Pancreatic Cancer and Tumor Microenvironment

Pancreatic Cancer and Tumor Microenvironment Editors Paul Grippo Hidayatullah Munshi

Pancreatic Cancer and Tumor Microenvironment

Editors

Paul J. Grippo Hidayatullah G. Munshi The Robert H. Lurie Comprehensive Cancer Center of Northwestern University Chicago, IL 60611, USA

Transworld Research Network, T.C. 37/661 (2), Fort P.O., Trivandrum-695 023 Kerala, India

Published by Transworld Research Network 2012; Rights Reserved Transworld Research Network T.C. 37/661(2), Fort P.O., Trivandrum-695 023, Kerala, India Editors Paul J. Grippo Hidayatullah G. Munshi Managing Editor S.G. Pandalai Publication Manager A. Gayathri Transworld Research Network and the Editors assume no responsibility for the opinions and statements advanced by contributors ISBN: 978-81-7895-548-3

Preface Pancreatic cancer, one of the deadliest of human cancers, is associated with pronounced collagen-rich stromal reaction. This ‘desmoplastic reaction’ is being increasingly recognized as an active contributor to pancreatic cancer progression. There is significant interplay between the stroma and the cancer cells to induce epithelial-mesenchymal transition (EMT), modulate microRNAs, affect the number and function of both the cancer stem cells and the mesenchymal stem cells, and also contribute to chemo-resistance. As our understanding of the role and regulation of this desomplastic reaction increases, it may provide therapeutic targets against this deadly cancer.

The goal of this monograph is to review the importance of the stromal reaction in pancreatic cancer. Particular emphasis is placed on mesenchymal cells and how their propensity to modulate the tumor microenvironment results in tumor growth, invasion and metastasis. The various chapters address key pathways that modulate mesenchymal cells and how these pathways function to modulate EMT, stem cells and contribute to the generation of the stromal compartment. These effects are likely contributors to the aggressive nature of pancreatic cancer, ultimately responsible for poor prognosis and reduced survival. These chapters also identify potential therapeutic targets and summarize ongoing clinical trials designed to specifically target the stromal reaction present in pancreatic cancer. Thus, there is potential for enhanced efficacy with current chemotherapeutics, even those defined as ineffective as single agents in pancreatic cancer, if we can successfully target the stromal reaction.

We feel privileged to have worked on this monograph with an outstanding group of co-authors from throughout the world. Many of them have a long-term interest in understanding the biology of pancreatic cancer, including the involvement of the ECM, and continue to be leaders in this area of cancer research. As this monograph would not have been possible without their contributions, we owe them a debt of gratitude. The quality of this monograph reflects their hard work and is to their credit.

We are also grateful to Mrs. A. Gayathri, Publication Manager, and Dr. S.G. Pandalai, Managing Editor at Transworld Research Network Publishers, who have worked diligently to bring this monograph to publication. Their assistance is much appreciated.

Paul J. Grippo, PhD. Hidayatullah G. Munshi, MD.

Foreword It takes two to tango : Pancreatic cancer and its microenvironment Pancreatic cancer is a devastating disease with a dismal prognosis, largely due to its propensity for early local and distant spread.

Understandably, researchers in the field initially concentrated on trying to elucidate the biology of cancer cells themselves. It was a reasonable approach to think that unraveling the cellular and molecular processes in the cancer cells would point towards possible ways of destroying such malignant cells.

However, despite significant advances in our knowledge regarding the molecular biology of pancreatic cancer cells, the success of treatment options was disappointing, with little improvement in the outlook for patients over the decades, even with combinations of therapies that included surgery, chemotherapy and radiotherapy.

While the research focus on cancer cell biology is not without merit, it has to be remembered that a tumor is much more than a clump of abnormal/malignant cells. It is a virtual ‘organ’ in itself since it contains not only cancer cells, but also has a blood and lymphatic supply, a neuronal network and infiltrating immune/inflammatory cells all embedded in a glycoprotein rich extracellular matrix. These structures form the ‘microenvironment’ of the tumor – it is an environment that most cancers cannot survive without since it provides nutrients, oxygen and a supportive milieu for abnormal growth.

In pancreatic cancer, pathologists and clinicians had long noted the prominent stromal/desmoplastic reaction which often appeared to ‘crowd out’ the cancer cells, so much so that some tumors appeared to be composed of 90% stroma and 10% tumor elements. However, up until 10 years ago, there was little research interest in the stroma surrounding pancreatic cancer cells, with the dense fibrosis often being dismissed as a mere epiphenomenon of the primary cancer. This situation changed drastically in the early 2000s, largely because of a significant breakthrough a few years earlier which provided unequivocal identification and characterization of the cell type that produces the fibrosis of both chronic pancreatitis and pancreatic cancer.

These cells, known as the pancreatic stellate cells, have been intensely studied in recent years and evidence is accumulating of their central role in interactions with not only pancreatic cancer cells but also other cell types in the stromal reaction such as endothelial cells and immune cells. Even more interesting is the observation that PSCs from the primary tumor travel to distant metastatic sites, which has prompted a rethink of our understanding of

tumor metastasis because it suggests that metastatic potential is not the sole preserve of cancer cells.

In view of the increased recognition of the key influence of tumorstromal interactions in local growth and distant spread of pancreatic cancer, the publication of this book is very timely. The Chapters have been written by senior investigators as well as up and coming researchers in the field who provide the readers with a state-of-the-art exposĂŠ on specific topics relevant to their experience and expertise. The Chapters describing the pathological characteristics (Chapter 1) as well as imaging findings (Chapter 2) of the desmoplastic reaction surrounding pancreatic cancer cells set the scene for subsequent Chapters that include a comprehensive discussion of current knowledge regarding the role of pancreatic stellate cells (Chapter 3) in the production of the stromal reaction and the critical place of these cells in stromal-tumor interactions facilitating cancer progression. A detailed discussion of animal models (Chapter 4), particularly mouse models that are the mainstays of in vivo research in this area, given the relative inaccessibility of the human pancreas during life, is highly useful in terms of underlining their importance in our current understanding of cancer development and progression. Chapters outlining epithelial-mesenchymal transition (Chapter 5), the role of the mesenchyme and the role of stem cells in progression and/or recurrence (Chapter 6) and the signaling pathways that mediate epithelial mesenchymal transition as well as other stromal-tumor interactions (Chapter 7), make an important contribution to the theme. This information prepares the ground for a discussion on the role of the tumor microenvironment in the well-known chemoresistance of pancreatic cancer (Chapter 8). Obviously, the ultimate goal of researchers in this field of tumor microenvironment interactions in pancreatic cancer is to be able to provide novel insights for improved approaches to pancreatic cancer treatment. The final Chapter discussing the development of potential strategies to target the stroma (Chapter 9) brings together the significant advances made by workers in recent years. There is no doubt that these novel approaches represent our best chance yet of winning the war against this deadly disease, which has claimed countless lives thus far.

Minoti Apte, MBBS., MMedSci., PhD., AGAF. Professor of Medicine Director, Pancreatic Research Group South Western Sydney Clinical School University of New South Wales Sydney, NSW 2052 Australia

Contents

Contributors Chapter 1 Pathology of pancreatic stroma in PDAC Zeshaan A. Rasheed, William Matsui and Anirban Maitra

1

Chapter 2 Imaging the pancreatic ECM Palamadai N. Venkatasubramanian

11

Chapter 3 Pancreatic stellate cells and fibrosis Phoebe Phillips

29

Chapter 4 Pancreatic cancer and the tumor microenvironment: Mesenchyme’s role in pancreatic carcinogenesis Laurent Bartholin Chapter 5 Epithelial-mesenchymal transition and pancreatic cancer progression Surabhi Dangi-Garimella, Seth B. Krantz, Mario A. Shields Paul J. Grippo and Hidayatullah G. Munshi Chapter 6 Pancreatic cancer stem cell and mesenchymal stem cell Shin Hamada and Tooru Shimosegawa

55

95

111

Chapter 7 Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer: Hedgehog, Notch and TGFβ Jennifer M. Bailey and Steven D. Leach Chapter 8 Desmoplasia and chemoresistance in pancreatic cancer Clifford J. Whatcott, Richard G. Posner, Daniel D. Von Hoff and Haiyong Han Chapter 9 Therapeutic targeting of pancreatic stroma Andrew S. Liss and Sarah P. Thayer

123

139

157

Contributors Jennifer M. Bailey Department of Surgery and The McKusick-Nathans Institute of Genetic Medicine, Johns Hopkins University School of Medicine Baltimore, MD, USA Laurent Bartholin TGFβ and Pancreatic Adenocarcinoma Lab, Centre de Recherche en Cancérologie de LYON (CRCL), UMR INSERM 1052 - CNRS 5286 – Lyon University, Lyon, France Surabhi Dangi-Garimella Division of Hematology/Oncology, Department of Medicine Feinberg School of Medicine, Northwestern University Chicago, IL, USA Paul J. Grippo Division of Surgical Oncology, Department of Surgery, Feinberg School of Medicine, Northwestern University; The Jesse Brown VA Medical Center; and The Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago, IL, USA Shin Hamada Division of Gastroenterology, Tohoku University Graduate School of Medicine, Aobaku, Sendai, Japan Haiyong Han Clinical Translational Research Division, The Translational Genomics Research Institute, Scottsdale, Arizona, USA Seth B. Krantz Division of Surgical Oncology, Department of Surgery, Feinberg School of Medicine, Northwestern University Chicago, IL, USA Steven D. Leach The Department of Surgery and The McKusick-Nathans Institute of Genetic Medicine, Johns Hopkins University School of Medicine Baltimore, MD, USA

Andrew S. Liss Department of Surgery and the Andrew L. Warshaw Institute for Pancreatic Cancer Research, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA Anirban Maitra Department of Oncology and Department of Pathology, Johns Hopkins University School of Medicine, Sol Goldman Pancreatic Cancer Research Center, Baltimore, Maryland, USA William Matsui Department of Oncology, Johns Hopkins University School of Medicine Sidney Kimmel Comprehensive Cancer Center, Baltimore Maryland, USA Hidayatullah G. Munshi Division of Hematology/Oncology, Department of Medicine, Feinberg School of Medicine, Northwestern University; The Jesse Brown VA Medical Center; and The Robert H. Lurie Comprehensive Cancer Center of Northwestern University Chicago, IL, USA Phoebe Phillips Pancreatic Cancer Translational Research Team, Pancreatic Research Group, South Western Sydney Clinical School and School of Medical Sciences, University of New South Wales Sydney, Australia Richard G. Posner Clinical Translational Research Division, The Translational Genomics Research Institute, Scottsdale, Arizona, USA Zeshaan A. Rasheed Department of Oncology, Johns Hopkins University School of Medicine Sidney Kimmel Comprehensive Cancer Center, Baltimore Maryland USA Mario A. Shields Division of Hematology/Oncology, Department of Medicine Feinberg School of Medicine, Northwestern University Chicago, IL, USA

Tooru Shimosegawa Division of Gastroenterology, Tohoku University Graduate School of Medicine, Aobaku, Sendai, Japan Sarah P. Thayer Department of Surgery and the Andrew L. Warshaw Institute for Pancreatic Cancer Research, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA Palamadai N. Venkatasubramanian Center for Basic M.R. Research, NorthShore University Health System, Evanston, IL, USA Daniel D. Von Hoff Clinical Translational Research Division, The Translational Genomics Research Institute, Scottsdale, Arizona, USA Clifford J. Whatcott Clinical Translational Research Division, The Translational Genomics Research Institute, Scottsdale, Arizona USA

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 1-10 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

1. Pathology of pancreatic stroma in PDAC Zeshaan A. Rasheed1, William Matsui1and Anirban Maitra1,2 Johns Hopkins University School of Medicine, 1Department of Oncology 2 Department of Pathology Sol Goldman Pancreatic Cancer Research Center, Baltimore, Maryland 21231, USA

Abstract. Pancreatic adenocarcinoma is a highly lethal disease that is histologically characterized by a dense desmoplastic reaction (DR) surrounding malignant epithelial cells. The DR is composed of extracellular matrix (ECM) proteins, fibroblasts, stellate cells, endothelial cells, immune cells, and neurons. Accumulating evidence indicates that the epithelial and stromal compartments interact to enhance the aggressive nature of this disease. Pancreatic cancer cells release various factors that stimulate the stroma.

Stromal cells, in turn, release mitogenic substances that stimulate tumor growth, invasion, and resistance to therapy. As we better understand the interactions between the stromal and epithelial cell compartments in pancreatic adenocarcinoma, it is becoming evident that anticancer therapies targeting the stroma, in addition to epithelial cells, may play a key role in improving clinical outcomes for patients with this deadly disease.

Introduction Pancreatic adenocarcinoma has one of the highest mortality rates of any malignancy. Most patients are initially diagnosed with unresectable Correspondence/Reprint request: Dr. Anirban Maitra, Professor of Pathology and Oncology, The Sol Goldman Pancreatic Cancer Research Center, Johns Hopkins University School of Medicine, 1550 Orleans St, CRB II Room 345, Baltimore, MD 21231, USA. E-mail: amaitra1@jhmi.edu

2

Zeshaan A. Rasheed et al.

(metastatic or locally advanced) disease and the median survival for those patients is only six to nine months. Even for patients who undergo surgery for localized disease, the five-year survival rate is only about 20% [1]. The poor prognosis for patients with pancreatic adenocarcinoma is largely due to its propensity to metastasize and its resistance to radiation and chemotherapy.

Most research has focused on better understanding the genetics and biology of pancreatic adenocarcinoma in order to develop better therapeutic strategies to treat patients with this malignancy. In recent years, one area of research has focused on the stromal compartment in pancreatic adenocarcinoma, and studies indicate that it contributes to the poor prognosis of patients with this malignancy.

A hallmark of pancreatic adenocarcinoma is the presence of a dense desmoplastic reaction (DR) that consist largely of fibroblasts, pancreatic stellate cells (PSCs), and extracellular matrix (ECM) proteins, including collagens I and III and fibronectin [2]. Other cells in the stroma include endothelial cells, immune cells, pericytes, and nerve fibers. In a number of malignancies, the presence of a DR in primary tumors has been associated with worse clinical outcomes [3, 4]. The DR in pancreatic adenocarcinoma is thought to contribute to the aggressive nature of this tumor by fostering tumor growth and metastatic spread and enhancing drug resistance. In this chapter we will review the biological importance of interactions between the stromal compartment and malignant epithelial cells of pancreatic adenocarcinoma. Better understanding these interactions are important for developing stroma-targeting therapies that could lead to improved patient outcomes.

Pancreatic cancer stellate cells and fibroblasts The major cellular constituents of the DR in pancreatic adenocarcinoma are PSCs and fibroblasts. The initial isolation and culture of PSCs from rats and humans was described in 1998 [5, 6]. The origin of PSCs is unclear, but they are believed to arise from mesenchymal, endodermal, and neuroectodermal origins. In normal tissue they can be identified based on the expression of glial fibrillary acidic protein (GFAP) and desmin. PSCs are thought to play an important role in the pathobiology of pancreatitis and pancreatic cancer, in which case they transform to an activated state and acquire characteristics of myofibroblasts and express Îą-smooth muscle actin (ÎąSMA). In the diseased organ, activated PSCs are postulated to arise from quiescent PSCs, fibroblasts, bone marrow-derived cells, or epithelialmesenchymal transition (EMT). Whatever their origins, PSCs are activated in

Stroma in PDAC

3

response to pancreatic injury and inflammation and play an active role in the progression of malignancy.

PSCs secrete and respond to a number of cytokines and have been found to actively proliferate, migrate, and produce ECM proteins, including type I collagen and fibronectin. A number of pathways have been implicated in this process including transforming growth factor beta (TGF-β), hepatocyte growth factor (HGF), fibroblast growth factor (FGF), and epidermal growth factor (EGF). Lohr et al. showed that TGF-β expressing Panc1 cells induced the proliferation of co-cultured fibroblasts with a concomitant increase in type I collagen and fibronectin expression. Furthermore, orthotopic injection of TGF-β expressing Panc1 cells in nude mice resulted in the formation of tumors with more desmoplasia and increased amounts of type I collagen and fibronectin [7]. A similar observation was made when PSCs were incubated in conditioned media from pancreatic cancer cell lines; PSCs cultured in conditioned media or with TGF-β displayed increased [3H] thymidine uptake and collagen synthesis [8]. Recent studies examining Hedgehog signaling in pancreatic adenocarcinoma have found that sonic hedgehog ligand secreted from malignant epithelial cells acts on fibroblasts and PSCs, promoting desmoplasia and increased metastasis in an orthotopic model [9, 10].

Secreted protein acidic and rich in cysteine (SPARC) is secreted by fibroblasts and found to be involved in cell migration and proliferation [11].

Furthermore, Infante et al. found that intratumoral SPARC expression was associated with worse prognosis in patients with resected disease [12].

Another protein that is secreted by PSCs upon stimulation by epithelial cells is periostin which modulates tumor cell invasion via AKT signaling and EMT [13, 14]. Global gene expression analysis of genes differentially expressed in fibroblasts cultured in the absence or presence of pancreatic cancer cell lines revealed 43 up-regulated and 31 down-regulated transcripts [15]. Among the most highly up-regulated genes were members of the CXC/CC chemokine family including MCP-1 (CCL2), interleukin- (IL)-8 (CXCL8), GRO1 (CXCL1), and GRO2 (CXCL2); all of which have been implicated in tumor invasion and angiogenesis.

Just as secreted factors from pancreatic cancer cells leads to activation of stromal cells, growth factors and proteins elicited from stromal cells have profound effects on the epithelial cell compartment. A number of studies have found that co-culturing fibroblasts or PSCs with pancreatic cancer cell lines leads to an increase in their invasive properties [15]. To explore the molecular mechanism for this phenomenon, Sato et al. studied global gene expression in cell lines co-cultured with fibroblasts and found differential expression of 143 genes [15]. The expression of five genes was validated:

4

Zeshaan A. Rasheed et al.

COX-2, hyaluronan synthase 2 (HAS2), and MMP-1 were up-regulated and Gravin was down-regulated [15]. All of the up-regulated genes have been linked with tumor cell invasion. In another study, Ohuchida found that coculture of fibroblasts with cancer cell lines or culturing cell lines in conditioned media from fibroblasts promoted invasiveness. They found that HGF secreted by the fibroblasts lead to increased c-Met phosphorylation and mitogen-activated protein kinase activity in the epithelial cell compartment [16].

PSCs, fibroblasts, and epithelial cells all contribute to regulating the composition of the ECM via proteolytic enzymes, or matrix metalloproteinases (MMPs), that are frequently over expressed in pancreatic cancer cells and are involved in the dynamic remodeling and turnover of ECM proteins [17, 18]. Specifically, MMP-9, MMP-2, and MMP-1 have been found to be expressed when cells come in contact with certain ECM proteins, such as type I collagen, thereby regulating the invasive properties of these cells [19-21]. Other studies have also identified tissue inhibitors of metalloproteinases (TIMPs), inhibitors of the extracellular proteinases, which are commonly over expressed in pancreatic adenocarcinoma and other malignancies. As an example, the serine protease inhibitor nexin-2 (SERPINE2) is secreted by epithelial cells and activates PSCs resulting in greater deposition of ECM proteins, increased tumor growth, and more invasive growth [22, 23].

A significant amount of research has focused on the effect of the ECM on tumor invasion, but it is also apparent that the ECM contributes to resistance to chemotherapy and radiation therapy. Cells cultured in the presence of conditioned media from PSCs and xenografts formed by coninjection of PSCs and pancreatic cancer cells lead to resistance to gemcitabine and radiation therapy [24]. In another study, type I collagen led to an increase in cell proliferation and relative protection from apoptosis [8].

Endothelial cells Though in vitro and animal studies have shown that pancreatic cancer cells induce angiogenesis by secreting molecules like VEGF and FGF, clinically, pancreatic tumors have been found to be hypovascular and hypoxic [25, 26]. Studies have linked hypoxia in patient tumors with worse clinical outcomes, including increased rates of tumor growth and metastasis [27-29].

Furthermore, clinical trials of the VEGF inhibitor, bevacizumab, in patients with pancreatic cancer have not shown clinical efficacy, suggesting that the microenvironment in pancreatic cancer is already hypoxic [30-32]. The hypoxic environment is largely thought to be secondary to the fibrotic

Stroma in PDAC

5

microenvironment produced by PSCs and the expression of a number of antiangiogenic substances, including endostatin and matrix metalloproteinase 12 [26, 33].

The hypoxic microenvironment in pancreatic cancer has been shown to induce the expression of hypoxia-inducible factor-1 (HIF1) in a number of cancers and has been shown to be overexpressed in pancreatic tumors from patients [34]. HIF1 expression has been associated with drug resistance and enhanced cell invasion in pancreatic cancer, which may be mediated through c-Met and Hedgehog signaling [35-38]. Recent work by Olive et al. also revealed that Hedgehog signaling might regulate intratumoral vascular density via effects on the tumor stroma [39].

Inflammatory cells Inflammatory cells are part of the stromal reaction found in pancreatic cancer and are thought to contribute to the development and progression of this disease. Supporting a role for inflammation in the development of pancreatic cancer is the finding that pancreatitis, or chronic inflammation of the pancreas, is a risk factor for developing pancreatic adenocarcinoma [40].

Several studies have shown that leukocytic infiltrates in pancreatic adenocarcinoma are largely immunosuppressive and associated with worse survival in humans [41, 42]. Furthermore, proinflammatory markers, such as IL-6, IL-8, IL-10, and IL-1 receptor antagonist, are elevated in the serum of patients with pancreatic cancer, and IL-6 has been associated with worse survival [43]. IL-6 has been shown to signal via signal transducer and activator of transcription 3 (Stat3), which is activated in pancreatic cancer and involved in tumor growth [44].

Pericytes and nerve cells The normal pancreas has an abundant nerve supply consisting of ganglia and myelinated and unmyelinated nerve cells. The degree of perineural invasion in the tumor has been associated with worse survival after resection and has been shown to be mediated by the chemokine receptor CX3CR1 [45, 46]. The mechanism for the association between perineural invasion and worse prognosis is not clear, but it may be a reflection of the degree of differentiation of the tumor [45]. In addition, the size and density of nerves have been shown to be increased in pancreatic cancers compared to normal tissue [47], but the impact of these cells on tumor progression and or resistance is not clear. It is possible that this process leads to significant patient morbidity caused by chronic pain.

6

Zeshaan A. Rasheed et al.

Clinically targeting the pancreatic stroma The stromal compartment and its interactions with malignant epithelial cells in pancreatic adenocarcinoma are clearly important in the pathogenesis of this deadly malignancy. Recent studies are beginning to show that targeting the stromal compartment in pancreatic cancer may have antitumor effects and may enhance sensitivity to radiation and chemotherapy. Several novel inhibitors of the Hedgehog pathway are being clinically tested in pancreatic adenocarcinoma (http://www.clinicaltrials.gov) and preclinical studies have shown that they are able to deplete the tumor of its stroma, resulting in increased intratumoral vascular density and a concomitant increase in intratumoral chemotherapy concentrations [39]. Furthermore, inhibition of the Hedgehog pathway has been shown to abrogate the formation of metastases in murine models [48, 49]. Preclinical studies of HIF inhibition in pancreatic cancer have shown sensitization of tumors to radiation with or without concurrent treatment with gemcitabine or 5fluorouracil [50]. Recent analogs of the active component in curcumin have been developed that inhibit Stat3 phosphorylation and have anti-tumor activities against pancreatic cancer [51, 52]. Likewise, an inhibitor of Src, one of the activators of Stat3 signaling, also abrogates Stat3 phosphorylation and reduces tumor growth in a murine model [53]. In a mouse model and in humans, activation of the tumor necrosis factor receptor using an agonist CD40 monoclonal antibody facilitated the infiltration of tumors with tumoricidal macrophages, resulting in depletion of tumor stroma and tumor regression [54].

Significant improvements in the survival of patients with pancreatic cancer have not been realized in more than four decades despite advances in our understanding of pancreatic cancer biology and genetics. Increasing evidence indicates that the DR in pancreatic cancer, consisting of stellate cells, fibroblasts, and a number of ECM proteins, plays an important role in the progression and resistance to chemotherapy of this malignancy. Through better understanding the interactions between the stromal compartment and malignant epithelial cells we are beginning to develop therapeutic strategies that target these interactions. It is yet to be determined if these approaches to treating pancreatic cancer will change the aggressive course of this malignancy.

References 1.

Winter, J.M., et al., 1423 pancreaticoduodenectomies for pancreatic cancer: A single-institution experience. J Gastrointest Surg, 2006. 10(9): p. 1199-210; discussion 1210-1.

Stroma in PDAC

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

7

Maitra, A. and R.H. Hruban, Pancreatic cancer. Annu Rev Pathol, 2008. 3: p.

157-88.

Erkan, M., et al., The activated stroma index is a novel and independent prognostic marker in pancreatic ductal adenocarcinoma. Clin Gastroenterol Hepatol, 2008. 6(10): p. 1155-61.

Hasebe, T., et al., Fibrotic focus in invasive ductal carcinoma of the breast: a histopathological prognostic parameter for tumor recurrence and tumor death within three years after the initial operation. Jpn J Cancer Res, 1997. 88(6): p.

590-9.

Apte, M.V., et al., Periacinar stellate shaped cells in rat pancreas: identification, isolation, and culture. Gut, 1998. 43(1): p. 128-33.

Bachem, M.G., et al., Identification, culture, and characterization of pancreatic stellate cells in rats and humans. Gastroenterology, 1998. 115(2): p. 421-32.

Lohr, M., et al., Transforming growth factor-beta1 induces desmoplasia in an experimental model of human pancreatic carcinoma. Cancer Res, 2001. 61(2): p.

550-5.

Armstrong, T., et al., Type I collagen promotes the malignant phenotype of pancreatic ductal adenocarcinoma. Clin Cancer Res, 2004. 10(21): p. 7427-37.

Bailey, J.M., A.M. Mohr, and M.A. Hollingsworth, Sonic hedgehog paracrine signaling regulates metastasis and lymphangiogenesis in pancreatic cancer.

Oncogene, 2009. 28(40): p. 3513-25.

Bailey, J.M., et al., Sonic hedgehog promotes desmoplasia in pancreatic cancer.

Clin Cancer Res, 2008. 14(19): p. 5995-6004.

Seux, M., et al., TP53INP1 decreases pancreatic cancer cell migration by regulating SPARC expression. Oncogene, 2011.

Infante, J.R., et al., Peritumoral fibroblast SPARC expression and patient outcome with resectable pancreatic adenocarcinoma. J Clin Oncol, 2007. 25(3): p. 319-25.

Erkan, M., et al., Periostin creates a tumor-supportive microenvironment in the pancreas by sustaining fibrogenic stellate cell activity. Gastroenterology, 2007.

132(4): p. 1447-64.

Kanno, A., et al., Periostin, secreted from stromal cells, has biphasic effect on cell migration and correlates with the epithelial to mesenchymal transition of human pancreatic cancer cells. Int J Cancer, 2008. 122(12): p. 2707-18.

Sato, N., N. Maehara, and M. Goggins, Gene expression profiling of tumorstromal interactions between pancreatic cancer cells and stromal fibroblasts.

Cancer Res, 2004. 64(19): p. 6950-6.

Ohuchida, K., et al., Radiation to stromal fibroblasts increases invasiveness of pancreatic cancer cells through tumor-stromal interactions. Cancer Res, 2004.

64(9): p. 3215-22.

Ottaviano, A.J., et al., Extracellular matrix-mediated membrane-type 1 matrix metalloproteinase expression in pancreatic ductal cells is regulated by transforming growth factor-beta1. Cancer Res, 2006. 66(14): p. 7032-40.

Mahadevan, D. and D.D. Von Hoff, Tumor-stroma interactions in pancreatic ductal adenocarcinoma. Mol Cancer Ther, 2007. 6(4): p. 1186-97.

8

Zeshaan A. Rasheed et al.

19. Zhang, K., et al., Slug enhances invasion ability of pancreatic cancer cells through upregulation of matrix metalloproteinase-9 and actin cytoskeleton remodeling. Lab Invest. 91(3): p. 426-38.

20. Shields, M.A., et al., Pancreatic Cancer Cells Respond to Type I Collagen by Inducing Snail Expression to Promote Membrane Type 1 Matrix Metalloproteinase-dependent Collagen Invasion. J Biol Chem. 286(12): p.

10495-504.

21. Binker, M.G., et al., TGF-beta1 increases invasiveness of SW1990 cells through Rac1/ROS/NF-kappaB/IL-6/MMP-2. Biochem Biophys Res Commun. 405(1): p.

140-5.

22. Neesse, A., et al., Pancreatic stellate cells potentiate proinvasive effects of SERPINE2 expression in pancreatic cancer xenograft tumors. Pancreatology, 2007. 7(4): p. 380-5.

23. Buchholz, M., et al., SERPINE2 (protease nexin I) promotes extracellular matrix production and local invasion of pancreatic tumors in vivo. Cancer Res, 2003.

63(16): p. 4945-51.

24. Hwang, R.F., et al., Cancer-associated stromal fibroblasts promote pancreatic tumor progression. Cancer Res, 2008. 68(3): p. 918-26.

25. Koong, A.C., et al., Pancreatic tumors show high levels of hypoxia. Int J Radiat Oncol Biol Phys, 2000. 48(4): p. 919-22.

26. Erkan, M., et al., Cancer-stellate cell interactions perpetuate the hypoxia-fibrosis cycle in pancreatic ductal adenocarcinoma. Neoplasia, 2009. 11(5): p. 497-508.

27. Chang, Q., et al., Hypoxia predicts aggressive growth and spontaneous metastasis formation from orthotopically grown primary xenografts of human pancreatic cancer. Cancer Res. 71(8): p. 3110-20.

28. Hiraoka, N., et al., Tumour necrosis is a postoperative prognostic marker for pancreatic cancer patients with a high interobserver reproducibility in histological evaluation. Br J Cancer. 103(7): p. 1057-65.

29. Reiser-Erkan, C., et al., Hypoxia-inducible proto-oncogene Pim-1 is a prognostic marker in pancreatic ductal adenocarcinoma. Cancer Biol Ther, 2008. 7(9): p.

1352-9.

30. Kindler, H.L., et al., Gemcitabine plus bevacizumab compared with gemcitabine plus placebo in patients with advanced pancreatic cancer: phase III trial of the Cancer and Leukemia Group B (CALGB 80303). J Clin Oncol. 28(22): p.

3617-22.

31. Small, W., Jr., et al., Phase II Trial of Full-Dose Gemcitabine and Bevacizumab in Combination With Attenuated Three-Dimensional Conformal Radiotherapy in Patients With Localized Pancreatic Cancer. Int J Radiat Oncol Biol Phys. 80(2): p. 476-82.

32. Fogelman, D., et al., Bevacizumab plus gemcitabine and oxaliplatin as first-line therapy for metastatic or locally advanced pancreatic cancer: a phase II trial.

Cancer Chemother Pharmacol.

33. Brammer, R.D., S.R. Bramhall, and M.C. Eggo, Endostatin expression in pancreatic tissue is modulated by elastase. Br J Cancer, 2005. 92(1): p. 89-93.

Stroma in PDAC

9

34. Zhong, H., et al., Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. Cancer Res, 1999. 59(22): p. 5830-5.

35. Akakura, N., et al., Constitutive expression of hypoxia-inducible factor-1alpha renders pancreatic cancer cells resistant to apoptosis induced by hypoxia and nutrient deprivation. Cancer Res, 2001. 61(17): p. 6548-54.

36. Niizeki, H., et al., Hypoxia enhances the expression of autocrine motility factor and the motility of human pancreatic cancer cells. Br J Cancer, 2002. 86(12): p.

1914-9.

37. Ide, T., et al., The hypoxic environment in tumor-stromal cells accelerates pancreatic cancer progression via the activation of paracrine hepatocyte growth factor/c-Met signaling. Ann Surg Oncol, 2007. 14(9): p. 2600-7.

38. Onishi, H., et al., Hypoxia activates the hedgehog signaling pathway in a ligandindependent manner by upregulation of Smo transcription in pancreatic cancer.

Cancer Sci. 39. Olive, K.P., et al., Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science, 2009. 324(5933): p. 1457-61.

40. Lowenfels, A.B., et al., Pancreatitis and the risk of pancreatic cancer.

International Pancreatitis Study Group. N Engl J Med, 1993. 328(20): p. 1433-7.

41. Clark, C.E., et al., Dynamics of the immune reaction to pancreatic cancer from inception to invasion. Cancer Res, 2007. 67(19): p. 9518-27.

42. De Monte, L., et al., Intratumor T helper type 2 cell infiltrate correlates with cancer-associated fibroblast thymic stromal lymphopoietin production and reduced survival in pancreatic cancer. J Exp Med. 208(3): p. 469-78.

43. Ebrahimi, B., et al., Cytokines in pancreatic carcinoma: correlation with phenotypic characteristics and prognosis. Cancer, 2004. 101(12): p. 2727-36.

44. Scholz, A., et al., Activated signal transducer and activator of transcription 3 (STAT3) supports the malignant phenotype of human pancreatic cancer.

Gastroenterology, 2003. 125(3): p. 891-905.

45. Hirai, I., et al., Perineural invasion in pancreatic cancer. Pancreas, 2002. 24(1): p. 15-25.

46. Marchesi, F., et al., The chemokine receptor CX3CR1 is involved in the neural tropism and malignant behavior of pancreatic ductal adenocarcinoma. Cancer Res, 2008. 68(21): p. 9060-9.

47. Ceyhan, G.O., et al., Nerve growth factor and artemin are paracrine mediators of pancreatic neuropathy in pancreatic adenocarcinoma. Ann Surg. 251(5): p.

923-31.

48. Feldmann, G., et al., Blockade of Hedgehog Signaling Inhibits Pancreatic Cancer Invasion and Metastases: A New Paradigm for Combination Therapy in Solid Cancers. Cancer Res, 2007. 67(5): p. 2187-2196.

49. Feldmann, G., et al., An orally bioavailable small-molecule inhibitor of Hedgehog signaling inhibits tumor initiation and metastasis in pancreatic cancer. Mol Cancer Ther, 2008. 7(9): p. 2725-2735.

10

Zeshaan A. Rasheed et al.

50. Schwartz, D.L., et al., Radiosensitization and stromal imaging response correlates for the HIF-1 inhibitor PX-478 given with or without chemotherapy in pancreatic cancer. Mol Cancer Ther. 9(7): p. 2057-67.

51. Bisht, S., et al., Systemic administration of polymeric nanoparticle-encapsulated curcumin (NanoCurc) blocks tumor growth and metastases in preclinical models of pancreatic cancer. Mol Cancer Ther. 9(8): p. 2255-64.

52. Lin, L., et al., Novel STAT3 phosphorylation inhibitors exhibit potent growthsuppressive activity in pancreatic and breast cancer cells. Cancer Res. 70(6): p.

2445-54.

53. Rajeshkumar, N.V., et al., Antitumor effects and biomarkers of activity of AZD0530, a Src inhibitor, in pancreatic cancer. Clin Cancer Res, 2009. 15(12): p. 4138-46.

54. Beatty, G.L., et al., CD40 agonists alter tumor stroma and show efficacy against pancreatic carcinoma in mice and humans. Science, 2011. 331(6024): p. 1612-6.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 11-27 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

2. Imaging the pancreatic ECM Palamadai N. Venkatasubramanian

Center for Basic M.R. Research, NorthShore University HealthSystem, Evanston, IL 60201, USA

Introduction The extracellular matrix (ECM) is the non-cellular component present within all tissues and organs and provides essential physical scaffolding for the cellular constituents in addition to initiating crucial biochemical and biomechanical cues that are required for tissue morphogenesis, differentiation and homeostasis [1]. The ECM is essentially composed of water, proteins, and polysaccharides. However, the composition and topology of ECM in each tissue is unique and is generated during tissue development through a dynamic interaction between the various cellular components and the evolving cellular and protein microenvironment. The physical and biochemical characteristics of the ECM generate the biochemical and mechanical properties of each organ, such as its tensile and compressive strength and elasticity. The ECM also mediates protection by a buffering action that maintains extracellular homeostasis and water retention. Further, the ECM directs essential morphological organization and physiological function by binding growth factors and interacting with cell-surface receptors to transduce signal and regulate gene transcription. Cell adhesion to the ECM Correspondence/Reprint request: Dr. Palamadai N. Venkatasubramanian, Center for Basic M.R. Research NorthShore University HealthSystem, Evanston, IL 60201, USA E-mail: PVenkatasubramanian@northshore.org

12

Palamadai N. Venkatasubramanian

is mediated by ECM receptors, such as integrins. Adhesion mediates cytoskeletal coupling to the ECM and is involved in cell migration through the ECM. The ECM is a highly dynamic structure that is constantly being remodeled. The properties of ECM vary not only from one tissue to another, but also from one physiological state to another, such as normal versus cancerous. However, the value of ECM changes induced by or related to pathology has yet to be fully targeted by imaging technology as a means to diagnose cancer and/or evaluate novel therapies.

ECM composition The ECM is composed of two main classes of macromolecules: fibrous proteins and proteoglycans [1-3]. The main fibrous proteins are collagens, elastins, fibronectins and laminins [4]. Collagen is the most abundant fibrous protein within the ECM and constitutes up to 30% of the total protein mass.

This scaffolding protein contributes to the tensile strength of tissue, regulates cell adhesion, supports chemotaxis and migration, and directs tissue development [1,5]. The main ECM component of interstitial tissue is fibrillar type I collagen. Type I collagen is the prototypical fibrillar collagen that represents up to 90% of the protein content of connective tissues. The bulk of interstitial collagen is transcribed and secreted by fibroblasts that either reside in the stroma or are from neighboring tissues. Type I collagen consists of two Îą1(I) chains and one Îą2(I) chain. The chains wrap around one another in a rope-like fashion to form the triple helix, nucleated at the N-terminus.

Collagen fibrils are strengthened by covalent crosslinking and deamination by the enzyme lysyl oxidase. Type IV collagen, a non-fibrillar collagen, is a major constituent of basement membranes, where cells adhere and interact extensively with Type IV collagen. Collagen XV and XVIII, the other nonfibrillar collagens, are also expressed in the basement membrane. Although collagen fibers are generally a heterogeneous mixture of different types, within a given tissue one type of collagen usually predominates. The fibrillar collagens influence multiple aspects of cell behavior by serving as ligands for integrin and non-integrin receptors, and as a reservoir of growth factors and peptide mediators. By their physical characteristics such as, fiber size, organization, density, stiffness, and pore size between fibers, they influence cell behavior. Recently, several studies have demonstrated that cancer progression results in structural changes in the collagenous stroma [6-11].

Elastin, another major ECM fiber, provides recoil to tissues that undergo repeated stretch. Elastin stretch is limited by tight association with collagen fibrils [12]. A third fibrous protein, fibronectin is involved in directing the organization of the interstitial ECM and plays a crucial role in mediating cell

Imaging the pancreatic ECM

13

attachment and function [1]. Fibronectin is also important for cell migration during development and has been implicated in cardiovascular disease and tumor metastasis [5,13].

Proteoglycans, the other type of macromolecular constituent of the ECM, fill the majority of the extracellular interstitial space within the tissue in the form of a hydrated gel [2]. Proteoglycans are composed of glycosaminoglycan chains covalently linked to a protein core [3,14]. The glycosaminoglycan chains on the protein core are unbranched polysaccharide chains composed of repeating disaccharide units that can be divided further into sulfated (chondroitin sulfate, heparin sulfate and keratin sulfate) and non-sulfated (hyaluronic acid) glycosaminoglycans [3]. These molecules are extremely hydrophilic and adopt highly extended conformations that are essential for hydrogel formation and that enable matrices that are formed by these molecules to withstand high compressive forces. Proteoglycans have a wide variety of functions that reflect their unique buffering, hydration, binding and force-resistance properties. Many genetic diseases have been linked to mutations in genes encoding proteoglycans [2].

Function of stroma in health and pathology Stroma surrounding the normal epithelial tissue is composed of nonactivated adipocytes and fibroblasts which secrete and organize fibrous proteins such as type I and type III collagens, elastin, fibronectin, and proteoglycans such as hyaluronic acid. Tissue homeostasis is achieved by a balance between metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs), controlled activity of crosslinking enzymes such as transglutaminases, and a number of ECM-bound growth factors that modulate cell growth and migration via a tightly controlled feedback circuit [1,15,16]. Aging is associated with a number of changes in the ECM. Thinning of the basement membrane occurs as a result of elevated MMPmediated degradation and reduced protein synthesis. Elevated levels of interleukins and cytokines result in inflammation. The collagen fiber network is modified because of inappropriate collagen crosslinking which renders the tissue less elastic, more weak and rigid. This compromised ECM organization is believed to promote age-related diseases such as cancer [1,17-19]. ECM changes occur also as a response to acute injury which induces would healing. Typically, this process involves the formation of a fibrin clot, stimulation of monocyte infiltration to the damaged ECM, and release of growth factors, MMPs and cytokines that promote angiogenesis and fibroblast migration and proliferation. The fibroblasts then synthesize and deposit large amounts of ECM proteins such as collagens, fibronectin

14

Palamadai N. Venkatasubramanian

and hyaluronic acid. In healthy tissue, this ECM remodeling is controlled by feedback mechanisms and tissue homeostasis is restored.

Tumor stroma displays characteristics expected of unhealed wound.

There is increasing evidence that ECM remodeling plays an active role in tumor progression and invasion. Inflammation and infiltrating T lymphocytes activate fibroblasts and induce their transdifferentiation into myofibroblasts [1] which deposit large quantities of ECM proteins and growth factors.

Aberrant crosslinking of newly deposited collagen and elastin fibers generate more rigid fibrils that stiffen the tissue. MMPs secreted and activated by tumor cells and myofibroblasts also remodel the basement membrane surrounding the tumor and release ECM-embedded growth factors. Vascular permeability and new vessel growth are promoted by these growth factors which facilitate tumor invasion and metastasis.

ECM in pancreatic cancer Pancreatic ductal adenocarcinoma is characterized by a fibrotic stroma with excessive connective tissue. The resistance of these tumors to all therapies has been traced to its pathological features [20]. Sparse vascularization coupled with copious deposition of extracellular components define the stroma in pancreatic cancer. The desmoplastic stroma surrounding the cancer cells participates in tumor-stroma interactions that promote pancreatic cancer cell invasion and metastasis [21]. The progression of precancerous PanIN lesions to pancreatic cancer is accompanied by changes in stroma. Early PanIN lesions may be associated with small amounts of normal stroma surrounding the normal pancreatic ducts. By contrast, PanIN 3 lesions begin to display enhanced stroma formation and invasive carcinoma is associated with extensive stroma. Although the cancer cells themselves are capable of synthesizing and releasing collagens, in PDAC, it is the stromal cells that are involved in ECM protein synthesis. Pancreatic stellate cells (PSCs) have emerged as important regulators of desmoplasia in pancreatic cancer [22] (see Chapter 3). The stroma in PDAC consists of proliferating fibroblasts and pancreatic stellate cells that produce and deposit fibronectin and collagens I and III [23]. Altered gene expression profile in the cancerassociated stroma might be responsible for the increased expression of collagen I. Many epithelial malignancies such as breast, prostate, ovarian and pancreatic cancers exhibit a significant stromal reaction around tumor cells.

Compared to other epithelial malignancies, pancreatic cancer displays the most prominent stromal reaction [24]; however, study of the epithelialstromal interaction in pancreatic cancer has been initiated only recently. This

Imaging the pancreatic ECM

15

lack of investigation of stroma in pancreatic cancer may be due to limited access to tumor tissue which may be alleviated by the availability of suitable animal models (see Chapter 4). Results of the few animal studies that have investigated the role of ECM in pancreatic cancer appear to mimic the situation in human pancreatic cancer. Tumors produced by injection of a mixture of PSCs and cancer cells had increased fibrosis and larger tumors (see Chapter 3). In another study by the same group, coinjection of PSCs with cancer cells overexpressing the serine protease inhibitor SERPINE2 into nude mice was reported to result in increased tumor growth [22]. These findings suggest that the increased ECM (particularly fibrillar collagen) deposition as a consequence of protease inhibition might facilitate cancer progression. A mutant Kras (G12D) progression mouse model of pancreatic ductal adenocarcinoma generated extensive ductal lesions and the acinar parenchyma was replaced by an intense desmoplastic reaction composed of collagen, fibroblasts, and inflammatory cells similar to human pancreatic cancer [25] (as discussed in Chapter 4).

Histopathological analyses of human pancreatic ductal adenocarcinoma (see Chapter 1) depicted dense collagen I and III bundles associated with fibroblasts with loss of basement membrane integrity and invasion of malignant cells into the interstitial matrix with exposure to collagens [26].

