lower motor neuron disease apparently caused by indigenous type C RNA virus in wild mice. J Natl Cancer Inst 51, 1243-1254. Gardner MB, Henderson BE, Rongey RW, et al (1973c) Spontaneous tumors of aging wild house mice. Incidence, pathology, and C-type virus expression. J Natl Cancer Inst 50, 719-734. Gardner MB, Officer JE, Rongey RW, Charman HP, Hartley JW, Estes JD, Huebner RJ (1973d) C-type RNA tumor virus in wild house mice (Mus musculus). In: Unifying Concepts of Leukemia, Bibl. Haemat. No 39, R.M. Dutcher and L. Chieco-Bianchi (eds.). Karger, Basel, pp. 335-344. Gardner MB, Henderson BE (1974a) Lower-motor-neuron disease in mice and amyotropic lateral sclerosis in man (letter). Lancet 2, 952 Gardner MB, Henderson BE, Menck H, Parker J, Estes JD, Huebner RJ (1974b) Spontaneous tumors and C-type virus in polyoma-infected aging wild house mice. J Natl Cancer Inst 52, 979-981. Gardner MB, Henderson Be, Estes JD, Rongey RW, Casagrande J, Pike M, HuebnerRJ (1976a) The epidemiology and virology of C-type virus-associated hematological cancers and related diseases in wild mice. Cancer Res 36, 574-581. Gardner MB, Klement V, Henderson BE, Meier H, Estes JD,, Huebner RJ (1976b) Genetic control of type C virus of wild mice. Nature 259, 143-145. Gardner MB, Klement V, Rongey RW, Henderson BE, McConahey P, Estes JD,, Huebner RJ (1976c) Type C virus expression in lymphoma-paralysis prone wild mice. J Natl Cancer Inst 57, 585-590 Gardner MB, Rasheed S, Shimizu S, Rongey RW, Henderson BE, McAllister RM Klement V, Charman HP, Gilden RV, Heberling RL, Huebner RL (1977) Search for RNA tumor virus in humans. Origins of Human Cancer 1235-1251. Gardner MB, Chiri A, Dougherty MF, Casagrande J, Estes JD (1979) Congenital transmission of murine leukemia virus from wild mice prone to the development of lymphoma and paralysis. J Natl Cancer Inst 62, 63-70.

Gardner MB (1980a) Historic Background (Chapter 1). In Molecular Biology of RNA Tumor Viruses JR Stephenson (ed.) Academic Press, New York, pp. 1-46. Gardner MB, Estes JD, Casagrande J, Rasheed S (1980b) Prevention of paralysis and suppression of lymphoma in wild mice by passive immunization to congenitally transmitted murine leukemia virus. J Natl Cancer Inst 64, 1251-1257. Gardner MB, Lund JK, Cardiff (1980c) Prevalence of distribution of mammary tumor antigen detectable by immunocytochemistry in spontaneous breast tumors of wild mice. J Natl Cancer Inst 64, 1251-1257. Gardner MB, Rasheed A, Pal BK, Estes JD, O’Brien SJ (1980c) Akvr-1, a dominant murine leukemia virus restriction gene, is polymorphic leukemia-prone wild mice. Proc Natl Acad Sci U S A 77, 531-535. Gardner MB, Rongey RW, Sarma P, Arnstein P (1981)Electron microscopic search for retrovirus particles in spontaneous tumors of the parakeet. Vet Pathol 18, 700-703. Gardner MB, Rasheed S (1982) Retroviruses in Feral Mice. Int Rev Exp Pathol 23, 209-265. Gardner MB, (1985) Retroviral spongiform polioencephalomyelopathy. Rev Infect Dis 7 Gardner MB, Kozal CA, O’Brien SJ (1991) The Lake Casitas wild mouse: Evolving genetic resistance to retroviral disease. Trends Genet 7, 22-27. Gardner MB Yamamoto J, Marthas M, Miller C, Jennings M, Rosenthal A, Luciw P, Planelles V, Yima T, Giavedoni L, Ahmed S, Steimer K, Haigwood N, Pedersen N (1992) SIV and FIV vaccine studies at UC Davis: 1991 Update. AIDS Res Hum Retroviruses 8, 1495-1498. Gardner MB, Carlos MP, Luciw PA (2004) Simian retroviruses. In: AIDS and Other Manifestations of HIV Infection, G.P. Wormser (ed.) Fourth Edition, Raven Press, LTD, New York, pp. 195-165. Griffiths DJ (2001) Endogenous retrovirus in the human genome sequence. Genome Biol 2, 1017.1-1017.5. Gross L (1951) “Spontaneous” leukemia developing in C3H mice following inoculation, in infancy, with AK-leukemia extracts, or AK-embryos. Proc Soc Exp Biol Med 76, 27-32. Hartley JW, Rowe WP (1976) Naturally occurring murine viruses in wild mice. Characterization of a new “amphotropic” class. J Virol 19, 