An in vitro investigation found that pancreatic cancer cells that were grown on type I collagen had a higher proliferative capacity most likely due to the overexpression of antiapoptotic protein Mcl-1 [27]. Another immunohistochemical study of human pancreatic cancer that used digital image analysis to quantify the expression of ECM proteins in tumor stroma [28] reported that the overall surface covered by collagen types I and III and fibronectin was significantly higher in tumor stroma than in normal tissue. In contrast to normal pancreas, other ECM proteins such as laminins, collagen type IV, and vitronectin were partially lost in the basement membranes of pancreatic cancer. In addition, laminins, fibronectin, and collagen type IV were found to increase tumor cell migration. It has been suggested that key ECM proteins such as collagen and fibronectin likely interact with cell surface integrin receptors to provide survival signals to pancreatic stellate cells and pancreatic cancer cells [26].

The role of stroma in the drug resistance of pancreatic cancer (see Chapter 8) has been experimentally verified using a mouse model of the disease [29]. In the context of the physiological resistance to drug treatment that solid tumors possess, it has been postulated that anomalous assembly of the collagen network component and its interaction with the proteoglycans component of the tumor ECM could greatly influence the physiological barrier to macromolecule motion posed by healthy tissue ECM [30]. Using

16

Palamadai N. Venkatasubramanian

in vitro experiments, it was verified that tumors with a well-defined collagen network are more resistant to penetration by macromolecular drugs compared with tumors that exhibit a loose collagen network. It was further determined that macromolecule access to tumor tissue is dominated by deficiencies in collagen assembly and is relatively insensitive to variations in glycosaminoglycan content. These results identify ECM characteristics of tumors, particularly pancreatic cancer, which could be useful in predicting penetration by therapeutic macromolecules. Imaging methods that are sensitive to the changing ECM characteristics of tumors are highly valuable as they are likely to have both diagnostic and predictive potential in assessing pancreatic cancer. In this context, it must be remembered that to date the microstructure of pancreas in its healthy state has not been explored in detail using any single imaging modality [31].

Imaging the ECM The extracellular matrix is a complex network of glycoproteins and proteoglycans that originated with the advent of multicellular organisms [4].

However, most direct imaging of ECM in tumors has been focused on imaging collagen, the most abundant protein in mammals. This is in part due the significance of collagen network remodeling in desmoplastic reaction associated with tumors, and to a large extent due to the availability of optical imaging methods to visualize changes in fibrillar collagen. Atomic resolution techniques such as atomic force microscopy and X-ray diffraction technique have been used to determine molecular structures of collagen fibrils from tendon [32,33] but are not useful for in vivo applications. The combined use of multiphoton microscopy (MPM) and second harmonic generation (SHG) imaging have been useful in monitoring the interaction of cancer cells with the surrounding stroma [34]. MPM is a variation of conventional laser scanning confocal microscopy and has become the method of choice for the investigation of single cell dynamics in living tissue specimens and in vivo [35]. MPM allows intrinsic contrast imaging, in addition to fluorescence based contrast. The primary sources of intrinsic contrast in MPM are the fibrillar collagen and elastin, which are the primary ECM constituents, as well as certain metabolites such as NAD and NADH. Collagen fibers and other highly repetitive chiral structures interact with femtosecond light pulses causing nonlinear light scattering and interference termed second harmonic generation (SHG), which produces visible light of exactly half the wavelength of the illumination [36,37]. The structure of ECM can thus be revealed by simultaneous imaging of collagen and elastin autofluorescence.

By using SHG and MPM, ECM was analyzed in esophageal cancer showing

Imaging the pancreatic ECM

17

that desmoplastic collagen fibers lost their typical fine structure [38]. In a breast cancer model, tumor-associated collagen signatures that could be used as characteristic markers of tumor expression were identified [39].

SHG was first demonstrated by Kleinman in crystalline quartz in 1962 [40]. Recent studies of the three-dimensional in vivo structures of wellordered protein assemblies such as collagen, microtubules, and muscle myosin are beginning to establish second harmonic imaging microscopy (SHIM) as a non-destructive imaging modality that holds promise for both basic research and clinical pathology. There are a number of detailed reviews on the theory, instrumentation and biological applications of SHG imaging and SHIM [36,41,42]. Therefore, we will discuss only the applications of SHG to image collagen in tumor tissues. SHG is a second-order nonlinear optical process that has symmetry constraints confining signal to regions lacking a center of symmetry. The basic instrumentation requirements for SHG microscopy are those for two photon fluorescence microscopy, namely, a scanning microscope coupled to a pulsed infrared laser [41,42]. When intense laser light passes through a highly polarizable material with a noncentrosymmetric molecular organization, the second harmonic light emerging from the material is at precisely half the wavelength of the light entering the material. This process, known as SHG, changes two nearinfrared incident photons into one emerging visible photon at exactly twice the energy (and half the wavelength). Similar to two-photon absorption, the amplitude of SHG is proportional to the square of the incident light intensity.

Therefore, SHG microscopy resembles two-photon microscopy in its intrinsic optical sectioning characteristics. As opposed to two-photon microscopy, SHG does not involve an excited state; as a result, energy is conserved and the coherence of laser light is preserved in SHG. Because of this coherence, most of the signal wave propagates with the laser. The exact ratio of the forward to backward signal is dependent upon the sample characteristics.

SGH signals are directly obtainable from several structural protein arrays, including collagen, without the use of exogenous molecular probes.

Biological materials can be highly polarizable and often assemble into large, ordered noncentrosymmetric structures. It is this property of collagen fibrils and their relevance in tumor microenvironment that we discussed earlier that has made collagen the target of several recent tumor studies.

In 1986, Freund et al. [43] used SHG microscopy to study the endogenous collagen structure in a rat tail tendon at ~50Îźm resolution. Four types of collagen (I, II, III and V) form fibrils, type IV forms sheets in basement membrane and types VI and IX bind collagen to other cell components [42]. Type I collagen is highly crystalline and is an effective generator of second harmonics. SHG has the potential to distinguish between

18

Palamadai N. Venkatasubramanian

different collagen types as well as the sensitivity to image collagen at high resolution, with diffraction-limited resolution near 300nm [44,45]. The relative alignment of fibrils and fibers is reflected in the magnitude of χ2, which is experimentally manifested in the SHG intensity [46]. The forwardbackward ratio carries information related to the sub-resolution size and packing of the fibrils and fibers. SHG has the advantage over TEM in being able to image noninvasively through intact 3D tissues with relatively large field of view.

Several recent studies have used SHG and multiphoton microscopy to image collagen in tumors and better understand how changes in density and matrix organization are related to tumor formation and progression. In an ex vivo imaging study of epithelial-stromal interactions in normal mammary glands, mammary tumors and tumor explants, local alterations in collagen density was clearly seen around tumors [10,39]. In addition to detecting alterations in collagen density, changes in the three-dimensional organization of collagen fibers was imaged in tumors using multiphoton laser-scanning microscopy to generate multiphoton excitation of endogenous fluorophores and SHG to image stromal collagen. Based on the observed changes in collagen organization, Keely, et al. have defined three tumor-associated collagen signatures (TACS) that could provide novel markers to diagnose and characterize breast cancer.

More importantly the SHG-imaged TACS-3, defined as collagen fiber bundles oriented perpendicular to the tumor boundary, has been reported to predict the long-term survival rate of human patients [47].

Remodeling of the ECM has been implicated in ovarian cancer as well.

Campagnola et al. [48] found that SHG emission attributes such as directionality and relative intensity which are related to the tissue structure were different in biopsies of human ovarian cancer, indicating changes in the collagen assembly. Based on their SHG observations they concluded that the malignant ovaries were characterized by denser collagen as well as higher regularity at both the fibril and fiber levels. Sahai et al. [49] further showed that invasive cancer cells used collagen fibers to facilitate migration in vivo, emphasizing the need to understand the role of collagen in tumor ECM remodeling and highlighting the value of imaging ECM changes using techniques such as SHG microscopy. Classical histological techniques cannot provide such specific information on ECM alterations in tumors. It has further been shown that SHG signal characteristics are sensitive to the type of collagen that is present and therefore, changes in ECM composition can be detected using this technique. In a detailed investigation using both model systems and mouse models to examine the capabilities of SHG for dynamic imaging of collagen modulation in tumors, Jain et al. [50] demonstrated SHG

Imaging the pancreatic ECM

19

Figure 1. Tumor-associated collagen signatures. (a)-(c) Example of TACS-1. A region of locally dense collagen (a) near (40 Οm 'above') a small tumor region (b) that is within the globally increased collagen region surrounding tumors, resulting from increased SHG (collagen) signal intensity; (c) three-dimensional surface plot of intensity showing an approximately three-fold signal increase at TACS-1. (d), (e) Example of TACS-2, showing straightened (taut) collagen fibers stretched around and constraining an expanded epithelial tumor volume. At regions of TACS-2, quantitative analysis [39] of fiber angles relative to the tumor boundary shows a distribution of fibers around 0° that correlates to non-invading regions of tumor cells. (f) Example of TACS-3, showing radially aligned collagen fibers, reorganized by tumor cells, at regions of tumor cell invasion. At regions of TACS-3, quantitative analysis [39] of fiber angles relative to the tumor boundary shows a distribution of fibers around 90° that correlates with local invasion of tumor cells. (Figure reproduced from [10]).

signal specifically came from fibrillar collagen I and there was no signal from collagen IV which forms the basement membranes. Up regulation of collagen V has been implicated in human breast cancer. Using the metrics of SHG intensity, fiber length, emission directionality and depth-dependent intensities, quantitative discrimination of type I and type V collagens was possible in self-assembled collagen gels that were model systems for ECM in invasive breast cancer [46]. To date there has been no SHG study of ECM

20

Palamadai N. Venkatasubramanian

changes in pancreatic cancer either in animal models or on human tissue although ELISA measurements have indicated changes in type IV collagen in pancreatic cancer. In vitro pancreatic cancer cells were found to produce and secrete more of type IV collagen [51]. The Swedish group also found that patients with pancreatic cancer had increased circulating levels of type IV collagen. Since collagen IV does not have a SHG signal, it is yet unclear how these promising findings could be exploited by nonlinear imaging techniques.

Structural differences between type I and type IV collagen in a number of biological tissues, including lens capsule, sclera, and tendon, have been investigated in vivo by another spectroscopic method called attenuated total reflection/Fourier transform infrared spectroscopy [52].

Both direct and indirect approaches have been employed to examine the pancreatic ECM using MR imaging. Time-signal intensity curve (TIC) obtained from dynamic magnetic resonance imaging was found to be useful in the evaluation of pancreatic fibrosis after pancreaticojejunostomy in patients undergoing pancreaticoduodenectomy [53]. A time-intensity curve (TIC) with a rapid rise to a peak followed by a rapid decline was characteristic of a normal pancreas without fibrosis. Pancreatic TICs with a slow rise to a peak followed by a slow decline or a plateau indicated a fibrotic pancreas. A study that prospectively analyzed dynamic contrast enhanced (DCE) MRI data from patients with focal pancreatic lesions found that histologically measured pancreatic fibrosis was negatively correlated with the rate of contrast agent uptake, and positively correlated with tissue volume fraction occupied by extravascular extracellular space [54]. In this study, rate constant of contrast agent uptake and extravascular extracellular space were calculated from DCE MR parameters, and fibrosis was semiquantitatively estimated by picrosirius staining. While these indirect imaging approaches have shown potential for evaluating pancreatic fibrosis, they are confounded by changes in vasculature associated with pathology, which was clearly illustrated in a contrast enhanced CT study [55].

Direct visualization of the mouse pancreas architecture has been reported using MR microscopic imaging [56]. In this approach, high spatial resolution images were acquired ex vivo from normal mouse pancreas without contrast agents on a high field strength imager. The resulting images were analyzed using volume rendering to resolve components in the pancreas such as islets, acinar cells, blood vessels and ECM. The three-dimensional architecture of the extracellular matrix which appeared as sheets could be characterized using MR microscopy. Using MR microscopy, extensive fibrosis was observed in older Pdx-Cre/LSL-Kras mouse pancreas, which displayed neoplastic lesions (Venkatasubramanian PN, et al. Unpublished results).

Imaging the pancreatic ECM

21

Figure 2. MR microimage of normal mouse pancreas. A. 2D view of a representative slice from a three-dimensionally acquired MR image. B-D. 3D volumerendered images of the boxed area in A. The color scale corresponds to the magnetic resonance signal intensity (green to red, low to high intensity). Cylindrical tube-like and planar sheet-like features can be identified in the high-intensity region of the volume image, in addition to islets, which appear ellipsoid. Note that these tubes and sheet-like features, both of which appear as high-intensity regions, cannot be clearly distinguished in the 2D view. B. A sheet (blue and cyan arrows) is seen underneath islets (red arrows). A tube appears in the bottom right corner (gold arrow). The sheet was measured to be ~116μm thick along the y axis (cyan arrows) and ~298μm in depth along the z axis (blue arrows). The total depth of this sheet through the tissue along the z axis (not shown) is ~1400μm. C. 3D image of the tube (gold arrow) shown in B in a different location with surrounding islets (red arrows). The diameter of the tube is ~88μm. The total length of the tube when tracked through the tissue (not shown) was measured to be ~5850μm. D. 3D image of a tube with a high intensity lumen (red) and the lower intensity wall (green). Note the clear distinction in morphology between the tube and the planar sheet. (Figure reproduced from [56]).

Another potential target for ECM imaging in abnormal stromal extracellular matrix remodeling can be hyaluronan metabolism. Hyaluronan in the ECM provides a favorable microenvironment for cell proliferation and migration, in addition to activating intracellular signals through interaction with cell surface receptors. Progression of breast cancer and many other cancers have been associated with elevated hyaluronan metabolism [57,58].

High levels of hyaluronan within tumor cells or in the peritumor stroma have been observed in many cancers and have been suggested to be prognostic indicators of poor outcome in breast, ovarian, gastric and colorectal cancers [59-61]. Changes in ECM hyaluronan levels are known to occur in pancreatitis [62] and pancreatic cancer [63]. Recently, many molecular

22

Palamadai N. Venkatasubramanian

imaging probes have been developed to report hyaluronan metabolic activity [64-69]. It has been suggested that visualization of the ECM in the MR microscopic images of normal mouse pancreas was based on the strong water-binding capacity of hyaluronan, a glycosaminoglycan [56]. Based on this study, MR microscopy can be a viable technique to indirectly image the ECM via changes in water content associated with hyaluronan levels in pancreatitis and pancreatic cancer.

Summary and conclusions ECM is a non-cellular component of tissues that provide physical scaffolding and participate in physiological function by interacting with cellspecific receptors to transduce various signals within cells. The properties of ECM vary from tissue to tissue and from one physiological state to another.

ECM, which is composed of proteins including collagen, polysaccharides such as proteoglycans, and water is a dynamic structure that is being constantly remodeled, especially in cancer. Many epithelial malignancies including pancreatic cancer exhibit prominent stromal reaction that involves alterations in ECM collagen characteristics. Non-linear optical imaging methods have been used to image this ECM remodeling in breast, ovarian and gastric cancers. Second Harmonic Generation or SHG imaging can directly detect signal from type I collagen which is the predominant component of ECM in healthy and cancerous tissues. SHG has been used to measure alterations in ECM composition such as increased levels of collagen in tumor stroma. Reorganization of type I collagen fibrils and fibers occurs in the tumor stroma in breast, ovarian and other epithelial cancers and SHG signal characteristics are sensitive to such architectural changes. Increased stromal reaction has been known in pancreatic cancer; however, it has not been examined using SHG. It has been established that pancreatic cancer is associated with increased levels of type IV collagen; but SHG cannot image this nonfibrillar collagen. The levels of hyaluronan, a proteoglycan of the ECM, has been known to change in pancreatitis and pancreatic cancer.

Molecular imaging using probes specific for hyaluronan metabolism can thus be used to image ECM changes in pancreatic diseases. Another technique that showed potential in imaging pancreatic ECM without the use of exogenous contrast agents is MR microscopy using high field strength imagers. Whereas SHG is limited in depth and field of view, MR microscopy has the ability to image the entire organ. While these molecular, optical and magnetic resonance imaging approaches have the potential for visualizing the ECM changes that are associated with pancreatic diseases such as pancreatic cancer, the promise these techniques hold has not been exploited. A better

Imaging the pancreatic ECM

23

understanding first of the specific ECM alterations at the molecular level would perhaps provide the knowledge needed to construct specific imaging tools for better diagnosis of pancreatic diseases.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Frantz C, Stewart KM, Weaver VM. The extracellular matrix at a glance.

Journal of Cell Science 123(24): 4195-4200 (2010).

Jarvelainen H, Sainio A, Koulu M, Wight TN, Penttinen R. Extracellular matrix molecules: potential targets in pharmacotherapy. Pharmacological Reviews 61: 198-223 (2009).

Schaefer L, Schaefer RM. Proteoglycans: from structural compounds to signaling molecules. Cell Tissue Res. 339: 237-246 (2010).

Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Molecular Biology of the Cell. Edn. 4. London: Garland; 2002.

Rozario T, DeSimone DW. The extracellular matrix in development and morphogenesis: a dynamic view. Dev. Biol. 341: 126-140 (2010).

Zutter MM. Collagens: they are “mostly� what we are. AACR Education Book American Association for Cancer Research 2010. http://educationbook. aacrjournals.org Ng MR, Brugge JS. A stiff blow from the stroma: collagen crosslinking drives tumor progression. Cancer Cell 16: 455-7 (2009).

Levental KR, Yu H, Kass L, et al. Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell 139: 891-906 (2009).

Weigelt B, Bissell MJ. Unraveling the microenvironmental influences on the normal mammary gland and breast cancer. Semin. Cancer Biol. 18: 311-21 (2008).

Provenzano PP, Inman DR, Eliceiri KW, Knittel JG, Yan L, Rueden CT, White JG, Keely PJ. Collagen density promotes mammary tumor initiation and progression. BMC Medicine 6: 11 (2008).

Provensano PP, Inman DR, Eliceiri KW, Keely PJ. Matrix density-induced mechanoregulationof breast cell phenotype, signaling and gene expression through a FAK-ERK linkage. Oncogene 28: 4326-43 (2009).

Wise SG, Weiss AS. Tropoelastin. Int. J. Biochem. Cell Biol. 41: 494-497 (2009).

Tsang KY, Cheung MC, Chan D, Cheah KS. The developmental roles of the extracellular matrix: beyond structure to regulation. Cell Tissue Res. 339: 93-110 (2010).

Iozz RV, Murdoch AD. Proteoglycans of the extracellular environment: clues from the gene and protein side offer novel perspectives in molecular diversity and function. FASEB J. 10: 598-614 (2009).

Mott JD, Werb Z. Regulation of matrix biology by matrix metalloproteinases.

Curr. Opin. Cell Biol. 16: 558-564 (2004).

Cruz-Munoz W, Khokha R. The role of tissue inhibitors of metalloproteinases in tumorigenesis and metastasis. Crit. Rev. Clin. Lab. Sci. 45: 291-338 (2008).

24

Palamadai N. Venkatasubramanian

17. Coppe JP, Desprez PY, Krtolica A, Campisi J. The secescence-associated secretory phenotype: the dark side of tumor suppression. Annu. Rev. Pathol. 5: 99-118 (2010).

18. Freund A, Orjalo AV, Desprez PY, Campisi J. Inflammatory networks during cellular senescence: causes and consequences. Trends Mol. Med. 16: 238-246 (2010).

19. Sprenger CC, Plymate SR, Reed MJ. Extracellular inflences on tumour angiogenesis in the aged host. Br. J. Cancer 98: 250-255 (2008).

20. Olson P, Hanahan D. Breaching the cancer fortress. Science 324: 1400-1401 (2009).

21. Korc M. Pancreatic cancer associated stroma production. Am. J. Surg. 194(4 Supplement 1): s84-s86 (2007).

22. Neesse A, et al. Stromal biology and therapy in pancreatic cancer. Gut 60: 861868 (2011).

23. Bachem MG, Schunemann M, Ramadani M, et al. Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells.

Gastroenterology 128: 907-921 (2005).

24. Vonlaufen A, Phillips PA, Xu Z, Goldstein D, Pirola RC, Wilson JS, Apte MV. Pancreatic stellate cells and pancreatic cancer cells: an unholy alliance. Cancer Res. 68(19): 7707-7710 (2008).

25. Hingorani SR, Petricoin EF, Maitra A, et al. Preinvasive and invasive ductal pancreatic cancer and its early detection in the mouse. Cancer Cell 4(6): 437-450 (2003).

26. Mahadevan D, Von Hoff DD. Tumor-stroma interactions in pancreatic ductal adenocarcinoma. Mol. Cancer Ther. 6(4): 1186-1197 (2007).

27. Armstrong T, Packham G, Murphy LB, et al. Type I collagen promotes the malignant phenotype of pancreatic ductal adenocarcinoma. Clin. Cancer Res. 10: 7427-37 (2004).

28. Ryschich E, Khamidjanov A, Kerkadze V, Buchler MW, Zoller M, Schmidt J. Promotion of tumor cell migration by extracellular matrix proteins in human pancreatic cancer. Pancreas 38(7): 804-810 (2009).

29. Olive KP, et al. Inhibitionof hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science 324: 1457-1461 (2009).

30. Netti PA, Berk DA, Swartz MA, Grodzinsky AJ, Jain RK. Role of extracellular matrix assembly ininterstitial transport in solid tumors. Cancer Research 60: 2497-2503 (2000).

31. Mori H. New insight of pancreatic imaging: from “unexplored” to “explored”.

Abdom. Imaging (2008).

32. Bozec L, van der Heijden G, Horton M. Collagen fibrils: nanoscale ropes.

Biophys. J. 92: 70-75 (2007).

33. Orgel JP, Irving TC, Miller A, Wess Tj. Microfibrillar structure of type I collagen in situ. Proc. Natl. Acad. Sci. USA 103: 9001-9005 (2006).

34. Zal T, Chodaczek G. Intravital imaging of anti-tumor immune response and the tumor microenvironment. Semin. Immunopathol. 32: 305-317 (2010).

Imaging the pancreatic ECM

25

35. Xu C, Zipfel W, Shear JB, Williams RM, Webb WW. Multiphoton fluorescence excitation: new spectral windows for biological nonlinear microscopy. Science 93: 10763-10768 (1996).

36. Zoumi A, Yeh A, Tromberg BJ. Imaging cells and extracellular matrix in vivo by using second-harmonic generation and two-photon excited fluorescence. Proc.

Natl. Acad. Sci. USA 99: 11014-11019 (2002).

37. Stoller P, Reiser KM, Celliers PM, Rubenchik AM. Polarization-modulated second harmonic generation in collagen. Biophys. J. 82: 3330-3342 (2002).

38. Zhuo S, Chen J, Xie S, Hong Z, Jiang X. Extracting diagnostic stromal organization features based on intrinsic two-photon excited fluorescence and second-harmonic generation signals. J. Biomed. Opt. 14: 020503 (2009).

39. Provenzano PP, Eliceiri KW, Campbell JM, Inman DR, White JG, Keely PJ. Collagen reorganization at the tumor-stromal interface facilitates local invasion.

BMC Medicine 4: 38 (2006).

40. Kleinman DA. Nonlinear dielectric polarization in optical media. Physical Review 126: 1977-79 (1962).

41. Campagnola PJ, Loew LM. Second-harmonic imaging microscopy for visualizing biomolecular arrays in cells, tissues and organisms. Nat. Biotechnology 21(11): 1356-60 (2003).

42. Cox G, Kable E. Second-harmonic imaging of collagen. In Method in Molecular Biology, Vol. 319: Cell Imaging Techniques: Methods and Protocols. DJ Taatjes and BT Mossman (Eds.) Humana Press: Totowa, NJ (2006).

43. Freund I, Deutsch M, Sprecher A. Connective tissue polarity. Optical secondharmonic microscopy, crossed-beam summation, and small-angle scattering in rat-tail tendon. Biophys. J. 50: 693-712 (1986).

44. Cox G, Kable E, Jones A, Fraser I, Manconi F, Gorrell M. 3-dimensional imaging of collagen using second harmonic generation. J. Struct. Biol. 141: 5362 (2002).

45. Cox GC, Manconi F, Kable E. Second harmonic imaging of collagen in mammalian tissue. Proc. SPIE 4620: 148-156 (2002).

46. Ajeti V, Nadiarnykh O, Ponik SM, Keely PJ, Eliceiri KW, Campagnola PJ. Structural changes in mixed Col I/Col V collagen gels probed by SHG microscopy: implications for probing stromal alterations in human breast cancer.

Biomedical Optics Express 2(8): 2307-16 (2011).

47. Conklin MW, Eickhoff JC, Riching KM, et al. Aligned collagen is a prognostic signature for survival in human breast carcinoma. Am. J. Pathol. 178(3): 1221-32 (2011).

48. Nadiarnykh O, LaComb RB, Brewer MA, Campagnola PJ. Alterations of the extracellular matrix in ovarian cancer studied by Second Harmonic Generation imaging microscopy. BMC Cancer 10: 94 (2010).

49. Sahai E, Wyckoff J, Philippar U, Segall JE, Gertler F, Condeelis J. Simultaneous imaging of GFP, CFT and collagen in tumors in vivo using multiphoton microscopy. BMC Biotechnology 5: 14 (2005).

26

Palamadai N. Venkatasubramanian

50. Brown E, McKee T, diTomaso E, Pluen A, Seed B, Boucher Y, Jain RK. Dynamic imaging of collagen and its modulation in tumors in vivo using secondharmonic generation. Nature Medicine 9: 796-800 (2003).

51. Ohlund D, Lundin C, Ardnor B, Oman M, Naredi P, Sund M. Type IV collagen is a tumour stroma-derived biomarker for pancreas cancer. Br. J. Cancer 101: 91-97 (2009).

52. Ozaki Y, Mizuno A, Kaneuchi F. Structural differences between type I and type IV collagen in biological tissues studied in vivo by attenduated total reflection/Fourier transform infrared spectroscopy. Applied Spectroscopy 46(4): 626-630 (1992).

53. Tajima Y, Matsuzaki S, Furui J, Isomoto I, Hayashi K, Kanematsu T. Use of the time-signal intensity curve from dynamic magnetic resonance imaging to evaluate remnant pancreatic fibrosis after pancreaticojejunostomy in patients undergoing pancreaticoduodenectomy. Br. J. Surg. 91(5): 595-600 (2004).

54. Bali MA, Metens T, Denolin V, Delhaye M, Demetter P, Closset J, Matos C. Tumoral and nontumoral pancreas: correlation between quantitative dynamic contrast-enhanced MR imaging and histopathologic parameters. Radiology 261(2): 456-468 (2011).

55. Hata H, Mori H, et al. Fibrous stroma and vascularity of pancreatic carcinoma: correlation with enhancement patterns on CT. Abdom. Imaging (2008).

56. Grippo PJ, Venkatasubramanian PN, et al. Visualization of mouse pancreas architecture using MR microscopy. Am. J. Pathol. 179(2): 610-618 (2011).

57. Tammi RH, Kultti A, Kosma VM, et al. Hyaluronan in human tumors: pathological and prognostic messages from cell-associated and stromal hyaluronan. Semin. Cancer Biol. 18: 288-95 (2008).

58. Veiseh M, Turley EA. Hyaluronan metabolism in remodeling extracellular matrix: probes for imaging and therapy of breast cancer. Integr. Biol. 3: 304-315 (2011).

59. Toole BP, Slomiany MG. Hyaluronan: a constitutive regulator of chemoresistance and malignancy in cancer cells. Semin. Cancer Biol. 2008 Aug;18(4):244-50.

60. Itano N, Zhuo L, Kimata K.Impact of the hyaluronan-rich tumor microenvironment on cancer initiation and progression. Cancer Sci. 2008 Sep; 99(9):1720-5.

61. Lokeshwar VB, Selzer MG. Hyalurondiase: both a tumor promoter and suppressor. Semin. Cancer Biol. 2008 Aug;18(4):281-7.

62. Lohr M, Fischer B, Weber H, Emmrich J, Nizze H, Liebe S, Klopper G. Release of hyaluronan and laminin into pancreatic secretions. Digestion 60: 48-55 (1999).

63. Theocharis AD, Tsara ME, Papageorgacopoulou N, Karavias DD, Theocharis DA. Pancreatic carcinoma is characterized by elevated content of hyluronan and chondroitin sulfate with altered disaccharide composition. Biochim. Biophys.

Acta 1502: 201-206 (2000).

64. Lee H, Lee K, Kim IK, Park TG. Synthesis, characterization, and in vivo diagnostic applications of hyaluronic acid immobilized gold nanoprobes.

Biomaterials 29: 4709-18 (2008).

Imaging the pancreatic ECM

27

65. Oh EJ, Park K, Kim KS, et al. Target specific and long acting delivery of protein, peptide, and nucleotide therapeutics using hyaluronic acid derivatives. J. Controlled Release 141: 2-12 (2010).

66. Lee Y, Lee H, Kim YB, et al. Bioinspired surface immobilization of hyaluronic acid on monodisperse magnetite nanocrystals for targeted cancer imaging. Adv. Mater. Deerfield 20: 4154-57 (2008).

67. Shiftan L, Israely T, Cohen M, Frydman V, Dafni H, Stern R, Neeman M. Magnetic resonance imaging visualization of hyaluronidase in ovarian carcinoma. Cancer Res. 65: 10316-23 (2005).

68. Kim J, Kim KS, Jiang G, et al. In vivo real-time bioimaging of hyaluronic acid derivatives using quantum dots. Biopolymers 89: 1144-53 (2008).

69. Kim KS, Hur W, Park SJ, Hong SW, Choi JE, Goh EJ, Yoon SK, Hahn SK. Bioimaging for targeted delivery of hyaluronic Acid derivatives to the livers in cirrhotic mice using quantum dots. ACS Nano. 2010 Jun 22;4(6):3005-14.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 29-53 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

3. Pancreatic stellate cells and fibrosis Phoebe Phillips

Pancreatic Cancer Translational Research Group, School of Medical Sciences University of New South Wales, Sydney, Australia

Abstract. Pancreatic cancer is characterised by a prominent desmoplastic/stromal reaction. It is now known that pancreatic stellate cells (PSCs) are the principal source of fibrosis in the stroma and interact closely with cancer cells to create a tumor facilitatory environment that stimulates local tumor growth and distant metastasis. Pancreatic fibrosis is initiated when PSCs become activated and undergo morphological and, more importantly, functional changes so that the rate of extracellular matrix (ECM) deposition exceeds the rate of ECM degradation in the gland. It is now well established that pancreatic cancer cells activate PSCs leading to increased fibrosis. This chapter summarises recent advances in our understanding of the role of fibrosis in pancreatic cancer, with particular reference to the central role played by pancreatic stellate cells. An improved knowledge of PSC biology has the potential to provide an insight into pathways that may be therapeutically targeted to inhibit PSC activation, thereby inhibiting the development of fibrosis in pancreatic cancer and interrupting stellate cell - cancer cell interactions so as to retard cancer progression.

Correspondence/Reprint request: Dr. Phoebe Phillips, Pancreatic Cancer Translational Research Group, School of Medical Sciences, University of New South Wales, Sydney, Australia. E-mail: p.phillips@unsw.edu.au

30

Phoebe Phillips

Introduction Pancreatic cancer is a devastating disease with a dismal prognosis. It is the fourth leading cause of cancer related death in Western societies, with a 5 year survival rate of less than 5% [1-2]. A major reason for the poor clinical outcome is its well known resistance to chemotherapeutic agents [3-5].

Despite aggressive treatment regimes, little improvement in patient survival has occurred in the last decade. In fact, the best chemotherapeutic treatments currently only prolong life by ~6-12 weeks [6]. Therefore, alternative approaches are urgently needed to improve the outcome of this condition.

Until recently, studies of the pathogenesis of pancreatic cancer have mainly focused on the molecular biology of the tumor cells themselves. However, there is now significant evidence showing that the intense stromal/ desmoplastic reaction around tumor elements (a feature of the majority of pancreatic cancers) plays an important role in tumor progression [7-12]. Our group established that the cells responsible for producing the stromal reaction in pancreatic cancer are pancreatic stellate cells (PSCs, the key fibrogenic cells in the pancreas) [13]. Our group and others have identified (using in vitro and in vivo approaches) significant bidirectional interactions between PSCs and cancer cells which facilitate local tumor growth and distant metastasis [13-18].

The microenvironment of pancreatic cancer A key histopathological feature of pancreatic cancer which is associated with its innate clinical and biological aggressiveness is its pronounced desmoplastic (stromal) reaction which is now considered an alternative therapeutic target in pancreatic cancer. Stroma production is stimulated by cancer-cell derived growth factors including transforming growth factor-β (TGFβ), hepatocyte growth factor (HGF), fibroblast growth factor (FGF), insulin-like growth factor 1 (IGF-1) and epidermal growth factor (EGF) [11].

The desmoplastic reaction is composed of extracellular matrix (ECM) proteins, primarily type I and III collagen, fibronectin and proteoglycans; small endothelium lined vessels; and a diverse population of cells including inflammatory cells, fibroblasts and stellate cells [19]. The stroma can form up to 90% of the tumor volume, a property which is unique to pancreatic cancer [12, 20]. Pancreatic cancer cells in vitro show a similar response to chemotherapeutic agents as cell lines derived from other solid tumors [20].

However, pancreatic cancer patients have a limited response to drugs such as paclitaxel compared to breast and prostate cancer patients, suggesting that the

Pancreatic stellate cells and fibrosis

31

unique tumor microenvironment in pancreatic cancer plays a role in chemoresistance [20].

Whilst stromal cells do not exhibit the genetic transformations seen in malignant pancreatic cancer cells, they are altered by cytokines and growth factors secreted by inflammatory cells and tumor cells [21]. Reciprocally, the stromal cells promote tumor cell migration, growth, invasion and resistance to drugs and apoptosis [21]. As mentioned above, PSCs have recently emerged as the principal regulator of the desmoplastic reaction in pancreatic cancer [13], as well as key effector cells in pancreatic fibrosis during necroinflammation in chronic pancreatitis [22-24] a known risk factor for pancreatic cancer [25]. Erkan et al [26] observed through staining pancreatic cancer tissue sections of patients for alpha smooth muscle actin (α-SMA the cytoskeletal protein marker for PSC activation) and collagen that a high activated stroma index (α-SMA/collagen) correlated with a poor prognosis.

Furthermore, the extensive ECM deposition by PSCs in pancreatic cancer causes distortion and compression of tumor vasculature by fibrous tissue which contributes to tumor hypoxia, a determinant of chemoresistance [27-30].

The biology of pancreatic stellate cells First isolated in 1998 [31-32], PSCs are resident cells of the pancreas predominantly located in the periacinar, periductal and perivascular spaces of the pancreas [31-33]. A recent study by Erkan et al [34], demonstrated that PSCs have a similar transcriptional fingerprint to hepatic stellate cells (HSCs, counterpart cells in the liver, key mediators of liver fibrosis) and therefore may have a similar origin. This study identified that collagen type 11α1 (COL11α1) may be a novel PSC specific marker with up to 65-fold higher expression levels in PSCs compared to HSCs [34]. Similar results were obtained by Buchholz et al [35] showing that PSCs only differed from HSCs with respect to at least 29 genes. Compared to HSC, PSCs exhibit increased periostin, serine protease 11, integrin- α7, connective tissue growth factor, cytoskeletal elements (actin γ2, desmoplakin) and hypoxia inducible factor 1α subunit. Despite these similarities, no consensus in the field has been reached as to the exact origin of PSCs, with both mesenchymal and neuroectodermal origin suggested.

In health, PSCs comprise ~4-7% of pancreatic cells and exist in their quiescent phenotype where they contain numerous cytoplasmic vitamin A containing lipid droplets, and express specific markers such as desmin and glial fibrillary acidic protein (GFAP) [13, 23]. In a normal pancreas, PSCs

32

Phoebe Phillips

may play a role in normal tissue architecture by regulating extracellular matrix turnover, given their known synthesis of ECM proteins and matrix degrading enzymes (matrix metalloproteinases [MMPs] and tissue inhibitors of metalloproteinases [TIMPs]) [36]. Recent novel evidence from our group also demonstrated that PSCs can synthesize and secrete the neurotransmitter acetylcholine, which may play a role in mediating enzyme secretion from acinar cells [37]. This feature also supports a potential neuroectodermal origin.

In response to pancreatic injury, PSCs transform into an active myofibroblast-like phenotype, under the influence of activating factors such as alcohol and oxidant stress [7, 22, 38] or from products of injured cells including pro-inflammatory cytokines and growth factors (Figure 1) [39-43].

This transdifferentiation is accompanied by a loss of cytoplasmic vitamin-A lipid droplets and increased cytoskeletal protein Îą-SMA expression [31].

Activated PSCs subsequently develop functional alterations including: 1) increased proliferation and migration [33, 39, 42, 44-45]; 2) synthesis of excessive ECM proteins (collagen, fibronectin, laminin) as well as matrix metalloproteinases and their inhibitors [13-14, 36, 39, 45]; and 3) secretion of growth factors and cytokines which exert both paracrine and autocrine effects that enhance cell growth and migration (Figure 1) [33, 46-50].

Transformation of PSCs from a quiescent to an activated phenotype has been the subject of intense study in recent years. Several signalling pathways/ molecules that mediate this process have been identified [51]. These include mitogen activated protein kinases (MAPK), phosphatidylinositol kinase (PI3K), protein kinase C (PKC), peroxisome proliferator activated receptor gamma (PPARÎł), the JAK-STAT pathway and the transcription factors nuclear factor-kappa B (NF-ÎşB) and activator protein-1 (AP-1).

Role of pancreatic stellate cells in pancreatic fibrosis Although this chapter focuses on the production of fibrous ECM by PSCs (the major cells mediating fibrosis in the pancreas), it is to be noted that other potential sources of fibrosis are present in the pancreas. The fibrotic matrix in pancreatic cancer was initially thought to be produced as a result of chronic injury as a host barrier against tumor invasion. However, evidence now indicates that the fibrotic reaction in pancreatic cancer is essential for tumor promotion and progression. The fibrotic ECM provides a physical scaffold and is a sink for soluble growth factors capable of influencing pancreatic cancer cell growth, survival and motility. Fibrogenesis is a dynamic process which is potentially reversible, at least in its early stages. A significant improvement in

Pancreatic stellate cells and fibrosis

33

Figure 1. Pancreatic stellate cell activation. In health, PSCs are in a quiescent state, where they contain numerous cytoplasmic vitamin A containing lipid droplets. In pancreatic cancer or chronic pancreatitis PSCs transform into an activated myofibroblast-like cell, which is accompanied by a loss of vitamin A lipid droplets.

PSCs are activated by growth factors and cytokines secreted by injured acinar cells and ductal cells, pancreatic cancer cells, endothelial cells and inflammatory cells.

Similarly, ethanol and its metabolite acetaldehyde as well as oxidant stress activate PSCs. PSC activation leads to production of fibrous extracellular matrix (ECM) proteins (collagen, laminin and fibronectin), increased migration to areas of injury and increased proliferation. Autocrine factors are also secreted by PSCs which perpetuate their activation. TGF = transforming growth factor; PDGF = platelet derived growth factor; VEGF = vascular endothelial growth factor; TNF = tumor necrosis factor; IL = interleukin; EMMPRIN = extracellular matrix metalloproteinase inducer; ET-1 = endothelin-1; CTGF = connective tissue growth factor; COX-2 = Cyclooxygenase.

in our understanding of the molecular mechanisms of pancreatic fibrosis were largely made possible by the identification, isolation and characterisation of PSCs [31-32].

A number of growth factors and proinflammatory cytokines including TGFβ1, platelet-derived growth factor (PDGF), tumor necrosis factor α (TNFα) and the interleukins 1 and 6 are known to be upregulated during

34

Phoebe Phillips

necroinflammation of the pancreas. Therefore, the concept that PSCs are activated by proinflammatory cytokines released during pancreatic injury has been tested by several investigators. It is well established that upon exposure to PDGF, cultured PSCs exhibit increased cell proliferation and migration [39, 41, 43, 45, 52-53]. The profibrogenic cytokine TGFβ has been shown to increase the expression of α-SMA and the ECM proteins collagen and fibronectin in PSCs [39, 41, 45, 50]. The proinflammatory cytokines TNFα, IL1 and IL6 have also been shown to activate PSCs [42]. PSCs are also activated by reactive oxygen species (generated within the pancreas by oxidative metabolism of alcohol and/or necroinflammation). Furthermore, PSCs can secrete TGFβ1 [50], PDGF [11], connective tissue growth factor (CTGF) and endothelin-1 [54]. These autocrine signals result in perpetuation of stellate cell activation and increased production of fibrous ECM. Dual staining (using Sirius red for collagen and immunostaining for αSMA) of sections from both human and rat pancreas has shown that there is co-localisation of αSMA positive staining with bands of fibrosis (collagen) suggesting the presence of activated PSCs in fibrotic areas [24]. Increased numbers of PSCs in areas of fibrosis have been reported using the TNBS model of pancreatic fibrosis. Haber et al [24] have demonstrated increased staining for the stellate cell marker desmin in areas of fibrosis indicating that PSC numbers are increased in areas of injury. This increase in PSC number may be due to local proliferation and or migration of PSCs to affected areas.

Studies using immunostaining for αSMA and in situ hybridization for collagen messenger RNA have also indicated that activated PSCs are the primary source of collagen in the fibrotic pancreas [39]. More recently, similar experimental approaches have indicated that activated PSCs are the principal source of collagen in the stromal reaction around pancreatic cancers [13, 55]. Thus evidence from in vivo studies (using human tissue and animal models) as well as in vitro studies (using cultured PSCs) supports the concept that PSCs are key players in the fibrogenesis process in the pancreas.

The role of cancer-stellate cell interactions in tumor progression As mentioned above, there is now a significant body of evidence showing that a bidirectional interaction exists between PSCs and pancreatic cancer cells which favours tumor progression (Figure 2) [13-18]. In summary, via production of mitogenic and fibrogenic mediators, pancreatic cancer cells attract and promote the activation, proliferation and motility of PSCs [13-14, 17-18, 56]. Pancreatic cancer cells also regulate the capacity of PSCs to

Pancreatic stellate cells and fibrosis

35

Figure 2. A schematic representation of the bi-directional interaction between pancreatic cancer cells and pancreatic stellate cells (PSCs). Pancreatic cancer cells secrete several factors which participate in the recruitment and activation of PSCs. When activated by cancer cells, PSCs have increased proliferation, ECM deposition (predominantly fibrous collagen) and migration. In turn, PSCs secrete factors which enhance the survival of pancreatic cancer cells (via increased proliferation and reduced apoptosis). PSCs secrete MMPs which increase the invasion of pancreatic cancer cells. While the fibrotic microenvironment produced by PSCs increases epithelial-mesenchymal transition (EMT) and chemoresistance of pancreatic cancer cells. This bi-directional interaction between PSCs and cancer cells favours tumor progression and metastasis. TGF = transforming growth factor; PDGF = plateletderived growth factor; VEGF = vascular endothelial growth factor; FGF = fibroblast growth factor; TNF = tumor necrosis factor; EMMPRIN = extracellular matrix metalloproteinase inducer; SDF = stromal derived factor; IGF = insulin growth factor; ECM = extracellular matrix; COX = cyclooxygenase; EGF = epidermal growth factor.

remodel the ECM [14]. Bachem et al [14] demonstrated that supernatant derived from human pancreatic cancer cell lines increased PSC proliferation and matrix synthesis, in a dose dependent manner. These two effects were suppressed when neutralising antibodies against the growth factors PDGF, FGF2, and TGFβ1 were introduced, confirming the influence of mediators released by pancreatic cancer cells on PSC activity [14]. Cancer cells also have the capacity to secrete the ECM metalloproteinase inducer (EMMPRIN), which results in increased MMP2 secretion by PSCs [57].

36

Phoebe Phillips

TGFβ1 has also been shown to stimulate MMP2 secretion by PSCs [36].

MMP2 has been associated with the invasive phenotype of pancreatic cancer cells and is essential for degradation of normal basement membrane for cancer progression.

Bachem et al [14] were the first to provide evidence of an interaction in vivo and reported that subcutaneous coinjection of PSCs and cancer cells into the flanks of nude mice resulted in larger tumors with a significant stromal reaction than those produced by injection of cancer cells alone. In another study by the same group, co-injection of PSCs with cancer cells overexpressing the serine protease inhibitor SERPINE2 into nude mice was reported to result in increased tumor growth [58]. These findings suggested that increased ECM (in particular, fibrillar collagen) deposition as a consequence of protease inhibition might facilitate cancer progression, a concept supported by the observation that ECM proteins protect pancreatic cancer cells from apoptosis. Although these studies provided important evidence of stromal-tumor interactions, subcutaneous models are not an ideal choice for studying pancreatic cancer because they do not allow the assessment of tumor behaviour within the appropriate microenvironment.

More recently, our laboratory developed an orthotopic model of pancreatic cancer which overcomes some of the limitations of subcutaneous models.

In vivo, mice receiving a co-injection of pancreatic cancer cells and human PSCs (hPSCs) into the pancreas, histologically exhibited dense bands of fibrosis which correlated with the presence of activated PSCs as assessed by Îą-SMA expression [17].

Simultaneously, pancreatic cancer co-opts PSCs to form a growth permissive and tumor facilitatory environment. By producing and sequestering multiple growth factors, cytokines and ECM components, PSCs actively participate in pancreatic cancer proliferation, migration, invasion, metastatic dissemination and survival [16-18]. Watanabe et al [59] reported that more extensive intratumoral fibroblastic cell proliferation correlates with a poorer disease outcome in pancreatic cancer patients. Using an in vivo orthotopic mouse model of pancreatic cancer, our laboratory demonstrated that intra-pancreatic co-injection of human PSCs with pancreatic cancer cells yielded larger tumors and more extensive local and distant metastases compared with injections of tumor cells alone [17]. Furthermore, PSCs co-migrated with pancreatic cancer cells to metastatic sites where they are postulated to play a role in seeding and supporting cancer growth [18].

Similarly Hwang et al [16] showed that the incidence of tumor formation, distant metastases and mean pancreatic weight were significantly augmented with increasing proportion of human PSC to pancreatic cancer cells. In line with these observations in vivo, human PSC conditioned medium stimulated

Pancreatic stellate cells and fibrosis

37

pancreatic tumor cell proliferation and colony formation in vitro [16]. Direct and indirect co-culture systems have further demonstrated increased migration, invasiveness and proliferation of pancreatic cancer cells in the presence of human PSCs [17, 60]. These findings support the notion that PSCs play a pivotal role in tumor progression.

The hypoxia fibrosis cycle An implication of the extensive desmoplastic reaction in pancreatic cancer is intratumoral hypoxia, a major determinant of pancreatic cancer chemoresistance [15, 29, 61]. Hypoxia arises when the proliferating tumor outgrows its ambient vascular supply. Although PSCs harbor proangiogenic abilities via the secretion of vascular endothelial growth factor (VEGF), FGF, periostin and collagen type I, it is predominantly fibrogenic, and the extensive periacinar deposition of ECM proteins has been shown to distort the normal parenchymal architecture and compress the fine capillary network therefore limiting oxygen diffusion [15, 30]. Concomitantly, PSCs amplify the production of endostatin by pancreatic cancer cells (an endogenous inhibitor of angiogenesis), which overwhelms the angiogenic capacity of both PSCs and pancreatic cancer and attenuates microvessel density [15]. Essentially, the vasculature in pancreatic cancer becomes highly disorganised, dysfunctional and permeable [62-63]. Moreover, as hypoxia has been shown to perpetuate PSC proliferation and activation in vitro and in vivo, intra-stromal hypoxia sustains the periacinar deposition of ECM [15, 64]. Consequently, at the invasive front of the reactive stroma, cytokines and growth factors secreted by pancreatic cancer cells, and fibrosis which leads to tumor hypoxia, perpetuates PSC activity and propagates fibrotic changes beyond the actual tumor itself (i.e. inducing a hypoxia-fibrosis cycle, Figure 3) [19].

In pancreatic cancer, the abnormal blood flow dynamics of the tumor vasculature and the dense stroma imposes a considerable diffusion barrier to systemic drug delivery (Figure 3). Using a transgenic mouse model of pancreatic cancer which is refractory to gemcitabine, Olive et al [30] demonstrated that pancreatic tumors had a dysfunctional vasculature and that delivery of autofluorescent drugs such as doxorubicin into the pancreatic cancer tissue was impaired compared with control tissues. Depletion of the pancreatic cancer stroma by inhibiting the paracrine hedgehog signalling by pancreatic cancer cells to PSCs, however, potentiated intratumoral vascular density, augmented the delivery of gemcitabine and transiently enhanced pancreatic cancer chemosensitivity [30, 65]. Recent clinical evidence by Komar et al [66] in human patients quantified blood flow and metabolic activity of pancreatic tumors using oxygen-15 labelled water [15O]-H2O and

38

Phoebe Phillips

Figure 3. Pancreatic stellate cell mediated hypoxia fibrosis cycle in the chemoresistance of pancreatic cancer. PSC activation by pancreatic cancer cells leads to increased fibrosis production, which results in decreased microvessel density and abnormal vasculature. This altered vasculature leads to intrastromal and intratumoral hypoxia. In response to hypoxia and cancer cell secretions, PSCs proliferate and sustain ECM deposition, inducing a hypoxia-fibrosis cycle. Hypoxia also increases PSC activation and epithelial-mesenchymal transition (EMT), both of which contribute to overall chemoresistance. In addition, the dense fibrosis and dysfunctional vasculature inhibit the delivery of chemotherapy agents into the tumor.

[18F]-fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT imaging. This study demonstrated reduced blood flow and high metabolic activity in pancreatic tumors compared to normal pancreatic tissue. In addition, a high ratio of glucose uptake to blood flow was a predictor of poor prognosis [66]. The authors suggested that a high metabolic activity combined with low blood flow is indicative of a hypoxic tumor microenvironment [66]. These findings strongly suggest that in pancreatic cancer, the fibrotic and highly avascular microenvironment synergistically: 1) reduce drug delivery via the poorly perfused blood into the tumor; 2) causes the sequestering of drugs in the peritumoral stroma; and 3) leads to a decline in the effective intracellular drug concentration within pancreatic cancer cells, compromising therapeutic success.

Concomitantly in this hypoxic environmental niche, there is a reduction in the delivery of nutrients and the clearance of by products from anaerobic metabolism thus inhibiting pancreatic cancer cell proliferation and limiting

Pancreatic stellate cells and fibrosis

39

the effectiveness of cytotoxic drugs which are incorporated during the S-phase of the cell cycle [20]. Furthermore, hypoxic stress stabilise the hypoxia regulated transcription factor, hypoxia inducible factor- 1Îą (HIF-1Îą) and upregulate genes associated with angiogenesis, glycolysis, enhanced survival and epithelial-mesenchymal transition, all of which have been shown to contribute to increased chemoresistance [67-70].

In addition to hypoxic mechanisms of chemoresistance, PSC-pancreatic cancer cell interactions impair responses to chemotherapy. Our laboratory has demonstrated that human PSC secretions are capable of conferring a chemoresistant phenotype by suppressing H2O2-induced apoptosis and increasing survival of pancreatic cancer cells [17]. Muerkoster et al [71] demonstrated that pancreatic cancer cells became less sensitive toward treatment with etoposide when co-cultured with PSCs. These results were supported by Hwang et al [16] who demonstrated that gemcitabine and radiation therapy were less effective in pancreatic cancer cells treated with conditioned medium of human PSCs. Interestingly the presence of ECM proteins produced by PSCs promoted resistance of pancreatic cancer cells to 5-fluorouracil (5-FU), cisplatin and doxorubicin [72]. These results suggest that tumor-stromal interactions may promote the survival of tumor cells in the presence of chemotherapy. However little is known about the proteins mediating PSC-induced chemoresistance of pancreatic cancer cells.

Role of pancreatic stellate cells and fibrosis in epithelialmesenchymal transition Epithelial-mesenchymal transition (EMT) is associated with metastatic spread and resistance to apoptosis. A characteristic feature of EMT is the switch of the epithelial-specific junction protein E-cadherin to mesenchymal N-cadherin. A recent study by Kikuta et al [73] showed that PSCs promoted EMT (decreased E-cadherin and increased vimentin and Snai-1) in pancreatic cancer cells using a co-culture assay. Although the exact mechanisms and in vivo evidence is lacking in this area, it is likely that the altered vascularity and increased fibrosis resulting from PSC activation leads to hypoxia, which is known to increase EMT in cancer cells (Figure 3).

Targeting pancreatic stellate cell mediated fibrosis: A potential therapy PSCs produce the stromal reaction in pancreatic cancer, provide a growth-permissive microenvironment for cancer cells and facilitate distant

40

Phoebe Phillips

spread, as outlined above. PSCs are activated (characterised by loss of vitamin A droplets in their cytoplasm and transformation into myofibroblasts) and found in increased numbers in the stromal/fibrotic areas of pancreatic cancer. This increase in PSC numbers may be due to increased activation, local proliferation and/or decreased apoptosis. Potential ways to deplete the stromal reaction are i) inhibit PSC activation or proliferation; ii) induce PSC apoptosis; and/or iii) inhibit fibrous tissue production by PSCs. Below evidence is provided on potential therapeutic targets for depletion of PSC mediated fibrosis, which may as a consequence, reduce cancer progression and increase drug delivery into the tumor.

Sonic hedgehog (SHH) has been shown to promote desmoplasia [65] and is over expressed in cancer cells of human pancreatic tumors [74]. Tumor cells secrete SHH, which in turn stimulates stromal cells. Bailey et al [65] demonstrated that injection of pancreatic cancer cells over expressing SHH into the pancreas of mice resulted in significantly increased desmoplasia.

This was attributable to increased SMA positive fibroblasts (PSCs) and the effect was blocked using a SHH antibody [65]. A recent ground breaking article, showed that pharmacological targeting of the desmoplastic stroma by inhibition of sonic hedgehog signalling, facilitated increased drug delivery into the pancreatic tumors of transgenic mice and increased survival [30]. It is highly likely that this therapeutic approach targeted PSCs. Evidence for this comes from the fact that Bailey et al [75] demonstrated that SMA-positive myofibroblasts (or PSCs) expressed Gli1 (hedgehog transcription factor) in tumours over expressing SHH, indicating that these cells are responsive to stimulation with SHH. Further support was recently provided by Walter et al [76] who demonstrated that the hedgehog receptor Smoothened was upregulated in human pancreatic cancer-associated fibroblasts compared to control fibroblasts and that the cancer associated fibroblasts responded to incubation with recombinant SHH by increased Gli1 expression.

Periostin is a matrix-specific protein which is secreted by PSCs and perpetuates their fibrogenic activity [28]. Periostin is induced by hypoxia and is a regulator of angiogenesis. This study showed that periostin increased collagen-I, fibronectin and TGFβ secreted by PSCs. Periostin expression in PSCs was induced by co-culture with pancreatic cancer cells. In cancer cells, periostin stimulated growth and conferred resistance to hypoxia. Therefore, periostin is a potential target to decrease fibrosis which contributes to chemoresistance in pancreatic cancer.

Halofuginone is a plant derived alkaloid and is an anti-fibrotic agent. In a thioacetamide-induced liver fibrosis rat model, administration of halofuginone orally before fibrosis induction, prevented the activation of most hepatic stellate cells and remaining cells expressed low levels of

Pancreatic stellate cells and fibrosis

41

collagen-1 [77]. More importantly, when halofuginone was given to rats with established fibrosis, complete resolution of fibrosis was observed [77]. In a mouse model of cerulein-induced pancreatitis, halofuginone prevented collagen synthesis via inhibiting Smad 3 phosphorylation [78]. In culture, halofuginone inhibited PSC proliferation [78]. Spector et al [79] recently demonstrated that inhibition of PSC activation by halofuginone before implantation with pancreatic cancer cells resulted in reductions in tumor growth and decreased collagen production (subcutaneous and orthotopic).

However, the authors did not examine if depletion of the fibrosis by halofuginone increases the delivery of chemotherapeutic agents.

Endothelin-1 (ET-1) and its receptors are expressed on both PSCs and pancreatic cancer cells [80]. Treatment of PSCs in vitro with Bosentan (ET-1 receptor antagonist) significantly decreased collagen synthesis in a dose dependent manner and inhibited PDGF-induced proliferation [80]. In addition, the same study showed that bosentan inhibited cancer cell-induced PSC proliferation (co-culture experiments) as well as PSC-induced cancer cell proliferation [80]. These results warrant further validation using preclinical models of pancreatic cancer, especially given the fact that ET-1 receptor antagonist interfered with the bidirectional interaction of PSCs and cancer cells.

Several studies have examined the effect of anti-oxidants on PSC activation. Vitamin E prevents PSC activation in vitro [22]. While in vivo studies involving experimental models of pancreatic fibrosis have reported that anti-oxidants, herbal products with antioxidant properties can significantly reduce pancreatic fibrosis. Îł-Tocotrienol, a novel, unsaturated form of vitamin E inhibited in vitro proliferation of pancreatic cancer cells and potentiated gemcitabine-induced apoptosis [81]. These effects were mediated by suppression of cyclin D1, c-myc, cyclooxygenase-2 (COX-2), Bcl-2, VEGF and chemokine receptor type 4 (CXCR4). Îł-Tocotrienol also inhibited tumor growth and enhanced gemcitabine activity in an orthotopic model of pancreatic cancer. In addition, tocotrienols induced activated PSC apoptosis, without affecting quiescent PSCs [82]. Therefore, tocotrienols may be an interesting therapeutic option because they may deplete both the stromal PSCs and pancreatic cancer cells. The use of tocotrienols needs to be further tested in a relevant pre-clinical animal model which has an extensive stromal reaction.

Our laboratory has also demonstrated that exogenous vitamin A and its metabolites induced PSC quiescence in vitro [83]. Vitamin A and its metabolites are also inhibitors of pancreatic cancer cell growth [84]. However, the vitamin A derivative 13-cis-retinoic acid has been tested in several phase II clinical trial in pancreatic cancer with varying results [85-87].

42

Phoebe Phillips

Cyclooxygenase-2 (COX-2) is an enzyme important in inflammation and the increased production of prostaglandins associated with tumorigenesis [88]. COX-2 is expressed in chronic pancreatitis and pancreatic cancer patients. Celecoxib (a selective COX-2 inhibitor) inhibited COX-2 activity, decreased TGFβ expression, induced metalloproteinase-2 activity and, consequently, prevented and reversed collagen accumulation in a model of liver fibrosis induced by carbon-tetrachloride administration [89]. Activated PSCs express COX-2 when stimulated with TGFβ [90] and co-cultured with pancreatic cancer cells [91]. Pharmacological inhibition of COX-2 in activated PSCs decreased the expression of COX-2, αSMA and collagen I, suggesting that COX-2 might be a relevant target in pancreatic cancer [90]. In addition, COX-2 inhibition also caused increased apoptosis in pancreatic cancer cells [92]. Further support that COX-2 is important in the PSC-cancer cell interaction was provided by Sato et al [93] who demonstrated that COX-2 expression is increased in both PSCs and cancer cells in response to co-culture. This study also showed that blockage of COX-2 (NS-398) partially inhibited PSC-induced invasion of pancreatic cancer cells [93].

Celecoxib has recently been tested in combination with gemcitabine and irinotecan in a phase II clinical trial in patients with advanced pancreatic cancer [94].

Galactin-3 is a member of the β-galactoside-binding protein family, which has been implicated in inflammation and cancer. Silencing the galectin-3 gene in vivo using siRNA inhibited myofibroblast activation and attenuated liver fibrosis in a carbon-tetrachloride liver injury model in mice [95]. A recent study by Jiang et al [96] demonstrated that pancreatic cancer cells express and secrete galectin-3. This study also demonstrated that pancreatic cancer cells stimulated PSC proliferation via galectin-3 (using an antibody for Galectin-3) [96]. The exact mechanism of action of galectin-3 is unknown and inhibitors of galectin 3 have not yet been tested in pre-clinical models of pancreatic cancer.

PDGF is a potent mitogen and chemoattractant for PSCs and therefore may be a potential therapeutic target for fibrosis in pancreatic cancer. PDGF and its receptor PDGFRβ is also upregulated in tumor cells. In fact, the PDGF-antagonist trapidil suppressed PDGF-induced ERK activation and decreased PSC proliferation in vitro [97]. While Masamune et al [98] showed that curcumin (polyphenol compound found in turmeric) resulted in decreased PDGF-induced proliferation of PSCs in vitro and also decreased αSMA expression and collagen-I production. In another study by the same group PDGF-induced PSC proliferation, migration and collagen production were inhibited by epigallocatechin-3-gallate (a polyphenol from green tea)

Pancreatic stellate cells and fibrosis

43

[99]. Rats with chronic liver injury (bile duct ligation model) were administered hepatic stellate cell (HSC)-specific PDGFR-β shRNA (linked to a GFAP promoter to reduce non-HSC mediated effects) which resulted in decreased liver injury and decreased hepatic fibrosis [100]. This example using the GFAP promoter to target stellate cells is a new powerful tool for cell-specific gene therapy. Furthermore, several PDGF receptor–related antagonists are being developed as potential anti-tumor and anti-stromal agents and have demonstrated promising antitumor activity in both preclinical and clinical settings including imatinib mesylate (Gleevec/ ST571), sunitinib malate (Sutent/SU11248), and CP-673,451.

Targeting the pro-fibrogenic TGFβ signalling pathway have showed promise in reducing experimental pancreatic fibrosis in rodents [101-102].

There are several drugs available which target TGFβ synthesis, ligand/receptor binding or receptor kinase signalling. Preclinical studies using TGFβ inhibitors have demonstrated efficacy in reducing metastasis and have shown improvements in cytotoxic drug delivery. Activin A, a member of the TGFβ family activates PSCs and promotes collagen secretion in an autocrine manner [49]. The authors also showed that activin A stimulates TGF-β secretion by PSCs and vice versa TGFβ stimulates activin A secretion by PSCs. Furthermore, follistatin (an endogenous activin A binding protein known to block the effect of activin A) [103] inhibited TGF-β secretion by PSCs and attenuated PSC activation and collagen secretion [49]. These data imply that activin A participates in pancreatic fibrosis and follistatin may be used as an anti-fibrotic agent [49]. Follistatin has not been tested in preclinical models of pancreatitis or pancreatic cancer.

A recent study showed that quiescent PSCs endogenously express albumin (localised in the cytoplasm with lipid droplets) [104]. Upon activation in culture and loss of vitamin A, albumin levels decreased. TGFβ treatment of PSCs also decreased albumin and stably transfected PSCs with albumin are resistant to TGFβ mediated activation [104]. On this point, albumin-bound paclitaxel (nab-paclitaxel;Abraxane) has been shown to bind to albumin receptors in tumor blood vessels and is released into the tumor microenvironment [105]. The albumin nanoparticles were used to deliver paclitaxel to avoid the previously toxic castor oil derivative Cremaphor. This drug is reported to result in "stromal collapse” [106] and increased drug delivery into tumors. Although scientists and clinicians are eagerly awaiting clinical trials with abraxane in pancreatic cancer, it may be interesting to examine the effect of albumin-bound paclitaxel on PSC activity in the stroma, especially given the results by Kim et al [104] described above.

Albumin itself may be a therapeutic option for reversion of PSCs back to

44

Phoebe Phillips

their quiescent phenotype and reduce production of pancreatic fibrosis. In addition, novel nanotechnologies may provide us with a way to target a specific cell type in the pancreatic tumor microenvironment (e.g. PSCs) and enhance the effectiveness of other anti-cancer therapies.

Summary Points • • • •

•

•

An extensive stromal/fibrous reaction and hypoxic microenvironment are characteristic features of pancreatic cancer.

Pancreatic stellate cells are key mediators of the fibrosis observed in the desmoplastic reaction of pancreatic cancer.

A bi-directional interaction exists between pancreatic stellate cells and pancreatic cancer cells which facilitates tumor progression and metastasis.

Therapeutic intervention in the last decade has had limited impact on improving patient survival in pancreatic cancer. Therefore, it is clear that targeting cancer cells alone is not enough. An alternative approach is to dual target the stromal PSCs and the tumor cells.

PSCs contribute significantly to the highly chemoresistant nature of pancreatic cancer by producing the extensive fibrous ECM which results in i) altered vasculature and decreased drug delivery to the tumor; ii) decreased sensitivity to chemotherapeutic agents; and iii) increased epithelial-mesenchymal transition.

Approaches to improve the efficacy of anti-cancer drugs include: i) restoration of vasculature to normal by decreasing the fibrosis; ii) the use of appropriate pre-clinical animal models which mimic human disease (i.e. poorly perfused and hypoxic microenvironment with extensive fibrotic reaction); and iii) targeting the cancer-stellate cell interactions.

Conclusion Recent developments in our understanding of the prominent stromal reaction in pancreatic cancer has highlighted the importance of the key fibrotic mediators, pancreatic stellate cells. Pancreatic cancer cells recruit PSCs to enhance their growth, modulate the ECM (to increase invasion and promote a chemoresistance phenotype) and aid in metastatic spread. Despite numerous clinical trials (conventional and targeted therapy) in pancreatic cancer over the past decade, only modest improvements in patient survival have been observed. Reasons for this include a lack of understanding chemoresistance mechanisms and inadequate pre-clinical animal models for testing potential therapies. In order to reliably predict outcomes in subsequent human

Pancreatic stellate cells and fibrosis

45

trials future research should consider the influence of the tumor microenvironment including the extensive fibrotic reaction. This chapter highlights the potential to therapeutically target PSCs which has the potential to decrease the production of fibrous tissue, slow down tumor progression and increase the delivery of chemotherapeutic agents to the tumor.

Acknowledgements Phoebe Phillips is currently supported by a Cancer Institute NSW Fellowship (2009-2012).

References 1.

2.

Hidalgo M. 2010, Pancreatic cancer. N Engl J Med. 362, 1605-1617.

Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, and Thun MJ. 2008, Cancer statistics, 2008. CA Cancer J Clin. 58, 71-96.

3. Andersson R, Aho U, Nilsson BI, Peters GJ, Pastor-Anglada M, Rasch W, and Sandvold ML. 2009, Gemcitabine chemoresistance in pancreatic cancer: molecular mechanisms and potential solutions. Scand J Gastroenterol. 44, 782-786.

4. Wang Z, Li Y, Ahmad A, Banerjee S, Azmi AS, Kong D, and Sarkar FH. 2011, Pancreatic cancer: understanding and overcoming chemoresistance. Nat Rev Gastroenterol Hepatol. 8, 27-33.

5. Zalatnai A and Molnar J. 2007, Review. Molecular background of chemoresistance in pancreatic cancer. In Vivo. 21, 339-347.

6. Trouilloud I, Dubreuil O, Boussaha T, Lepere C, Landi B, Zaanan A, Bachet JB, and Taieb J. 2011, Medical treatment of pancreatic cancer: New hopes after 10 years of gemcitabine. Clin Res Hepatol Gastroenterol. 35, 367-374.

7. Apte M, Pirola R, and Wilson J. 2009, New insights into alcoholic pancreatitis and pancreatic cancer. J Gastroenterol Hepatol. 24 Suppl 3, S51-56.

8. Duner S, Lopatko Lindman J, Ansari D, Gundewar C, and Andersson R. 2010, Pancreatic cancer: the role of pancreatic stellate cells in tumor progression.

Pancreatology. 10, 673-681.

9. Farrow B, Albo D, and Berger DH. 2008, The role of the tumor microenvironment in the progression of pancreatic cancer. J Surg Res. 149, 319-328.

10. Habisch H, Zhou S, Siech M, and Bachem MG. 2010, Interaction of Stellate Cells with Pancreatic Carcinoma Cells. Cancers. 2, 1661-1682.

11. Mahadevan D and Von Hoff DD. 2007, Tumor-stroma interactions in pancreatic ductal adenocarcinoma. Mol Cancer Ther. 6, 1186-1197.

12. Neesse A, Michl P, Frese KK, Feig C, Cook N, Jacobetz MA, Lolkema MP, Buchholz M, Olive KP, Gress TM, and Tuveson DA. 2010, Stromal biology and therapy in pancreatic cancer. Gut. 60, 861-880.

46

Phoebe Phillips

13. Apte MV, Park S, Phillips PA, Santucci N, Goldstein D, Kumar RK, Ramm GA, Buchler M, Friess H, McCarroll JA, Keogh G, Merrett N, Pirola R, and Wilson JS. 2004, Desmoplastic reaction in pancreatic cancer: role of pancreatic stellate cells. Pancreas. 29, 179-187.

14. Bachem MG, Sch端nemann M, Ramadani M, Siech M, Beger H, Buck A, Zhou S, Schmid-Kotsas A, and Adler G. 2005, Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells.

Gastroenterology. 128, 907-921.

15. Erkan M, Reiser-Erkan C, Michalski CW, Deucker S, Sauliunaite D, Streit S, Esposito I, Friess H, and Kleeff J. 2009, Cancer-stellate cell interactions perpetuate the hypoxia-fibrosis cycle in pancreatic ductal adenocarcinoma.

Neoplasia. 11, 497-508.

16. Hwang RF, Moore T, Arumugam T, Ramachandran V, Amos KD, Rivera A, Ji B, Evans DB, and Logsdon CD. 2008, Cancer-associated stromal fibroblasts promote pancreatic tumor progression. Cancer Res. 68, 918-926.

17. Vonlaufen A, Joshi S, Qu C, Phillips PA, Xu Z, Parker NR, Toi CS, Pirola RC, Wilson JS, Goldstein D, and Apte MV. 2008, Pancreatic stellate cells: partners in crime with pancreatic cancer cells. Cancer Res. 68, 2085-2093.

18. Xu Z, Vonlaufen A, Phillips PA, Fiala-Beer E, Zhang X, Yang L, Biankin AV, Goldstein D, Pirola RC, Wilson JS, and Apte MV. 2010, Role of pancreatic stellate cells in pancreatic cancer metastasis. Am J Pathol. 177, 2585-2596.

19. Erkan M, Reiser-Erkan C, Michalski CW, and Kleeff J. 2010, Tumor microenvironment and progression of pancreatic cancer. Exp Oncol. 32, 128-131.

20. Li J, Wientjes MG, and Au JL. 2010, Pancreatic cancer: pathobiology, treatment options, and drug delivery. AAPS J. 12, 223-232.

21. Mueller MM and Fusenig NE. 2004, Friends or foes - bipolar effects of the tumour stroma in cancer. Nat Rev Cancer. 4, 839-849.

22. Apte MV, Phillips PA, Fahmy RG, Darby SJ, Rodgers SC, McCaughan GW, Korsten MA, Pirola RC, Naidoo D, and Wilson JS. 2000, Does alcohol directly stimulate pancreatic fibrogenesis? Studies with rat pancreatic stellate cells.

Gastroenterology. 118, 780-794.

23. Apte MV, Pirola RC, and Wilson JS. 2006, Battle-scarred pancreas: role of alcohol and pancreatic stellate cells in pancreatic fibrosis. J Gastroenterol Hepatol. 21 Suppl 3, S97-S101.

24. Haber PS, Keogh GW, Apte MV, Moran CS, Stewart NL, Crawford DH, Pirola RC, McCaughan GW, Ramm GA, and Wilson JS. 1999, Activation of pancreatic stellate cells in human and experimental pancreatic fibrosis. Am J Pathol. 155, 1087-1095.

25. Motoo Y, Shimasaki T, Ishigaki Y, Nakajima H, Kawakami K, and Minamoto T. 2011, Metabolic Disorder, Inflammation, and Deregulated Molecular Pathways Converging in Pancreatic Cancer Development: Implications for New Therapeutic Strategies. Cancers. 3, 446-460.

26. Erkan M, Michalski CW, Rieder S, Reiser-Erkan C, Abiatari I, Kolb A, Giese NA, Esposito I, Friess H, and Kleeff J. 2008, The activated stroma index is a

Pancreatic stellate cells and fibrosis

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

47

novel and independent prognostic marker in pancreatic ductal adenocarcinoma.

Clin Gastroenterol Hepatol. 6, 1155-1161.

Couvelard A, O'Toole D, Leek R, Turley H, Sauvanet A, Degott C, Ruszniewski P, Belghiti J, Harris AL, Gatter K, and Pezzella F. 2005, Expression of hypoxiainducible factors is correlated with the presence of a fibrotic focus and angiogenesis in pancreatic ductal adenocarcinomas. Histopathology. 46, 668-676.

Erkan M, Kleeff J, Gorbachevski A, Reiser C, Mitkus T, Esposito I, Giese T, Buchler MW, Giese NA, and Friess H. 2007, Periostin creates a tumor-supportive microenvironment in the pancreas by sustaining fibrogenic stellate cell activity.

Gastroenterology. 132, 1447-1464.

Koong AC, Mehta VK, Le QT, Fisher GA, Terris DJ, Brown JM, Bastidas AJ, and Vierra M. 2000, Pancreatic tumors show high levels of hypoxia. Int J Radiat Oncol Biol Phys. 48, 919-922.

Olive KP, Jacobetz MA, Davidson CJ, Gopinathan A, McIntyre D, Honess D, Madhu B, Goldgraben MA, Caldwell ME, Allard D, Frese KK, Denicola G, Feig C, Combs C, Winter SP, Ireland-Zecchini H, Reichelt S, Howat WJ, Chang A, Dhara M, Wang L, Ruckert F, Grutzmann R, Pilarsky C, Izeradjene K, Hingorani SR, Huang P, Davies SE, Plunkett W, Egorin M, Hruban RH, Whitebread N, McGovern K, Adams J, Iacobuzio-Donahue C, Griffiths J, and Tuveson DA. 2009, Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science. 324, 1457-1461.

Apte MV, Haber PS, Applegate TL, Norton ID, McCaughan GW, Korsten MA, Pirola RC, and Wilson JS. 1998, Periacinar Stellate Shaped Cells in Rat Pancreas - Identification, Isolation, and Culture. Gut. 43, 128-133.

Bachem MG, Schneider E, Gross H, Weidenbach H, Schmid RM, Menke A, Siech M, Beger H, Grunert A, and Adler G. 1998, Identification, Culture, and Characterization of Pancreatic Stellate Cells in Rats and Humans.

Gastroenterology. 115, 421-432.

Omary MB, Lugea A, Lowe AW, and Pandol SJ. 2007, The pancreatic stellate cell: a star on the rise in pancreatic diseases. J Clin Invest. 117, 50-59.

Erkan M, Weis N, Pan Z, Schwager C, Samkharadze T, Jiang X, Wirkner U, Giese NA, Ansorge W, Debus J, Huber PE, Friess H, Abdollahi A, and Kleeff J. 2010, Organ-, inflammation- and cancer specific transcriptional fingerprints of pancreatic and hepatic stellate cells. Mol Cancer. 9, 88.

Buchholz M, Kestler HA, Holzmann K, Ellenrieder V, Schneiderhan W, Siech M, Adler G, Bachem MG, and Gress TM. 2005, Transcriptome analysis of human hepatic and pancreatic stellate cells: organ-specific variations of a common transcriptional phenotype. J Mol Med. 83, 795-805.

Phillips PA, McCarroll JA, Park S, Wu MJ, Pirola R, Korsten M, Wilson JS, and Apte MV. 2003, Rat pancreatic stellate cells secrete matrix metalloproteinases: implications for extracellular matrix turnover. Gut. 52, 275-282.

Phillips PA, Yang L, Shulkes A, Vonlaufen A, Poljak A, Bustamante S, Warren A, Xu Z, Guilhaus M, Pirola R, Apte MV, and Wilson JS. 2010, Pancreatic stellate cells produce acetylcholine and may play a role in pancreatic exocrine secretion. Proc Natl Acad Sci U S A. 107, 17397-17402.

48

Phoebe Phillips

38. Masamune A, Satoh A, Watanabe T, Kikuta K, Satoh M, Suzuki N, Satoh K, and Shimosegawa T. 2010, Effects of ethanol and its metabolites on human pancreatic stellate cells. Dig Dis Sci. 55, 204-211.

39. Apte MV, Haber PS, Darby SJ, Rodgers SC, McCaughan GW, Korsten MA, Pirola RC, and Wilson JS. 1999, Pancreatic stellate cells are activated by proinflammatory cytokines : implications for pancreatic fibrogenesis. Gut. 44, 534-541.

40. Gao R and Brigstock DR. 2005, Connective tissue growth factor (CCN2) in rat pancreatic stellate cell function: integrin alpha5beta1 as a novel CCN2 receptor.

Gastroenterology. 129, 1019-1030.

41. Kordes C, Brookmann S, Haussinger D, and Klonowski-Stumpe H. 2005, Differential and synergistic effects of platelet-derived growth factor-BB and transforming growth factor-beta1 on activated pancreatic stellate cells. Pancreas.

31, 156-167.

42. Mews P, Phillips P, Fahmy R, Korsten M, Pirola R, Wilson J, and Apte M. 2002, Pancreatic stellate cells respond to inflammatory cytokines: potential role in chronic pancreatitis. Gut. 50, 535-541.

43. Vonlaufen A, Phillips PA, Yang L, Xu Z, Fiala-Beer E, Zhang X, Pirola RC, Wilson JS, and Apte MV. 2010, Isolation of quiescent human pancreatic stellate cells: a promising in vitro tool for studies of human pancreatic stellate cell biology. Pancreatology. 10, 434-443.

44. Phillips PA, Wu MJ, Kumar RK, Doherty E, McCarroll JA, Park S, Pirola RC, Wilson JS, and Apte MV. 2003, Cell migration: a novel aspect of pancreatic stellate cell biology. Gut. 52, 677-682.

45. Schneider E, Schmid-Kotsas A, Zhao J, Weidenbach H, Schmid RM, Menke A, Adler G, Waltenberger J, Grunert A, and Bachem MG. 2001, Identification of mediators stimulating proliferation and matrix synthesis of rat pancreatic stellate cells. American Journal of Physiology: Cell Physiology. 281, C532-543.

46. Aoki H, Ohnishi H, Hama K, Ishijima T, Satoh Y, Hanatsuka K, Ohashi A, Wada S, Miyata T, Kita H, Yamamoto H, Osawa H, Sato K, Tamada K, Yasuda H, Mashima H, and Sugano K. 2006, Autocrine loop between TGF-beta1 and IL1beta through Smad3- and ERK-dependent pathways in rat pancreatic stellate cells. Am J Physiol Cell Physiol. 290, C1100-1108.

47. Aoki H, Ohnishi H, Hama K, Shinozaki S, Kita H, Yamamoto H, Osawa H, Sato K, Tamada K, and Sugano K. 2006, Existence of autocrine loop between interleukin-6 and transforming growth factor-beta1 in activated rat pancreatic stellate cells. J Cell Biochem. 99, 221-228.

48. Jiang X, Abiatari I, Kong B, Erkan M, De Oliveira T, Giese NA, Michalski CW, Friess H, and Kleeff J. 2009, Pancreatic islet and stellate cells are the main sources of endocrine gland-derived vascular endothelial growth factor/ prokineticin-1 in pancreatic cancer. Pancreatology. 9, 165-172.

49. Ohnishi N, Miyata T, Ohnishi H, Yasuda H, Tamada K, Ueda N, Mashima H, and Sugano K. 2003, Activin A is an autocrine activator of rat pancreatic stellate cells: potential therapeutic role of follistatin for pancreatic fibrosis. Gut. 52, 1487-1493.

Pancreatic stellate cells and fibrosis

49

50. Shek FW, Benyon RC, Walker FM, McCrudden PR, Pender SL, Williams EJ, Johnson PA, Johnson CD, Bateman AC, Fine DR, and Iredale JP. 2002, Expression of Transforming Growth Factor-b1 by Pancreatic Stellate Cells and Its Implications for Matrix Secretion and Turnover in Chronic Pancreatitis.

American Journal of Pathology. 160, 1787-1798.

51. Masamune A and Shimosegawa T. 2009, Signal transduction in pancreatic stellate cells. J Gastroenterol. 44, 249-260.

52. Masamune A, Satoh M, Kikuta K, Suzuki N, and Shimosegawa T. 2005, Activation of JAK-STAT pathway is required for platelet-derived growth factor-induced proliferation of pancreatic stellate cells. World J Gastroenterol.

11, 3385-3391.

53. McCarroll JA, Phillips PA, Kumar RK, Park S, Pirola RC, Wilson JS, and Apte MV. 2004, Pancreatic stellate cell migration: role of the phosphatidylinositol 3kinase(PI3-kinase) pathway. Biochem Pharmacol. 67, 1215-1225.

54. Klonowski-Stumpe H, Reinehr R, Fischer R, Warskulat U, Luthen R, and Haussinger D. 2003, Production and effects of endothelin-1 in rat pancreatic stellate cells. Pancreas. 27, 67-74.

55. Yen TW, Aardal NP, Bronner MP, Thorning DR, Savard CE, Lee SP, and Bell RH, Jr. 2002, Myofibroblasts are responsible for the desmoplastic reaction surrounding human pancreatic carcinomas. Surgery. 131, 129-134.

56. Vonlaufen A, Phillips PA, Xu Z, Goldstein D, Pirola RC, Wilson JS, and Apte MV. 2008, Pancreatic stellate cells and pancreatic cancer cells: an unholy alliance. Cancer Res. 68, 7707-7710.

57. Schneiderhan W, Diaz F, Fundel M, Zhou S, Siech M, Hasel C, Moller P, Gschwend JE, Seufferlein T, Gress T, Adler G, and Bachem MG. 2007, Pancreatic stellate cells are an important source of MMP-2 in human pancreatic cancer and accelerate tumor progression in a murine xenograft model and CAM assay. J Cell Sci. 120, 512-519.

58. Neesse A, Wagner M, Ellenrieder V, Bachem M, Gress TM, and Buchholz M. 2007, Pancreatic stellate cells potentiate proinvasive effects of SERPINE2 expression in pancreatic cancer xenograft tumors. Pancreatology. 7, 380-385.

59. Watanabe I, Hasebe T, Sasaki S, Konishi M, Inoue K, Nakagohri T, Oda T, Mukai K, and Kinoshita T. 2003, Advanced pancreatic ductal cancer: fibrotic focus and beta-catenin expression correlate with outcome. Pancreas. 26, 326-333.

60. Fujita H, Ohuchida K, Mizumoto K, Egami T, Miyoshi K, Moriyama T, Cui L, Yu J, Zhao M, Manabe T, and Tanaka M. 2009, Tumor-stromal interactions with direct cell contacts enhance proliferation of human pancreatic carcinoma cells.

Cancer Sci. 100, 2309-2317.

61. Evans SM and Koch CJ. 2003, Prognostic significance of tumor oxygenation in humans. Cancer Lett. 195, 1-16.

62. Jain RK. 2005, Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy. Science. 307, 58-62.

63. Tredan O, Galmarini CM, Patel K, and Tannock IF. 2007, Drug resistance and the solid tumor microenvironment. J Natl Cancer Inst. 99, 1441-1454.

50

Phoebe Phillips

64. Masamune A, Kikuta K, Watanabe T, Satoh K, Hirota M, and Shimosegawa T. 2008, Hypoxia stimulates pancreatic stellate cells to induce fibrosis and angiogenesis in pancreatic cancer. Am J Physiol Gastrointest Liver Physiol. 295, G709-717.

65. Bailey JM, Swanson BJ, Hamada T, Eggers JP, Singh PK, Caffery T, Ouellette MM, and Hollingsworth MA. 2008, Sonic hedgehog promotes desmoplasia in pancreatic cancer. Clin Cancer Res. 14, 5995-6004.

66. Komar G, Kauhanen S, Liukko K, Seppanen M, Kajander S, Ovaska J, Nuutila P, and Minn H. 2009, Decreased blood flow with increased metabolic activity: a novel sign of pancreatic tumor aggressiveness. Clin Cancer Res. 15, 5511-5517.

67. Akakura N, Kobayashi M, Horiuchi I, Suzuki A, Wang J, Chen J, Niizeki H, Kawamura K, Hosokawa M, and Asaka M. 2001, Constitutive expression of hypoxia-inducible factor-1alpha renders pancreatic cancer cells resistant to apoptosis induced by hypoxia and nutrient deprivation. Cancer Res. 61, 6548-6554.

68. Arumugam T, Ramachandran V, Fournier KF, Wang H, Marquis L, Abbruzzese JL, Gallick GE, Logsdon CD, McConkey DJ, and Choi W. 2009, Epithelial to mesenchymal transition contributes to drug resistance in pancreatic cancer.

Cancer Res. 69, 5820-5828.

69. Schwartz DL, Bankson JA, Lemos R, Jr., Lai SY, Thittai AK, He Y, Hostetter G, Demeure MJ, Von Hoff DD, and Powis G. 2010, Radiosensitization and stromal imaging response correlates for the HIF-1 inhibitor PX-478 given with or without chemotherapy in pancreatic cancer. Mol Cancer Ther. 9, 2057-2067.

70. Shah AN, Summy JM, Zhang J, Park SI, Parikh NU, and Gallick GE. 2007, Development and characterization of gemcitabine-resistant pancreatic tumor cells. Ann Surg Oncol. 14, 3629-3637.

71. Muerkoster S, Wegehenkel K, Arlt A, Witt M, Sipos B, Kruse ML, Sebens T, Kloppel G, Kalthoff H, Folsch UR, and Schafer H. 2004, Tumor stroma interactions induce chemoresistance in pancreatic ductal carcinoma cells involving increased secretion and paracrine effects of nitric oxide and interleukin-1beta. Cancer Res. 64, 1331-1337.

72. Miyamoto H, Murakami T, Tsuchida K, Sugino H, Miyake H, and Tashiro S. 2004, Tumor-stroma interaction of human pancreatic cancer: acquired resistance to anticancer drugs and proliferation regulation is dependent on extracellular matrix proteins. Pancreas. 28, 38-44.

73. Kikuta K, Masamune A, Watanabe T, Ariga H, Itoh H, Hamada S, Satoh K, Egawa S, Unno M, and Shimosegawa T. 2010, Pancreatic stellate cells promote epithelial-mesenchymal transition in pancreatic cancer cells. Biochem Biophys Res Commun. 403, 380-384.

74. Thayer SP, di Magliano MP, Heiser PW, Nielsen CM, Roberts DJ, Lauwers GY, Qi YP, Gysin S, Fernandez-del Castillo C, Yajnik V, Antoniu B, McMahon M, Warshaw AL, and Hebrok M. 2003, Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature. 425, 851-856.

Pancreatic stellate cells and fibrosis

51

75. Bailey JM, Mohr AM, and Hollingsworth MA. 2009, Sonic hedgehog paracrine signaling regulates metastasis and lymphangiogenesis in pancreatic cancer.

Oncogene. 28, 3513-3525.

76. Walter K, Omura N, Hong SM, Griffith M, Vincent A, Borges M, and Goggins M. 2010, Overexpression of smoothened activates the sonic hedgehog signaling pathway in pancreatic cancer-associated fibroblasts. Clin Cancer Res. 16, 1781-1789.

77. Bruck R, Genina O, Aeed H, Alexiev R, Nagler A, Avni Y, and Pines M. 2001, Halofuginone to prevent and treat thioacetamide-induced liver fibrosis in rats.

Hepatology. 33, 379-386.

78. Zion O, Genin O, Kawada N, Yoshizato K, Roffe S, Nagler A, Iovanna JL, Halevy O, and Pines M. 2009, Inhibition of transforming growth factor beta signaling by halofuginone as a modality for pancreas fibrosis prevention.

Pancreas. 38, 427-435.

79. Spector I, Honig H, Kawada N, Nagler A, Genin O, and Pines M. 2010, Inhibition of pancreatic stellate cell activation by halofuginone prevents pancreatic xenograft tumor development. Pancreas. 39, 1008-1015.

80. Fitzner B, Brock P, Holzhuter SA, Nizze H, Sparmann G, Emmrich J, Liebe S, and Jaster R. 2009, Synergistic growth inhibitory effects of the dual endothelin-1 receptor antagonist bosentan on pancreatic stellate and cancer cells. Dig Dis Sci. 54, 309-320.

81. Kunnumakkara AB, Sung B, Ravindran J, Diagaradjane P, Deorukhkar A, Dey S, Koca C, Yadav VR, Tong Z, Gelovani JG, Guha S, Krishnan S, and Aggarwal BB. 2010, {Gamma}-tocotrienol inhibits pancreatic tumors and sensitizes them to gemcitabine treatment by modulating the inflammatory microenvironment.

Cancer Res. 70, 8695-8705.

82. Rickmann M, Vaquero EC, Malagelada JR, and Molero X. 2007, Tocotrienols induce apoptosis and autophagy in rat pancreatic stellate cells through the mitochondrial death pathway. Gastroenterology. 132, 2518-2532.

83. McCarroll JA, Phillips PA, Santucci N, Pirola RC, Wilson JS, and Apte MV. 2006, Vitamin A inhibits pancreatic stellate cell activation: implications for treatment of pancreatic fibrosis. Gut. 55, 79-89.

84. Guo J, Xiao B, Lou Y, Yan C, Zhan L, Wang D, and Zhao W. 2006, Antitumor effects of all-trans-retinoic acid on cultured human pancreatic cancer cells. J Gastroenterol Hepatol. 21, 443-448.