19-25. Hartley JW, Wolford NK, Old LJ, Rowe WP (1977) A new class of murine leukemia virus associated with development of spontaneous lymphoma. Proc Natl Acad Sci U S A 74, 789792. Hanes B, Gardner MB, Loosli CG, Heidbreder G, Kogan B, Marylander H, Huebner RJ (1970) Pet association with selected human cancers: A household questionnaire survey. J Natl Cancer Inst 45, 115-1162 Hayward WS, Neel BG,, Astrin SM (1981) Activation of Cellular onc gene by promoter insertion in ALV-induced lymphoid leukosis. Nature 290, 475-480. Hersh EM, Hanna MG, Gutterman JU, Mavligit G, Yurconic M, Gschwind CR (1974) Human immune response to active immunization with Rauscher leukemia virus I. Cellmeditated and cell-associated immunity. J Natl Cancer Inst 53, 317. Howard TM, Sheng Z, Wang M, Wu T, Rasheed S (2006) Molecular and phylogenetic analysis of a new amphotropic murine leukemia virus (MuLV-1313). Virology Journal 3, 101-118. Huebner RJ, Todaro GT (1969) Oncogenes of RNA tumor viruses as determinants of cancer. Proc Natl Acad Sci U S A 64, 1087-1094. Hughes JF, Coffin JM (2001) Evidence for genomic rearrangements mediated by human endogenous retroviruses during primate evolution. Nature Genetics 29, 487-489.

Gardner: Search for oncogenic retroviruses in wild mice and man Ikeda H, Sugimura H (1989) Fv-4 resistance gene: a truncated endogenous murine leukemia virus with ecotropic interference properties. J Virol 63, 5405-5412. Johnson W,, Coffin J (1999) Constructing primate phylogenies from ancient retrovirus sequences. Proc Natl Acad Sci U S A 96, 10254-10260. Jolicoeur P, Nicolaiew N, Des Groseillers L, Rassart E (1983) Molecular cloning of infectious viral DNA from ecotropic neurotropic wild mouse retrovirus. J Virol 45, 146-150. Jung YT, Lyu MS, Buckler-White A, Kozak CA (2002) Characterization of a polytropic murine leukemia virus proviral sequence associated with the virus resistance gene Rmcf of DBA/2 mice. J Virol 76, 8218-8224. Kay DG, Gravel C, Pothier F, Laperriere A, Robitaille Y, Jolicoeur P (1993) Neurological disease induced by transgenic mice expressing the env gene of the Cas-Br-E murine retrovirus. Proc Natl Acad Sci U S A 90, 4538-4542. Kelly PF, Vandergriff J, Bathwani A, Nienhuis AW, Vanin EF (2000) Highly efficient gene transfer into cord blood nonbese diabetic/severe combined immunodeficiency repopulating cells by oncoretroviral vector particles pseudotyped with the feline endogenous retrovirus (RD114) envelope protein. Blood 96, 1206-1214. Klement V, Gardner MB, Henderson BE, Ihle JN, Estes JD, Stanley AG, Gilden RV (1976) Inefficient humoral immune response of wild mice to persistent leukemia virus infection. J Natl Cancer Inst 57, 1169-1173. Kozak C, O’Neill R (1987) Diverse wild mouse origin of xenotropic, mink cell focus-forming, and two types of ecotropic proviral genes. J Virol 61, 3082-3088. Kuper H, Adami H-O, Trichopoulos D (2000) Infections as a major preventable cause of human cancer. J Intern Med 248, 171-183. Lai MMC, Shimizu S, Rasheed S, Pal BK, Gardner MB (1982) Characterizartion of genome structure of amphotropic and ecotropic wild mouse retroviruses. J Virol 41, 605-614. Langdon WY, Hartley JW, Klinken SP, Ruscetti SK, Morse HC III (1989) v-cbl, and oncogene from a dual-recombinant murine retrovirus that induced early B-lineage lymphomas. Proc Natl Acad Sci U S A 86, 1168-1172. Law LW, Moloney JB (1961) Studies of congenital transmission of leukemia virus in mice. Proc Soc Exp Biol Med 108, 715-723. Levy JA (1976) Endogenous C-type viruses: double agents in natural life processes. Biomedicine 24, 84-93. Löwer R, Löwer J, Reinhard K (1996) The viruses in all of us: Characteristics and biological significance of human endogenous retrovirus sequences. Proc Natl Acad Sci U S A 93, 5177-5184. Lwoff A (1960) Tumor viruses and cancer problem: a summation of the conference. Cancer Res 20, 407-414. Mant C, Gillett C, D’Arrigo C, Cason J (2004) Human murine mammary tumor virus-like agents are genetically distinct from endogenous retroviruses and are not detectable in breast cancer cell lines or biopsies. Virology 318, 393-404. McAllister RM, Nicolson M, Gardner MB, Rongey RW, Rashees S, Sarma PS, Huebner RJ, Hatanaka M, Oroszlan S, Gilden RV, Kabigting A, Vernon L (1972) C-type virus released from cultured human rhabdomyosarcoma cells. Nat New Biol 235, 3-6. Medstrand P, Mager DL (1998) Human-specific integrations of the HERV-K endogenous retrovirus family. J Virol 72, 9782-9787. Miller DG, Edwards RH, Miller AD (1994) Cloning of the cellular receptor for amphotropic murine retrovirus reveals homology to that for gibbon ape leukemia virus. Proc Natl Acad Sci U S A 91, 78-82

O’Brien SJ, Berman EJ, Estes JD, Gardner MB (1983) Murine retroviral restriction genes Fv-4 and Akvr-1 are alleles of a single locus. J Virol 47, 649-651. Odaka TH, Ikeda H, Yoshikura H, Moriwaka K, Suzuki S (1981) Fv-4: gene controlling resistance to NB-tropic Friend murine leukemia virus. Distribution in wild mice. Introduction into genetic background of BALB/c mice, and mapping of chromosomes. J Natl Cancer Inst 67, 1123-1127. Officer JE, Tecson N, Estes JD, Fontanilla E, Rongey RW, Gardner MB (1973) Isolation of a neurotropic type C virus. Science 181 Peters G, Brooks S, Smith R, Dickson C (1983) Tumorigenesis by mouse mammary tumor virus: Evidence for a common region for provirus integration in mammary tumors. Cell 33, 369-377. Portis JL (1990a) Wild mouse retrovirus: Pathogenesis. Curr Top Microbiol Immunol 160, 11-27. Rasheed S, Nelson-Rees WA, Toth EM, Arstein P, Gardner MB (1974) Characterization of newlyderived human sarcoma cell line (HT-1080). Cancer 33, 1027-1033. Rasheed S, Gardner M, Chan E (1976) Amphotropic host range of naturally occurring wild mouse leukemia viruses. J Virol 19, 13-18. Rasheed S, Toth E, Gardner MB (1977) Characterization of purely ecotropic and amphotropic naturally occurring wild mouse leukemia viruses. Intervirology 8, 323-335. Rasheed S, Charman HP, Gardner MB (1978a) Wild rat type C virus: Isolation and characterization. Virology 89, 605-609. Rasheed S, Gardner MB, Huebner RJ (1978b) In vitro isolation of stable rat sarcoma viruses. Proc Natl Acad Sci U S A 75, 2972-2776. Rasheed S, Gardner MB (1979) Herpesvirus in Orang-utan (H. pongo) and search for retrovirus in nonhuman primates. Comp Immunol Microbiol Infect Dis 2, 265-274. Rasheed S, Gardner MB (1981) Growth properties and susceptibility to viral transformation of skin fibroblasts from individuals at high genetic risk to colorectal cancer. J Natl Cancer Inst 66, 43-49. Rasheed S, Pal BK, Gardner MB (1982) Characterization of highly oncogenic murine leukemia virus from wild mice. Int J Cancer 29, 345-350. Relander T, Johansson M, Olsson K, Ikeda Y, Takeuchi Y, Collins M, Richter J (2005) Gene transfer to repopulating human CD34+ cells using amphotropic-GALV-, or RD114Pseudotyped HIV-1-Based vectors from stable producer cells. Mol Ther 11, 452-459. Rhim JS, Cho HC, Rabstein L, RJ Gordon, Bryan RJ, Gardner MB, Huebner RJ (1972) Transformation of mouse cells infected with AKR leukemia virus induced by smog extracts. Nature 239, 103-107. Rettig RA (1977) Cancer crusade. The story of the National Cancer Act of 1971. Princeton University Press, Princeton, New Jersey. Rice MC, Gardner MB, O’Brien SJ (1980) Genetic diversity in feral house mice which harbor and suffer from infectious murine leukemia virus. Biochem Genet 18 Robinson HL, Astrin SM, Senior AM, Salazar FH (1981) Host susceptibility to endogenous viruses: defective, gygoproteinexpressiong proviruses interfere with infections. J Virol 40, 745-751. Rongey RW, Hlavackova A, Lara S, Estes J, Gardner MB (1973) Types B and C RNA virus in breast tissue and milk of wild mice. J Natl Cancer Inst. 50, 1581. Rosenberg N, Jolicoeur P (1997) Retroviral pathogenesis. In: Retroviruses edited by: Coffin JM, Hughes SH and Varmus He. Cold Spring Harbor, Cold Spring Harbor Laboratory Press pp. 475-585.