85. Brembeck FH, Schoppmeyer K, Leupold U, Gornistu C, Keim V, Mossner J, Riecken EO, and Rosewicz S. 1998, A phase II pilot trial of 13-cis retinoic acid and interferon-alpha in patients with advanced pancreatic carcinoma. Cancer. 83, 2317-2323.

86. Michael A, Hill M, Maraveyas A, Dalgleish A, and Lofts F. 2007, 13-cisRetinoic acid in combination with gemcitabine in the treatment of locally advanced and metastatic pancreatic cancer--report of a pilot phase II study. Clin Oncol (R Coll Radiol). 19, 150-153.

87. Moore DF, Jr., Pazdur R, Sugarman S, Jones D, 3rd, Lippman SM, Bready B, and Abbruzzese JL. 1995, Pilot phase II trial of 13-cis-retinoic acid and

52

Phoebe Phillips

interferon-alpha combination therapy for advanced pancreatic adenocarcinoma.

Am J Clin Oncol. 18, 525-527.

88. Dang CT, Shapiro CL, and Hudis CA. 2002, Potential role of selective COX-2 inhibitors in cancer management. Oncology (Williston Park). 16, 30-36.

89. Chavez E, Segovia J, Shibayama M, Tsutsumi V, Vergara P, Castro-Sanchez L, Salazar EP, Moreno MG, and Muriel P. 2010, Antifibrotic and fibrolytic properties of celecoxib in liver damage induced by carbon tetrachloride in the rat.

Liver Int. 30, 969-978.

90. Aoki H, Ohnishi H, Hama K, Shinozaki S, Kita H, Osawa H, Yamamoto H, Sato K, Tamada K, and Sugano K. 2007, Cyclooxygenase-2 is required for activated pancreatic stellate cells to respond to proinflammatory cytokines. Am J Physiol Cell Physiol. 292, C259-268.

91. Yoshida S, Ujiki M, Ding XZ, Pelham C, Talamonti MS, Bell RH, Jr., Denham W, and Adrian TE. 2005, Pancreatic stellate cells (PSCs) express cyclooxygenase-2 (COX-2) and pancreatic cancer stimulates COX-2 in PSCs. Mol Cancer. 4, 27.

92. Ding XZ, Tong WG, and Adrian TE. 2000, Blockade of cyclooxygenase-2 inhibits proliferation and induces apoptosis in human pancreatic cancer cells.

Anticancer Res. 20, 2625-2631.

93. Sato N, Maehara N, and Goggins M. 2004, Gene expression profiling of tumorstromal interactions between pancreatic cancer cells and stromal fibroblasts.

Cancer Res. 64, 6950-6956.

94. Lipton A, Campbell-Baird C, Witters L, Harvey H, and Ali S. 2010, Phase II trial of gemcitabine, irinotecan, and celecoxib in patients with advanced pancreatic cancer. J Clin Gastroenterol. 44, 286-288.

95. Henderson NC, Mackinnon AC, Farnworth SL, Poirier F, Russo FP, Iredale JP, Haslett C, Simpson KJ, and Sethi T. 2006, Galectin-3 regulates myofibroblast activation and hepatic fibrosis. Proc Natl Acad Sci U S A. 103, 5060-5065.

96. Jiang HB, Xu M, and Wang XP. 2008, Pancreatic stellate cells promote proliferation and invasiveness of human pancreatic cancer cells via galectin-3.

World J Gastroenterol. 14, 2023-2028.

97. Jaster R, Sparmann G, Emmrich J, and Liebe S. 2002, Extracellular signal regulated kinases are key mediators of mitogenic signals in rat pancreatic stellate cells. Gut. 51, 579-584.

98. Masamune A, Suzuki N, Kikuta K, Satoh M, Satoh K, and Shimosegawa T. 2006, Curcumin blocks activation of pancreatic stellate cells. J Cell Biochem. 97, 1080-1093.

99. Masamune A, Kikuta K, Satoh M, Suzuki N, and Shimosegawa T. 2005, Green tea polyphenol epigallocatechin-3-gallate blocks PDGF-induced proliferation and migration of rat pancreatic stellate cells. World J Gastroenterol. 11, 3368-3374.

100. Chen SW, Chen YX, Zhang XR, Qian H, Chen WZ, and Xie WF. 2008, Targeted inhibition of platelet-derived growth factor receptor-beta subunit in hepatic stellate cells ameliorates hepatic fibrosis in rats. Gene Ther. 15, 1424-1435.

101. Nagashio Y, Ueno H, Imamura M, Asaumi H, Watanabe S, Yamaguchi T, Taguchi M, Tashiro M, and Otsuki M. 2004, Inhibition of transforming growth

Pancreatic stellate cells and fibrosis

53

factor beta decreases pancreatic fibrosis and protects the pancreas against chronic injury in mice. Lab Invest. 84, 1610-1618.

102. Yoo BM, Yeo M, Oh TY, Choi JH, Kim WW, Kim JH, Cho SW, Kim SJ, and Hahm KB. 2005, Amelioration of pancreatic fibrosis in mice with defective TGFbeta signaling. Pancreas. 30, e71-79.

103. Nakamura T, Takio K, Eto Y, Shibai H, Titani K, and Sugino H. 1990, Activinbinding protein from rat ovary is follistatin. Science. 247, 836-838.

104. Kim N, Yoo W, Lee J, Kim H, Lee H, Kim YS, Kim DU, and Oh J. 2009, Formation of vitamin A lipid droplets in pancreatic stellate cells requires albumin. Gut. 58, 1382-1390.

105. Garber K. 2004, Improved Paclitaxel formulation hints at new chemotherapy approach. J Natl Cancer Inst. 96, 90-91.

106. Garber K. 2010, Stromal Depletion Goes on Trial in Pancreatic Cancer. Journal of the National Cancer Institute. 102, 448-450.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 55-93 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

4. Pancreatic cancer and the tumor microenvironment: Mesenchyme’s role in pancreatic carcinogenesis Laurent Bartholin - TGFβ and Pancreatic Adenocarcinoma – Lab, Centre de Recherche en Cancérologie de LYON (CRCL), UMR INSERM 1052 - CNRS 5286 – Lyon I University UMR S 1052 Centre Léon Bérard, 28, rue Laënnec, 69373 LYON Cedex 08, France

Abstract. Mouse models of cancers represent an almost obligatory step along the sinuous road toward the design of new drugs for clinical application. There are three distinct uses for such mouse models, the ultimate goal of which, in a combined effort, is to provide new anti-cancer treatments. Mouse models of cancer are particularly well-adapted for understanding the biological principles of cancer (identification of new markers of the disease, putative targets, and the origin of cancer cells), testing anticancer drugs in preclinical studies (neoadjuvant, adjuvant, and antimetastatic treatments), and assessing the role of environmental conditions (tobacco, diet or environmental stress). The objective of this chapter is to present the diverse mouse models developed to date and, wherever possible, to define what they have taught us about pancreatic tumor-stroma interactions. A thorough overview of mouse models of pancreatic cancers currently Correspondence/Reprint request: Dr. Laurent Bartholin, - TGFβ and Pancreatic Adenocarcinoma – Lab, Centre de Recherche en Cancérologie de LYON (CRCL), UMR INSERM 1052 - CNRS 5286 – Lyon I University UMR S 1052, Centre Léon Bérard, 28, rue Laënnec, 69373 LYON Cedex 08, France E-mail: laurent.bartholin@unicancer.fr

56

Laurent Bartholin

available is presented (section I): chemically-induced models, transplanted models, and genetically engineered mouse (GEM) models. Presenting the broad spectrum and variety of mouse models will illustrate the extreme difficulty of selecting the model that is most suitable for answering a precisely delineated question (response to drug, stepwise tumorigenesis, environmental effects, etc.). The knowledge that has been acquired from these mouse models of pancreatic cancers will then be addressed, both at the fundamental and the clinical level (section II.). A particular attention will be paid to how these models have contributed to a better understanding of the cancer cell-stroma interactions. In addition, this chapter will explore how this integrated approach, considering the tumor bulk as a complex multicellular entity, could be used to develop new anti-cancer strategies targeting the complex dialogue between the different cell populations, rather than focusing only on the epithelial cancer cell components.

Introduction Pancreatic ductal adenocarcinoma (PDAC) is a very aggressive neoplasm; it is always fatal and is the fifth most common cause of death from cancer in the Western world [1]. The median age of patients diagnosed with a PDAC is 65-70 years. PDAC is therefore considered to be a cancer of the elderly (with the exception of familial/inherited forms, which are believed to represent less than 10% of patients). PDAC, which affects the exocrine pancreas, accounts for the vast majority of pancreatic cancers (more than 80%).

PDAC arises from precursor lesions with ductal differentiated features.

Indeed, both human patients and mouse models with this type of lesions have an increased risk of developing a PDAC. The precursors of the disease are divided into three groups: PanIN (pancreatic intraepithelial neoplasm), IPMN (intraductal papillary mucinous neoplasm), and MCN (mucinous cystic neoplasm) [2]. Despite the general microscopic ductal architecture of precursor lesions and invasive tumors, the cellular origin of PDAC has been extensively debated in the last decade. A growing body of evidence, mainly resulting from the use of genetically engineered mouse (GEM) models of pancreatic cancers, suggests that PDAC could arise from non-ductal pancreatic cells (acinar, centroacinar and endocrine cells).

PDAC is characterized by the presence of an abundant reactive stroma called ‘desmoplastic stroma’. Cancer cells may account for less than 10% of the total tumor volume. The desmoplastic stroma is composed of different cell types. The most abundant cells are myofibroblasts, also called ‘cancer associated fibroblasts’ (CAFs), ‘activated fibroblasts’, or ‘activated pancreatic stellate cells’ (activated PSC) [3]. Myofibroblasts mainly result from the activation of the ‘normal’ resident pancreatic fibroblasts or

Modeling pancreatic cancer in mice

57

‘quiescent PSC’, which have acquired increased proliferative properties. They also have undergone cytoskeleton rearrangements, and secrete extra-cellular matrix components that will ‘shape’ the desmoplastic stroma. Immune cells are the second most abundant category of cells composing the desmoplastic stroma, which is also composed of vascular cells, adipocytes, and neurons [4].

The main functional consequence of this abundant stroma is to ‘isolate’ the cancer cells from the surrounding normal pancreatic tissue, forming a barrier between cancer cells and non-transformed surrounding cells. Compared to other tumors, PDAC are poorly vascularized, which forces the cancer cells to adapt their metabolism to a very hostile environment (low oxygen and nutrient concentrations). These features actively contribute to the above-normal resistance of pancreatic cancers to classic chemotherapies and the associated poor prognosis of patients with PDAC. Indeed, besides the physical barrier provided by the desmoplastic stroma, anti-cancer treatments must also face the extreme resistance of pancreatic cancer cells used to surviving under stressful conditions, a ‘skill’ that helps them to overcome the cytotoxic effect of the usual anti-cancer drugs.

As no efficient treatment is currently available, it is essential to develop new strategies to fight pancreatic cancer. Given that the desmoplastic stroma is a keystone of pancreatic cancer aggressiveness, it is tempting to speculate that future anti-cancer treatments will need to target the cancer cell-stroma interactions (in association with drugs targeting the cancer cells themselves).

Such an approach requires a better understanding of the complex dialogue between stromal and cancer cells. The aim of this chapter is to present the different mouse models of pancreatic cancer and, where relevant, to specify what they have taught us about pancreatic tumor-stroma interactions. In this chapter, tumors affecting other organs (such as the breast) will eventually be mentioned briefly where they involve pertinent processes featuring cancer cell-stroma interactions (e.g. metastasis), which could reasonably be extrapolated to pancreatic cancer biology. Indeed, pancreatic cancer remains a poorly-known neoplasm compared to other tumors, and it is possible that work already done on other tumors may be applied directly to pancreatic cancers. A particular attention will be paid to the fibroblasts, the most abundant cell type present in the desmoplastic stroma, although other stromal components, such as immune or vascular cells, will also be referred to where they provide information leading to a better understanding of cancer cellstroma interactions. Finally, in view of the extraordinary diversity of signaling pathways involved in pancreatic cancers and their complex interactions, this chapter will not present an exhaustive description of these pathways. Two pathways, potentially acting paracrinally (and thus suitable for studying cancer cell-stroma interactions), will be explored in more detail.

58

Laurent Bartholin

The first of these, the Hedgehog (Hh) pathway, has provided the most advanced mouse models of pancreatic cancer involving cancer cell-stroma interactions as well as initial exploratory preclinical trials. The second, the Transforming Growth Factor Beta (TGFβ) pathway, illustrates the complex interactions between cancer cells and stroma and affects virtually all cell types inside the tumors possibly producing contradictory outputs (pro- or anti-oncogenic). Its signaling pathways represent attractive targets for the development of new anti-cancer drugs. This choice does not reflect underestimation of the crucial role played by other core autocrine or paracrine signaling pathways identified in human pancreatic cancer, such as PDGF, HGF, Wnt, Notch or integrins signaling [5].

I. Different types of mouse models of pancreatic cancers This section presents most of the mouse models of pancreatic cancers that have been developed so far. Most of them have not been created specifically to study cancer cell-stroma interactions. However, it is important to know as much as possible about these models as they may display interesting features and inspire future studies that will address cancer cell-stroma interactions.

Indeed, knowledge of these existing models could well encourage the researcher to use them to explore cancer cell-stroma interactions.

Historically, the first models of pancreatic cancer were obtained by using chemical carcinogens on rats. Later, as progress was made in anesthesiology, surgery and the development of immune-depressed mice, transplanted models rapidly became the gold standard of pancreatic cancer modeling, as they did for other tumors. Over the last decade, a tremendous amount of work has been done on identifying recurrent mutations in human PDAC and progress in bioengineering allowed the generation of genetically engineered mouse (GEM) models bearing the human mutations. GEM models represent a significant improvement in mouse modeling of pancreatic cancer, notably because they recapitulate the stepwise progression of the human disease from precursor lesions to metastatic dissemination through locally invasive lesions.

For many applications, GEM models have been supplanting or are expected soon to supplant the transplanted models that do not faithfully reproduce some aspects of the tumor’s biology, in particular the complex dialogue between the cancer cells and the stroma. It is, however, important to underscore that all types of models have their own individual uses; none of them should be abandoned as these different approaches offer the possibility of being combined in the same mouse model. It should be noted that the cell lines established from human tumors [6] and material provided from the tumors resected from the patients or the different mouse models may

Modeling pancreatic cancer in mice

59

represent very valuable material for in vitro and ex vivo experiments [7] (2D, 3D and organotypic culture models). These models fall beyond the scope of this review but need to be mentioned because of the interesting information they can provide, in particular in order to understand better the behavior of tumor cells (mesenchymal or cancerous) in an extracellular matrix with a controlled composition. Some extend orthotopic models, using human tissue to extrapolate the relationship between human epithelial cells and mouse mesenchymal cells.

1. Chemically-induced models The first animal models of adenocarcinoma were developed in the late 1970s in rats, using 7,12-dimethylbenzanthracene (DMBA) (long known to be carcinogens in the skin and liver) [8] and azaserine [9]. Intraperitoneal injection of N-nitroso-bis (2-oxypropyl) amine gave promising results with adenocarcinoma with ductal differentiation [10-13] and the induction of mutations found in humans, such as Kras-activating mutations and TP53inactivating mutations. However, these models developed tumors with an acinar-cell differentiation rather than tumors with a ductal differentiation, which still represent the vast majority of pancreatic cancers in humans. These models also developed tumors in lungs and liver, these ‘off-target effects’ probably being due to the intraperitoneal injection of the carcinogen drugs, which diffuse freely throughout the body. To limit this effect, Rivera et al. directly implanted DMBA into the rats’ pancreas or perfused the drugs into the main pancreatic duct of rats [14]. Their strategy succeeded in inducing ductal lesions, recapitulating the progression of the human disease (other classic carcinogens such as ethylnitrisoguanidine -ENNG- and methylnitrisoguanidine -MNNG- were shown to be inefficient). This model was later successfully reproduced in mice [15]. In the end, the chemicallyinduced mouse models of pancreatic cancer were not used extensively because of the dangers posed by the manipulated carcinogens, the induction of tumors in other organs, the inconvenience of drug perfusion in small animals and the emergence of transplanted models. More specifically, these approaches did not allow for the gathering of fundamental information regarding stromal components and/or desmoplasia.

2. Transplanted models Transplanted mouse models of pancreatic cancer have been used extensively to understand pancreatic tumor biology and to test anti-cancer treatments. The “host” into which the material is transplanted varies; immune-

60

Laurent Bartholin

deficient mice are most often used [16] but experiments have also been done with immune-competent mice [17], using tumoral material from immunecompatible mice. The “material” that is transplanted can be cells, tumor explants or even retroviral vectors encoding oncogenes [18]. The grafted material can come from diverse sources: a syngeneic donor, which is genetically identical to the recipient, an allogeneic donor, which is not genetically identical to the recipient but belongs to the same species, or a xenogeneic donor, which is from a different species to that of the recipient. The “anatomical site for grafting” also varies: a heterotopic graft into an abnormal anatomical position (usually subcutaneous), represents a convenient approach but is usually quite far removed from what really happens in vivo. An orthotopic graft into the normal anatomical position (e.g. in the pancreas) mimics better the real situation but this technique is more difficult to perform since it requires anesthesia and a laparotomy, and it presents a higher risk of morbidity (induced pancreatitis due to lesions resulting from surgery) and mortality (as a result of anesthesia, surgery or reactive pancreatitis).

Many models using the techniques and material described above succeeded in developing mouse models with pancreatic tumors. Very interestingly, it has been shown that Pancreatic Stellate Cells (PSC), characterized in 1998 [19,20], play a crucial role in creating a suitable environment for PDAC development. Transplanted models of mixtures composed of CAFs and cancer cells have turned out to be a powerful aid to understanding the relationship between stromal and cancer cells. However, transplanted models clearly have limitations. First, immune-depressed mice such as Nude mice are very fragile and very sensitive to opportunistic bacterial infections that can, unfortunately, jeopardize the preclinical trials.

Furthermore, they do not allow investigation of the early steps of tumor transformation and the heterogeneity of the experimental approaches (transplanted material, nature of the host, and site of the graft) have recurrently provided conflicting results in terms of drug efficiency, for example, a problem less often encountered with other tumors. Indeed, pancreatic cancer biology presents properties (desmoplastic stroma, poor vascularization, and abundant immune infiltrates) that cannot be faithfully reproduced in the transplanted models. This is the rationale for developing GEM (Genetically Engineered Mouse) models – more robust models offering the possibility of obtaining reproducible results.

3. Genetically Engineered Mouse (GEM) models To overcome some of the limitations of the chemically-induced and transplanted mouse models of cancers, a lot of work has been done over the

Modeling pancreatic cancer in mice

61

last decade to develop GEM models of pancreatic cancers. These models consist in animals developing pancreatic lesions arising from their own pancreas (‘autochthonous’ models) due to mutations (activation of oncogenes and/or inactivation of tumor suppressor genes, alone or in combination) that have been introduced by genetic engineering inside the genome of the mouse and are then transmissible to their progeny. This was made possible by advances in bioengineering techniques to develop genetically modified mice and the identification of genetic mutant signatures of PDAC in humans. The general principle underlying the generation of a GEM model of pancreatic cancer is the introduction inside the mouse genome of mutations recurrently observed in human pancreatic cancers.

Schematically, we can distinguish three approaches permitting the expression of a mutated allele, approaches that are not mutually exclusive and are possibly combined: i) constitutive expression (whole body (somatic) mutations or mutations in a specific organ, depending on the promoter chosen to drive the expression of the mutation), ii) conditional expression (Cre-lox system limiting the expression of the mutation inside the cell compartments expressing Cre-recombinase) or iii) inducible expression (tet on/off and tamoxifen induction allowing the expression of the mutation at a specific time of embryonic or adult life). Transgenic mice usually express transgenes that can either be inserted randomly in the genome, usually several copies with the associated risk of altering the expression of endogenous target genes located at the vicinity of the insertion site (‘pronucleus injection’ of the transgene) or inserted in specific sites (‘homologous recombination’ at the Hprt or Rosa26 loci in ES cells). Homologous recombination also permits inactivation (‘knockout’) or activation (‘knockin’) of a specific gene at its endogenous locus, to preserve its normal expression levels and patterns.

Most of the recurrent mutations observed in human PDAC turned out to induce either embryonic lethal phenotypes, severe developmental defects or multiple cancers when expressed in the whole body of GEM models. As a consequence, results of GEM models of pancreatic cancer with somatic mutations were fairly unsuccessful in terms of efficiently modeling the human pancreatic disease. Several examples are presented below. For instance, mouse models bearing somatic mutations found in the familial form of the human disease (germinal mutations), such as the inactivation of STK11/LKB1, a gene that is mutated in the Peutz-Jeghers (PJS) syndrome, develop gastrointestinal polyps and occasionally benign serous cystadenomas, contrasting with PJS patients presenting a very high risk of developing pancreatic cancer in addition to gastrointestinal polyps [21]. A second example is provided by Transforming Growth Factor Alpha (TGFα), which is highly expressed in human pancreatic cancers. Under the control of

62

Laurent Bartholin

metallothionein 1 ubiquitous promoter (a gene physiologically involved in heavy metals detoxification in GEMs), TGFα accumulates in many tissues, blood and urine and has pleiotropic effects in various tissues [22]. The pancreas ducts show pre-neoplastic transformation, but mice also develop a wide spectrum of tumors in other organs. A third example to illustrate the limitation of modeling pancreatic cancer through somatic mutations is the expression of the recurrent activating Kras mutation found in more than 90% of the sporadic forms of pancreatic cancers. Even using the elegant ‘hit-andrun’ techniques [23] that permit the expression of the mutation only in a proportion of somatic cells, this approach failed to induce pancreatic tumors [24]. In this study, the authors created a new mouse strain harboring a latent oncogenic allele of Kras (KrasLA) carrying the Kras exon 1 with an activating glycine to aspartic acid mutation at codon 12 (G12D). This allele is capable of spontaneous activation in vivo in animals carrying the targeted insertion allele. It is expected that the recombination frequency will ensure that animals carrying the KrasLA allele present widely distributed cells expressing the Kras oncogene in a surrounding environment mainly composed of cells that will not express the mutation, thus forming chimeric animals. Such mice mainly develop lung adenocarcinoma, thymic lymphoma and papilloma tumors, but no pancreatic tumors. This wide and variable spectrum may result from rearrangement events occurring at varying frequencies in different tissues of the mouse. Alternatively, some cells (lung, thymus and skin) might be especially sensitive to the ability of oncogenic Kras to induce abundant tumor proliferation. These early effects mask the potentially damaging effect that Kras activation could have on other organs, such as the pancreas presenting failsafe programs more efficiently. Together, the three models with somatic mutations presented here, e.g. i) STK11/LKB1 knockout to mimic a familial form of polyposis/pancreatic cancer, ii) TGFα transgenic mice to reproduce the overexpression of an oncogenic cytokine, and iii) KrasLA ‘hit and run’ mice expressing the most commonly observed mutation observed in human tumors – even in a limited proportion of somatic cells – clearly illustrate that somatic alterations expected to induce pancreatic tumors failed to mimic the human disease perfectly. These results demonstrate the necessity of targeting mutations in the pancreas in order to stop other defects from developing, which compromise the use of these models for a thorough analysis of pancreatic tumorigenesis.

Hence, the next objective of GEM models was to target the expression of genetic alterations inside the pancreas. The first successful approaches were designed in the late 1980s using acinar-specific promoters (elastase), driving the expression of viral or cellular proteins with oncogenic properties (KrasG12V, myc, SV40, TGFα, KrasG12D) [25-32]. Unfortunately, these models

Modeling pancreatic cancer in mice

63

developed precursor lesions without clear adenocarcinoma differentiation or with latency hardly compatible with experimental approaches, especially preclinical trials. For instance, mice overexpressing TGFα under the control of the elastase promoter will eventually develop pancreatic tumors with tubular differentiation after reaching of the age of one year.

The urgent task was then to develop ‘robust’ mouse models of PDAC with short latency and high penetrance, which could be used conveniently and efficiently in research laboratories for fundamental and preclinical applications. The first description of a genetically defined mouse model of pancreatic adenocarcinoma was published in 2001 [33]. This report shows that ectopic expression of TGFα in murine pancreatic acinar cells, using the elastase promoter [27], cooperates with TP53 deficiency [34] to induce, within 45 days, pancreatic adenocarcinoma with ductal differentiation according to CK19 positive status of the lesions. This observation contrasts with other work describing cooperative interactions between TGFα, Ink4a/Arf, and TP53 heterozygous inactivation (under the methallothionein-1 promoter) to induce a pancreatic neoplasm defined as serous cyst adenoma (SCA) [35]. This is a less aggressive pancreatic cystic neoplasm of ducts compared with pancreatic ductal adenocarcinoma (PDAC). It is rare in the general population but occurs frequently in Von Hippel-Lindau (VHL) patients. This observation clearly illustrates the importance of the cellular compartment that is targeted and gives a glimpse of the importance of genetic backgrounds and micro-environments in driving pancreatic tumorigenesis.

The generation of Mist1-KrasG12D Knock-In mice expressing the oncogenic mutant form of Kras at earlier stage of development resulted in metastatic exocrine pancreatic carcinomas with a mixed differentiation [36].

At this point, the new task was to model the stepwise progression of precursor lesions leading to PDAC. The arrival of Cre-lox technology offered a new opportunity to improve mouse models of pancreatic cancer by targeting the expression of conditional genetic alterations inside the pancreas.

Several models were designed, mainly using two mouse lines expressing Cre in the epithelial lineages of the pancreas: Pdx1-Cre [37] and Ptf1/p48-Cre [38]. Both Cre-expressing lines drive the expression of virtually any floxed alleles in all the pancreatic lineages with an epithelial origin (Pdx1-Cre ‘leaks’ in the distal stomach and the proximal duodenum [39]). This then provides the opportunity to study the effect of any Cre-inducible mutation inside the pancreas, whereas previous strategies required the creation of a new strain for each mutant allele under the control of a pancreas specific promoter. This system literally allowed a boom in mouse models developed in subsequent years. To overcome a possible bias that could result from abnormally high oncogene activation or oncogene activation in inappropriate

64

Laurent Bartholin

cell types or developmental stages, the lox-stop-lox KrasG12D (LSL-KrasG12D) knock-in mouse strain was created [40]. It bears a Cre-inducible conditional allele targeted to the endogenous Kras locus with the most common alteration found in pancreatic cancers. The LSL-KrasG12D allele is expressed at endogenous levels after Cre-mediated excision of a transcriptional stopper element in the appropriate cell compartment. Associated with Pdx1-Cre and Ptf1-Cre alleles [41], this new Cre-inducible activated mutant Kras allele induces the development of age-dependent precursor lesions (mPanINs) (according to the standard classification resulting from a workshop held in Park City, Utah, September 16-19, 1999 [2] http://pathology.jhu.edu/ pancreas_panin/) (in 100% of animals) and PDAC after one year (in about 10% of mice). Because of its insertion at the endogenous locus and because it mimics the most common genetic alteration found in human pancreatic cancers, LSL-KrasG12D became the most frequently used allele for modeling pancreatic cancers in mice.

To reduce latency and increase penetrance of aggressive pancreatic lesions described by Hingorani et al. in their pioneering work in [41], the LSL-KrasG12D allele was combined with inactivation of the tumor suppressor Ink4A/Arf, known to be inactivated in a large proportion of pancreatic cancers. Pdx1-Cre; LSL-KrasG12D; Ink4A/ArfL/L, and Ptf1-Cre;LSL-KrasG12D; Ink4A/ArfL/L developed very aggressive PDAC at a few weeks of age (median survival approximately two months), representing the most robust and aggressive mouse model of PDAC developed until then [42,43]. This model is a real achievement, since it recapitulates the different steps of PDAC development from precursor lesions (PanINs) to aggressive and metastatic tumors. However, the PDAC that develop in the LSL-KrasG12D; Ink4A/ArfL/L models presented a poorly differentiated (sarcomatoid) phenotype that does not represent the majority of pancreatic tumors in humans. This work and the unique opportunity offered by the Cre-lox system prompted researchers to combine the KrasG12D allele with tumor suppressor inactivation or oncogenic activations observed in humans, such as TP53 [43,44], SMAD4 [45-47], LKB1 [48], TβRII [49], TIF1γ [50], SMO [51], BRCA2 [52], Notch [53], activated β-Catenin [54] and MMP1 [55]. The TP53R172H (structural mutant) or TP53R270H (contact mutant) gain of function mutants observed in Li-Fraumeni syndrome, observed at codons 175 and 273 respectively in humans [56,57], were both engineered into the endogenous TP53 locus in mice (‘knockin’ mice). When associated with LSL-KrasG12D allele, these provide the model that best mimics human pancreatic cancers [43,44]: aggressive well-differentiated PDAC, with a desmoplastic stroma resembling human tumors, with 100% penetrance after a few weeks of age (median survival approximately five months). In contrast to TP53-/- knockout

Modeling pancreatic cancer in mice

65

mice with large gene deletion[34,58,59], TP53 gain of function mutants presented a different spectrum of tumors and were shown to support metastatic dissemination more efficiently. It should be noted that either inactivation of these tumor suppressors or activation of oncogenes is generally not sufficient on its own to induce aggressive pancreatic adenocarcinoma. Eventually, expression of KrasG12D mutant will induce the formation of PanINs, which will rarely evolve toward aggressive PDAC after a long latency in a small proportion of animals [41]. STK11/LKB1 (Pdx1Cre;LKB1L/L) inactivation will eventually lead to the formation of serous cyst adenomas [60]. PTEN inactivation leads to IPMNs [61] and β-catenin activation is sufficient to induce a rare tumor of low malignancy called solid pseudopapillary neoplasms (SPN) [54].

The next generation of GEM models of pancreatic cancers aimed to develop inducible systems (doxycycline, tamoxifen), in combination with previously described conditional organ-specific systems (Pdx1-Cre and Ptf1Cre), to explore the effect of mutations in the adult rather than the embryo (Pdx1-Cre and Ptf1-Cre drive the expression of Cre in the embryonic epithelial lineages of pancreas; for review see [62]). Several inducible mouse models have been designed bearing an inducible activated KrasG12D allele in the acinar compartment mouse strains. For instance, Guerra et al. show that elastase-tTA; tetO-Cre; LSL-KrasG12V mice in which the expression of the KrasG12V mutant is induced after doxycycline treatment will develop PanINs only under certain stress conditions modifying the pancreatic parenchyma [63]. Interestingly, other studies using other systems inducible by tamoxifen (CreER) pancreas-specific promoters (Pdx1-CreERTM [37]; Rip-CreERTM [64]; proCPA1-CreERT2 [65]) observed the development spontaneous PanINs in the adult [53,66,67]. It is thus noteworthy that, depending on the system, all these data obtained with activation of Kras in the adult produced quite diverse results, with different phenotypes, ranging from no lesions to highgrade PanINs formation.

A promising model was recently published by Scott Lowe lab, which may facilitate mouse modeling [68]. They have developed a new ‘inducible and reversible’ strategy using conditional RNA-interference, offering the opportunity of developing multiallelic shRNA transgenic animals to evaluate in parallel the function of many mammalian genes. This system relies on the insertion in ES cells of a targeting construct developed as a recipient vector for any miR30-based shRNA linked to a fluorescent reporter, enabling easy tracking and isolation of cells. Using this innovative approach, they succeeded in generating mice that mimicked the pre-existing lung adenocarcinoma mouse models obtained after intranasal injection of adenoviral Cre vector in the LSL-KrasG12D strain [40]. To test whether gene

66

Laurent Bartholin

suppression of Arf, which has been proposed as a way of bringing about this oncogenic effect of Kras, could produce similar results using RNAi, the authors produced mice harboring four alleles: (i) TG-p19.157 (Encoding a shRNA targeting Arf), (ii) LSL-KrasG12D (encoding a Cre-inducible oncogenic isoform of Kras), (iii) rCCSP-rtTA (encoding the reverse tettransactivator targeted to the CCSP locus specific to the lung epithelium), and (iv) R26-LSL-luciferase (for lineage tracking). After Cre inoculation and doxycycline treatment, longitudinal surveillance of animals by luciferase imaging demonstrated that sh-Arf/KrasG12D mice show an increased tumor burden compared to Kras-only controls. One can envision such an approach being used soon to target the pancreas.

In conclusion, two important observations can be made regarding GEM models of pancreatic cancers. The first is that modeling pancreatic tumors presenting general features of human tumors, in particular the presence of desmoplastic stroma, high penetrance and a short latency, has required considerable effort over several years from a number of laboratories that are world leaders in the fields of pancreatic cancer and mouse genetic engineering. This illustrates the complexity of pancreatic genetics. Secondly, it is necessary to associate several mutations in order to obtain aggressive tumors (generally one oncogene activation, coupled with a tumor suppressor inactivation to allow further genetic events resulting from genomic instability or absence of efficient failsafe programs). This latter aspect should be related to recent data provided by the “Pancreatic Cancer Genome Project�, which analyzed the protein-determining exons of all coding genes in 24 pancreatic tumors and reported an average of 63 somatic mutations per tumor [5]. These observations reinforce the idea that pancreatic cancer arises from very complex combinations of genetic alterations resulting from a high resistance to oncogenic stresses. Failsafe programs are probably very efficient in resisting so many alterations and this would also explain why mice do not spontaneously develop PDAC and why it takes a long time (median age of 70 years) to develop a pancreatic cancer in humans. However, the effects are devastating when the correct cocktail of genetic alterations is present. Like an ancient building that took a considerable time to complete (such as a medieval fortress), it is a very resistant edifice!

II. Contribution of mouse models in the understanding of pancreatic cancer cell-stroma interactions In this section, I will explore how the aforementioned mouse models of pancreatic cancers have contributed to a better understanding of the biology and genetics of this devastating disease. In particular, I will look at how they

Modeling pancreatic cancer in mice

67

have increased our understanding of the complex dialogue between the cancer and the stromal cells that compose the bulk of pancreatic tumors, and ultimately their usefulness in testing anticancer agents targeting the stroma in preclinical studies.

1. Cellular origin of pancreatic cancer The cellular origin of pancreatic cancer remains an open question (the appellation ‘Tumor initiating cells’ is preferred to the appellation ‘Cancer Stem Cell’, which is both ambiguous and improper regarding the definition of a genuine embryonic or somatic stem cell). However, mouse models of pancreatic cancers have improved our understanding of this problem and have allowed us to address several hypotheses. The first hypothesis, which was commonly accepted initially with respect to the ductal differentiation of the vast majority of human PDAC, was that PDAC arises from epithelial cells lining the ducts. However, this hypothesis has to a great extent been questioned and contested in the light of results obtained with GEM models.

Indeed, the observation that the expression of KrasG12D under the control of the CK19 ductal promoter is not sufficient to induce lesions argues strongly against this hypothesis [69]. A second hypothesis to explain the cellular origin of PDAC is that it arises from other cell lineages present inside the pancreas, through trans-differentiation processes, as supported by Kras oncogene expression in distinct cell populations (as mentioned in the previous section). A third hypothesis is that pancreatic cancer arises from injuries, which mobilize progenitors that form, under these stress conditions, more ductal tissue rather than acinar tissue, a process known as acinar-toductal metaplasia (ADM). This possibility is very well illustrated by pancreatitis, in which the injured acinar tissue is replaced by ductal structures with a significantly increased risk of developing pancreatic cancers. A fourth hypothesis to explain the cellular origin of PDAC is that cancer cells arise from embryonic pluripotent cells before they are committed into the ductal lineages. The use of Cre recombinase under the control of promoters specific to the embryonic pancreatic lineages (Pdx1-Cre and Ptf1-Cre) clearly supports this hypothesis. It is also possible that pancreatic cancers arise from a stem niche composed of a few pluripotent cells in the adult. It has recently been shown that adult progenitor cells of the exocrine pancreas were located in a Sox9-expressing progenitor zone [70]. Using internal ribosome entry site (IRES)-enhanced GFP (eGFP) mice or IRES-LacZ ‘knockin’ mice in the Sox9 locus, the authors showed that Sox9 expressing cells (e.g. cytokeratinpositive duct cells including centroacinar cells) continuously supplied the acinar compartment during physiological organ maintenance. The fifth

68

Laurent Bartholin

hypothesis is rather provocative. What if the tumor initiating cells in the pancreas were not initially resident cells in this organ? Pancreatic cancer, as discuss later in this chapter, does indeed often emerge from a reactive injured pancreas that is colonized by cells of different origins (hypothetically from Sox9 positive cells that are contiguous and anatomically connected to pancreatic ducts through the duodenum). Interestingly, an ectopic origin for mesenchymal cells in the desmoplastic stroma has been clearly demonstrated.

Recipient mice transgenic for the rat insulin promoter II gene linked to the large-T antigen of SV40 (RIPTag) developed solid beta-cell tumors of the endocrine pancreas when they were irradiated to ablate their bone marrow and injected (tail vein) with /GFP-positive whole-bone marrow cells from a donor [71]. Surprisingly, GFP detection in the RIPTag-induced endocrine tumors revealed that 25% PSC came from the donor bone marrow. Along the same lines, Yauch et al. produced models by xenografting surgical biopsies from human patients in Nude mice. They observed that the human-derived stroma was replaced by host mouse stroma in growing tumors [72]. These observations raise interesting questions about the functional significance in tumorigenesis of having a non-homogenous PSC population. This dimension should be taken into account when new anti-cancer drugs are designed to target the mesenchymal cells present inside the tumor. In the final analysis, we must acknowledge that none of these models definitively determined the cellular origin of PDAC. This is perhaps because there are from several origins or it could be that the mouse model that will allow us to find a definitive answer to this question has not yet been created! The difficulty we have in understanding the cellular origin of pancreatic cancer should be linked to the extraordinary plasticity of the different cells composing this organ, involving trans-differentiation processes that are largely regulated by the state of differentiation in the stromal environment, which influences or even dictates the future of pancreatic cells.

2. Stepwise tumorigenesis Tumor transformation is commonly considered to be a multi-step process characterized by initial transformation (associated with immortalization properties), amplification (associated with increased proliferation), local invasion (associated with cell migration locally) and metastatic dissemination (associated with intravasation, survival in the blood stream, extravasation, and survival in the distant host organ). As described in 1889 by Stephen Paget in his “Seed & Soil� theory [73], cancer cells need to grow in environmental conditions that promote their survival and their expansion to the distant sites of metastases. This concept is also true for the primary site of

Modeling pancreatic cancer in mice

69

the tumor. Indeed, the inherent biological and genetic properties of cancer cells are not necessarily sufficient to allow tumor initiation and local growth.

Depending on the type of cancer, the length of time between these different steps ranges from days to decades. It is even believed that some of them may occur simultaneously, but this aspect is beyond the scope of this chapter. The stromal environment influences the three main steps of tumorigenesis: tumor initiation, local invasion, and metastatic dissemination.

a- Pancreatitis, a permissive environment for tumor initiation Probably the best-documented example of environment promoting pancreatic transformation is pancreatitis. Schematically, pancreatitis is characterized by the destruction of the pancreatic parenchyma associated with immune infiltrates (acute pancreatitis). Ultimately, the damaged pancreatic tissue will be replaced by fibrotic tissue (chronic pancreatitis). The main etiology of human acute pancreatitis results from inappropriate zymogen activation inside the acini (instead of their activation when discharged in the digestive tract). The pancreas is then literally auto-digested, releasing inflammatory cytokines responsible for the recruitment of immune cells.

Repeated damage or uninterrupted aggression will lead to chronic pancreatitis. Pancreatitis represents a major risk of PDAC (for review see [74]). In this context, it is interesting to mention that pancreatitis is also characterized by a desmoplastic stroma and acinar-to-ductal metaplasia (ADM). The two pathologies share strong similarities regarding their gene array expression patterns and serum biomarkers. Also, pancreatitis is almost always accompanied by precursor lesions of pancreatic cancer. The two pathologies are thus clearly related and differential diagnosis between pancreatitis and low grade tumors of the pancreas is sometimes difficult.

Paradoxically, the increased number of immune cells has an anti-immune effect and even contributes actively towards tumor progression, a process commonly known as suppression of antitumor immunity. Tumors literally ‘corrupt’ the immune system to use it as a tumor promoter facilitating local invasion.

The use of mouse models eventually enabled researchers to demonstrate the functional relationship between pancreatitis and the susceptibility to develop pancreatic cancers, thus establishing a clear link between the inflamed status of the pancreatic parenchyma and cancer. The first mouse model of pancreatitis was obtained using chemical treatments. Administration of very high doses of cerulean, a cholecystokinin (CCK) analogue, has been used for decades to induce acute pancreatitis in rodents [75,76] through a mechanism that was further demonstrated to be dependent on activation of

70

Laurent Bartholin

the NF-kappaB/Rel pathway [77]. Another convenient and efficient way to induce pancreatitis is to inject L-Arginine, which acts via a mechanism that is still not clearly understood [78]. Alternative ways of inducing pancreatitis have been described (alcohol-induced, diet-induced, and immune-induced), presenting advantages and disadvantages (see review for detailed description of these models [79]). However, these mouse models of pancreatitis did not go on to develop pancreatic cancers, suggesting that other predisposing criteria, such as a genetic background, may be important for the onset of PDAC when there is pancreatitis.

GEM models with incomplete pancreatitis features have also been developed. This was possible due to a better understanding of the biochemistry of pancreatitis. Schematically, the PRSS1 gene encodes trypsinogen, which represents the inactive precursor of trypsin. Activation by autocatalytic proteolysis is inhibited by SPINK1, preventing auto digestion of the pancreas, the main cause of pancreatitis in humans. GEM models presenting a homozygous inactivation of SPINK3, the mouse homologue of SPINK1, lead to autophagic degeneration of the pancreas and early postnatal mortality, due to pancreatic insufficiency without obvious signs of immune infiltrates reminiscent of pancreatitis [80], all of which have also been correlated with increased trypsin activity [81]. It is possible that in their model, Ohmuraya et al. generated a phenotype that was so severe that the rapid destruction of the pancreas parenchyma masked a possible pancreatitis.

Transgenic mice expressing a trypsinogen with R122H mutation under the control of the elastase promoter were documented to present an increased tendency to auto-activation [82,83]. This model presents a mild phenotype (slight increase in the serum levels of lipase without any obvious histological modifications) that can be exacerbated through repeated treatments with cerulean. This mild phenotype can probably be attributed to a low expression level of R122H mutated trypsinogen. Finally, the expression of IL1β under the control of the elastase promoter has been shown to induce chronic pancreatitis [84]. It is noteworthy that none of these chemically-induced or genetically-induced models of pancreatitis were sufficient to induce the onset of aggressive pancreatic cancers until the generation of cLGL-KRasG12V mice, which express an Cre-inducible Kras G12V mutant under the CMV promoter express high levels of Kras leading to PDAC through a rather predominant pancreatitis-like phenotype [85]. Our recent laboratory work [86] showed that TGFβ signaling activation in the epithelial compartment cooperates with KrasG12D to induce acute pancreatitis. These two models are interesting since they report the induction of pancreatitis induced solely by genetic alteration, limiting the side effects and out-of-range effects observed with models requiring the use of chemicals.

Modeling pancreatic cancer in mice

71

It was eventually shown that chemical treatments such as that using cerulean may facilitate KrasG12D -induced PanINs formation and their eventual progression towards more aggressive lesions [87-89]. Two papers published recently demonstrate the importance of tumor-elicited inflammation and the molecular pathway inflammation responsible in KrasG12D-induced pancreatic transformation [90,91]. These studies show that the absence of IL-6/Stat3 could significantly slow down the formation of KrasG12D-induced PanINs. Guerra et al. showed that inducible expression in the adult of the mutant activated form of KrasG12V under the control of the elastase acinar-specific promoter was not sufficient to induce transformation [63]. It is very interesting that Guerra et al., aware of the possible cancer-driving effect of pancreatitis, demonstrated that when their animals were treated with cerulean (to induce pancreatitis), they developed abundant pancreatic advanced PanINs and PDAC within a few months (less than eight months). This study confirmed something that has long been observed in clinics, i.e. patients with pancreatitis present a higher risk of developing pancreatic cancer (for review see [74]). A recent study emphasized the critical role of pancreatitis in pancreatic cancer development, showing that endocrine cells are completely refractory to KrasG12D transformation, except in a context of chronic pancreatic injury [67]. In contrast, other groups, using other systems, showed that inducible activation of Kras in the adult acinar compartment was sufficient to induce neoplastic pancreatic precursor lesions [53,66]. This important observation underlined the fact that results can differ widely according to the mouse models used and/or the time at which the mutation is activated. Different explanations can be proposed for the apparent discrepancies between these models: i) endocrine, acinar cell or centroacinar cell origin, ii) the Kras+/LSLG12Vgeo allele being a weaker allele because of the absence of 3’-UTR, which is known to bind let7 miRNA tumor suppressor, iii) KrasG12D and KrasG12V have previously been reported to have slightly different oncogenic properties, and iv) doxycycline and tamoxifen may not be equally efficient in recombining the alleles in vivo (the different genetic backgrounds may also contribute to the differences observed).

b- Pancreatic Stellate Cells (PSC), a catalyzer of aggressiveness Pancreatic Stellate Cells (PSC) were characterized in 1998 [19] [20]. The early co-transplantation experiments involving HPSC (human pancreatic stellate cells) and pancreatic cancer cell lines (MiaPaCa-2, PANC-1, and SW850) involved subcutaneous transplantation into nude mice. Some of the transplants also included myofibroblast cells purified from pancreatic tissue

72

Laurent Bartholin

blocks removed during surgery on human patients with resectable pancreatic cancers or pancreatitis (illustrating again the close relationship between the two diseases) [92]. There was an increased deposition of connective tissue in pancreatic carcinoma by stellate cells in response to paracrine signals (secretion of TGFβ, FGF2 and PDGF) from carcinoma cells. This strongly supported tumor growth, demonstrated by at least a threefold increase in tumor volume after 11 days. Other studies have demonstrated that the presence of pancreatic stellate cells increases the growth of pancreatic cancer cells in subcutaneous xenograft models [93,94]. The early models showing increased tumor growth after co-injection of HPSCs and cancer cells in orthotopic models of pancreatic cancer were described in 2008 [95,96]. I have already stated that CAFs represent heterogeneous populations of different origins. Among them, a sub-population of stromal cells, apparently of mesenchyme origin and enriched in tumors, expresses the fibroblast activation protein α (FAP). To examine the role of FAP in cancer, Kraman et al. [97] showed that depletion of FAP-expressing cells in subcutaneous models of pancreatic ductal adenocarcinoma induced tumor regression. This resulted from acute, hypoxic death of both cancer and stromal cells, observed after FAP+ cell ablation. The mechanism of cell death involved TNFα and IFNγ, two cytokines already known to have immunosuppressive and antiangiogenic properties.

c- Stromal environment of distant organs dictates site-specific metastatic dissemination Different experimental approaches are commonly used to study the metastatic dissemination process: i) tracking metastatic cells that will ‘escape’ from the primary tumor and enter the blood stream (intravasation), ii) analyzing lung metastatic nodules formed after injection of cancer cells into the tail vein, iii) evaluation of different sites of metastases (such as bone or brain) resulting from intra-arterial inoculation (intra-cardiac). In all cases, the metastatic cells must escape from the vasculature system (extravasation), colonize, establish inside the distant organ, and then survive. After latency that may last weeks, years and even decades, the metastatic cells will start to proliferate again, invading the organ and eventually jeopardizing its function.

It is then well accepted, even if not fully understood, that a complex chemical dialogue is established between the metastatic cells and resident stromal cells in the colonized organ. This dialogue is responsible for the colonization, survival and ‘awakening’ of metastases.

In the past decade, transplanted models have offered an adequate approach for better understanding organ-specific dissemination. Indeed, work

Modeling pancreatic cancer in mice

73

in Massague’s laboratory has dissected the molecular basis responsible for the specificity of breast cancer cells bearing cognate genetic alterations for migrating to, and settling in, a specific organ. It is interesting that genes responsible for the specific tropism of a cell for a target organ are generally related to their capacity for enabling the survival of the cancer cell in the new environment. This illustrates, once again, the importance of cancer cellstroma interactions in cancers. These studies were of the capacity of breast cancer cells to metastasize in distant organs such as bone [98-100], lungs [101-104], and brain [104,105]. Breast cancer cells that express CXCR4 (C-X-C chemokine receptor 4), for example, are disposed to establish metastases in the bone marrow. The bone marrow-resident mesenchyme cells highly express the pro-survival chemokine SDF1/CXCL12 (stromal cellderived factor 1/ C-X-C chemokine ligand 12). Then, in a vicious cycle, metastatic cells and bone marrow cells mutually stimulate each other and so facilitate metastatic colonization. Such paracrine loops facilitating the colonization of distant organs by pancreatic metastasis have not yet been characterized. They therefore represent an important challenge for the near future, because it is well documented that pancreatic cancers are highly metastatic. The only information available on the metastatic process of pancreatic cancers concerns the early steps of the process, e.g. local invasion that leads to cancer cells entering the blood stream. This information comes from experiments that show the importance of adhesion, EMT and migration processes. Different studies have reported that PSC may facilitate metastatic dissemination) [95,96], however, to my knowledge, little is known about what dictates towards which organ metastases will migrate (peritoneal, lymphatic, gastric or hepatic).

3. Implication of external and familial factors in pancreatic cancer tumorigenesis The vast majority of pancreatic cancers arise from sporadic mutations or metabolic dysfunctions without a clear etiology. Individual susceptibility to pancreatic cancers may differ significantly depending on environment, lifestyle (diet and exposure to carcinogens), genetic background or ethnicity.

Indeed, clinical data collected over decades clearly show that certain human populations have higher probabilities of developing pancreatic cancers than others. The progress that has been made over the most recent decade in identifying the epidemiological cohorts at high risk, combined with the emergence of mouse models for pancreatic cancers, has permitted the better understanding of the molecular and environmental influences on the onset of the disease. I do not here exhaustively list all factors so far identified as

74

Laurent Bartholin

predisposing to pancreatic cancer (for a review see [106]). Instead, I focus on what mouse models have taught us about such predisposing factors, some of them resulting from perturbations in the interactions between pancreatic epithelial cells and the other cell types present in the pancreas (inflammation for instance). Some studies provide evidence that mouse genetic background and diet could promote development of the disease. For instance, alteration of strain background and a high omega-6 fat diet induces earlier onset of pancreatic neoplasia in Elastase-KrasG12D transgenic mice [107].

Retrospective epidemiological studies over the last 50 years suggest that chronic exposure to chemicals could increase the risk of pancreatic cancers [108]. Although the epidemiological data provide unquestionable evidence for the effect of cigarette smoking, other risk sources are more open to dispute because of the presence of other predisposing factors. In such cases, mouse models of pancreatic cancer offer unique opportunities for testing the carcinogenicity of certain compounds and for ultimately deciphering the molecular basis of their tumorigenic effects. Using the DMBA mouse model of pancreatic cancer, Wendt et al. showed that alcohol consumption (independently of its possible induction of pancreatitis, a major risk for pancreatic cancer), promotes the development of precursor PanIN lesions and pancreatic cancers, whereas caffeine has no effect [109]. Using the same mouse model, Bersh et al. demonstrated that tobacco is responsible for the promotion of pancreatic cancers in DMBA-treated rodents [110]. The carcinogenic effects of tobacco in a GEM model of chronic pancreatitis were also evaluated in 2010 [111]. In their study, Song et al. used the elastase-IL1β GEM mouse model of chronic pancreatitis [84]. These mice overexpress the pro-inflammatory IL1β cytokine, with a pancreas-specific promoter, and develop chronic pancreatitis that only rarely evolves into pancreatic cancer after the mice are 24 months old. After four months, tobacco-treated mice develop significant pancreatic ductal epithelial flattening and severe glandular atrophy, considered to be early signs of transformation, compared to untreated transgenic mice. This represents a clear example of the progression from exposure to chemicals to the onset of pancreatitis, major modification of the stromal environment, and the development of PDAC. Exposure to chemicals is not the only source of pancreatitis. Indeed, it is commonly accepted that 5-10% of all pancreatic cancers are familial.

However, the germ line mutations responsible are known for only a few of them (extensively reviewed elsewhere [106,112]). Several syndromes are linked to the mutations that have been identified (such as Familial atypical mole-multiple melanoma (FAMMM)/Melanoma-pancreatic cancer syndrome (CDKN2A), Peutz-Jeghers (PJS) syndrome (STK11/LKB1), Familial/hereditary pancreatitis (PRSS1, PINK1, CFTR, CTRC, PRSS2),

Modeling pancreatic cancer in mice

75

Cystic fibrosis (CFTR), Lynch syndrome/ HNPCC human non-polyposis colorectal cancer (MMR), Familial breast–ovarian cancer (BRCA1, BRCA2), Li-Fraumeni syndrome (TP53), Familial adenomatous polyposis (FAP) (APC), Multiple endocrine neoplasia (MEN), and Von Hippel-Lindau syndrome (VHL)). Each syndrome is associated with a particular risk of developing pancreatic cancer, ranging from 20% for VHL and CDKN2A to a few per cent for Cystic Fibrosis. Interestingly, GEM models bearing these mutations have been created but most of them failed to reproduce the human disease, again illustrating the complex genetics of pancreatic cancers. Clear alterations involving an impaired stromal environment have not yet been demonstrated in these models, although they are highly probable.

4. Relevant paracrine pathways in pancreatic cancers and their targeting a. General considerations Because of the urgent need to find new therapies against pancreatic cancers, most studies have been clinical. Many of them associated Gemcitabine with drugs already approved even if not necessarily designed specifically to target pancreatic cancers. These disappointing trials are also an indirect consequence of the lack of convenient mouse models that mimic human cancers well enough to allow the identification of new targets and the testing of new compounds against PDAC. Indeed, even if the early models of pancreatic cancer developed provided precious information for understanding pancreatic cancer development, especially for exploring the role of the recurrent mutations found in humans, it is only recently that mouse models have been developed for the specific investigation of these questions. It is unfortunate (but not really surprising) that this strategy has yielded rather dubious results. For instance, chemically-induced models have not been used extensively for preclinical studies, because the induced disease is generally not confined to the pancreas but also affects other organs. The genetic alterations are poorly controlled, making such models inadequate for guiding drug design and raising complications in preclinical tests. The greatest disappointment arises from the use of transplanted models, the gold standard during the last two decades for testing new drugs against a number of different tumors in preclinical studies. Even if xenograft models of pancreatic cancers respond fairly well to treatment, they systematically showed only modest efficiency in clinical trials against pancreatic cancers. Antiangiogenics or gemcitabine [113,114], for example, showed very promising effects in transplanted models but turned out to be inconclusive in clinical

76

Laurent Bartholin

application [115,116]. This occurs because, contrary to autochthonous tumors, transplanted tumors are highly vascularized, usually lack a reactive stroma, and have no immune infiltrates (recipient mice usually being immune deficient). These artificial features distorted the clinical effect of the tested drugs.

In this context, researchers have been concentrating their efforts on two priorities. The first is identifying new targets and designing innovative treatments. Tremendous progress has been made to this end over the most recent decade in characterizing the core pathways (PDGF, HGF, Wnt/βCatenin, Notch, and Hedgehog), impaired in all tumors [5]. Second, in order to test the validity of these ‘new’ pathways as putative anti-cancer opportunities by targeting them with innovative treatments, researchers needed to develop mouse models appropriately mimicking the human disease. Extraordinary common endeavor has been devoted to creating such models. They have to develop tumors with biological architecture (abundant desmoplastic stroma and poor vascularization), appropriate metabolism, and genetic alterations closely similar to those of human tumors. The GEM models previously described in this chapter thus clearly represent a very promising opportunity for predicting effective anti-cancer treatments [117], because they closely mimic the real physiology of the tumor and faithfully reproduce the stromal environment, an advantage in preclinical trials. The overall progress in both pancreatic cancer genetics and in developing mouse models of the disease are expected to provide new therapies in coming years.

These therapeutics will ultimately give significant improvements in patient morbidity and survival.

To limit the scope of this section, I will focus on two pathways, Hedgehog (Hh) and Transforming Growth Factor Beta (TFGβ), which act in a paracrine manner (cell non-autonomous) since they are mediated by secreted factors and imply a complex dialogue between cancer and stromal cells. Hedgehog undoubtedly represents the most accomplished and suitable target for preclinical studies resulting from the development of mouse models in the last 10 years. The TGFβ pathway is a good example of an integrated system, illustrating fairly well the problematic interconnection between pathways. Even if mouse models are presently less advanced than those for Hh, targeting the TGFβ pathway may offer attractive and promising opportunities for therapies against pancreatic cancer. Other pathways may also offer very interesting and promising targets for anti-cancer targets, but they will not be described here so as to limit the length of the chapter. For instance restoration of anti-tumoral immune response will not be treated but provide very encouraging results. A recent study illustrated that the restoration of tumor immune surveillance by targeting tumor-infiltrating

Modeling pancreatic cancer in mice

77

macrophages could rely on modification of the stromal microenvironment [118]. Using the Pdx1-Cre;LSL-KrasG12D;TP53R172H mouse model of pancreatic cancer, Beatty et al. have demonstrated that an agonist CD40 antibody was associated with the infiltration of tumor-associated macrophages (TAMs), with anti-tumoral properties that were unexpectedly associated to reduced stroma rather than a T-Cell reaction.

b. The Hedgehog pathway Improved understanding of the Hh pathway greatly contributed to a better appreciation of the complex interaction between cancer and stromal cells and the identification of new therapeutic targets. Investigation of this pathway has also yielded promising results in preclinical trials. Elegant mouse models have been developed to decipher, step by step, the precise role of this pathway in pancreatic cancer. Research using these models has led to the identification of innovative drug targets and they represent a signpost for approaching other pathways. Indeed, anti Hh strategies offer new opportunities to target the mesenchymal compartment of the desmoplastic stroma in pancreatic cancer, opening new perspectives for treatment. These results are largely attributable to mouse models having consistently evolved over the last 10 years so as to first allow us to understand tumor biology and then to test anti-Hh drugs.

Hedgehog is a secreted signaling protein discovered in Drosophila. Three analogous proteins have been identified: Sonic hedgehog (Shh), Indian hedgehog (Ihh) and Desert hedgehog (Dhh). Hh proteins interact with the receptor Patched (Ptch), derepressing the activity of the receptor Smoothened (Smo), which activate transcription factors of the Glioma (Gli) family, resulting in the activation of specific genes involved in proliferation and differentiation processes [119]. Thayer et al. show that aberrant activation of the Hh pathway occurs in the majority of human pancreatic cancers and that overexpression of Shh in the epithelial compartment of pancreas (Pdx-Shh mice) induces formation of PanIN-like lesions [120]. Mao et al. further developed a mouse model allowing the expression of tamoxifen-inducible activated Smo receptor in the adult whole body in CAG-CreER; LSL-Rosa26SmoM2-YFP mice (CAG is a composite ubiquitous promoter containing the chicken beta-actin promoter and cytomegalovirus enhancer). They also observed pancreatic precursor lesions with short latency and high penetrance, confirming that increased Hh signaling could drive tumorigenesis [121].

Morton et al. prepared pancreatic duct epithelial cells from three GEM models of pancreatic cancers (Ptf1-Cre;TP53L/L;K19-tv-a (avian leukosis virus subgroup) or Ptf1-Cre;Ink4a/ArfL/L;K19-tv-a or Ptf1-Cre;TP53L/L;

78

Laurent Bartholin

Ink4a/ArfL/L;K19-tv-a) [122]. They then infected cultured primary cells prepared from the tumors developed in those animals with a virus encoding Shh and demonstrated that expression of Shh enhances proliferation in a MAPK- and PI3-kinase-dependent manner. Shh expression also protects explanted cells from apoptosis and efficiently accelerates tumor progression in mouse orthotopic xenotransplants. Together, these studies clearly demonstrate an active oncogenic role for activated Hh signaling in pancreatic tumors.

Because Hh is a secreted factor, it was crucial to identify the target cells inside the tumor and distinguish possible autocrine (autonomous) from paracrine (non-autonomous) effects. Previous studies were not designed to answer this question, even if they clearly demonstrated the oncogenic role of the Hh pathway. Effort has consequently been invested in developing mouse models in which to attempt to answer the crucial question: by which mechanism does Hh facilitate pancreatic tumor initiation and progression?

The paracrine oncogenic effect was first suggested on the basis of a study by Pasca di Magliano et al., indicating that activation of the Hh signaling pathway in Pdx1-Cre;LSL-CLEG2 encoding an active form of the GLI2 transcription factor (CLEG2 transgene) does not develop precursor pancreatic lesions [123]. Bailey et al. showed that orthotopical grafted genetically modified Capan-2 overexpressing Shh induced the formation of tumors with greater desmoplastic reaction than that observed in ‘normal’ Capan-2 not overexpressing Shh [124]. They thus provided the first evidence showing that Shh produced by epithelial pancreatic cancer cells clearly stimulates desmoplastic reaction and enhances tumor progression. Simultaneously, Yauch et al. developed a mouse model that, allowed testing of the paracrine requirement of Hh signaling in pancreatic cancer. They established a new paradigm supporting a role of Hh secreted by cancer cells in activating Hh signaling in stromal cells and so, ultimately, stimulating tumor growth [72].

They identified a subset of human tumor epithelial cell lines expressing Hh ligands, xenografted these cells in Ptch1-lacZ; Rag2-/- mice and observed strong β-galactosidase activity only in the stromal cells edging the epithelial cancer cells. This observation demonstrates that Hh ligand produced by human cancer cell lines acted on murine stromal cells. Yauch et al. also generated xenograft models in nude mice using biopsies from human patients. They observed that human-derived stroma was replaced by host mouse stroma in growing tumors. This observation offered a unique opportunity to differentiate Hh pathway activity in the tumor from that in stromal compartments by species-specific primer sets. This system allowed them to show i) that human cancer cells were the main source of Hh ligand, whereas mouse stromal cells were the main target of Hh, and ii) inhibition of

Modeling pancreatic cancer in mice

79

the Hh pathway with chemicals or blocking antibodies down regulated Hh target genes in the stromal compartment but not within the tumor epithelium.

In a further approach, Yauch et al. co-injected mouse embryonic fibroblasts (MEFs) prepared from CAG-CreER;SmoL/L GEM embryos deficient for Shh signaling. They observed that Smo inactivation in fibroblasts after tamoxifendependent induction of Cre significantly reduced tumor growth, thus validating the idea that Hh activity in the stromal microenvironment constitutes a growth signal for epithelial tumor cells. The paracrine (e.g. cells with an activated Hh pathway) hypothesis was demonstrated for the first time in an autochthonous GEM model using an active mutant form of Smoothened (SmoM2) by Tian et al. [125]. Expression of SmoM2-YFP (fused to YFP for lineage tracking) within the pancreatic epithelium using Pdx1-Cre; Rosa26LSL-SmoM2-YFP mice was not sufficient to induce lesions. Neither does it potentate KrasG12D-driven pancreatic adenocarcinoma progression in Pdx1Cre; LSL-KrasG12D; Rosa26-LSL-SmoM2-YFP mice. This strongly suggests that activation of the Hh signal inside the epithelial compartment has no oncogenic effect. In order to explore in which compartments activated SmoM2 mutant expression may induce the activation of Hh signaling pathway, Tian et al also generated a Pdx1-Cre; Rosa26-LSL-SmoM2-YFP; Ptch1-lacZ strain. Despite the fact that SmoM2-YFP was detectable in all epithelial pancreatic lineages (acini, ducts, and islets), as predicted by the expression profile of Pdx1-Cre, no activated Hh signaling was identified in those cells (β-galactosidase negative). In contrast, a very strong β-galactosidase signal was observed in mesenchymal cells, probably reflecting the reception of Hh signal from ligands expressed by the epithelial cells. Tian et al. performed laser-capture micro dissection of pancreatic tumors in three GEM models (Pdx1-Cre;LSL-KrasG12D, Pdx1-Cre;LSLKrasG12D;TP53R270H, and PdxCre;LSL-KrasG12D;SmoM2) and further separation of tumor epithelium from the surrounding stroma revealed a 13fold Gli increased expression in stroma compared to 150-fold in epithelial cancer cells (these results were confirmed in human tumors). It is noteworthy that Hh has a paracrine effect on other cell types in pancreatic tumors. Bayley et al., for example, using an orthotopic mouse model of pancreatic cancer, showed that the Shh paracrine signal could stimulate lymphangiogenesis and facilitate metastasis [126].

Desmoplastic stroma represents a crucial feature explaining the aggressiveness and drug resistance of pancreatic cancer, and the Hh pathway seems to be a pivotal point supporting the development and the maintenance of this desmoplastic stroma. Hh signaling, therefore, has naturally become an attractive target for drugs against this deadly disease. As early as 2003, Berman et al. showed that treatment with cyclopamine, an Hh inhibitor, of

80

Laurent Bartholin

freshly-resected human pancreatic carcinomas with identified high level of Hh activity, transplanted into nude mice, resulted in a significant reduction of tumor burden [127]. Cyclopamine has also been shown to synergize with gemcitabine to reduce metastases in an orthotopic xenograft model of pancreatic cancer [128]. An orally bioavailable small-molecule (IPI-269609) of Hh signaling inhibits systemic metastases in orthotopic xenografts established from human pancreatic cancer cell lines [129]. Feldmann et al.

[130] further showed that Hh pathway inhibition by cyclopamine in the autochthonous Pdx1-Cre;LSL-KrasG12D;Ink4A/ArfL/L mouse model [42] modestly but significantly prolonged survival (67 rather than 61 days). Bailey et al., using their orthotopic Capan-2 model overexpressing Shh, showed that desmoplastic tumor size decreased when treated with a Shh-blocking antibody [124]. It is currently unclear why anti-oncogenic effects of anti-Hh strategies are revealed in some of the experiments addressing the autocrine role of Hh in epithelial compartments. Among suggested reasons, however, are out of target effects due to high drug concentrations, the use of reporters, and effects on host fibroblasts that might have colonized the tumors of human epithelial origin.

In a pioneering work, Olive et al. demonstrated that low efficiency of an anti-cancer drug is the result of inefficient drug delivery to the tumor [131].

In a first approach, they observed that the active metabolite of Gemcitabine is present in transplanted models but not in GEM models (Pdx1-Cre;LSLKrasG12D;TP53R270H and Pdx1-Cre;LSL-KrasG12D;TP53R172H), providing the first validation in an ‘autochthonous’ mouse model. They attributed this observation to the presence of the abundant desmoplastic stroma in GEM models, compared to transplanted models. They then hypothesized that targeting the stroma may represent a prerequisite for allowing a drug to reach cancer cells and may ultimately make ‘conventional’ chemotherapies more efficient. To that end, Olive et al. showed that the IPI-926 Shh inhibitor efficiently shrinks the desmoplastic stroma and increases delivery of chemotherapy in Pdx1-Cre;LSL-KrasG12D;TP53R270H and Pdx1-Cre;LSLKrasG12D;TP53R172H mice. Unfortunately, reducing the desmoplastic was not sufficient to achieve real clinical benefits. In fact, it increased drug delivery via increased angiogenesis, which also increased the fueling of the tumor with oxygen and nutrients, illustrating once again the extraordinary capacity of pancreatic tumors to adapt to environmental changes, probably as a result of cancer cell plasticity and/or the ability to mobilize new cancer cells progenitors. However, this study opens a very interesting perspective in which combinations of anti Hh may give promising results when associated with Gemcitabine or oxyplatin, two drugs of the cancer cells themselves, along with of anti-angiogenic drugs, which could then be given a new lease

Modeling pancreatic cancer in mice

81

of life. Of course such innovative therapeutic approaches can only be tested at the preclinical level, given the potential risk of increasing tumor angiogenesis. A new generation of drugs targeting the Hh pathway, such as terpenoids and a flavonoid glycoside from Acacia pennata leaves as hedgehog/GLI-mediated transcriptional inhibitors, is also in the pipeline [132].

IV. The TGFβ pathway MT1-MMP cooperates with KrasG12D to promote pancreatic fibrosis through increased TGFβ signaling The Transforming Growth Factor Beta (TGFβ) is a secreted polypeptide belonging to a wide family of cytokines and growth factors including TGFβs, Bone morphogenetic Proteins (BMPs), and activins. TGFβ has many roles during embryonic development and adult life (polarity of the embryo, immunosuppression, and wound healing). At the early stages of tumorigenesis, TGFβ has tumor-suppressive functions (anti-proliferative and pro-apoptotic effects). As the tumor progresses, the TGFβ ‘loses’ its protective properties and stimulates tumor progression (Epithelial to Mesenchymal Transition – EMT, angiogenesis, extracellular matrix remodeling, and immunosuppression). Upon binding to its receptors, TGFβ triggers phosphorylation of the SMAD2 and SMAD3 transcription factors.

Phosphorylated SMAD2 and SMAD3 then interact with SMAD4. The SMAD2/3/4 complex accumulates within the nucleus, binds to DNA and activates the transcription of target genes, leading to proliferative arrest or apoptosis of epithelial cells. The TGFβ pathway appears to be of particular importance to PDAC tumor suppression, since it is impaired in virtually all cases of this malignancy [5], and since SMAD4 (also known as DPC4 for Deleted in Pancreatic Carcinoma), classically considered as the main effector mediating the anti-proliferative effect of TGFβ, is deleted in about 50% of pancreatic adenocarcinomas [133,134]. This suggests that loss of Smad4mediated cell growth inhibition is crucial for PDAC progression. To illustrate the key role of the tumor suppressive functions that are ablated in pancreatic cancers, GEM models with an inactivated TGFβ pathway (SMAD4 knockout, TβRII knockout, overexpressed inhibitory SMAD7) have been shown to accelerate Kras-mediated tumor transformation [45-47,49,135]. More recently, it has been shown that inactivation of TIF1γ, a factor involved in TGFβ signaling [136-139], cooperates with KrasG12D to induce cystic tumors of the pancreas, although a functional relationship between TGFβ and TIF1γ has not yet been established [50]. Our recent results reveal that TIF1γ tumor-

82

Laurent Bartholin

suppressor activity is at least partially independent of SMAD4 activity [140].

In apparent contradiction to the loss of sensitivity to the ‘protective effects’ of TGFβ observed in pancreatic tumors (as well as many other types of tumors), it is also well-documented that TGFβ is found at very high concentrations in pancreatic tumors (and other tumors), suggesting that TGFβ also facilitates tumor progression. TGFβ plays a crucial role in conferring aggressiveness to pancreatic cancers (and cancers in general). TGFβ production, which increases as the tumor progresses, stimulates tumor progression (Epithelial to Mesenchymal Transition – EMT, angiogenesis, extracellular matrix remodeling, and immune suppressive effect). Among the oncogenic properties of TGFβ that may be targeted by anti-cancer drugs, there is EMT: TGFβ signaling inactivation reverts the mesenchymal phenotype of KrasG12D; Ink4A/ArfL/L tumors [45]. We then face a question that still requires an answer: Why is it that loss of SMAD4 is associated with tumor progression, while SMAD4 is also needed to confer aggressiveness by inducing EMT? Interestingly, a growing body of experimental evidence suggests that a dual role of TGFβ during tumorigenesis is closely related to the cellular context in which the TGFβ signal occurs, which dictates the specific signaling pathways and specific genes that are induced within particular cells to drive either protective or damaging effects. The dual role of TGFβ during carcinogenesis would then largely result from its pleiotropic role in virtually all cell types present in the tumor. For example, contrary to the well-known immunosuppressive effect of TGFβ on immune cells, we recently showed, in a mouse model that expresses a constitutively active type I TGFβ receptor [141,142], that co-activation of TGFβ signaling and the KrasG12D oncogene, specifically in the epithelial lineage of pancreas (Pdx1-Cre), induces pancreatitis and PDAC [143]. It is also important to mention that TGFβ contributes to local invasion in breast cancer. Indeed, localized and reversible TGFβ signaling, in ‘leading fibroblasts’ at the invasive front, switches breast cancer cells from cohesive to single cell motility [144]. It would therefore be very interesting to explore thoroughly whether the same mechanism is observed in pancreatic tumors.

As for Hh, the TGFβ pathway represents an attractive candidate for further therapeutic approaches that will attempt to target the desmoplastic stroma. Therapeutic targeting of the TGFβ pathway in tumors is based on the rationale that blocking TGFβ function might i) empower the immune system against the tumor, ii) block EMT and cancer cell migration, iii) inhibit the production of autocrine metastatic and survival factors, and iv) inhibit matrix remodeling promoting tumor progression and angiogenesis. Most of these effects are expected to influence the stromal environment of the cancer cells

Modeling pancreatic cancer in mice

83

rather than the cancer cells themselves. Several classes of compounds can be used: antisense molecules, ligand trap molecules, such as TGFβ neutralizing antibodies and soluble TGFβ receptors, and small synthetic molecules inhibiting kinase activity of TGFβ receptors. Despite the strong therapeutic potential, it is striking that relatively few clinical trials have been performed to date using these molecules [145]. If we specifically consider trials performed on pancreatic cancers, they can be counted on the fingers of one hand. This is all the more surprising considering both the crucial role of TGFβ in these tumors and the absence of efficient treatment against PDAC. This tentative behavior probably results, with some justification, from the dual role of TGFβ during tumor progression, which has compromised its targeting by anti-cancer drugs. It is quite rightly feared that inhibition of TGFβ could lead to chronic inflammatory syndromes and autoimmune reactions, and could enhance the progression of premalignant lesions in relation to what is observed in SMAD, TGFβ and TGFβ receptor null mouse models. As a line of evidence, it has been reported that blocking TGFβ signaling could enhance tumor growth at the primary site in breast cancer [146-149]. Surprisingly, systemic administration of TGFβ blockers has reportedly not caused any of these possible side effects in animal models in the few pre-clinical trials performed so far to treat several types of tumors, especially metastasis. A soluble TGFβ receptor attenuates expression of metastasis-associated genes and suppresses pancreatic cancer cell metastasis in transplanted models (xenograft or orthograft) of the human pancreatic cancer cell line (PANC-1) in immune-depressed mice [150]. Reduced metastasis was also reported in orthotopic pancreatic cancer xenograft models after treatment with TGFβ receptor kinase inhibitors SD-208 [151] and LY2109761 [152]. Interestingly, Kano et al. show (xenografted TGFβinsensitive BxPC3 human pancreatic adenocarcinoma cell line in Nude mice as a disease model) that the small-molecule TGFβ receptor inhibitor (LY364947) can modify the tumor microenvironment (alteration in cancerassociated neo-vasculature) and modify permeability to drugs [153,154].

Very recently, Trabedersen (AP12009), an antisense molecule directed against TGFβ2 in an orthotopic mouse model of metastatic pancreatic cancer, significantly reduced tumor growth, lymph node metastasis and angiogenesis [155]. Phase I/II clinical trials using AP12009 in pancreatic cancer are currently in progress. Because of the potential risk associated with anti-TGFβ inhibitors, their use requires rigorous investigation in preclinical models and in particular they need to be better targeted. Recent progress in identifying the mechanisms responsible for the functional switch of TGFβ from a tumorsuppressor to an oncogene stimulated renewed interest in developing therapeutics that would specifically attempt to inhibit the oncogenic

84

Laurent Bartholin

properties and preserve/restore the onco-suppressive properties of TGFβ. The ‘double-edged sword’ of TGF-β exerts both cell-specific and contextdependent effects, and TGFβ represents the archetypal factor acting in cellautonomous or non-cell autonomous mechanisms. In this context, it is expected that further progress will be made towards a better understanding of the complex role of this cytokine through the use of mouse models of pancreatic cancers, and that this will provide new therapeutic strategies.

Conclusion-perspectives Pancreatic cancer is probably at the same time one of the most aggressive neoplasms and one of the neoplasms for which the least progress has been made in the last decades. Several reasons can be found for this: the fact that it is a tumor affecting the elderly (so not a public health priority), late diagnosis, the inherent aggressiveness of the tumors, and the absence of adapted mouse models to study the stepwise progression of the disease and new drugs. The increase in life expectancy in many countries, together with recent progress in the fields of medical imaging, genetics, and mouse modeling, should make it possible to propose new therapeutics in coming years.

Pancreatic cancer perfectly illustrates the fact that new mouse models and therapeutics will need to focus both on targeting cancer cells and targeting stromal cells. To that end, we need to come up with innovative approaches. What should we expect from the ‘next generation’ of mouse models of pancreatic cancer? Ideally, these models should present many properties, such as allowing modulation in different compartments at a precise time point, being possibly reversible, and allowing the expression of different mutations. They should also be cost-effective and not timeconsuming.

In order to offer the possibility of new treatments to fight pancreatic cancer, mouse models of pancreatic cancers must face several challenges, such as deciphering the respective role of signaling pathways in the different compartments composing the tumor and eventually identifying genetic and epigenetic alterations that would specifically affect the stromal compartment.

We must also determine the cellular origin of PDAC and understand the molecular basis of the diverse genetic alterations that contribute to the development of cancer. This integrated view of the disease is crucial in refining the targeting of treatments that will be proposed to patients (individualized therapy) to limit out of targets and side effects of new anticancer drugs. Such out of targets and side effects result from the blockage of pathways, which has multiple effects, sometimes antagonistic, depending on

Modeling pancreatic cancer in mice

85

the cellular compartment (TGFβ signaling). It is important to bear in mind that primary tumor and metastasis may not respond in the same way to treatments, as exemplified by preclinical trials performed using anti-TGFβ strategies.

In parallel, we need to improve our capacity to monitor routinely, in longitudinal studies, tumor development in mouse models using live-imaging technics (high resolution ultrasonography, PET, and in vivo fluorescence).

More specifically, since the stromal compartment is now considered to be the ‘Achilles tendon’ of pancreatic cancer, future imaging techniques should be capable of addressing subtle changes in this compartment after genetic alterations or treatments aimed at targeting stroma.

A critical point for research is ultimately to determine the criteria that make it possible to anticipate and assess a significant response, to identify the targets of chemoprevention, including the relevant molecular pathways, cell types, and environmental conditions promoting onset and progression of the different lesions. The main difficulty is then to choose the right balance between the resolution power of a given model and the question we are asking. Indeed, it is expected that the same mouse models will not be used to address both tumor initiation and metastatic dissemination. In conclusion, progress in treating pancreatic cancer relies on a precise match between mouse models and the techniques for monitoring the different cell compartments of the tumor.

References 1.

2.

3.

4.

5.

Jemal, A., Murray, T., Samuels, A., Ghafoor, A., Ward, E., Thun, M. J.2003, Ca: a Cancer Journal for Clinicians, 53, 5.

Hruban, R. H., Adsay, N. V., Albores-Saavedra, J., Compton, C., Garrett, E. S., Goodman, S. N., Kern, S. E., Klimstra, D. S., Kloppel, G., Longnecker, D. S., Luttges, J., Offerhaus, G. J.2001, American Journal of Surgical Pathology, 25, 579.

Bachem, M. G., Zhou, S., Buck, K., Schneiderhan, W., Siech, M.2008, Langenbecks Arch Surg, 393, 891.

Chu, G. C., Kimmelman, A. C., Hezel, A. F., DePinho, R. A.2007, J Cell Biochem, 101, 887.

Jones, S., Zhang, X., Parsons, D. W., Lin, J. C., Leary, R. J., Angenendt, P., Mankoo, P., Carter, H., Kamiyama, H., Jimeno, A., Hong, S. M., Fu, B., Lin, M. T., Calhoun, E. S., Kamiyama, M., Walter, K., Nikolskaya, T., Nikolsky, Y., Hartigan, J., Smith, D. R., Hidalgo, M., Leach, S. D., Klein, A. P., Jaffee, E. M., Goggins, M., Maitra, A., Iacobuzio-Donahue, C., Eshleman, J. R., Kern, S. E., Hruban, R. H., Karchin, R., Papadopoulos, N., Parmigiani, G., Vogelstein, B., Velculescu, V. E., Kinzler, K. W.2008, Science, 321, 1801.

86

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Laurent Bartholin

Monti, P., Marchesi, F., Reni, M., Mercalli, A., Sordi, V., Zerbi, A., Balzano, G., Carlo, V., Allavena, P., Piemonti, L.2004, Virchows Archiv, 445, 236.

Froeling, F. E., Marshall, J. F., Kocher, H. M.2010, J Biotechnol, 148, 16.

Dissin, J., Mills, L. R., Mains, D. L., Black, O., Jr., Webster, P. D., 3rd.1975, J Natl Cancer Inst, 55, 857.

Lilja, H. S., Hyde, E., Longnecker, D. S., Yager, J. D., Jr.1977, Cancer Res, 37, 3925.

Pour, P., Althoff, J., Kruger, F. W., Mohr, U.1977, Cancer Lett, 2, 233.

Pour, P., Althoff, J., Kruger, F. W., Mohr, U.1977, J Natl Cancer Inst, 58, 1449.

Pour, P., Gingell, R., Langenbach, R., Nagel, D., Grandjean, C., Lawson, T., Salmasi, S.1980, Cancer Res, 40, 3585.

Naito, Z., Takahashi, M., Furukawa, F., Toyoda, K., Hasegawa, R., Hayashi, Y.1986, Acta Pathol Jpn, 36, 1359.

Rivera, J. A., Graeme-Cook, F., Werner, J., Z'Graggen, K., Rustgi, A. K., Rattner, D. W., Warshaw, A. L., Fernandez-del Castillo, C.1997, Surgery, 122, 82.

Osvaldt, A. B., Wendt, L. R., Bersch, V. P., Backes, A. N., de Cassia, A. S. R., Edelweiss, M. I., Rohde, L.2006, Surgery, 140, 803.

Caldas, C., Hahn, S. A., da Costa, L. T., Redston, M. S., Schutte, M., Seymour, A. B., Weinstein, C. L., Hruban, R. H., Yeo, C. J., Kern, S. E.1994, Nat Genet, 8, 27.

Tseng, W. W., Winer, D., Kenkel, J. A., Choi, O., Shain, A. H., Pollack, J. R., French, R., Lowy, A. M., Engleman, E. G.2010, Clin Cancer Res, 16, 3684.

Yoshida, T., Hanahan, D.1994, American Journal of Pathology, 145, 671.

Apte, M. V., Haber, P. S., Applegate, T. L., Norton, I. D., McCaughan, G. W., Korsten, M. A., Pirola, R. C., Wilson, J. S.1998, Gut, 43, 128.

Bachem, M. G., Schneider, E., Gross, H., Weidenbach, H., Schmid, R. M., Menke, A., Siech, M., Beger, H., Grunert, A., Adler, G.1998, Gastroenterology, 115, 421.

Bardeesy, N., Sinha, M., Hezel, A. F., Signoretti, S., Hathaway, N. A., Sharpless, N. E., Loda, M., Carrasco, D. R., DePinho, R. A.2002, Nature, 419, 162.

Jhappan, C., Stahle, C., Harkins, R. N., Fausto, N., Smith, G. H., Merlino, G. T.1990, Cell, 61, 1137.

Hasty, P., Ramirez-Solis, R., Krumlauf, R., Bradley, A.1991, Nature, 350, 243 Johnson, L., Mercer, K., Greenbaum, D., Bronson, R. T., Crowley, D., Tuveson, D. A., Jacks, T.2001, Nature, 410, 1111 .

Quaife, C. J., Pinkert, C. A., Ornitz, D. M., Palmiter, R. D., Brinster, R. L.1987, Cell, 48, 1023.

Ornitz, D. M., Hammer, R. E., Messing, A., Palmiter, R. D., Brinster, R. L.1987, Science, 238, 188.

Sandgren, E. P., Luetteke, N. C., Palmiter, R. D., Brinster, R. L., Lee, D. C.1990, Cell, 61, 1121.

Sandgren, E. P., Quaife, C. J., Paulovich, A. G., Palmiter, R. D., Brinster, R. L.1991, Proc Natl Acad Sci U S A, 88, 93.

Glasner, S., Memoli, V., Longnecker, D. S.1992, Am J Pathol, 140, 1237.

Modeling pancreatic cancer in mice

87

30. Sandgren, E. P., Luetteke, N. C., Qiu, T. H., Palmiter, R. D., Brinster, R. L., Lee, D. C.1993, Mol Cell Biol, 13, 320.

31. Wagner, M., Luhrs, H., Kloppel, G., Adler, G., Schmid, R. M.1998, Gastroenterology, 115, 1254.

32. Grippo, P. J., Nowlin, P. S., Demeure, M. J., Longnecker, D. S., Sandgren, E. P.2003, Cancer Research, 63, 2016.

33. Wagner, M.2001, Genes & Development, 15, 286.

34. Jacks, T., Remington, L., Williams, B. O., Schmitt, E. M., Halachmi, S., Bronson, R. T., Weinberg, R. A.1994, Curr Biol, 4, 1.

35. Bardeesy, N., Morgan, J., Sinha, M., Signoretti, S., Srivastava, S., Loda, M., Merlino, G., DePinho, R. A.2002, Molecular and Cellular Biology, 22, 635.

36. Tuveson, D. A.2006, Cancer Research, 66, 242.

37. Gu, G., Dubauskaite, J., Melton, D. A.2002, Development, 129, 2447.

38. Kawaguchi, Y., Cooper, B., Gannon, M., Ray, M., MacDonald, R. J., Wright, C. V. E.2002, Nature Genetics, 32, 128.

39. Wu, K. L., Gannon, M., Peshavaria, M., Offield, M. F., Henderson, E., Ray, M., Marks, A., Gamer, L. W., Wright, C. V., Stein, R.1997, Mol Cell Biol, 17, 6002.

40. Jackson, E. L., Willis, N., Mercer, K., Bronson, R. T., Crowley, D., Montoya, R., Jacks, T., Tuveson, D. A.2001, Genes & Development, 15, 3243.

41. Hingorani, S. R., Petricoin, E. F., Maitra, A., Rajapakse, V., King, C., Jacobetz, M. A., Ross, S., Conrads, T. P., Veenstra, T. D., Hitt, B. A., Kawaguchi, Y., Johann, D., Liotta, L. A., Crawford, H. C., Putt, M. E., Jacks, T., Wright, C. V., Hruban, R. H., Lowy, A. M., Tuveson, D. A.2003, Cancer Cell, 4, 437.

42. Aguirre, A. J.2003, Genes & Development, 17, 3112.

43. Bardeesy, N., Aguirre, A. J., Chu, G. C., Cheng, K. H., Lopez, L. V., Hezel, A. F., Feng, B., Brennan, C., Weissleder, R., Mahmood, U., Hanahan, D., Redston, M. S., Chin, L., Depinho, R. A.2006, Proc Natl Acad Sci U S A, 103, 5947.

44. Hingorani, S. R., Wang, L., Multani, A. S., Combs, C., Deramaudt, T. B., Hruban, R. H., Rustgi, A. K., Chang, S., Tuveson, D. A.2005, Cancer Cell, 7, 469 45. Bardeesy, N., Cheng, K. H., Berger, J. H., Chu, G. C., Pahler, J., Olson, P., Hezel, A. F., Horner, J., Lauwers, G. Y., Hanahan, D., DePinho, R. A.2006, Genes Dev, 20, 3130.

46. Izeradjene, K., Combs, C., Best, M., Gopinathan, A., Wagner, A., Grady, W. M., Deng, C. X., Hruban, R. H., Adsay, N. V., Tuveson, D. A., Hingorani, S. R.2007, Cancer Cell, 11, 229.

47. Kojima, K., Vickers, S. M., Adsay, N. V., Jhala, N. C., Kim, H. G., Schoeb, T. R., Grizzle, W. E., Klug, C. A.2007, Cancer Res, 67, 8121.

48. Morton, J. P., Jamieson, N. B., Karim, S. A., Athineos, D., Ridgway, R. A., Nixon, C., McKay, C. J., Carter, R., Brunton, V. G., Frame, M. C.2010, Gastroenterology, 139, 586.

49. Ijichi, H., Chytil, A., Gorska, A. E., Aakre, M. E., Fujitani, Y., Fujitani, S., Wright, C. V., Moses, H. L.2006, Genes Dev, 20, 3147.

50. Vincent, D. F., Yan, K. P., Treilleux, I., Gay, F., Arfi, V., Kaniewski, B., Marie, J. C., Lepinasse, F., Martel, S., Goddard-Leon, S., Iovanna, J. L., Dubus, P.,

88

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

Laurent Bartholin

Garcia, S., Puisieux, A., Rimokh, R., Bardeesy, N., Scoazec, J. Y., Losson, R., Bartholin, L.2009, PLoS Genet, 5, e1000575.

Nolan-Stevaux, O., Lau, J., Truitt, M. L., Chu, G. C., Hebrok, M., FernandezZapico, M. E., Hanahan, D.2009, Genes Dev, 23, 24.

Skoulidis, F., Cassidy, L. D., Pisupati, V., Jonasson, J. G., Bjarnason, H., Eyfjord, J. E., Karreth, F. A., Lim, M., Barber, L. M., Clatworthy, S. A., Davies, S. E., Olive, K. P., Tuveson, D. A., Venkitaraman, A. R.2010, Cancer Cell, 18, 499.

De La O, J.-P., Emerson, L. L., Goodman, J. L., Froebe, S. C., Illum, B. E., Curtis, A. B., Murtaugh, L. C.2008, Proc Natl Acad Sci U S A., 105, 18907.