Cancer Therapy Vol 6-A.2, page 71 Rowe WP (1973) Genetic factors in the natural history of murine leukemia virus infection. GHA Clowes Memorial Lecture. Cancer Res 33, 3061-3068. Roy-Burman P, Rogney RW, Henderson BE, Gardner MB (1973) Attempts to detect RNA tumor virus in human milk. Nat New Biol 244, 146. Roy-Burman P, Pal BK,, Gardner MB (1972) Inhibitor of the DNA-dependent DNA polymerase of some RNA tumor viruse3s in feline sera. Nat New Biol 237, 45-47. Roy-Burman P, Dougherty M, Pal BK, Charman HP, Klement V, Gardner MB (1976) Assay for type C virus in mouse sera based on particulate reverse transriptase activity. J Virol 19, 1107-1110. Sarma PS, Gazdar AF (1974) Recent progress in studies of mouse type-c virus. Curr Top Microbiol Immunol 68, 128. Sharpe AH, Jaenisch R, Ruprecht RM (1987) Retroviruses and mouse embryos: A rapid model for neurovirulence and transplacental antiviral therapy. Science 236, 1671-1674. Stehelin D, Varmus HE, Bishop JM,, Vogt PK (1976) DNA related to transforming gene(s) of avian sarcoma viruses is present in normal avian DNA. Nature 260, 170-173. Tailor CS, Nouri A, Zhao Y, Takeuchi Y, Kabat D (1999) A sodium-dependent neutral-amino-acid transporter mediates infections of feline and baboon endogenous retrovirus and simian type D retroviruses. J Virol 73, 4470-7774. Tarlinton RE, Meers J,, Young PR (2006) Retroviral invasion of the koala genome. Nature 442, 79-81. Temin HM (1964) Nature of the provirus of Rous sarcoma. J Natl Cancer Inst 17, 557-570. Temin HM, Mizutani (1970) RNA-dependant DNA polymerase in virons of Rous sarcoma virus. Nature 226, 1211-1213.

Tsuji Y, Doi H, Yamabe T, Ishimaru T, Miyamoto T, Hino S (1990) Prevention of mother to child transmission of human T-lymphotropic virus type-I. Pediatrics 96, 11-17. Urisman A, Molinaro RJ, Fischer N, Plummer SJ, Casey G, Klein EA, Malathi K, Magi-Galluzzi C, Tubbs RR, Ganem D, Silverman RH, DeRisi JL (2006) Identification of a novel gammaretrovirus in prostate tumors of patients homozygous for R462Q RNASEL variant. PLoS Pathogens 2, e25 doi:10.1371/journal.ppat.0020025. Villensen P, Aagaard L, Wiuf C, Pedersen FS (2004) Identification of endogenous retroviral reading frames in the human genome. Retrovirology 1, 32. Wang Y, Holland JF, Bleiweiss IJ, Melana S, Lui X, Pelisson I, Cantarella A, Stellrecht K, Mani S, Pogo BG-T (1995) Detetion of mammary tumor virus env gene-like sequences in human breast cancer. Cancer Res 55, 5173-5179. Weiss RA (2006) The discovery of endogenous retroviruses. Retrovirology 3, 67. Wiley CA, Gardner MB, (1993) The pathogenesis of murine retroviral infection of the central nervous system. Brain Pathol 3, 123-128. Wu T, Yan Y, Kozak C (2005) Rmcf2, a xenotropic provirus in the Asian mouse species Mus castaneus, blocks infections by polytropic mouse gammaretroviruses. J Virol 79, 96779684. Zachary JF, Knupp CJ, Wong PKY (1986) Noninflammatory spongiform polioencephalopathy caused by a neurotropic temperature-sensitive mutant of Moloney murine leukemia virus TB. Am J Pathol 114, 457-468.

Gardner: Search for oncogenic retroviruses in wild mice and man