Heiser, P. W., Cano, D. A., Landsman, L., Kim, G. E., Kench, J. G., Klimstra, D. S., Taketo, M. M., Biankin, A. V., Hebrok, M.2008, Gastroenterology, 135, 1288.

Krantz, S. B., Shields, M. A., Dangi-Garimella, S., Cheon, E., Barron, M. R., Hwang, R. F., Rao, M. S., Grippo, P. J., Bentrem, D. J., Munshi, H. G.2011, Mol Cancer Res, Lang, G. A., Iwakuma, T., Suh, Y. A., Liu, G., Rao, V. A., Parant, J. M., Valentin-Vega, Y. A., Terzian, T., Caldwell, L. C., Strong, L. C., El-Naggar, A. K., Lozano, G.2004, Cell, 119, 861.

Olive, K. P., Tuveson, D. A., Ruhe, Z. C., Yin, B., Willis, N. A., Bronson, R. T., Crowley, D., Jacks, T.2004, Cell, 119, 847.

Donehower, L. A., Harvey, M., Slagle, B. L., McArthur, M. J., Montgomery, C. A., Jr., Butel, J. S., Bradley, A.1992, Nature, 356, 215.

Purdie, C. A., Harrison, D. J., Peter, A., Dobbie, L., White, S., Howie, S. E., Salter, D. M., Bird, C. C., Wyllie, A. H., Hooper, M. L., et al.1994, Oncogene, 9, 603.

Hezel, A. F., Gurumurthy, S., Granot, Z., Swisa, A., Chu, G. C., Bailey, G., Dor, Y., Bardeesy, N., Depinho, R. A.2008, Mol Cell Biol, 28, 2414.

Stanger, B. Z., Stiles, B., Lauwers, G. Y., Bardeesy, N., Mendoza, M., Wang, Y., Greenwood, A., Cheng, K.-h., McLaughlin, M., Brown, D.2005, Cancer Cell, 8, 185.

Jorgensen, M. C., Ahnfelt-Ronne, J., Hald, J., Madsen, O. D., Serup, P., Hecksher-Sorensen, J.2007, Endocr Rev, 28, 685.

Guerra, C., Schuhmacher, A. J., Canamero, M., Grippo, P. J., Verdaguer, L., Perez-Gallego, L., Dubus, P., Sandgren, E. P., Barbacid, M.2007, Cancer Cell, 11, 291.

Dor, Y., Brown, J., Martinez, O. I., Melton, D. A.2004, Nature, 429, 41.

Zhou, Q., Brown, J., Kanarek, A., Rajagopal, J., Melton, D. A.2008, Nature, 455, 627.

Habbe, N., Shi, G., Meguid, R. A., Fendrich, V., Esni, F., Chen, H., Feldmann, G., Stoffers, D. A., Konieczny, S. F., Leach, S. D., Maitra, A.2008, Proceedings of the National Academy of Sciences, 105, 18913.

Friedlander, S. Y. G., Chu, G. C., Snyder, E. L., Girnius, N., Dibelius, G., Crowley, D., Vasile, E., DePinho, R. A., Jacks, T.2009, Cancer Cell, 16, 379.

Modeling pancreatic cancer in mice

89

68. Premsrirut, P. K., Dow, L. E., Kim, S. Y., Camiolo, M., Malone, C. D., Miething, C., Scuoppo, C., Zuber, J., Dickins, R. A., Kogan, S. C., Shroyer, K. R., Sordella, R., Hannon, G. J., Lowe, S. W.2011, Cell, 145, 145.

69. Brembeck, F. H., Schreiber, F. S., Deramaudt, T. B., Craig, L., Rhoades, B., Swain, G., Grippo, P., Stoffers, D. A., Silberg, D. G., Rustgi, A. K.2003, Cancer Research, 63, 2005.

70. Furuyama, K., Kawaguchi, Y., Akiyama, H., Horiguchi, M., Kodama, S., Kuhara, T., Hosokawa, S., Elbahrawy, A., Soeda, T., Koizumi, M., Masui, T., Kawaguchi, M., Takaori, K., Doi, R., Nishi, E., Kakinoki, R., Deng, J. M., Behringer, R. R., Nakamura, T., Uemoto, S.2011, Nat Genet, 43, 34.

71. Direkze, N. C., Hodivala-Dilke, K., Jeffery, R., Hunt, T., Poulsom, R., Oukrif, D., Alison, M. R., Wright, N. A.2004, Cancer Res, 64, 8492.

72. Yauch, R. L., Gould, S. E., Scales, S. J., Tang, T., Tian, H., Ahn, C. P., Marshall, D., Fu, L., Januario, T., Kallop, D., Nannini-Pepe, M., Kotkow, K., Marsters, J. C., Rubin, L. L., de Sauvage, F. J.2008, Nature, 455, 406.

73. Paget, S.1989, Cancer Metastasis Rev, 8, 98.

74. Raimondi, S., Lowenfels, A. B., Morselli-Labate, A. M., Maisonneuve, P., Pezzilli, R.2010, Best Pract Res Clin Gastroenterol, 24, 349.

75. Lampel, M., Kern, H. F.1977, Virchows Arch A Pathol Anat Histol, 373, 97.

76. Adler, G., Hupp, T., Kern, H. F.1979, Virchows Arch A Pathol Anat Histol, 382, 31.

77. Steinle, A. U., Weidenbach, H., Wagner, M., Adler, G., Schmid, R. M.1999, Gastroenterology, 116, 420.

78. Mizunuma, T., Kawamura, S., Kishino, Y.1984, J Nutr, 114, 467.

79. Su, K. H., Cuthbertson, C., Christophi, C.2006, HPB (Oxford), 8, 264.

80. Ohmuraya, M., Hirota, M., Araki, M., Mizushima, N., Matsui, M., Mizumoto, T., Haruna, K., Kume, S., Takeya, M., Ogawa, M., Araki, K., Yamamura, K.2005, Gastroenterology, 129, 696.

81. Ohmuraya, M., Hirota, M., Araki, K., Baba, H., Yamamura, K.2006, Pancreas, 33, 104.

82. Selig, L., Sack, U., Gaiser, S., Kloppel, G., Savkovic, V., Mossner, J., Keim, V., Bodeker, H.2006, BMC Gastroenterol, 6, 30.

83. Archer, H., Jura, N., Keller, J., Jacobson, M., Barsagi, D.2006, Gastroenterology, 131, 1844.

84. Marrache, F., Tu, S. P., Bhagat, G., Pendyala, S., Osterreicher, C. H., Gordon, S., Ramanathan, V., Penz-Osterreicher, M., Betz, K. S., Song, Z., Wang, T. C.2008, Gastroenterology, 135, 1277.

85. Ji, B., Tsou, L., Wang, H., Gaiser, S., Chang, D. Z., Daniluk, J., Bi, Y., Grote, T., Longnecker, D. S., Logsdon, C. D.2009, Gastroenterology, 137, 1072.

86. Vincent, D. F., Arfi, V. A., Kaniewski, B., Martel, S., Goddard-Leon, S., Rimokh, R., Dubus, P., Treilleux, I., Bartholin, L.In preparation, 87. De La, O. J., Murtaugh, L. C.2009, Cell Cycle, 8, 1860.

88. Carriere, C., Young, A. L., Gunn, J. R., Longnecker, D. S., Korc, M.2009, Biochem Biophys Res Commun, 382, 561.

90

Laurent Bartholin

89. Morris, J. P. t., Cano, D. A., Sekine, S., Wang, S. C., Hebrok, M.2010, J Clin Invest, 120, 508.

90. Lesina, M., Kurkowski, M. U., Ludes, K., Rose-John, S., Treiber, M., Kloppel, G., Yoshimura, A., Reindl, W., Sipos, B., Akira, S., Schmid, R. M., Algul, H.2011, Cancer Cell, 19, 456.

91. Fukuda, A., Wang, S. C., Morris, J. P. t., Folias, A. E., Liou, A., Kim, G. E., Akira, S., Boucher, K. M., Firpo, M. A., Mulvihill, S. J., Hebrok, M.2011, Cancer Cell, 19, 441.

92. Bachem, M. G., Schunemann, M., Ramadani, M., Siech, M., Beger, H., Buck, A., Zhou, S., Schmid-Kotsas, A., Adler, G.2005, Gastroenterology, 128, 907.

93. Neesse, A., Wagner, M., Ellenrieder, V., Bachem, M., Gress, T. M., Buchholz, M.2007, Pancreatology, 7, 380.

94. Schneiderhan, W., Diaz, F., Fundel, M., Zhou, S., Siech, M., Hasel, C., Moller, P., Gschwend, J. E., Seufferlein, T., Gress, T., Adler, G., Bachem, M. G.2007, J Cell Sci, 120, 512.

95. Hwang, R. F., Moore, T., Arumugam, T., Ramachandran, V., Amos, K. D., Rivera, A., Ji, B., Evans, D. B., Logsdon, C. D.2008, Cancer Res, 68, 918.

96. Vonlaufen, A., Joshi, S., Qu, C., Phillips, P. A., Xu, Z., Parker, N. R., Toi, C. S., Pirola, R. C., Wilson, J. S., Goldstein, D., Apte, M. V.2008, Cancer Res, 68, 2085.

97. Kraman, M., Bambrough, P. J., Arnold, J. N., Roberts, E. W., Magiera, L., Jones, J. O., Gopinathan, A., Tuveson, D. A., Fearon, D. T.2010, Science, 330, 827.

98. Kang, Y., Siegel, P. M., Shu, W., Drobnjak, M., Kakonen, S. M., Cordon-Cardo, C., Guise, T. A., Massague, J.2003, Cancer Cell, 3, 537.

99. Zhang, X. H., Wang, Q., Gerald, W., Hudis, C. A., Norton, L., Smid, M., Foekens, J. A., Massague, J.2009, Cancer Cell, 16, 67.

100. Lu, X., Wang, Q., Hu, G., Van Poznak, C., Fleisher, M., Reiss, M., Massague, J., Kang, Y.2009, Genes Dev, 23, 1882.

101. Minn, A. J., Gupta, G. P., Siegel, P. M., Bos, P. D., Shu, W., Giri, D. D., Viale, A., Olshen, A. B., Gerald, W. L., Massague, J.2005, Nature, 436, 518.

102. Gupta, G. P., Nguyen, D. X., Chiang, A. C., Bos, P. D., Kim, J. Y., Nadal, C., Gomis, R. R., Manova-Todorova, K., Massague, J.2007, Nature, 446, 765.

103. Padua, D., Zhang, X. H., Wang, Q., Nadal, C., Gerald, W. L., Gomis, R. R., Massague, J.2008, Cell, 133, 66.

104. Nguyen, D. X., Chiang, A. C., Zhang, X. H., Kim, J. Y., Kris, M. G., Ladanyi, M., Gerald, W. L., Massague, J.2009, Cell, 138, 51.

105. Bos, P. D., Zhang, X. H. F., Nadal, C., Shu, W., Gomis, R. R., Nguyen, D. X., Minn, A. J., van de Vijver, M. J., Gerald, W. L., Foekens, J. A., MassaguĂŠ, J.2009, Nature, 459, 1005.

106. Landi, S.2009, Mutation Research/Reviews in Mutation Research, 681, 299.

107. Cheon, E. C., Strouch, M. J., Barron, M. R., Ding, Y., Melstrom, L. G., Krantz, S. B., Mullapudi, B., Adrian, K., Rao, S., Adrian, T. E., Bentrem, D. J., Grippo, P. J.2011, Int J Cancer, 128, 2783.

108. Lankisch, P. G.2009, Pancreas, 38, 494.

Modeling pancreatic cancer in mice

91

109. Wendt, L. R., Osvaldt, A. B., Bersch, V. P., Schumacher Rde, C., Edelweiss, M. I., Rohde, L.2007, Acta Cir Bras, 22, 202.

110. Bersch, V. P., Osvaldt, A. B., Edelweiss, M. I., Schumacher Rde, C., Wendt, L. R., Abreu, L. P., Blom, C. B., Abreu, G. P., Costa, L., Piccinini, P., Rohde, L.2009, Pancreas, 38, 65.

111. Song, Z., Bhagat, G., Quante, M., Baik, G. H., Marrache, F., Tu, S. P., Zhao, C. M., Chen, D., Dannenberg, A. J., Wang, T. C.2010, Lab Invest, 90, 426.

112. Hruban, R. H., Canto, M. I., Goggins, M., Schulick, R., Klein, A. P.2010, Adv Surg, 44, 293.

113. Bruns, C. J., Shrader, M., Harbison, M. T., Portera, C., Solorzano, C. C., Jauch, K. W., Hicklin, D. J., Radinsky, R., Ellis, L. M.2002, Int J Cancer, 102, 101.

114. Bocci, G., Danesi, R., Marangoni, G., Fioravanti, A., Boggi, U., Esposito, I., Fasciani, A., Boschi, E., Campani, D., Bevilacqua, G., Mosca, F., Del Tacca, M.2004, Eur J Pharmacol, 498, 9.

115. Kindler, H. L., Friberg, G., Singh, D. A., Locker, G., Nattam, S., Kozloff, M., Taber, D. A., Karrison, T., Dachman, A., Stadler, W. M., Vokes, E. E.2005, J Clin Oncol, 23, 8033.

116. Kindler, H. L., Friberg, G., Skoog, L., Wade-Oliver, K., Vokes, E. E.2005, Am J Clin Oncol, 28, 340.

117. Singh, M., Lima, A., Molina, R., Hamilton, P., Clermont, A. C., Devasthali, V., Thompson, J. D., Cheng, J. H., Bou Reslan, H., Ho, C. C., Cao, T. C., Lee, C. V., Nannini, M. A., Fuh, G., Carano, R. A., Koeppen, H., Yu, R. X., Forrest, W. F., Plowman, G. D., Johnson, L.2010, Nat Biotechnol, 28, 585.

118. Beatty, G. L., Chiorean, E. G., Fishman, M. P., Saboury, B., Teitelbaum, U. R., Sun, W., Huhn, R. D., Song, W., Li, D., Sharp, L. L., Torigian, D. A., O'Dwyer, P. J., Vonderheide, R. H.2011, Science, 331, 1612.

119. Dosch, J. S., Pasca di Magliano, M., Simeone, D. M.2010, Pancreatology, 10, 151.

120. Thayer, S. P., di Magliano, M. P., Heiser, P. W., Nielsen, C. M., Roberts, D. J., Lauwers, G. Y., Qi, Y. P., Gysin, S., Fernandez-del Castillo, C., Yajnik, V., Antoniu, B., McMahon, M., Warshaw, A. L., Hebrok, M.2003, Nature, 425, 851.

121. Mao, J., Ligon, K. L., Rakhlin, E. Y., Thayer, S. P., Bronson, R. T., Rowitch, D., McMahon, A. P.2006, Cancer Res, 66, 10171.

122. Morton, J. P., Mongeau, M. E., Klimstra, D. S., Morris, J. P., Lee, Y. C., Kawaguchi, Y., Wright, C. V., Hebrok, M., Lewis, B. C.2007, Proc Natl Acad Sci U S A, 104, 5103.

123. Pasca di Magliano, M., Sekine, S., Ermilov, A., Ferris, J., Dlugosz, A. A., Hebrok, M.2006, Genes Dev, 20, 3161.

124. Bailey, J. M., Swanson, B. J., Hamada, T., Eggers, J. P., Singh, P. K., Caffery, T., Ouellette, M. M., Hollingsworth, M. A.2008, Clin Cancer Res, 14, 5995.

125. Tian, H., Callahan, C. A., DuPree, K. J., Darbonne, W. C., Ahn, C. P., Scales, S. J., de Sauvage, F. J.2009, Proceedings of the National Academy of Sciences, 106, 4254.

126. Bailey, J. M., Mohr, A. M., Hollingsworth, M. A.2009, Oncogene, 28, 3513.

92

Laurent Bartholin

127. Berman, D. M., Karhadkar, S. S., Maitra, A., Montes De Oca, R., Gerstenblith, M. R., Briggs, K., Parker, A. R., Shimada, Y., Eshleman, J. R., Watkins, D. N., Beachy, P. A.2003, Nature, 425, 846.

128. Feldmann, G., Dhara, S., Fendrich, V., Bedja, D., Beaty, R., Mullendore, M., Karikari, C., Alvarez, H., Iacobuzio-Donahue, C., Jimeno, A., Gabrielson, K. L., Matsui, W., Maitra, A.2007, Cancer Res, 67, 2187.

129. Feldmann, G., Fendrich, V., McGovern, K., Bedja, D., Bisht, S., Alvarez, H., Koorstra, J. B., Habbe, N., Karikari, C., Mullendore, M., Gabrielson, K. L., Sharma, R., Matsui, W., Maitra, A.2008, Mol Cancer Ther, 7, 2725.

130. Feldmann, G., Habbe, N., Dhara, S., Bisht, S., Alvarez, H., Fendrich, V., Beaty, R., Mullendore, M., Karikari, C., Bardeesy, N., Ouellette, M. M., Yu, W., Maitra, A.2008, Gut, 57, 1420.

131. Olive, K. P., Jacobetz, M. A., Davidson, C. J., Gopinathan, A., McIntyre, D., Honess, D., Madhu, B., Goldgraben, M. A., Caldwell, M. E., Allard, D., Frese, K. K., Denicola, G., Feig, C., Combs, C., Winter, S. P., Ireland-Zecchini, H., Reichelt, S., Howat, W. J., Chang, A., Dhara, M., Wang, L., Ruckert, F., Grutzmann, R., Pilarsky, C., Izeradjene, K., Hingorani, S. R., Huang, P., Davies, S. E., Plunkett, W., Egorin, M., Hruban, R. H., Whitebread, N., McGovern, K., Adams, J., Iacobuzio-Donahue, C., Griffiths, J., Tuveson, D. A.2009, Science, 324, 1457.

132. Rifai, Y., Arai, M. A., Koyano, T., Kowithayakorn, T., Ishibashi, M.2010, J Nat Prod, 73, 995.

133. Hahn, S. A., Schutte, M., Hoque, A. T., Moskaluk, C. A., da Costa, L. T., Rozenblum, E., Weinstein, C. L., Fischer, A., Yeo, C. J., Hruban, R. H., Kern, S. E.1996, Science, 271, 350.

134. Hansel, D. E., Kern, S. E., Hruban, R. H.2003, Annu Rev Genomics Hum Genet, 4, 237.

135. Kuang, C., Xiao, Y., Liu, X., Stringfield, T. M., Zhang, S., Wang, Z., Chen, Y.2006, Proc Natl Acad Sci U S A, 103, 1858.

136. Dupont, S., Zacchigna, L., Cordenonsi, M., Soligo, S., Adorno, M., Rugge, M., Piccolo, S.2005, Cell, 121, 87.

137. He, W., Dorn, D. C., Erdjument-Bromage, H., Tempst, P., Moore, M. A., Massague, J.2006, Cell, 125, 929.

138. Dupont, S., Mamidi, A., Cordenonsi, M., Montagner, M., Zacchigna, L., Adorno, M., Martello, G., Stinchfield, M. J., Soligo, S., Morsut, L., Inui, M., Moro, S., Modena, N., Argenton, F., Newfeld, S. J., Piccolo, S.2009, Cell, 136, 123.

139. Doisne, J. M., Bartholin, L., Yan, K. P., Garcia, C. N., Duarte, N., Le Luduec, J. B., Vincent, D., Cyprian, F., Horvat, B., Martel, S., Rimokh, R., Losson, R., Benlagha, K., Marie, J. C.2009, J Exp Med, 206, 1365.

140. Vincent, D. F., Gout, J., Chuvin, N., Arfi, V., Pommier, R., Bertolino, P., Jonckheere, N., Ripoche, D., Kaniewski, B., Martel, S., J.B., L., S., G.-L., Colombe, A., Janier, M., Van Seuningen, I., Valcourt, U., Treilleux, I., Dubus, P., Bardeesy, N., Bartholin, L.Submitted,

Modeling pancreatic cancer in mice

93

141. Bartholin, L., Cyprian, F. S., Vincent, D., Garcia, C. N., Martel, S., Horvat, B., Berthet, C., Goddard-LĂŠon, S., Treilleux, I., Rimokh, R., Marie, J. C.2008, Genesis, 46, 724.

142. Vincent, D. F., Kaniewski, B., Powers, S. E., Havenar-Daughton, C., Marie, J. C., Wotton, D., Bartholin, L.2010, Genesis, 48, 559.

143. Vincent, D. F., Bartholin, L.2009, FASEB Summer Research Conferences, "The TGFβ Superfamily: Signaling in Development and Disease", Carefree, Arizona, 144. Giampieri, S., Manning, C., Hooper, S., Jones, L., Hill, C. S., Sahai, E.2009, Nature Cell Biology, 11, 1287.

145. Meulmeester, E., Ten Dijke, P.2011, J Pathol, 223, 205.

146. Muraoka, R. S., Koh, Y., Roebuck, L. R., Sanders, M. E., Brantley-Sieders, D., Gorska, A. E., Moses, H. L., Arteaga, C. L.2003, Mol Cell Biol, 23, 8691.

147. Muraoka-Cook, R. S., Dumont, N., Arteaga, C. L.2005, Clin Cancer Res, 11, 937s.

148. Muraoka-Cook, R. S., Shin, I., Yi, J. Y., Easterly, E., Barcellos-Hoff, M. H., Yingling, J. M., Zent, R., Arteaga, C. L.2006, Oncogene, 25, 3408.

149. Siegel, P. M.2003, Proceedings of the National Academy of Sciences, 100, 8430 150. Rowland-Goldsmith, M. A., Maruyama, H., Matsuda, K., Idezawa, T., Ralli, M., Ralli, S., Korc, M.2002, Mol Cancer Ther, 1, 161.

151. Gaspar, N. J., Li, L., Kapoun, A. M., Medicherla, S., Reddy, M., Li, G., O'Young, G., Quon, D., Henson, M., Damm, D. L., Muiru, G. T., Murphy, A., Higgins, L. S., Chakravarty, S., Wong, D. H.2007, Mol Pharmacol, 72, 152.

152. Melisi, D., Ishiyama, S., Sclabas, G. M., Fleming, J. B., Xia, Q., Tortora, G., Abbruzzese, J. L., Chiao, P. J.2008, Mol Cancer Ther, 7, 829.

153. Kano, M. R., Bae, Y., Iwata, C., Morishita, Y., Yashiro, M., Oka, M., Fujii, T., Komuro, A., Kiyono, K., Kaminishi, M., Hirakawa, K., Ouchi, Y., Nishiyama, N., Kataoka, K., Miyazono, K.2007, Proc Natl Acad Sci U S A, 104, 3460.

154. Kano, M. R., Komuta, Y., Iwata, C., Oka, M., Shirai, Y. T., Morishita, Y., Ouchi, Y., Kataoka, K., Miyazono, K.2009, Cancer Sci, 100, 173.

155. Schlingensiepen, K. H., Jaschinski, F., Lang, S. A., Moser, C., Geissler, E. K., Schlitt, H. J., Kielmanowicz, M., Schneider, A.2011, Cancer Sci.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 95-110 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

5. Epithelial-mesenchymal transition and pancreatic cancer progression 1

Surabhi Dangi-Garimella1, Seth B. Krantz2, Mario A. Shields1 Paul J. Grippo2,4 and Hidayatullah G. Munshi1,3,4

Division of Hematology/Oncology, Department of Medicine and 2Division of Surgical Oncology Department of Surgery, Feinberg School of Medicine, Northwestern University; 3The Jesse Brown VA Medical Center; and 4The Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago, IL 60611, USA

Abstract. Pancreatic ductal adenocarcinoma (PDAC) continues to be one of the most lethal human malignancies, with median survival of less than one year and overall 5-year survival of less than 5%.

There is increasing evidence for contribution of epithelialmesenchymal transition (EMT) to pancreatic cancer metastasis and to treatment resistance. In this chapter we will review the role of EMT in pancreatic cancer progression, focusing particularly on the transcription factors and microRNAs involved in EMT. We will examine how EMT is involved in the generation and maintenance of stem cells, and detail the role of EMT in modulating resistance of PDAC cells to drug therapies. Finally, we will identify putative EMT-targeting agents that may help to reduce the morbidity and mortality associated with pancreatic cancer.

Correspondence/Reprint request: Dr. Surabhi Dangi-Garimella, Hidayatullah G. Munshi, Division of Hematology/Oncology, Department of Medicine, Chicago, IL 60611, USA E-mails: s-dangi-garimella@northwestern.edu or h-munshi@northwestern.edu

96

Surabhi Dangi-Garimella et al.

1. Introduction Despite the use of aggressive combination therapy in highly selected patients (1), pancreatic ductal adenocarcinoma (PDAC) continues to be one of the most lethal human malignancies and remains a daunting challenge for patients, clinicians, and researchers alike. There are approximately 43,000 new cases each year in the United States, with over 36,000 deaths, making it the fourth leading cause of cancer death (2). Median survival is less than one year and overall 5-year survival is less than 5% (3, 4). Additionally, over 80% of patients present with advanced disease not amenable to surgical resection, and even for those who do undergo surgery, treatment remains difficult with a 5-year survival of only 20% (5-7). Several factors are thought to contribute to the aggressive nature of pancreatic cancer. Anatomically, the location of the pancreas means patients are often asymptomatic until the disease is advanced, when they present with jaundice from obstruction of the bile ducts, or pain from invasion of the surrounding nerves. Histologically, PDAC is associated with a dense fibrotic reaction, known as the desmoplastic reaction, which is thought to contribute to disease progression and chemoresistance (8, 9). Despite improvements in surgical technique, enhanced imaging, and new chemotherapeutic agents, outcomes for patients remain extremely poor, and a better understanding of the cellular and biochemical factors that contribute to this terrible disease is essential if we are to make any significant improvements in the treatment of PDAC.

2. Epithelial-mesenchymal transition and PDAC Epithelial-mesenchymal transition (EMT) is a developmental process that allows cells that are a part of a rigid architecture to escape and spread to distant sites (10-12). As cells undergo EMT, they lose their epithelial features including loss of their sheet-like architecture, loss of polarity, and down regulation of E-cadherin (Fig. 1). The cells also develop a mesenchymal phenotype, taking on a spindle-like, fusiform morphology, become motile, and start expressing mesenchymal markers, e.g. N-cadherin, fibronectin, vimentin (11, 13). Although the features of EMT were initially characterized in vitro, studies from patients with a variety of cancers have provided evidence for EMT in vivo (14-17). In human pancreatic tumor samples, fibronectin and vimentin are increased in high-grade tumors and within poorly differentiated areas of low-grade tumors, with a corresponding decrease in E-cadherin expression. Significantly, these patients have worse survival than those patients whose tumors demonstrate less evidence of EMT. In a study based on a rapid autopsy program for patients with pancreatic cancer,

Epithelial-mesenchymal transition and pancreatic cancer progression

97

Figure 1. EMT regulation in pancreatic cancer. Some of the main drivers of EMT in pancreatic cancer are transcription factors Snail, Slug and Zeb1, which are in turn regulated by cytokines (NF-α) and growth factors (TGF-β) as well as microRNAs. These signaling molecules transform non-invasive and chemo-sensitive cells into motile and invasive chemo-resistant cells that have stem-cell-like properties.

75% of the primary tumors with mesenchymal features developed metastatic lesions to liver and lung (14).

The key regulators of EMT include Snail, Slug, Zeb1, and Twist, which are zinc finger transcription factors that repress genes responsible for the epithelial phenotype (10-12). In resected PDAC specimens nearly 80% have moderate to strong Snail expression while only 50% show similar Slug expression, with very few having strong Twist expression (15). Snail expression is inversely correlated with E-cadherin expression, with decreased E-cadherin expression associated with higher tumor grade. Similar results were seen in pancreatic cell lines, with poorly differentiated lines showing higher levels of Snail and lower levels of E-cadherin compared with moderately differentiated cell lines (15). Zeb1 expression in pathologic specimens also correlates with advanced tumor grade and worse outcomes (16-18). In one study, tissue microarray analysis of pancreatic cancer showed an inverse relationship between Zeb1 and E-cadherin expression (17).

Furthermore, silencing of Zeb1 in pancreatic cancer cell lines leads to the upregulation of E-cadherin and restoration of an epithelial phenotype (17).

Interestingly, Zeb1 is primarily responsible for the acquisition of an EMT phenotype, along with increased migration and invasion in response to NF-κB signaling in pancreatic cancer cells (18).

Transforming growth factor-β (TGF-β), one of the primary drivers of EMT (9, 11), can increase expression of Snail, Slug and Zeb1 in a variety of cancers. Recently, we published that pancreatic cancer cells on encountering type I collagen induce Snail expression through increased TGF-β signaling.

98

Surabhi Dangi-Garimella et al.

Collagen-induced Snail expression was abrogated using siRNA against TGF-β type I receptor or against Smad4. TGF-β2 expression of Snail in endothelial cells and subsequent endothelial-mesenchymal transition was also shown to involve Smad signaling. However, in contrast to ERK1/2 regulation of TGF-β2-induced Snail expression in endothelial cells, we have found that collagen-induced Snail expression does not involve ERK1/2 signaling and is primarily mediated by Smad signaling. Although TGF-β can promote EMT, it is important to note that TGF-β has both tumor suppressive and tumor promoting affects on pancreatic cancer (19-21). Loss of Smad4 early in tumor development leads to loss of TGF-β growth inhibition and unchecked tumor growth in mouse models of pancreatic cancer. These tumors, however, are generally well differentiated (22, 23). Tumors with intact Smad4 signaling, meanwhile, are associated with an increase in EMT and subsequently are poorly differentiated (23). Furthermore, these advanced tumors that have undergone EMT show increased tumor proliferation and migration in response to TGF-β (23).

Consistent with TGF-β driven EMT leading to tumor proliferation in advanced tumors, EMT is also associated with cancers becoming oncogene independent. Induction of EMT with TFG-β causes previously K-ras dependent cells to become K-ras independent; conversely, K-ras independent cells forced to undergo mesenchymal to epithelial transformation (24) by targeting Zeb1 with shRNA subsequently become K-ras dependent (25). As cells undergo EMT, tumors that once may have responded to interruption of oncogenic signaling pathways may become unresponsive (25), which has important implications for drugs specifically designed to target these growth pathways, such as epidermal growth factor receptor (EGFR) inhibitors.

Inflammation plays a significant role in pancreatic cancer (26, 27), and inflammatory signaling through NF-κB has been shown to increase both EMT and cancer cell invasion. Snail activity is increased via stabilization at the protein level in response to TNF-α driven NF-κB signaling (28). Additionally, knockdown of Snail in this system abrogates TNF-α driven cancer cell migration and invasion (28). A similar interaction between NF-κB and EMT is seen in pancreatic cancer cells following TNF-α treatment. Transfection with a dominant negative form of IκBα abrogates the effect of TNF-α (18).

Interestingly, TGF-β-induced EMT is also dependent on NF-κB signaling (18).

Chronic pancreatitis, which is associated with ongoing inflammation and fibrosis (29), has been identified as a risk factor for pancreatic cancer in humans and contributes to PDAC progression in mouse models of pancreatic cancer (30, 31). In addition, acute pancreatitis can accelerate the progression of precursor pancreatic intraepithelial neoplastic (PanIN) lesions to PDAC in

Epithelial-mesenchymal transition and pancreatic cancer progression

99

mutant K-ras-driven mouse models of pancreatic cancer (32, 33).

Interestingly, although expression of embryonic mutant K-ras has been shown to be sufficient for tumor initiation in various mouse models of pancreatic cancer, expression of K-ras in adult mouse pancreas does not result in any obvious phenotypic changes (31). Induction of chronic pancreatitis is essential for PDAC development in these adult mice (31). Stat3, one of the key mediators of inflammatory signaling (34), was recently shown to be required for initiation and progression of PDAC following cerulein-induced pancreatitis in mutant K-ras mice (32, 33). Increased Stat3 signaling was shown as a result of increased IL-6 expression by both cancer cells and inflammatory cells (32, 33). Importantly, Stat3 signaling is aberrantly activated in human PDAC tumor samples and controls proliferation and invasion of human pancreatic cancer cells (32, 33). Human PDAC patients have increased circulating IL-6 levels and human PDAC tumor specimens stain for increased IL-6 expression (32, 33). Interestingly, Snail can also modulate inflammatory signaling in vivo through upregulation of chemokines and cytokines (35-37). Snail overexpression in keratinocytes increases production of cytokines IL-6, IL-8 and the chemokine CXCL1 (35). Moreover, Snail overexpression in epidermal keratinocytes in a transgenic mouse model promotes cutaneous inflammation that is associated with increased IL-6 production by keratinocytes and increased Stat3 signaling (36).

3. Role of microRNAs in modulating EMT in pancreatic cancer MicroRNAs are small single-stranded non-coding RNAs that have been reported in many cancers, including pancreatic cancer (38-41). They serve as either tumor promoters or suppressors depending on their downstream effects (38-41). MicroRNAs of the miR-200 family (miR-200a, b, c, miR-141 and miR-429) and miR-205 have been identified as key negative regulators of both EMT and the metastatic ability of cancer cells (42, 43). These microRNAs are downregulated in high grade and poorly differentiated tumors, while forced expression of miR-200 microRNAs has been shown to inhibit TGF-β1-induced EMT in MDCK cells. In lung cells, forced miR-200 expression abrogates the ability of the cells to become invasive and metastatic. The miR-200 family targets the key regulators of EMT including Zeb1 and Sip1 (also known as Zeb2), and as such leads to increased E-cadherin levels (42, 43).

Recent surveys of global microRNA expression patterns in pancreatic cancer cell lines have shown that 39 microRNAs, including the miR-200

100

Surabhi Dangi-Garimella et al.

family, are deregulated and have at least 2-fold differential expression in PDAC cell lines compared to control non-transformed pancreatic ductal cell lines (44). Expression of miR-200 family members correlates positively with E-cadherin expression and negatively with the miR-200 target Zeb1 (44).

High levels of miR-200c expression strongly correlate with E-cadherin levels in resected human pancreatic tumor samples and are associated with significantly better survival rates compared with patients whose tumors have low levels of miR-200c expression (45). Interestingly, Zeb1 can also directly suppress transcription of miR-200 family members miR-141 and miR-200c (46), indicating a significant interplay between Zeb1 and miR-200 family microRNAs that contributes to the differentiation state of pancreatic cancer cells. Several microRNAs have been shown to be overexpressed in pancreatic cancer, one of which is miR-21. miR-21 was shown to be overexpressed in 79% of pancreatic cancers as opposed to 27% of chronic pancreatitis (75).

Another study indicated that pancreatic tumors with elevated expression of miR-155, miR-203, miR-210 and miR-222 have a much higher risk of tumorrelated death as compared to tumors with a lower expression (Greither, T., L. F. Grochola, et al. (2010). “Elevated expression of microRNAs 155, 203, 210 and 222 in pancreatic tumors is associated with poorer survival. Int J Cancer 126(1): 73-80.

4. Contribution of EMT to stem cells in pancreatic cancer There is increasing interest in the subpopulation of cells within tumors that have stem cell-like properties (47-49). These cells are frequently associated with metastatic foci and chemoresistance and are increasingly linked to an EMT phenotype (50-52). The resistance to chemotherapy of this subpopulation prevents eradication of cancer, and presents the looming threat of recurrence.

There is ever more evidence to suggest the existence of cancer stem cells in pancreatic cancer (53, 54). These cells are CD44-high/CD24-high and express epithelial specific antigen (ESA) (53). Although the CD44high/CD24-high/ESA-high cells comprise a small population of any particular pancreatic tumor, these cells have the ability to self-renew and reproduce the original tumor heterogeneity. Moreover, gemcitabine-resistant cells isolated from established cancer cell lines were also found to be CD44high (55). Pancreatic cancer stem cells can also be identified by high expression of CD133 and these cells are highly tumorigenic and resistant to standard therapy (56). Recently, using aldehyde dehydrogenase (ALDH) activity as a more specific marker of cancer stem cells, it was shown that ALDH-high cells comprise an even more select subpopulation of cells in

Epithelial-mesenchymal transition and pancreatic cancer progression

101

human pancreatic cancers that are tumorigenic and capable of producing tumors at very low numbers (14). These ALDH-high cells have reduced E-cadherin expression and increased Slug expression (14). Interestingly, overexpression of Snail in pancreatic cancer cells leads to increased ALDH expression. ALDH-high cells with a mesenchymal phenotype have also been found in metastatic lesions of patients with pancreatic cancer (14).

MicroRNAs that are associated with EMT, such as the miR-200 family, also regulate stem cell behavior (57, 58). miR-200c cooperates with other microRNAs to suppress expression of stem cell factors, such as Bmi1, Sox2 and KLF4 in cancer cells and mouse embryonic stem cells (57, 58). Stem cells isolated from a number of different tumor types also show increased expression of miR-21 along with its upstream regulator AP-1, and these molecules were found to be responsible for chemoresistance of the tumors. A small molecular inhibitor of AP-1 and anti-miR-21 is able to sensitize the cells to topotecan and decrease colony formation (59). In glioma cells, ID4mediated suppression of miR-9* results in induction of the stem cell protein SOX2, making the cells chemoresistant by upregulating the ABC family of transport proteins (60).

5. Importance of EMT in enhancing drug resistance in pancreatic cancer Pancreatic cancer remains extremely lethal in large part due to the poor response to existing treatments (61, 62). EMT has been shown to be a significant contributor to chemo-resistance in several cancers, including in pancreatic cancer (17, 63-65). Induction of gemcitabine resistance in previously sensitive cell lines results in development of cells with an EMT phenotype that is associated with an increased migratory and invasive ability compared to gemcitabine sensitive cells (65). Moreover, gene expression profiling of chemoresistant cell lines has shown a strong association between expression of genes associated with EMT and chemotherapy resistance (17).

Specifically, the EMT transcription factor Zeb1 is upregulated in resistant cell lines and correlates with decreased expression of E-cadherin. Silencing of Zeb1 with siRNA causes mesenchymal to epithelial transition (24) and restores chemosensitivity (17). Significantly, maintenance of chemoresistance in cell lines that have undergone EMT is dependent on Notch and NF-ÎşB signaling (64). Inhibition of Notch-2 down regulates Zeb1, Snail and Slug expression, attenuates NF-ÎşB signaling, and reduces the migratory and invasive capacity of the gemcitabine resistant cells. (64). Interestingly, the heparin binding growth factor Midkine, which is overexpressed in

102

Surabhi Dangi-Garimella et al.

chemoresistant PDAC, can interact with and activate Notch-2 to promote EMT (66).

EMT plays a role in modulating resistance not only to traditional chemotherapies, but to targeted biologic therapies as well. Cells that express either mutated E-cadherin, or have high levels of Snail, Zeb1, and vimentin, and thus a mesenchymal phenotype, show significantly decreased growth inhibition in response to treatment with the EGFR inhibitor erlotinib than cells with an epithelial phenotype (16). Interestingly, cells from the same patient have been shown to have differential response to drug treatment, with cells from the primary tumor being responsive while cells isolated from a liver metastases and demonstrating a mesenchymal phenotype being resistant to erlotinib (16).

Transgenic mouse models have established that pancreatic cancer cells may not be inherently chemoresistant (65). The pronounced fibrotic reaction, primarily generated by myofibroblast-like stellate cells (67-69), can limit the delivery of current chemotherapeutic agents to the cancer cells. While quiescent fibroblasts within the microenvironment are activated by TGF-β (70), a significant number of myofibroblasts have in fact been shown to arise from epithelial cells that have undergone EMT (71). In the adult kidney activation of Snail is sufficient to cause renal fibrosis (72), while Hedgehog signaling, which has been shown to contribute to EMT (73), was recently shown in pancreatic cancer to contribute to resistance to gemcitabine through modulation of the tumor microenvironment, specifically by affecting the stroma and type I collagen (8, 14, 64). Thus, EMT may modulate chemoresistance not only within cancer cells themselves, but also by modulating the tumor microenvironment through generation of desmoplastic reaction.

MicroRNAs have also been identified as mediators of chemo-resistance in various cancers. Although we showed that let-7 does not mediate gemcitabine resistance in pancreatic cancer either on 2D surfaces or in 3D collagen microenvironment, other microRNAs have been demonstrated to mediate chemo-resistance. For example, miR-21, which is overexpressed in PDAC tumors and predicts for poor outcome (74), contributes to gemcitabine resistance in part through modulation of PI3-kinase-Akt signaling (75, 76). miR-21 increases pro-survival PI3-kinase signaling through repression of PTEN, upregulates the pro-survival Bcl-2 protein and inhibits the proapoptotic Bax protein in PDAC cells to promote gemcitabine resistance. The effect of miR-21 on PI3-kinase-Akt signaling and Bcl2 is not unique to pancreatic cancer cells as it can also mediate doxorubicin resistance in bladder cancer cells (77). Recently, miR-15a and miR-214 were reported to be dysregulated in pancreatic cancer. miR-15a was identified as a suppressor

Epithelial-mesenchymal transition and pancreatic cancer progression

103

of pancreatic tumor growth while miR-214 was found to promote chemoresistance (78). It has also been hypothesized that microRNAs that contribute to chemo-resistance of pancreatic cancer might be overexpressed in the stem cell population of the cancer, thus helping the cells to resist apoptosis and aid in cancer recurrence.

6. Targeting EMT in pancreatic cancer Given the role of EMT in chemo-resistance and tumor progression specifically targeting EMT could improve the survival rates of pancreatic cancer patients. Although increasing expression of miR-200 family microRNAs could restore the epithelial state and make the tumors more sensitive to therapeutic agents, delivery of microRNAs have yet to be translated to the in vivo environment due to a number of technical barriers related to safety, delivery and efficacy (79, 80). Consequently, there is increasing interest in using compounds that can modulate EMT-inducing microRNAs or transcription factors. Curcumin analogue CDF can restore miR-200 levels and sensitize pancreatic cancer cells to gemcitabine treatment in vitro (81). The naturally occurring flavanoid Silibinin can downregulate Zeb1 and Slug expression and thus attenuate EMT in prostate cancer cells (82). The anti-diabetic drug metformin can also decrease expression of the Zeb1, Twist1 and Slug in breast cancer cells. Metformin also decreases the ability of breast cancer stem cells to form mammospheres through reduction in the CD44-high/CD24-low population (83). Salinomycin, which was discovered as part of a drug screen designed to find compounds effective against EMT, can reduce the population of cancer stem cells.

Clinical trials targeting Hedgehog, Wnt and Notch signaling, known EMT pathways that have been implicated in cancer stem cells and chemoresistance (8, 14, 64), are also underway. For example, the Hedgehog inhibitor GDC-0449 is being evaluated in combination with chemotherapy in patients with metastatic pancreatic cancer (http://clinicaltrials.gov/ct2/show /NCT01088815). GDC-0449 is also being evaluated to determine whether it can specifically target cancer stem cells (http://clinicaltrials.gov/ct2/show/ NCT01195415). There are also a number of ongoing clinical trials using Notch inhibitors in patients with locally advanced or metastatic pancreatic cancer. For example, the Notch inhibitor RO4929097 is being evaluated as both neoadjuvant therapy (http://clinicaltrials.gov/ct2/show/NCT01192763) and in patients with metastatic pancreatic cancer (http://clinicaltrials.gov/ct2/ show/NCT01232829). Finally, the Wnt inhibitor PRI-724 is being evaluated in patients with advanced solid tumors, including in patients with unresectable pancreatic cancer (http://clinicaltrials.gov/ct2/show /NCT01302405).

104

Surabhi Dangi-Garimella et al.

In summary, targeting EMT holds significant promise in treating pancreatic cancer patients. Targeting EMT could contribute to increased sensitivity to standard chemotherapy and to growth factor directed therapies, such as those against EGFR signaling. By attenuating fibrosis, it can also increase delivery of drugs to cancer cells. Targeting EMT can also reduce the population of cancer stem cells that are thought to contribute to metastatic disease and treatment resistance. As our understanding of the role and regulation of EMT in pancreatic cancer increases and as we identify how best to target EMT, we will be able to improve the outcomes of patients with pancreatic cancer.

Acknowledgements This work was supported by National Institutes of Health/National Cancer Institute grant number R01CA126888 (H.G.M.), the Baseball Charities Foundation (M.A.S.) and the Association for Academic Surgery Foundation (S.B.K).

References 1.

2.

3.

4.

5.

6.

7.

8.

Conroy, T., Desseigne, F., Ychou, M., Bouche, O., Guimbaud, R., Becouarn, Y., Adenis, A., Raoul, J.L., Gourgou-Bourgade, S., de la Fouchardiere, C., et al.

FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med 364:1817-1825.

Jemal, A., Siegel, R., Xu, J., and Ward, E. 2010. Cancer statistics, 2010. CA Cancer J Clin 60:277-300.

Hidalgo, M. Pancreatic cancer. N Engl J Med 362:1605-1617.

Vincent, A., Herman, J., Schulick, R., Hruban, R.H., and Goggins, M. Pancreatic cancer. Lancet, 378:607-620.

Altekruse, S.F., Kosary, C.L., Krapcho, M., Neyman, N., Aminou, R., Waldron, W., Ruhl, J., Howlader, N., Tatalovich, Z., Cho, H., et al. 2010. SEER Cancer Statistics Review, 1975-2007, NCI. Bethesda, MD: National Cancer Institute.

Stojadinovic, A., Hoos, A., Brennan, M.F., and Conlon, K.C.P. 2002.

Randomized clinical trials in pancreatic cancer. Surg Oncol Clin N Am 11: 207-229, x.

Winter, J.M., Cameron, J.L., Campbell, K.A., Arnold, M.A., Chang, D.C., Coleman, J., Hodgin, M.B., Sauter, P.K., Hruban, R.H., Riall, T.S., et al. 2006.

1423 pancreaticoduodenectomies for pancreatic cancer: A single-institution experience. J Gastrointest Surg 10:1199-1210; discussion 1210-1191.

Olive, K.P., Jacobetz, M.A., Davidson, C.J., Gopinathan, A., McIntyre, D., Honess, D., Madhu, B., Goldgraben, M.A., Caldwell, M.E., Allard, D., et al.

2009. Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science 324:1457-1461.

Epithelial-mesenchymal transition and pancreatic cancer progression

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

105

Ottaviano, A.J., Sun, L., Ananthanarayanan, V., and Munshi, H.G. 2006.

Extracellular matrix-mediated membrane-type 1 matrix metalloproteinase expression in pancreatic ductal cells is regulated by transforming growth factorbeta1. Cancer Research 66:7032-7040.

Lee, J.M., Dedhar, S., Kalluri, R., and Thompson, E.W. 2006. The epithelialmesenchymal transition: new insights in signaling, development, and disease. J Cell Biol 172:973-981.

Thiery, J.P., Acloque, H., Huang, R.Y.J., and Nieto, M.A. 2009. Epithelialmesenchymal transitions in development and disease. Cell 139:871-890.

Kalluri, R., and Weinberg, R.A. 2009. The basics of epithelial-mesenchymal transition. The Journal of clinical investigation 119:1420-1428.

Yang, A.D., Camp, E.R., Fan, F., Shen, L., Gray, M.J., Liu, W., Somcio, R., Bauer, T.W., Wu, Y., Hicklin, D., et al. 2006. Vascular endothelial growth factor receptor-1 activation mediates epithelial to mesenchymal transition in human pancreatic carcinoma cells. Cancer Research 66:46-51.

Rasheed, Z.A., Yang, J., Wang, Q., Kowalski, J., Freed, I., Murter, C., Hong, S.M., Koorstra, J.-B., Rajeshkumar, N.V., He, X., et al. 2010. Prognostic significance of tumorigenic cells with mesenchymal features in pancreatic adenocarcinoma. J Natl Cancer Inst 102:340-351.

Hotz, B., Arndt, M., Dullat, S., Bhargava, S., Buhr, H.-J., and Hotz, H.G. 2007.

Epithelial to mesenchymal transition: expression of the regulators snail, slug, and twist in pancreatic cancer. Clin Cancer Res 13:4769-4776.

Buck, E., Eyzaguirre, A., Barr, S., Thompson, S., Sennello, R., Young, D., Iwata, K.K., Gibson, N.W., Cagnoni, P., and Haley, J.D. 2007. Loss of homotypic cell adhesion by epithelial-mesenchymal transition or mutation limits sensitivity to epidermal growth factor receptor inhibition. Mol Cancer Ther 6:532-541.

Arumugam, T., Ramachandran, V., Fournier, K.F., Wang, H., Marquis, L., Abbruzzese, J.L., Gallick, G.E., Logsdon, C.D., Mcconkey, D.J., and Choi, W. 2009. Epithelial to Mesenchymal Transition Contributes to Drug Resistance in Pancreatic Cancer. Cancer Research 69:5820-5828.

Maier, H.J., Schmidt-Strassburger, U., Huber, M.A., Wiedemann, E.M., Beug, H., and Wirth, T. 2010. NF-kappaB promotes epithelial-mesenchymal transition, migration and invasion of pancreatic carcinoma cells. Cancer Lett 295:214-228.

Derynck, R., and Zhang, Y.E. 2003. Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature 425:577-584.

Shi, Y., and MassaguĂŠ, J. 2003. Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell 113:685-700.

MassaguĂŠ, J. 2000. How cells read TGF-beta signals. Nat Rev Mol Cell Biol 1:169-178.

Izeradjene, K., Combs, C., Best, M., Gopinathan, A., Wagner, A., Grady, W.M., Deng, C.-X., Hruban, R.H., Adsay, N.V., Tuveson, D.A., et al. 2007.

Kras(G12D) and Smad4/Dpc4 haploinsufficiency cooperate to induce mucinous cystic neoplasms and invasive adenocarcinoma of the pancreas. Cancer Cell 11:229-243.

106

Surabhi Dangi-Garimella et al.

23. Bardeesy, N., Cheng, K.-H., Berger, J.H., Chu, G.C., Pahler, J., Olson, P., Hezel, A.F., Horner, J., Lauwers, G.Y., Hanahan, D., et al. 2006. Smad4 is dispensable for normal pancreas development yet critical in progression and tumor biology of pancreas cancer. Genes Dev 20:3130-3146.

24. Conroy, T., Paillot, B., Francois, E., Bugat, R., Jacob, J.H., Stein, U., Nasca, S., Metges, J.P., Rixe, O., Michel, P., et al. 2005. Irinotecan plus oxaliplatin and leucovorin-modulated fluorouracil in advanced pancreatic cancer--a Groupe Tumeurs Digestives of the Federation Nationale des Centres de Lutte Contre le Cancer study. J Clin Oncol 23:1228-1236.

25. Singh, A., Greninger, P., Rhodes, D., Koopman, L., Violette, S., Bardeesy, N., and Settleman, J. 2009. A gene expression signature associated with "K-Ras addiction" reveals regulators of EMT and tumor cell survival. Cancer Cell 15:489-500.

26. Gidekel Friedlander, S.Y., Chu, G.C., Snyder, E.L., Girnius, N., Dibelius, G., Crowley, D., Vasile, E., Depinho, R.A., and Jacks, T. 2009. Context-dependent transformation of adult pancreatic cells by oncogenic K-Ras. Cancer Cell 16:379-389.

27. Guerra, C., Schuhmacher, A.J., CaĂąamero, M., Grippo, P.J., Verdaguer, L., PĂŠrez-Gallego, L., Dubus, P., Sandgren, E.P., and Barbacid, M. 2007. Chronic pancreatitis is essential for induction of pancreatic ductal adenocarcinoma by KRas oncogenes in adult mice. Cancer Cell 11:291-302.

28. Wu, Y., Deng, J., Rychahou, P.G., Qiu, S., Evers, B.M., and Zhou, B.P. 2009.

Stabilization of Snail by NF-κB Is Required for Inflammation-Induced Cell Migration and Invasion. Cancer Cell 15:416-428.

29. Braganza, J.M., Lee, S.H., McCloy, R.F., and McMahon, M.J. 2011. Chronic pancreatitis. Lancet 377:1184-1197.

30. Lowenfels, A.B., Maisonneuve, P., Cavallini, G., Ammann, R.W., Lankisch, P.G., Andersen, J.R., Dimagno, E.P., Andren-Sandberg, A., and Domellof, L. 1993. Pancreatitis and the risk of pancreatic cancer. International Pancreatitis Study Group. N Engl J Med 328:1433-1437.

31. Guerra, C., Schuhmacher, A.J., Canamero, M., Grippo, P.J., Verdaguer, L., Perez-Gallego, L., Dubus, P., Sandgren, E.P., and Barbacid, M. 2007. Chronic pancreatitis is essential for induction of pancreatic ductal adenocarcinoma by KRas oncogenes in adult mice. Cancer Cell 11:291-302.

32. Lesina, M., Kurkowski, M.U., Ludes, K., Rose-John, S., Treiber, M., Kloppel, G., Yoshimura, A., Reindl, W., Sipos, B., Akira, S., et al. 2011. Stat3/Socs3 activation by IL-6 transsignaling promotes progression of pancreatic intraepithelial neoplasia and development of pancreatic cancer. Cancer Cell 19:456-469.

33. Fukuda, A., Wang, S.C., Morris, J.P.t., Folias, A.E., Liou, A., Kim, G.E., Akira, S., Boucher, K.M., Firpo, M.A., Mulvihill, S.J., et al. 2011. Stat3 and MMP7 Contribute to Pancreatic Ductal Adenocarcinoma Initiation and Progression.

Cancer Cell 19:441-455.

34. Grivennikov, S.I., Greten, F.R., and Karin, M. 2010. Immunity, inflammation, and cancer. Cell 140:883-899.

Epithelial-mesenchymal transition and pancreatic cancer progression

107

35. Lyons, J.G., Patel, V., Roue, N.C., Fok, S.Y., Soon, L.L., Halliday, G.M., and Gutkind, J.S. 2008. Snail up-regulates proinflammatory mediators and inhibits differentiation in oral keratinocytes. Cancer Res 68:4525-4530.

36. Du, F., Nakamura, Y., Tan, T.L., Lee, P., Lee, R., Yu, B., and Jamora, C. 2010.

Expression of snail in epidermal keratinocytes promotes cutaneous inflammation and hyperplasia conducive to tumor formation. Cancer Res 70:10080-10089.

37. Rowe, R.G., Lin, Y., Shimizu-Hirota, R., Hanada, S., Neilson, E.G., Greenson, J.K., and Weiss, S.J. 2011. Hepatocyte-Derived Snail1 Propagates Liver Fibrosis Progression. Mol Cell Biol. 31:2392-2403.

38. Lewis, B.P., Burge, C.B., and Bartel, D.P. 2005. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120:15-20.

39. Lim, L.P., Lau, N.C., Garrett-Engele, P., Grimson, A., Schelter, J.M., Castle, J., Bartel, D.P., Linsley, P.S., and Johnson, J.M. 2005. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 433:769-773.

40. McManus, M.T., and Sharp, P.A. 2002. Gene silencing in mammals by small interfering RNAs. Nat Rev Genet 3:737-747.

41. Dillhoff, M., Wojcik, S.E., and Bloomston, M. 2009. MicroRNAs in solid tumors. J Surg Res 154:349-354.

42. Gregory, P.A., Bert, A.G., Paterson, E.L., Barry, S.C., Tsykin, A., Farshid, G., Vadas, M.A., Khew-Goodall, Y., and Goodall, G.J. 2008. The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1. Nat Cell Biol 10:593-601.

43. Peter, M.E. 2009. Let-7 and miR-200 microRNAs: guardians against pluripotency and cancer progression. Cell Cycle 8:843-852.

44. Kent, O.A., Mullendore, M., Wentzel, E.A., L贸pez-Romero, P., Tan, A.C., Alvarez, H., West, K., Ochs, M.F., Hidalgo, M., Arking, D.E., et al. 2009. A resource for analysis of microRNA expression and function in pancreatic ductal adenocarcinoma cells. Cancer biology & therapy 8:2013-2024.

45. Yu, J., Ohuchida, K., Mizumoto, K., Sato, N., Kayashima, T., Fujita, H., Nakata, K., and Tanaka, M. 2010. MicroRNA, hsa-miR-200c, is an independent prognostic factor in pancreatic cancer and its upregulation inhibits pancreatic cancer invasion but increases cell proliferation. Mol Cancer 9:169.

46. Burk, U., Schubert, J., Wellner, U., Schmalhofer, O., Vincan, E., Spaderna, S., and Brabletz, T. 2008. A reciprocal repression between ZEB1 and members of the miR-200 family promotes EMT and invasion in cancer cells. EMBO Rep 9:582-589.

47. Gupta, P.B., Onder, T.T., Jiang, G., Tao, K., Kuperwasser, C., Weinberg, R.A., and Lander, E.S. 2009. Identification of selective inhibitors of cancer stem cells by high-throughput screening. Cell 138:645-659.

48. Al-Hajj, M., Wicha, M.S., Benito-Hernandez, A., Morrison, S.J., and Clarke, M.F. 2003. Prospective identification of tumorigenic breast cancer cells.

Proceedings of the National Academy of Sciences of the United States of America 100:3983-3988.

108

Surabhi Dangi-Garimella et al.

49. Korkaya, H., and Wicha, M.S. 2010. Cancer stem cells: nature versus nurture.

Nat Cell Biol 12:419-421.

50. Santisteban, M., Reiman, J.M., Asiedu, M.K., Behrens, M.D., Nassar, A., Kalli, K.R., Haluska, P., Ingle, J.N., Hartmann, L.C., Manjili, M.H., et al. 2009.

Immune-induced epithelial to mesenchymal transition in vivo generates breast cancer stem cells. Cancer Research 69:2887-2895.

51. Gupta, P.B., Chaffer, C.L., and Weinberg, R.A. 2009. Cancer stem cells: mirage or reality? Nat Med 15:1010-1012.

52. Mani, S.A., Guo, W., Liao, M.-J., Eaton, E.N., Ayyanan, A., Zhou, A.Y., Brooks, M., Reinhard, F., Zhang, C.C., Shipitsin, M., et al. 2008. The epithelialmesenchymal transition generates cells with properties of stem cells. Cell 133:704-715.

53. Li, C., Heidt, D.G., Dalerba, P., Burant, C.F., Zhang, L., Adsay, V., Wicha, M., Clarke, M.F., and Simeone, D.M. 2007. Identification of pancreatic cancer stem cells. Cancer Research 67:1030-1037.

54. Lee, C.J., Dosch, J., and Simeone, D.M. 2008. Pancreatic cancer stem cells. J Clin Oncol 26:2806-2812.

55. Hong, S.P., Wen, J., Bang, S., Park, S., and Song, S.Y. 2009. CD44-positive cells are responsible for gemcitabine resistance in pancreatic cancer cells. Int J Cancer 125:2323-2331.

56. Hermann, P.C., Huber, S.L., Herrler, T., Aicher, A., Ellwart, J.W., Guba, M., Bruns, C.J., and Heeschen, C. 2007. Distinct populations of cancer stem cells determine tumor growth and metastatic activity in human pancreatic cancer. Cell Stem Cell 1:313-323.

57. Shimono, Y., Zabala, M., Cho, R.W., Lobo, N., Dalerba, P., Qian, D., Diehn, M., Liu, H., Panula, S.P., Chiao, E., et al. 2009. Downregulation of miRNA-200c links breast cancer stem cells with normal stem cells. Cell 138:592-603.

58. Wellner, U., Schubert, J., Burk, U.C., Schmalhofer, O., Zhu, F., Sonntag, A., Waldvogel, B., Vannier, C., Darling, D., zur Hausen, A., et al. 2009. The EMTactivator ZEB1 promotes tumorigenicity by repressing stemness-inhibiting microRNAs. Nat Cell Biol 11:1487-1495.

59. Misawa, A., Katayama, R., Koike, S., Tomida, A., Watanabe, T., and Fujita, N. AP-1-Dependent miR-21 expression contributes to chemoresistance in cancer stem cell-like SP cells. Oncol Res 19:23-33.

60. Jeon, H.M., Sohn, Y.W., Oh, S.Y., Kim, S.H., Beck, S., Kim, S., and Kim, H. ID4 Imparts Chemoresistance and Cancer Stemness to Glioma Cells

by Derepressing miR-9*-Mediated Suppression of SOX2. Cancer Res 71: 3410-3421.

61. Li, Y., VandenBoom, T.G., Kong, D., Wang, Z., Ali, S., Philip, P.A., and Sarkar, F.H. 2009. Up-regulation of miR-200 and let-7 by natural agents leads to the reversal of epithelial-to-mesenchymal transition in gemcitabine-resistant pancreatic cancer cells. Cancer Research 69:6704-6712.

62. Hidalgo, M. 2010. Pancreatic cancer. N Engl J Med 362:1605-1617.

63. Yang, A.D., Fan, F., Camp, E.R., van Buren, G., Liu, W., Somcio, R., Gray, M.J., Cheng, H., Hoff, P.M., and Ellis, L.M. 2006. Chronic oxaliplatin resistance

Epithelial-mesenchymal transition and pancreatic cancer progression

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

109

induces epithelial-to-mesenchymal transition in colorectal cancer cell lines. Clin Cancer Res 12:4147-4153.

Wang, Z., Li, Y., Kong, D., Banerjee, S., Ahmad, A., Azmi, A., Ali, S., Abbruzzese, J.L., Gallick, G.E., and Sarkar, F. 2009. Acquisition of epithelialmesenchymal transition phenotype of gemcitabine-resistant pancreatic cancer cells is linked with activation of the notch signaling pathway. Cancer Research 69:2400-2407.

Shah, A.N., Summy, J.M., Zhang, J., Park, S., Parikh, N., and Gallick, G.E. 2007.

Development and characterization of gemcitabine-resistant pancreatic tumor cells. Annals of Surgical Oncology 14:3629-3637.

Gungor, C., Zander, H., Effenberger, K.E., Vashist, Y.K., Kalinina, T., Izbicki, J.R., Yekebas, E., and Bockhorn, M. Notch signaling activated by replication stress-induced expression of Midkine drives Epithelial-Mesenchymal Transition and Chemoresistance in Pancreatic Cancer. Cancer Res. 71:5009-5019.

Hwang, R.F., Moore, T., Arumugam, T., Ramachandran, V., Amos, K.D., Rivera, A., Ji, B., Evans, D.B., and Logsdon, C.D. 2008. Cancer-associated stromal fibroblasts promote pancreatic tumor progression. Cancer Research 68:918-926.

Apte, M.V., Park, S., Phillips, P.A., Santucci, N., Goldstein, D., Kumar, R.K., Ramm, G.A., Buchler, M., Friess, H., McCarroll, J.A., et al. 2004. Desmoplastic reaction in pancreatic cancer: role of pancreatic stellate cells. Pancreas 29: 179-187.

Menke, A., and Adler, G. 2002. TGFbeta-induced fibrogenesis of the pancreas.

Int J Gastrointest Cancer 31:41-46.

Lewis, M.P., Lygoe, K.A., Nystrom, M.L., Anderson, W.P., Speight, P.M., Marshall, J.F., and Thomas, G.J. 2004. Tumour-derived TGF-beta1 modulates myofibroblast differentiation and promotes HGF/SF-dependent invasion of squamous carcinoma cells. Br J Cancer 90:822-832.

Iwano, M., Plieth, D., Danoff, T.M., Xue, C., Okada, H., and Neilson, E.G. 2002.

Evidence that fibroblasts derive from epithelium during tissue fibrosis. The Journal of clinical investigation 110:341-350.

Boutet, A., De Frutos, C.A., Maxwell, P.H., Mayol, M.J., Romero, J., and Nieto, M.A. 2006. Snail activation disrupts tissue homeostasis and induces fibrosis in the adult kidney. EMBO J 25:5603-5613.

Wang, Z., Li, Y., Kong, D., and Sarkar, F.H. 2010. The role of Notch signaling pathway in epithelial-mesenchymal transition (EMT) during development and tumor aggressiveness. Curr Drug Targets 11:745-751.

Giovannetti, E., Funel, N., Peters, G.J., Del Chiaro, M., Erozenci, L.A., Vasile, E., Leon, L.G., Pollina, L.E., Groen, A., Falcone, A., et al. MicroRNA-21 in pancreatic cancer: correlation with clinical outcome and pharmacologic aspects underlying its role in the modulation of gemcitabine activity. Cancer Res 70:4528-4538.

Dillhoff, M., Liu, J., Frankel, W., Croce, C., and Bloomston, M. 2008.

MicroRNA-21 is overexpressed in pancreatic cancer and a potential predictor of survival. J Gastrointest Surg 12:2171-2176.

110

Surabhi Dangi-Garimella et al.

76. Moriyama, T., Ohuchida, K., Mizumoto, K., Yu, J., Sato, N., Nabae, T., Takahata, S., Toma, H., Nagai, E., and Tanaka, M. 2009. MicroRNA-21 modulates biological functions of pancreatic cancer cells including their proliferation, invasion, and chemoresistance. Mol Cancer Ther. 8:1067-1074.

77. Tao, J., Lu, Q., Wu, D., Li, P., Xu, B., Qing, W., Wang, M., Zhang, Z., and Zhang, W. microRNA-21 modulates cell proliferation and sensitivity to doxorubicin in bladder cancer cells. Oncol Rep 25:1721-1729.

78. Zhang, X.J., Ye, H., Zeng, C.W., He, B., Zhang, H., and Chen, Y.Q. Dysregulation of miR-15a and miR-214 in human pancreatic cancer. J Hematol Oncol 3:46.

79. Brown, B.D., and Naldini, L. 2009. Exploiting and antagonizing microRNA regulation for therapeutic and experimental applications. Nat Rev Genet 10: 578-585.

80. Jackson, A.L., and Linsley, P.S. 2010. Recognizing and avoiding siRNA offtarget effects for target identification and therapeutic application. Nat Rev Drug Discov 9:57-67.

81. Ali, S., Ahmad, A., Banerjee, S., Padhye, S., Dominiak, K., Schaffert, J.M., Wang, Z., Philip, P.A., and Sarkar, F.H. 2010. Gemcitabine sensitivity can be induced in pancreatic cancer cells through modulation of miR-200 and miR-21 expression by curcumin or its analogue CDF. Cancer Research 70:3606-3617.

82. Wu, K., Zeng, J., Li, L., Fan, J., Zhang, D., Xue, Y., Zhu, G., Yang, L., Wang, X., and He, D. 2010. Silibinin reverses epithelial-to-mesenchymal transition in metastatic prostate cancer cells by targeting transcription factors. Oncol Rep 23:1545-1552.

83. Vazquez-Martin, A., Oliveras-Ferraros, C., CufĂ­, S., Del Barco, S., MartinCastillo, B., and Menendez, J.A. 2010. Metformin regulates breast cancer stem cell ontogeny by transcriptional regulation of the epithelial-mesenchymal transition (EMT) status. Cell cycle (Georgetown, Tex) 9.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 111-122 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

6. Pancreatic cancer stem cell and mesenchymal stem cell Shin Hamada and Tooru Shimosegawa Division of Gastroenterology, Tohoku University Graduate School of Medicine Sendai, Japan

Abstract. Pancreatic cancer is one of the most life-threatening cancers and its prognosis has not been improved despite advances in diagnostic and therapeutic strategies. The reasons for resistance against conventional therapy and re-growth of untreatable tumor are now attributed to the existence of cancer stem cells (CSC), which occupy only a small part of the entire cancer tissue. CSC has characteristic features such as chemoresistance, establishment of metastasis and reconstruction of a hierarchical population of tumor cells. Specific surface markers have been identified, enabling the purification of CSC fractions from cell lines or clinical samples, serving to dissect the nature of CSCs. In addition, tumor stromal cells also contribute to the malignant behavior of cancer cells by promoting an invasive phenotype and the development of metastasis. Among those stromal cells, a mesenchymal stem cell (MSC) has unique contributions to the cancer tissue since this type of cell could accumulate in the tumor from distant organs such as bone marrow. MSC also acts as a defender of cancer cells by infiltrating into the tumor. This unique feature is now applied to generating cancer-specific delivery systems by utilizing MSC as a vehicle of therapeutic agents. The tumor microenvironment, which is Correspondence/Reprint request: Dr. Shin Hamada, Division of Gastroenterology, Tohoku University Graduate School of Medicine, Sendai, Japan. E-mail: hamadas@med.tohoku.ac.jp

112

Shin Hamada & Tooru Shimosegawa

created by the complex interaction between cancer cells and stromal cells, yields cancer promoting effects, especially induction of a hypoxic environment. Hypoxiainduced signals activate cellular adaptation machinery such as increased cell survival or enhanced stemness in pancreatic cancer cells. Tumor stromal cells also support engraftment of metastatic nodules which provide a partial CSC niche. Targeting the tumor stroma, CSCs, and MSCs will possibly lead to the development of novel therapeutic methods against pancreatic cancer.

Introduction Pancreatic cancer is a characteristic cancer by its dismal prognosis and resistance against conventional therapies such as chemotherapy or radiotherapy. Only ~20 % of pancreatic cancer patients are eligible for surgical resection, which is the only curative therapy, though the 5-year survival in these patients is less than 25 % [1]. Gemcitabine is a standard chemotherapeutic agent against pancreatic cancer, while the efficacy of treatment stays palliative in patients with unresectable disease [2]. The reasons for such clinical course are due to the regrowth of therapy-resistant tumor after therapeutic interventions. The concept of CSC, which attracts great attention in the past decade, accounts for these phenomena as a small subset of therapy-resistant cells which give rise to the untreatable tumors. In pancreatic cancer, several CSC markers such as CD24/CD44/ESA triple positive [3] or CD133 [4] are identified and these CSC fractions revealed higher tumorigenicity, metastatic capability and chemoresistance.

Pancreatic cancer manifests distinct histological characteristics referred to as a desmoplastic reaction which consists of dense fibrotic stroma surrounding tumor cells [5]. Establishment of this fibrotic tissue results from inflammation caused by cancer cells which activate intra- and/or extra-pancreatic proinflammatory cells [6]. It is noteworthy that the inflammatory process can recruit bone marrow derived cells into the tumor tissue [7,8]. It has been suspected that the existence of a desmoplastic reaction might contribute to pancreatic cancer progression. Recent research uncovered a significant role of the tumor stromal cells (including MSC) in pancreatic cancer such as protection from chemotherapeutic agents [9] or establishment of distant metastasis [10].

Interaction between cancer cells and stromal cells is a dynamic process which involves production of multiple cytokines and activation of various signaling pathways. Stromal cells also contribute to the maintenance of CSC function [11] which is the latest target for novel therapeutic intervention against CSC. This chapter reviews the characteristic feature of CSC in pancreatic cancer, interaction of cancer cells and tumor stromal cells, and contribution of stromal cells during cancer progression.

Pancreatic cancer stem cell and mesenchymal stem cell

113

Normal pancreatic stem cell and pancreatic CSC Recent research has uncovered that tissue patterning of normal organ development requires normal tissue stem cells to give rise to a wide variety of cells [12]. These normal tissue stem cells are capable of self-renewal while providing progenitor cells for asymmetrical division [13]. The CSC concept was established as a counterpart to normal tissue stem cells. As shown in figure 1, normal tissue stem cells give rise to various types of cells. In contrast, CSC has self-renewal capacity and manifests several characteristics such as initiation of metastasis, chemoresistance, and hierarchical reconstitution of the entire tumor. The origin of cancer is possibly attributed to the normal tissue stem cell due to the similarity of normal stem cells and CSCs in regards to possessing varying degrees of differentiation [14].

However, normal tissue stem cells and CSC are not always identical. Normal tissue stem cells reside in the assigned tissue structure while CSC are capable of migrating from the original tumor to the establishment of distant metastasis (“migrating CSC� concept) [15]. This concept also indicates the partial independence of CSC from stem-cell niche which plays an indispensable role in the maintenance of the normal tissue stem cell [16].

Figure 1. Comparison of the normal pancreatic stem cell function and CSC function.

Marker of normal pancreatic stem cell and CSC The precise location of normal tissue stem cells within pancreas is still controversial. The centroacinar cell is considered to be a possible pancreatic

114

Shin Hamada & Tooru Shimosegawa

tissue stem cell according to its anatomical location and the expression of the transcriptional factor Hes1 which marks multipotent progenitor in the mouse pancreas [17]. Recent research identified that the intestinal stem cell marker DCAMKL-1 was successfully isolated from a normal pancreatic stem cell fraction [18]. Isolated cells formed spheroids in the non-adherent culture condition and expressed markers of exocrine pancreas in this study.

However, above mentioned molecules are not utilized to purify the CSC fraction from pancreatic cancer cell lines or surgical specimen. Instead distinct surface markers have been used to isolate CSCs. For example, side population cell is defined by the ability to excrete Hoechst dye from cell and identified as a hematopoietic stem cell marker [19]. Side population cells from pancreatic cancer cell lines also contain CSC fraction evidenced by increased tumorigenicity and chemoresistance [20]. The CD24/CD44/ESA triple positive cell population isolated from the surgical specimen also depicted the increased tumorigenicity in another study [3]. The CD133/Prominin positive cell population revealed the capability to reconstruct the hierarchical population of tumor cells in an orthotopic implantation model [4]. Furthermore, CD133/CXCR4 double positive subpopulation was identified to be essential for the establishment of metastatic lesions. Currently the CSC phenotype can be summarized by its tumorigenicity, chemoresistance, metastatic ability and reconstruction of hierarchy. These characteristics were based on the resultant cancer cell behavior after the isolation of specific population from entire cancer cells. Definitive CSC functions need to be elucidated for clarifying the nature of CSC.

Pancreatic cancer stem cell and therapy-resistance Therapeutic strategies against pancreatic cancer suffer from frequent recurrence after radical surgery and resistance against conventional therapies such as chemotherapy and radiotherapy [21]. A small number of CSCs can reconstruct the entire cancer tissue, leading to the recurrence after surgery.

CSCs are reported to be highly resistant to the current chemotherapeutic agents [22]. Currently available chemotherapeutic agents mainly target proliferating cancer cells and are unable to eliminate CSCs. The entire mechanism of chemoresistance in CSC is cryptic but recent researches clarified part of this picture.

Side population cells are reported to express high level of ATP-binding cassette (ABC) transporters. Expression levels of ABC transporters in cancer cells determine chemoresistant phenotype since chemotherapeutic agents such as etoposide, doxorubicin, vincristine and paclitaxel are direct substrates of ABC transporters [23,24]. Side population cells in pancreatic cancer also

Pancreatic cancer stem cell and mesenchymal stem cell

115

express ABCB1 and ABCG2 which play important role in chemoresistance of tumors in other organs [25]. Expression of ABCG2 induces resistance against gemicitabine [26] though there is no clear evidence that ABC transporters directly efflux gemcitabine or its metabolites in pancreatic cancer cells. The cellular uptake of gemcitabine is mediated by nucleoside transporters including ENT1 [27] and expression levels of these molecules in side population cells or other CSC fractions are not fully explored. Detailed analysis will identify the mechanism of resistance in CSC of pancreatic cancer.

These reports mainly focused on the existing therapy-resistant population in cancer cells. Another report indicated that conventional chemotherapy itself could propagate the CSC population in pancreatic cancer. Long-term culture of pancreatic cancer cell line L3.6pl and AsPC-1 with increasing doses of gemcitabine enhances the CD24/CD44/ESA triple positive CSC fraction [28]. These cells also revealed spindle-shaped appearance and decreased expression of E-cadherin which are the specific features of epithelial-mesenchymal transition (EMT) suggesting the acquisition of a metastatic phenotype. These observations indicate that conventional therapies could create selective pressure for more malignant phenotype of cancer cells.

Interaction of cancer cells and stromal cells Pancreatic cancer cells affect stromal cell function and vice-versa. The interactions between pancreatic cancer cells and stromal cells were extensively studied but little was known about the detailed molecular mechanism so far. Detailed characterization of cell surface markers and/or cytokines has exposed intriguing mechanisms involved in these interactions.

Numerous factors produced in pancreatic cancer stromal cells have tumorpromoting effect as inducers of cell migration and invasion of cancer cells.

For instance, CD133 positive pancreatic cancer cells showed enhanced cellular migration and invasion after co-culture with stromal cells [11]. The siRNA-based knockdown of CXCR4 in CD133 positive cells attenuated cellular migration and invasion which suggest that CSC-supporting function of stromal cells could be mediated by humoral factor such as SDF-1, the CXCR4 ligand. In addition, SPARC is also a secreted protein from tumor stromal cells and higher SPARC expression in pancreatic cancer stroma inversely correlated with overall survival. SPARC enhanced the cellular invasion of pancreatic cancer cells under the co-culture with human pancreatic stellate cells (hPSC) in vitro [29]. These findings clearly demonstrate that cancer stromal cells enhance the metastatic potential of cancer cells which may also be a CSC function.

116

Shin Hamada & Tooru Shimosegawa

Pancreatic cancer cells are also capable of modulating stromal cell functions. Pancreatic stellate cells (PSCs) play a central role during the establishment of fibrosis after inflammation in the pancreas, though they normally reside within the pancreas in a quiescent state [30]. PSC is activated by cytokines, reactive oxygen species and other stimuli caused by inflammation. Pancreatic cancer cells also activate PSC via humoral factors in a co-culture system [31] which indicates that pancreatic cancer cells can create a suitable microenviroment for their survival by activating stromal cells. A recent study indicated that the novel ligand named PAUF (pancreatic adenocarcinoma upregulated factor) from cancer cells could activate the Tolllike receptor 2 (TLR-2) and induce extracellular signal-regulated kinase phosphorylation in THP-1 cells (a human acute monocytic leukemia cell line) [32]. This signaling pathway facilitates the production of tumor-promoting cytokines while inhibiting innate immune surveillance against cancer cells.

Hedgehog signal is also reported to have a protective role in pancreatic cancer. This signaling pathway affects the neovascularization of pancreatic tumor via the alteration of Ang-1 and IGF-1 expression in bone marrowderived pro-angiogenic cells [33]. Further investigation should be carried out since CSC has a distinct expression profile of these stroma-modifying factors from other types of cells.

Role of MSC in pancreatic cancer MSC was first identified within the bone marrow cells and their localization was confirmed around the perivascular area in various organs [34]. MSCs can differentiate into mesenchymal tissues such as bone, adipose tissue, and cartilage [35]. Recent research suggested its possible role during inflammation, immune response, wound healing, and cancer progression [3639]. MSCs, as well as other types of stromal cells, have been reported to contribute to pancreatic cancer progression. Tumor-derived growth factors such as platelet-derived growth factor, epidermal growth factor and VEGF induced migration of MSCs in in vitro experiments [40]. Migrated MSCs could form spheroids with pancreatic cancer cells in vitro and these spheroids can also reside within orthotopically implanted tumors in an in vivo study.

Increase in the vascularity was observed in these orthotopic pancreatic tumors with MSCs. This effect was attributed to the VEGF production by MSCs. Current knowledge suggests MSCs recruited into tumor facilitate cancer therapy resistance by promoting tumor growth.

However, active incorporation of MSCs into pancreatic tumors has now drawn attention to the possibility that this phenomenon could be a novel therapeutic target for specific delivery of anticancer agents. Genetically

Pancreatic cancer stem cell and mesenchymal stem cell

117

engineered labeled MSCs injected into tumor-bearing mice efficiently accumulate within the pancreas tumor [41]. Another group indicated the possibility of adipose-derived MSC utilization for the tumor-specific delivery of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) [42]. A similar approach was also made by combining the TRAIL expressing MSC and knockdown of X-linked inhibitor of apoptosis protein in cancer cells by RNA interference [43]. These approaches seem promising as tumor-specific delivery system which consist from sustainable resource by utilizing patients cells. More efficacious methods need to be developed to accumulate “engineered” MSCs into the tumor for clinical application.

Microenvironment of pancreatic cancer; Role of hypoxia Pancreatic cancer is generally recognized as a less enhanced lesion by contrast-enhanced computed tomography which reflects hypovascularity in cancer tissue including dense stromal cells and a few cancer cells as mentioned earlier. Poor blood perfusion results in significant hypoxia of cancer cells compared to the normal tissue, and this status triggers cellular adaptation. The transcriptional factor HIF1α contributes to the gene expression required for the adaptation for hypoxia and angiogenesis upon hypoxic conditions, which rapidly degrade normoxia [44]. HIF1α is frequently expressed in pancreatic cancer and higher expression correlates with poor prognosis [45]. The HIF1α mediated signal influences both cancer cells and stromal cells. The HIF1α contributes to the production of VEGF which acts as a potent inducer of angiogenesis [46]. In addition, the microRNA miR-210 is induced by hypoxic conditions in various cancer cells including pancreatic cancer in HIF1α dependent manner [47]. Cellular adaptation to the hypoxic condition known as the “Pasteur effect”, switches energy production to glycolysis from oxidative phosphorylation. This is regulated by miR-210 which directly targets ironsulfur cluster assembly proteins (ISCU1/2) in primary endothelial cells and transformed epithelial cells [48].

The detailed relationship between hypoxia and normal tissue stem cell or CSCs remained unclear until this decade. Recent progress in this field of research indicated the essential role of hypoxia during the maintenance of normal stem cell or CSC function. Long-term culture of human embryonic stem cell (hESC) under normoxic conditions frequently results in spontaneous differentiation while hypoxic culture condition significantly lowered this ratio, with sustained expression levels of Nanog and Oct4 [49].

Hypoxia-mediated signals contribute to the maintenance of CSCs or MSCs in several reports. For example, CD133 expression levels in cancer cells turned out to be regulated by HIF1α as a downstream effector of the mTOR pathway

118

Shin Hamada & Tooru Shimosegawa

[50]. Furthermore, short term exposure of MSCs to hypoxic conditions resulted in increased expression of CX3CR1and CXCR, which increased the ability of cells to migrate after stimulation and xenotypic grafting into chicken embryo [51]. This line of evidence indicates that hypoxia could contribute to the propagation of cancer cells with the CSC phenotype and expansion of cancer-promoting MSCs within a tumor.

Pancreatic cancer stromal cells as a possible CSC niche Establishment of distant metastasis is a critical step of pancreatic cancer since it is deemed as an unresectable disease with extremely poor prognosis.

Cancer metastasis requires multiple biological processes such as migration from the site of origin, invasion beyond the basement membrane, extravasation into the blood vessel or lymphatic vessel, and engraftment in distant organ. These processes also require dynamic cancer cell-stromal cell interactions. Contribution of stromal cells to the cancer cell migration and invasion was already discussed in the previous section. Several reports indicated that stromal cells also have a beneficial role in engraftment of cancer cells. Injection of pancreatic cancer cells combined with gendermismatch PSC into mouse pancreas resulted in increased number of distant metastasis to liver, mesentery, or mediastinum [10]. Metastatic nodules contained exogenous PSCs which were identified by the presence of Y chromosome. These results clearly demonstrated that stromal cells yield survival advantages to CSCs, particularly during metastasis as a niche since metastatic cancer cells are considered to harbor the CSC phenotype [15].

Inhibition of the cancer-stromal cell interactions revealed favorable capabilities to reduce tumor invasion and metastasis in experimental models.

Contribution of hedgehog pathway to the cancer-stromal cell interactions were extensively studied for many years since small molecule inhibitors such as cyclopamine and its derivatives were available [52]. A novel inhibitor of the hedgehog pathway, IPI-269609, was reported to reduce metastasis and invasion in an orthotopic implantation model [53]. Treatment of xenograft tumor by this agent resulted in the decrease of aldehyde dehydrogenasepositive cells which contain CSCs. Another study indicated that alteration of tumor stromal structures by hedgehog inhibition [54] showed that CSCs are sensitive to modulating the pancreatic cancer stroma.

Closing remarks Characteristic features of CSCs, cancer stromal cells, and MSCs were discussed in this chapter. Tumor stroma in pancreatic cancer is not a wasteland

Pancreatic cancer stem cell and mesenchymal stem cell

119

but a formidable defender of cancer cells and CSCs. Conventional therapies which ta rget cancer cells reach their limit, stressing the importance of targeting cancer associated tissues such as tumor stroma or vasculature.

Possible therapeutic targets within the cancer cell-stromal cell interactions are summarized in figure 2. Molecular targeting of cancer storma or vasculature is now carried out and several compounds are undergoing clinical trials.

Further investigation will provide novel therapeutic strategies into the pancreatic cancer treatment.

Figure 2. Therapeutic targets within the cancer-stroma interaction.

References 1.

2.

3.

4.

5.

6.

Hidalgo M. 2010, N Engl J Med, 362, 1605-1617.

Eckel F, Schneider G, Schmid RM. 2006, Expert Opin Investig Drugs, 15, 1395-1410.

Li C, Heidt DG, Dalerba P, Burant CF, Zhang L, Adsay V, Wicha M, Clarke MF, Simeone DM. 2007, Cancer Res, 67, 1030-1037.

Hermann PC, Huber SL, Herrler T, Aicher A, Ellwart JW, Guba M, Bruns CJ, Heeschen C. 2007, Cell Stem Cell, 1, 313-323.

Masamune A, Shimosegawa T. 2009, J Gastroenterol, 44, 249-260.

Mahadevan D, Von Hoff DD. 2007, Mol Cancer Ther, 6, 1186-1197.

120

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Shin Hamada & Tooru Shimosegawa

Marrache F, Pendyala S, Bhagat G, Betz KS, Song Z, Wang TC. 2008, Gut, 57, 1113-1120.

Watanabe T, Masamune A, Kikuta K, Hirota M, Kume K, Satoh K, Shimosegawa T. 2009, Am J Physiol Gastrointest Liver Physiol, 297, G1138-G1146.

Muerkoster S, Wegehenkel K, Arlt A, Witt M, Sipos B, Kruse ML, Sebens T, Kloppel G, Kalthoff H, Folsch UR, Schafer H. 2004, Cancer Res, 64, 1331-1337.

Xu Z, Vonlaufen A, Phillips PA, Fiala-Beer E, Zhang X, Yang L, Biankin AV, Goldstein D, Pirola RC, Wilson JS, Apte MV. 2010, Am J Pathol, 177, 2585-2596.

Moriyama T, Ohuchida K, Mizumoto K, Cui L, Ikenaga N, Sato N, Tanaka M. 2010, Cancer, 116, 3357-3368.

Katsumoto K, Shiraki N, Miki R, Kume S. 2010, Dev Growth Differ, 52, 115-129.

Brittan M, Wright NA. 2002, J Pathol, 197, 492-509.

Brabletz S, Schmalhofer O, Brabletz T. 2009, J Pathol, 217, 307-317.

Wellner U, Schubert J, Burk UC, Schmalhofer O, Zhu F, Sonntag A, Waldvogel B, Vannier C, Darling D, zur Hausen A, Brunton VG, Morton J, Sansom O, Schuler J, Stemmler MP, Herzberger C, Hopt U, Keck T, Brabletz S, Brabletz T. 2009, Nat Cell Biol, 11, 1487-1495.

Fellous TG, McDonald SA, Burkert J, Humphries A, Islam S, De-Alwis NM, Gutierrez-Gonzalez L, Tadrous PJ, Elia G, Kocher HM, Bhattacharya S, Mears L, El-Bahrawy M, Turnbull DM, Taylor RW, Greaves LC, Chinnery PF, Day CP, Wright NA, Alison MR. 2009, Stem Cells, 27, 1410-1420.

Kopinke D, Brailsford M, Shea JE, Leavitt R, Scaife CL, Murtaugh LC. 2011, Development, 138, 431-441.

May R, Sureban SM, Lightfoot SA, Hoskins AB, Brackett DJ, Postier RG, Ramanujam R, Rao CV, Wyche JH, Anant S, Houchen CW. 2010, Am J Physiol Gastrointest Liver Physiol, 299, G303-G310.

Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC. 1996, J Exp Med, 183, 1797-1806.

Yao J, Cai HH, Wei JS, An Y, Ji ZL, Lu ZP, Wu JL, Chen P, Jiang KR, Dai CC, Qian ZY, Xu ZK, Miao Y. 2010, Oncol Rep, 23, 1375-1382.

Li D, Xie K, Wolff R, Abbruzzese JL. 2004, Lancet, 363, 1049-1057.

Scopelliti A, Cammareri P, Catalano V, Saladino V, Todaro M, Stassi G. 2009, Expert Opin Biol Ther, 9, 1005-1016.

Stordal B, Davey R. 2009, Curr Cancer Drug Targets, 9, 354-365.

Burkhart CA, Watt F, Murray J, Pajic M, Prokvolit A, Xue C, Flemming C, Smith J, Purmal A, Isachenko N, Komarov PG, Gurova KV, Sartorelli AC, Marshall GM, Norris MD, Gudkov AV, Haber M. 2009, Cancer Res, 69, 6573-6580.

Zhou J, Wang CY, Liu T, Wu B, Zhou F, Xiong JX, Wu HS, Tao J, Zhao G, Yang M, Gou SM. 2008, World J Gastroenterol, 14, 925-930. de Wolf C, Jansen R, Yamaguchi H, de Haas M, van de Wetering K, Wijnholds J, Beijnen J, Borst P. 2008, Mol Cancer Ther, 7, 3092-3102.

Pancreatic cancer stem cell and mesenchymal stem cell

121

27. Mackey JR, Yao SY, Smith KM, Karpinski E, Baldwin SA, Cass CE, Young JD. 1999, J Natl Cancer Inst, 91, 1876-1881.

28. Shah AN, Summy JM, Zhang J, Park SI, Parikh NU, Gallick GE. 2007, Ann Surg Oncol, 14, 3629-3637.

29. Mantoni TS, Schendel RR, Rodel F, Niedobitek G, Al-Assar O, Masamune A, Brunner TB. 2008, Cancer Biol Ther, 7, 1806-1815.

30. Masamune A, Watanabe T, Kikuta K, Shimosegawa T. 2009, Clin Gastroenterol Hepatol, 7, S48-S54.

31. Apte MV, Park S, Phillips PA, Santucci N, Goldstein D, Kumar RK, Ramm GA, Buchler M, Friess H, McCarroll JA, Keogh G, Merrett N, Pirola R, Wilson JS. 2004, Pancreas, 29, 179-187.

32. Park HD, Lee Y, Oh YK, Jung JG, Park YW, Myung K, Kim KH, Koh SS, Lim DS. 2011, Oncogene, 30, 201-211.

33. Nakamura K, Sasajima J, Mizukami Y, Sugiyama Y, Yamazaki M, Fujii R, Kawamoto T, Koizumi K, Sato K, Fujiya M, Sasaki K, Tanno S, Okumura T, Shimizu N, Kawabe J, Karasaki H, Kono T, Ii M, Bardeesy N, Chung DC, Kohgo Y. 2010, PLoS One, 5, e8824.

34. Shi S, Gronthos S. 2003, J Bone Miner Res, 18, 696-704.

35. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, Moorman MA, Simonetti DW, Craig S, Marshak DR. 1999, Science, 284, 143-147.

36. Bouffi C, Bony C, Courties G, Jorgensen C, Noel D. 2010, PLoS One, 5, e14247.

37. Kode JA, Mukherjee S, Joglekar MV, Hardikar AA. 2009, Cytotherapy, 11, 377-391.

38. Hocking AM, Gibran NS. 2010, Exp Cell Res, 316, 2213-2219.

39. Quante M, Tu SP, Tomita H, Gonda T, Wang SS, Takashi S, Baik GH, Shibata W, Diprete B, Betz KS, Friedman R, Varro A, Tycko B, Wang TC. 2011, Cancer Cell, 19, 257-272.

40. Beckermann BM, Kallifatidis G, Groth A, Frommhold D, Apel A, Mattern J, Salnikov AV, Moldenhauer G, Wagner W, Diehlmann A, Saffrich R, Schubert M, Ho AD, Giese N, Buchler MW, Friess H, Buchler P, Herr I. 2008, Br J Cancer, 99, 622-631.

41. Zischek C, Niess H, Ischenko I, Conrad C, Huss R, Jauch KW, Nelson PJ, Bruns C. 2009, Ann Surg, 250, 747-753.

42. Grisendi G, Bussolari R, Cafarelli L, Petak I, Rasini V, Veronesi E, De Santis G, Spano C, Tagliazzucchi M, Barti-Juhasz H, Scarabelli L, Bambi F, Frassoldati A, Rossi G, Casali C, Morandi U, Horwitz EM, Paolucci P, Conte P, Dominici M. 2010, Cancer Res, 70, 3718-3729.

43. Mohr A, Albarenque SM, Deedigan L, Yu R, Reidy M, Fulda S, Zwacka RM. 2010, Stem Cells, 28, 2109-2120.

44. Kaelin WG Jr, Ratcliffe PJ. 2008, Mol Cell, 30, 393-402.

45. Miyake K, Yoshizumi T, Imura S, Sugimoto K, Batmunkh E, Kanemura H, Morine Y, Shimada M. 2008, Pancreas, 36, e1-e9.

46. Damert A, Ikeda E, Risau W. 1997, Biochem J 327, , 419-423.

122

Shin Hamada & Tooru Shimosegawa

47. Huang X, Ding L, Bennewith KL, Tong RT, Welford SM, Ang KK, Story M, Le QT, Giaccia AJ. 2009, Mol Cell, 35, 856-867.

48. Chan SY, Zhang YY, Hemann C, Mahoney CE, Zweier JL, Loscalzo J. 2009, Cell Metab, 10, 273-284.

49. Zachar V, Prasad SM, Weli SC, Gabrielsen A, Petersen K, Petersen MB, Fink T. 2010, In Vitro Cell Dev Biol Anim, 46, 276-283.

50. Matsumoto K, Arao T, Tanaka K, Kaneda H, Kudo K, Fujita Y, Tamura D, Aomatsu K, Tamura T, Yamada Y, Saijo N, Nishio K. 2009, Cancer Res, 69, 7160-7164.

51. Hung SC, Pochampally RR, Hsu SC, Sanchez C, Chen SC, Spees J, Prockop DJ. 2007, PLoS One, 2, e416.

52. Cengel KA. 2004, Cancer Biol Ther, 3, 165-166.

53. Feldmann G, Fendrich V, McGovern K, Bedja D, Bisht S, Alvarez H, Koorstra JB, Habbe N, Karikari C, Mullendore M, Gabrielson KL, Sharma R, Matsui W, Maitra A. 2008, Mol Cancer Ther, 7, 2725-2735.

54. Olive KP, Jacobetz MA, Davidson CJ, Gopinathan A, McIntyre D, Honess D, Madhu B, Goldgraben MA, Caldwell ME, Allard D, Frese KK, Denicola G, Feig C, Combs C, Winter SP, Ireland-Zecchini H, Reichelt S, Howat WJ, Chang A, Dhara M, Wang L, Ruckert F, Grutzmann R, Pilarsky C, Izeradjene K, Hingorani SR, Huang P, Davies SE, Plunkett W, Egorin M, Hruban RH, Whitebread N, McGovern K, Adams J, Iacobuzio-Donahue C, Griffiths J, Tuveson DA. 2009, Science, 324, 1457-1461.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 123-138 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

7. Signaling pathways mediating epithelialmesenchymal crosstalk in pancreatic cancer: Hedgehog, Notch and TGFβ Jennifer M. Bailey and Steven D. Leach

The Department of Surgery and The McKusick-Nathans Institute of Genetic Medicine Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

Abstract. Pancreatic cancer is an extremely aggressive disease characterized by a complex stromal or desmoplastic response. This dense desmoplastic response is in part the result of pancreatic stellate cell activation in response to paracrine signals from preneoplastic and neoplastic pancreatic epithelium. Activated stellate cells secrete a large array of soluble factors that directly influence malignant epithelial, vascular and inflammatory cells, and also alter the composition of the extracellular matrix. Thus, the signaling interactions between tumor and stellate cells represent a large, diverse spectrum of bidirectional and reciprocal influences necessary for the initiation and promotion of pancreatic cancer. A number of pathways including Hedgehog, Notch and Transforming growth factor-beta are involved in mediating cross-talk between the malignant pancreatic epithelium and its associated stroma, making them unique pathways to target for the treatment of pancreatic cancer.

Correspondence/Reprint request: Dr. Jennifer M. Bailey, The Department of Surgery and The McKusickNathans Institute of Genetic Medicine, Johns Hopkins University School of Medicine, 733 N Broadway Broadway Research Building Rm. 472, Baltimore, MD 21205, USA. E-mail: jbaile45@jhmi.edu

124

Jennifer M. Bailey & Steven D. Leach

Introduction Pancreatic cancer is a devastatingly lethal disease characterized by intense desmoplasia. Desmoplasia is a histological term used to describe an abundant tumor stroma (host cells) that consists of activated stellate cells, myofibroblasts, immune cells, inflammatory cells, blood and lymphatic vessels and a complex extracellular matrix. The desmoplastic response can be initiated even by the earliest premalignant lesions in pancreatic cancer, known as Pancreatic Intraepithelial Neoplasia (PanIN). Tumor-associated desmoplasia increases in the progression of disease to the point where the majority of pancreatic tumor volume is comprised of activated stroma. While once thought mechanistically to reflect a host response against the tumor cells, desmoplasia is now being investigated as a dynamic series of autocrine and paracrine signaling interactions between host cells and tumor cells that both enhances the desmoplastic response and accelerates pancreatic cancer initiation, progression and metastasis. Factors secreted from neoplastic pancreatic cells such as transforming growth factor-beta (TGFβ), vascular endothelial growth factor (VEGF), hepatocyte growth factor (HGF), Sonic hedgehog (SHH), epidermal growth factor (EGF), fibroblast growth factors (FGF’s) and insulin-like growth factors (IGF’s) all signal to the adjacent microenvironment, which constitutes a direct paracrine mechanism by which pre-neoplastic and neoplastic cells communicate with other cell types (1, 2).

These paracrine factors specifically signal to pancreatic stellate cells, promoting the activated pancreatic stellate cell phenotype. Once activated, pancreatic stellate cells secrete excess extracellular matrix proteins and also secrete a number of factors including matrix metalloproteases (MMPs), PDGF, FGF, TGFβ, IGF1, small leucine-rich proteoglycans, periostin, collagen I, EGF and heparin sulfate proteoglycans. These factors all influence the progression of pancreatic cancer. Better insight into the critical factors secreted by both tumor cells and activated stellate cells will aid in the development of targeted, stromal-directed therapies. These therapies will hopefully be effective in minimizing the amount of tumor-associated desmoplasia, allowing improved cytotoxic drug delivery and a reduction in tumor burden and tumor progression.

Pancreatic stellate cells promote tumor progression Pancreatic stellate cells play a critical role in the promotion of pancreatic cancer. A number of experiments have demonstrated these tumor-promoting effects. When conditioned media from pancreatic stellate cells is added to pancreatic cancer cells in vitro, the medium induces the proliferation,

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

125

migration and invasion of pancreatic cancer cells (3, 4). Furthermore, coculture experiments in vitro show that direct co-cultures increase cell proliferation more significantly than indirect co-culture conditions (5).

Experiments performed in vivo have demonstrated that, when injected into nude mice, combinations of pancreatic cancer cells and pancreatic stellate cells have a higher frequency of tumor formation, increased tumor volume, proliferation and demoplasia when compared to pancreatic cancer cell lines injected alone (4, 6-8). Human pancreatic stellate cells have also been visualized in the metastatic sites of mice implanted with human pancreatic cancer cells and pancreatic stellate cells (9). The sites of metastasis noted from the orthotopic model also showed increased angiogenesis in the tumors derived from the co-cultured cells, which strengthens the hypothesis that pancreatic stellate cells are not only mediators of tumor initiation, but they are also critical regulators of metastasis and angiogenesis. In vitro experiments have shown that pancreatic stellate cells can induce cancer cell resistance to both gemcitabine and radiation, a mechanism that might be mediated through the stellate cell’s secretion of both laminin and fibronectin, which have antiapoptotic effects on tumor cells (3, 4).

Pancreatic stellate cells also contribute to pancreatic cancer progression by secreting factors that play a critical role in altering the extracellular matrix (ECM). Pancreatic stellate cells secrete collagen I, which is associated with increased migration and proliferation of pancreatic cancer cells (10). Stromal cells also synthesize and secrete SPARC (secreted protein acidic and rich in cystein), which is a protein involved in cell migration and wound repair and is associated with worse prognosis in pancreatic cancer patients (11). Recent data reveals that SPARC levels are higher in pancreatic stellate cells than in pancreatic tumor cells, and that conditioned medium from pancreatic stellate cells contains SPARC (12, 13). MMPs are proteolytic enzymes that promote cancer invasion by degrading the extracellular matrix. MMP-2 and MMP-9 are expressed in pancreatic cancer. Pancreatic stellate cells synthesize both MMP-2 and its inhibitors, tissue inhibitors of metalloproteinases 1 and 2 (14, 15).

Thrombospondin (TSP), secreted by pancreatic stellate cells, increases the production of MMP-9 in pancreatic tumor cells (16). In this way, tumor and stromal cells contribute to enhanced tumor invasion and metastasis by secreting factors that degrade the extracellular matrix and facilitate the extravasation of tumor cells. Periostin is a matricellular protein expressed by pancreatic stellate cells in response to co-culture conditions with pancreatic tumor cells. At high concentrations, periostin can induce phosphorylation of Akt and promote cell migration (17).

126

Jennifer M. Bailey & Steven D. Leach

Epithelial-mesenchymal crosstalk: Focus on Hedgehog, Notch and TGFβ Hedgehog signaling The hedgehog signaling pathway is a critical morphogenic signaling pathway, both during development and in adult tissues. The hedgehog pathway is tightly regulated by two twelve-pass transmembrane receptors, Patched1 (Ptch1) and Patched2 (Ptch2), which are localized to the primary cilium (18), a microtubule-rich structure emanating from the apical surface of virtually all epithelial cells. In the absence of hedgehog ligands, Ptch inhibits the localization and lateral transport of a serpentine protein, Smoothened (Smo) to the primary cilium (19). The transcription factors Gli1, Gli2 and Gli3 also localize to the primary cilium (20, 21) along with Suppressor of Fused (SuFu), which mediates Gli activity in mammalian cells (21). Upon Sonic hedgehog (Shh), Indian hedgehog (Ihh) or desert hedgehog (Dhh) ligand binding to Ptch, the repression of Ptch on Smo is relieved and Smo translocates along the primary cilium. Smo translocation along the primary cilium reorganizes SuFu and alters the subcellular localization of Gli2 and Gli3 such that Gli2 is localized to the ciliary tip and the nucleus where it initiates the transcriptional activation of hedgehog pathway target genes including Gli1 and Ptch (22).

In the early stages of gut formation, both Shh and Ihh are expressed in the endodermal epithelium (23, 24). Despite their expression in the primitive gut, both Shh and Ihh are absent in from the early endoderm that is designated to become pancreas (25, 26); however, the receptor Patched1 (Ptch1) is expressed in mesenchyme adjacent to the pancreas anlage (25).

Gain of function experiments using transgenic mouse models have been instrumental in demonstrating the effects of expressing Shh during pancreatic development. When Shh is expressed in early pancreatic development under the regulation of the Pdx-1 promoter, the result is almost uniform loss of pancreas (25). Ectopic expression of either Shh or Ihh promotes the formation of intestinal smooth muscle within the overlying mesenchyme, with associated conversion of pre-pancreatic endoderm to an intestinal fate (27). These data indicate a necessity for tight regulation of the hedgehog pathway to ensure proper patterning of the pancreas and gut. In the mature pancreas, expression of Ihh, Dhh, Hhip, Ptch1 and Smo have been observed in islet and ductal cells (28-30), and ligands are also detected in pancreatic duct glands (31). These glands may represent an adult ductal progenitor compartment, perhaps reflected in the fact that the hedgehog pathway is also required for proper pancreatic regeneration after injury (32).

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

127

The hedgehog pathway is now a recognized therapeutic target for the treatment of pancreatic cancer. A number of experiments have confirmed the pathway to be a critical mediator of pancreatic cancer initiation and metastasis. In addition to a number of other reports on ligand-dependent Hh pathway signaling within tumor epithelium, early experiments in pancreatic cancer showed ligand-dependent initiation of pancreatic tumors (33, 34).

Treatment of pancreatic cell lines and xenograft tumors with cyclopamine, a Smo antagonist, suppressed tumor cell growth in vitro and caused regression of xenograft tumors in vivo (33, 34). Specifically, treatment with cyclopamine decreased the growth of tumor cell lines expressing Ptch and treatment with cyclopamine significantly reduced the Ptch mRNA levels in first passage pancreatic xenografts (34). Treatment with 5E1, a monoclonal antibody that binds to and neutralizes the effects of hedgehog ligands, also decreased the growth of pancreatic xenografts and decreased Ptch mRNA levels (34). In a genetically engineered mouse model of pancreatic cancer, Pdx1-Cre;LsL-Kras(G12D);Ink4AArf(lox/lox), mutant Kras increased Shh mRNA levels and treatment with cyclopamine prolonged survival (35).

Perhaps even more significant and relevant to therapeutic targeting of this pathway for pancreatic cancer is the defined paracrine mechanism by which the hedgehog pathway influences the tumor microenvironment to enhance initiation, angiogenesis and metastasis this deadly form of cancer.

Shh and Ihh are both expressed in and secreted from PanIN and PDAC cells and they target the stromal cells to promote desmoplasia (7, 33, 34, 36-39).

Since the initial experiments with cyclopamine, a number of experiments have been done to address the relative contribution of Hh ligands secreted from pancreatic epithelial cells to enhance tumor growth in vivo. In one set of experiments, co-culture conditions were used to define the interactions between tumor cells and stromal cells. In this experiment, human pancreatic cancer xenografts were implanted into mice along with mouse embryonic fibroblasts (MEFs) that were isolated from transgenic mice containing a tamoxifen-inducible deletion of Smo. When human pancreatic cancer cells were implanted at sub-optimal numbers, the human xenografts did not form measureable tumors, but co-injection with MEFs led to the formation of tumors. The tumors were significantly reduced in size when the xenografts were co-cultured with wild-type MEFs that no longer expressed Smo (40). In another set of experiments, human cancer cell lines were transduced with a retrovirus that expressed Shh. When the Shh overexpressing cell line was co-cultured with human pancreatic fibroblasts, a significant reduction in time-to-tumor progression was observed compared to co-cultures of control human cancer cell lines with control human pancreatic fibroblasts (36). The requirement for paracrine signaling in pancreatic cancer was further

128

Jennifer M. Bailey & Steven D. Leach

elucidated by in a series of experiments using a transgenic mouse model that had an activating Smo mutation in the pancreatic epithelium (PdxCre; SmoM2). Activation of Smo alone in the pancreatic epithelium was not enough to drive neoplastic transformation. PdxCre;SmoM2 mice were crossed to a well 窶田haracterized mouse model of pancreatic cancer, PdxCre; KrasG12D. SmoM2 alone did not potentiate Kras-driven transformation. To study cellular responses to hedgehog ligand expression from pancreatic epithelium, the PdxCre;SmoM2;KrasG12D mice were crossed to a Ptc1-LacZ reporter line. LacZ expression was only detected in the stromal cells of pancreatic tumors, indicating these cells were expressing Ptch1. In contrast, no Lacz expression was ever detected in the pancreatic epithelium of PdxCre;SmoM2 or PdxCre;KrasG12D;SmoM2 mice, supporting a model of paracrine hedgehog signaling in pancreatic cancer (41).

In vitro co-culture experiments have further delineated the effects of hedgehog ligand stimulation on cultured human pancreatic fibroblasts. In response to recombinant human Shh, the cultured fibroblasts increased expression of vascular endothelial growth factor (VEGF), a positive regulator of angiogenesis, along with matrix metalloprotease 9 (MMP9) and Ptch(36).

Similarly, when pancreatic ductal cells were isolated from the PdxCre; KrasG12D mice, the cells failed to express Gli1 in response to recombinant hedgehog ligand stimulation, whereas recombinant hedgehog ligand stimulation led to a 10-50 fold induction of Gli1 in human pancreatic fibroblasts (41). To analyze which compartments are expressing downstream hedgehog pathway components, quantitative RT-PCR has been performed on tumors isolated from PdxCre;KrasG12D, PdxCre;KrasG12D;SmoM2 and PdxCre;KrasG12D;p53R270H mice. Using laser capture microdissection, the epithelial compartments were isolated from the stromal compartments. In all three mouse models, there was a 13-150 fold increase in stromal levels of Gli1 compared to the epithelial compartment (41). These data were consistent when a similar experiment was performed on human pancreatic tumors.

Smo antagonists have been shown to decrease tumor associated desmoplasia leading to an increase in tumor vasculature and an increase in drug delivery (42). These data are interesting, as multiple orthotopic pancreatic models have addressed the hypothesis that other stromal cell types, including endothelial cells and bone marrow cells, also respond to hedgehog ligands secreted from pancreatic tumors. To address the question of hedgehog pathway recruitment of tumor-associated vasculature and the recruitment of bone marrow cells into the tumor, KP-1N murine tumors were grown in chimeric mice with GFP-labeled bone marrow. When treated with cyclopamine, the investigators noted a significant decline in the recruitment of bone marrow derived cells into the xenografts (43). To further characterize

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

129

the nature of hedgehog recruitment of bone marrow cells, Nakamra et al set up a series of in vitro experiments. In these experiments, the use of conditioned media led to a significant increase in the number of host cells migrating into the matrigel and the cells formed a cord-like structure. Further in vitro experiments confirmed a role for the hedgehog pathway in the migration of precursor bone marrow cells. Transwell migration assays confirmed a Smo-dependent migratory capacity of bone marrow precursor cells. On a molecular level, IGF-1 was identified as a target of hedgehog signaling in bone marrow derived pro-angiogenic cells. Similarly, when Capan-2 (human pancreatic cancer cell line) orthotopically challenged mice were treated with a hedgehog neutralizing antibody, a significant reduction was noted in the number of Lyve-1+ tumor associated lymphatic vessels and a significant decrease was noted in the number of lymph nodes containing metastatic tumor cells (36). Thus, there is a critical role for hedgehog signaling in the regulation of tumor vasculature and lymphatics, but the exact nature and mechanism still remains to be defined.

Notch signaling The Notch signaling pathway is another important regulator of embryonic pancreatic development, where it is required for the expansion of pancreatic progenitor cells. In the adult pancreas, active Notch signaling is only observed in the centroacinar cells, as revealed by Hes1 staining of both human and mouse pancreas (44, 45). More recently, a Notch-responsive reporter strain confirmed active Notch signaling in the adult mouse centroacinar cells (46) and also confirmed active Notch signaling in PanIN and tumor cells of mice with advanced pancreatic ductal adenocarcinoma (PDAC). The Notch pathway consists of transmembrane proteins that serve as both ligands and receptors. There are four Notch receptors (Notch -1, -2, -3 and -4) and there are two families of Notch ligands, Delta (Delta -1, -3 and -4) and Jagged (Jagged -1 and -2). Notch signaling is initiated through ligandreceptor interactions between neighboring cells. The ligand-receptor interactions initiate two cleavage events, the first of which is the ADAM protease TACE, which cleaves the receptor on the extracellular side. The second event is a critical target for therapy and involves cleavage of the receptor within the transmembrane domain and is mediated by Îł-secretase activity. The second cleavage event effectively liberates the intracellular domain of Notch (ICD), which allows for ICD nuclear translocation and association with transcription factors to induce the expression of members of the Hes, Herp and Hey family of transcription factors.

130

Jennifer M. Bailey & Steven D. Leach

The Notch pathway has three defined modes of activity (reviewed in (47)), which mechanistically increases the relevance of this pathway for tumor-stromal interactions. One mechanism is lateral inhibition, which defines a mechanism where Notch is expressed in a population of cells and amplifies small or weak differences to regulate cell fate. After cell fate differentiation, only one cell within an initially equivalent field of cells continues to express Notch receptors. Once the cells take on a more differentiated cell fate, Notch determines lineage decisions, where two daughter cells signal through asymmetric inheritance of ligands. An example of this mechanism is in Drosophila, where differentiating cells express higher levels of Delta to inhibit differentiation of other progeny. The third mechanism is inductive signaling. In this situation, Notch ligands and receptors are expressed on different cell types, such that Notch can only be activated in the receptor-bearing cell. This occurs often between two populations of cells and can help to serve as a boundary between two groups.

This has been modeled in tumor and stromal cells and their supportive and surrounding niche. In breast cancer, Jagged1 is expressed at higher levels in human breast cancer cell lines with higher metastatic potential to the bone.

Furthermore, the incidence of bone metastasis and relapse is higher in patients with high Jagged1 expression. When Sethi et al overexpressed Jagged1 in a mildly metastatic murine breast cancer cell line, Jagged1 promoted osteolytic bone metastasis through Jagged1-Notch interactions in bone cells (48).

Assessment of Notch during pancreatic regeneration, acute and chronic pancreatitis and pancreatic cancer has been analyzed using both microarray and immunohistochemistry (45, 50). Microarray data comparing normal human pancreatic tissue to human pancreatic cancer tissue showed an increase in Notch pathway components (Jagged1, Jagged2, Notch1, Notch2, Notch3, Notch4, Hes1 and Hey1) in pancreatic cancer tissue relative to normal. These data were confirmed using Real-time RT-PCR on an additional set of human pancreatic cancer versus normal tissues. Miyamoto et al further characterized Notch pathway activation by employing in situ hybridization and immunohistochemistry. In situ hybridization revealed Jagged1 transcript levels in intralobular ducts of the pancreas and also in invasive pancreatic cancer. Immunohistochemistry revealed Jagged2 expression in normal interlobular ductal epithelium and in pancreatic cancer and confirmed high expression of Notch1, Notch2 and Hes1 in pancreatic cancer relative to normal pancreas (45).

Experiments done by Plentz et al defined a signature for Notch pathway activation in the Pdx1-Cre LSL-KrasG12D p53Lox/+ mouse model. For these experiments, Plentz et al crossed the Pdx1-Cre LSL-KrasG12D p53Lox/+ mouse

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

131

model with a Notch-responsive GFP reporter strain (51). Immunohistochemical analysis for GFP revealed active Notch signaling in 8 out of 9 pancreatic tumors from the mouse model. Further analysis using RT-PCR showed that mRNA levels of Notch ligands, receptors and target genes are also increased in murine pancreatic cancer tissue relative to the normal mouse pancreas (46). They further defined a tumor promoting mechanism by Notch signaling using a gamma sectretase inhibitor, which led to a decrease in the levels of Hes1 expression and a reduction in the proliferation of human PanIN and PDAC cell lines in vitro. Gamma secretase inhibition also reduced the formation of Kras-driven PanIN in vivo. These studies conclusively demonstrate that the Notch signaling pathway plays a crucial role in pancreatic cancer.

In an effort to study pathways critical in tumor-stromal interactions, Fujita et al isolated pancreatic fibroblasts from normal human pancreas and using RT-PCR showed expression of Notch1 and Jagged1 in normal cultured pancreatic fibroblasts. When Fujita et al performed co-culture experiments with pancreatic cancer and pancreatic myofibroblast cell lines, they determined by RT-PCR that direct co-culture of pancreatic cancer cells with pancreatic myofibroblasts increased the mRNA levels of Hes1 in both cell types (5). Thus, while the precise locations of all relevant Notch receptors and ligands in pancreatic cancer and its associated stroma have not yet been established, the available data support both an intra-epithelial axis and a stromal-to-epithelial axis of Notch signaling in pancreatic malignancy.

TGFβ signaling Transforming growth factor β (TGF-β) controls a plethora of cellular processes including cellular proliferation, differentiation, apoptosis and specification of cell fate. The pathway has been studied in a diverse set of contexts from embryogenesis to mature organ homeostasis, regeneration and cancer. Specifically, the TGFβ signaling pathway is another critical regulator of pancreatic development and morphogenesis. The signaling pathway is initiated by ligand-receptor interactions. The TGFβ superfamily consists of a number of ligands including the TGF-β family (TGF-β1, TGF-β2 and TGF-β3), Activin, Nodal and bone morphogenetic proteins (BMPs). The signaling cascade is activated when TGFβ ligands bind to the type II receptor, which is a serine/threonine kinase that catalyzes the phosphorylation of the type I receptor, initiating a signaling cascade which culminates in the phosphorylation of Smad proteins. There are eight Smad proteins that constitute three functional classes: the receptor regulated Smad (R-Smad),

132

Jennifer M. Bailey & Steven D. Leach

the Co-mediator Smad (Co-Smad) and the inhibitory Smad (I-Smad). In general, the phosphorylated R-Smads (Smads-1,-2,-3,-5 and 8) undergo homotrimerization and form heteromeric complexes with the Co-Smad, Smad-4. These multimeric Smad complexes also associate with additional cofactors, further fine-tuning their transcriptional activity. The I-Smads (Smads 6 and 7) exert their inhibitory function by competing with R-Smads for Co-Smad or receptor interaction.

The TGFβ pathway is an established regulator of pancreatic fibrosis and TGFβ serum levels are elevated in patients with chronic pancreatitis (52).

TGF-β regulates pancreatic fibrosis through both paracrine and autocrine signaling mechanisms involving pancreatic stellate cells. With respect to autocrine signaling, pancreatic stellate cells also express Smad2, Smad3, Smad4 and the TGF-β receptors type I and II (53). Pancreatic stellate cells express TGF-βI, which constitutes 2-5% of the total amount of TGF-β ligands secreted by pancreatic stellate cells. Work by Shek, et al. demonstrated that pancreatic stellate cells express procollagen I, TIMP-1, TIMP-2 and MMP-2.

When stimulated with TGF-βI, collagen protein synthesis increased by 34%.

TGF-β1 decreased the proliferative rate of cultured pancreatic stellate cells and a pan-TGFβ neutralizing antibody increased the proliferative rate by 40%.

TGF-β1 decreased the levels of MMP-9 and MMP-3 by 34%, while the panneutralizing TGFβ antibody increased MMP-9 by 39% (15). Thus, TGFβ autocrine stimulation of pivotal extracellular matrix factors along with autocrine inhibition of metalloproteases represent mechanisms by which pancreatic stellate cells likely modulate pancreatic fibrosis.

Evidence for a paracrine mechanism exists in data showing TGF-β1, TGF-β2 and TGF-β3 are all expressed and secreted from pancreatic cancer cells, and potentially capable of activating the pathway in the adjacent stroma. TGF-β2 expression, as assessed by immunohistochemistry, is correlated with advanced tumor stage, while the absence of TGF-β ligands in tumors is correlated with longer postoperative survival (54). When compared to the normal pancreas, pancreatic adenocarcinomas have increased mRNA expression of TGF-β1 (11 fold), TGF-β2 (7 fold) and TGF-β3 (9 fold)(54).

Notably, TGF-βRII mRNA is also increased in the pancreatic cancer tissue relative to the normal pancreas. In contrast, TGF-βRIII mRNA levels are increased mostly in the surrounding stroma (55). Genetic analysis of human pancreatic cancers reveals that a mutation in at least one of the TGFβ family members (i.e. TGFβR2, SMAD3, SMAD4) is present in 100% of pancreatic cancers (56). The most common of these mutations is the loss of SMAD4, which is deleted in 50% of human pancreatic cancers (57). These findings

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

133

suggest a non-functional axis of TFGβ signaling within the malignant epithelium, underscoring probable epithelial-to-stromal paracrine signaling.

Transgenic mouse models have begun to be utilized to investigate the role of TGFβ signaling in pancreatic cancer progression. In these models, loss of SMAD4 leads to more aggressive disease in the setting of mutant Kras-driven cancer. In two KrasG12D-driven pancreatic cancer models, homozygous or heterozygous deletion of SMAD4 led to more aggressive pancreatic cancer, suggesting that SMAD4 normally restrains progression of pancreatic cancer (58, 59). More recent work has specifically interrogated the role of TGFβR1 and TGFβR2 in pancreatic cancer. Initial work demonstrated that reduced levels of TFGβR2 led to a more aggressive form of pancreatic cancer (60). Adrian, et al crossed a TGFβR1 haploinsufficient mouse to a mouse model involving acinar-specific activation of mutant Kras (ElastaseKrasG12D mouse). They discovered a significant (4 fold) decrease in the number

Figure 1. A schematic of pathways regulating pancreatic cancer desmoplasia.

Pancreatic tumor cells and stellate cells are involved in dynamic paracrine and autocrine signaling cascades that have been shown to enhance tumor initiation and progression.

134

Jennifer M. Bailey & Steven D. Leach

of mice that developed precancerous lesions. Furthermore, the investigators noted a decrease in local fibrosis and inflammation (61). Thus, the ratio of TGFβR1/TGFβR2 may be a factor influencing tumor initiation and progression, associated with paracrine modulation of the desmoplastic response.

The effects of TGFβ on tumor initiation and progression are clearly complex. TGFβ has profound growth suppressive effects in the earliest stages of human cancer, and yet levels of TGFβ ligands and receptors are elevated in more advanced stage and metastatic human pancreatic cancer. The autocrine and paracrine signaling mechanisms by which TGFβ regulates cancer and stromal cells requires additional study in order to determine the most effective way to therapeutically target this pathway.

Conclusion Pancreatic cancer-associated desmoplasia represents a hypovascular environment consisting of a plethora of cell types, forming a complex barrier to the delivery of cytotoxic drugs. This hypovascular tumor microenvironment is rich in extracellular matrix, myofibroblasts, immune cells and endothelial cells. Specifically, paracrine interactions between tumor cells and pancreatic stellate cells enhances the mitogenic and invasive properties of tumor cells. Pancreatic cancer cells produce SHH and TGFβ, and also express Notch ligands and receptors. Pancreatic cancer cells also express PDGF, ET-1 and serpine-2, all of which promote the activated pancreatic stellate cell phenotype. In return, pancreatic stellate cells release a number of factors and stimuli such as PDGF, FGF, SPARC, MMPs, EGF, HGF and IGF, all of which modulate epithelial tumor growth. Pancreatic stellate cells also sectrete a variety of factors that alter the extracellular matrix including collagen type I, fibronectin, periostin and small leucine-rich proteoglycans. These alterations create a reactive microenvironment that aids in tumor growth, invasion and metastasis, and increases the resistance of pancreatic tumor cells to chemotherapy. The complexity of signaling interactions between all the different cell types present in the pancreatic cancer microenvironment presents an ongoing challenge to scientists to discover pathways critical for the maintenance of not only tumor cells, but also their associated stroma. Targeting individual pathways that are essential paracrine regulators of cancer cell survival, proliferation, differentiation and invasion may represent an effective strategy for providing pancreatic cancer patients with improved survival and eventual long-term cure.

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

135

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Ide T, Kitajima Y, Miyoshi A, et al. Tumor-stromal cell interaction under hypoxia increases the invasiveness of pancreatic cancer cells through the hepatocyte growth factor/c-Met pathway. Int J Cancer 2006; 119: 2750-9.

Mahadevan D, Von Hoff DD. Tumor-stroma interactions in pancreatic ductal adenocarcinoma. Mol Cancer Ther 2007; 6: 1186-97.

Hwang RF, Moore T, Arumugam T, et al. Cancer-associated stromal fibroblasts promote pancreatic tumor progression. Cancer Res 2008; 68: 918-26.

Vonlaufen A, Joshi S, Qu C, et al. Pancreatic stellate cells: partners in crime with pancreatic cancer cells. Cancer Res 2008; 68: 2085-93.

Fujita H, Ohuchida K, Mizumoto K, et al. Tumor-stromal interactions with direct cell contacts enhance proliferation of human pancreatic carcinoma cells. Cancer Sci 2009; 100: 2309-17.

Schneiderhan W, Diaz F, Fundel M, et al. Pancreatic stellate cells are an important source of MMP-2 in human pancreatic cancer and accelerate tumor progression in a murine xenograft model and CAM assay. J Cell Sci 2007; 120: 512-9.

Bailey JM, Swanson BJ, Hamada T, et al. Sonic hedgehog promotes desmoplasia in pancreatic cancer. Clin Cancer Res 2008; 14: 5995-6004.

Bachem MG, Schunemann M, Ramadani M, et al. Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells.

Gastroenterology 2005; 128: 907-21.

Xu Z, Vonlaufen A, Phillips PA, et al. Role of pancreatic stellate cells in pancreatic cancer metastasis. Am J Pathol; 177: 2585-96.

Grzesiak JJ, Bouvet M. The alpha2beta1 integrin mediates the malignant phenotype on type I collagen in pancreatic cancer cell lines. Br J Cancer 2006; 94: 1311-9.

Infante JR, Matsubayashi H, Sato N, et al. Peritumoral fibroblast SPARC expression and patient outcome with resectable pancreatic adenocarcinoma. J Clin Oncol 2007; 25: 319-25.

Mantoni TS, Schendel RR, Rodel F, et al. Stromal SPARC expression and patient survival after chemoradiation for non-resectable pancreatic adenocarcinoma.

Cancer Biol Ther 2008; 7: 1806-15.

Chen G, Tian X, Liu Z, et al. Inhibition of endogenous SPARC enhances pancreatic cancer cell growth: modulation by FGFR1-III isoform expression. Br J Cancer; 102: 188-95.

Phillips PA, McCarroll JA, Park S, et al. Rat pancreatic stellate cells secrete matrix metalloproteinases: implications for extracellular matrix turnover. Gut 2003; 52: 275-82.

Shek FW, Benyon RC, Walker FM, et al. Expression of transforming growth factor-beta 1 by pancreatic stellate cells and its implications for matrix secretion and turnover in chronic pancreatitis. Am J Pathol 2002; 160: 1787-98.

Farrow B, Albo D, Berger DH. The role of the tumor microenvironment in the progression of pancreatic cancer. J Surg Res 2008; 149: 319-28.

136

Jennifer M. Bailey & Steven D. Leach

17. Kanno A, Satoh K, Masamune A, et al. Periostin, secreted from stromal cells, has biphasic effect on cell migration and correlates with the epithelial to mesenchymal transition of human pancreatic cancer cells. Int J Cancer 2008; 122: 2707-18.

18. Rohatgi R, Milenkovic L, Scott MP. Patched1 regulates hedgehog signaling at the primary cilium. Science 2007; 317: 372-6.

19. Corbit KC, Aanstad P, Singla V, Norman AR, Stainier DY, Reiter JF. Vertebrate Smoothened functions at the primary cilium. Nature 2005; 437: 1018-21.

20. Zeng H, Jia J, Liu A. Coordinated translocation of mammalian Gli proteins and suppressor of fused to the primary cilium. PLoS One; 5: e15900.

21. Haycraft CJ, Banizs B, Aydin-Son Y, Zhang Q, Michaud EJ, Yoder BK. Gli2 and Gli3 localize to cilia and require the intraflagellar transport protein polaris for processing and function. PLoS Genet 2005; 1: e53.

22. Kim J, Kato M, Beachy PA. Gli2 trafficking links Hedgehog-dependent activation of Smoothened in the primary cilium to transcriptional activation in the nucleus. Proc Natl Acad Sci U S A 2009; 106: 21666-71.

23. Bitgood MJ, McMahon AP. Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 1995; 172: 126-38.

24. Aubin J, Dery U, Lemieux M, Chailler P, Jeannotte L. Stomach regional specification requires Hoxa5-driven mesenchymal-epithelial signaling.

Development 2002; 129: 4075-87.

25. Apelqvist A, Ahlgren U, Edlund H. Sonic hedgehog directs specialised mesoderm differentiation in the intestine and pancreas. Curr Biol 1997; 7: 801-4.

26. Hebrok M, Kim SK, Melton DA. Notochord repression of endodermal Sonic hedgehog permits pancreas development. Genes Dev 1998; 12: 1705-13.

27. Kawahira H, Scheel DW, Smith SB, German MS, Hebrok M. Hedgehog signaling regulates expansion of pancreatic epithelial cells. Dev Biol 2005; 280: 111-21.

28. Hebrok M, Kim SK, St Jacques B, McMahon AP, Melton DA. Regulation of pancreas development by hedgehog signaling. Development 2000; 127: 4905-13.

29. Thomas MK, Rastalsky N, Lee JH, Habener JF. Hedgehog signaling regulation of insulin production by pancreatic beta-cells. Diabetes 2000; 49: 2039-47.

30. Kawahira H, Ma NH, Tzanakakis ES, McMahon AP, Chuang PT, Hebrok M. Combined activities of hedgehog signaling inhibitors regulate pancreas development. Development 2003; 130: 4871-9.

31. Strobel O, Rosow DE, Rakhlin EY, et al. Pancreatic duct glands are distinct ductal compartments that react to chronic injury and mediate Shh-induced metaplasia. Gastroenterology; 138: 1166-77.

32. Fendrich V, Esni F, Garay MV, et al. Hedgehog signaling is required for effective regeneration of exocrine pancreas. Gastroenterology 2008; 135: 621-31.

33. Thayer SP, di Magliano MP, Heiser PW, et al. Hedgehog is an early and late mediator of pancreatic cancer tumorigenesis. Nature 2003; 425: 851-6.

34. Berman DM, Karhadkar SS, Maitra A, et al. Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours. Nature 2003; 425: 846-51.

Signaling pathways mediating epithelial-mesenchymal crosstalk in pancreatic cancer

137

35. Feldmann G, Habbe N, Dhara S, et al. Hedgehog inhibition prolongs survival in a genetically engineered mouse model of pancreatic cancer. Gut 2008; 57: 1420-30.

36. Bailey JM, Mohr AM, Hollingsworth MA. Sonic hedgehog paracrine signaling regulates metastasis and lymphangiogenesis in pancreatic cancer. Oncogene 2009; 28: 3513-25.

37. Feldmann G, Dhara S, Fendrich V, et al. Blockade of hedgehog signaling inhibits pancreatic cancer invasion and metastases: a new paradigm for combination therapy in solid cancers. Cancer Res 2007; 67: 2187-96.

38. Feldmann G, Fendrich V, McGovern K, et al. An orally bioavailable smallmolecule inhibitor of Hedgehog signaling inhibits tumor initiation and metastasis in pancreatic cancer. Mol Cancer Ther 2008; 7: 2725-35.

39. Prasad NB, Biankin AV, Fukushima N, et al. Gene expression profiles in pancreatic intraepithelial neoplasia reflect the effects of Hedgehog signaling on pancreatic ductal epithelial cells. Cancer Res 2005; 65: 1619-26.

40. Yauch RL, Gould SE, Scales SJ, et al. A paracrine requirement for hedgehog signalling in cancer. Nature 2008; 455: 406-10.

41. Tian H, Callahan CA, DuPree KJ, et al. Hedgehog signaling is restricted to the stromal compartment during pancreatic carcinogenesis. Proc Natl Acad Sci U S A 2009; 106: 4254-9.

42. Olive KP, Jacobetz MA, Davidson CJ, et al. Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer.

Science 2009; 324: 1457-61.

43. Nakamura K, Sasajima J, Mizukami Y, et al. Hedgehog promotes neovascularization in pancreatic cancers by regulating Ang-1 and IGF-1 expression in bone-marrow derived pro-angiogenic cells. PLoS One; 5: e8824.

44. Kopinke D, Brailsford M, Shea JE, Leavitt R, Scaife CL, Murtaugh LC. Lineage tracing reveals the dynamic contribution of Hes1+ cells to the developing and adult pancreas. Development; 138: 431-41.

45. Miyamoto Y, Maitra A, Ghosh B, et al. Notch mediates TGF alpha-induced changes in epithelial differentiation during pancreatic tumorigenesis. Cancer Cell 2003; 3: 565-76.

46. Plentz R, Park JS, Rhim AD, et al. Inhibition of gamma-secretase activity inhibits tumor progression in a mouse model of pancreatic ductal adenocarcinoma.

Gastroenterology 2009; 136: 1741-9 e6.

47. Bray SJ. Notch signalling: a simple pathway becomes complex. Nat Rev Mol Cell Biol 2006; 7: 678-89.

48. Sethi N, Dai X, Winter CG, Kang Y. Tumor-derived JAGGED1 promotes osteolytic bone metastasis of breast cancer by engaging notch signaling in bone cells. Cancer Cell; 19: 192-205.

49. Stanger BZ, Stiles B, Lauwers GY, et al. Pten constrains centroacinar cell expansion and malignant transformation in the pancreas. Cancer Cell 2005; 8: 185-95.

138

Jennifer M. Bailey & Steven D. Leach

50. Jensen JN, Cameron E, Garay MV, Starkey TW, Gianani R, Jensen J. Recapitulation of elements of embryonic development in adult mouse pancreatic regeneration. Gastroenterology 2005; 128: 728-41.

51. Duncan AW, Rattis FM, DiMascio LN, et al. Integration of Notch and Wnt signaling in hematopoietic stem cell maintenance. Nat Immunol 2005; 6: 314-22.

52. Fogar P, Pasquali C, Basso D, et al. Transforming growth factor beta, fibrogenesis and hyperglycemia in patients with chronic pancreatitis. J Med 1998; 29: 277-87.

53. Ohnishi H, Miyata T, Yasuda H, et al. Distinct roles of Smad2-, Smad3-, and ERK-dependent pathways in transforming growth factor-beta1 regulation of pancreatic stellate cellular functions. J Biol Chem 2004; 279: 8873-8.

54. Friess H, Yamanaka Y, Buchler M, et al. Enhanced expression of transforming growth factor beta isoforms in pancreatic cancer correlates with decreased survival. Gastroenterology 1993; 105: 1846-56.

55. Friess H, Yamanaka Y, Buchler M, et al. Enhanced expression of the type II transforming growth factor beta receptor in human pancreatic cancer cells without alteration of type III receptor expression. Cancer Res 1993; 53: 2704-7.

56. Jones S, Zhang X, Parsons DW, et al. Core signaling pathways in human pancreatic cancers revealed by global genomic analyses. Science 2008; 321: 1801-6.

57. Schutte M, Rozenblum E, Moskaluk CA, et al. An integrated high-resolution physical map of the DPC/BRCA2 region at chromosome 13q12. Cancer Res 1995; 55: 4570-4.

58. Bardeesy N, Cheng KH, Berger JH, et al. Smad4 is dispensable for normal pancreas development yet critical in progression and tumor biology of pancreas cancer. Genes Dev 2006; 20: 3130-46.

59. Kojima K, Vickers SM, Adsay NV, et al. Inactivation of Smad4 accelerates Kras(G12D)-mediated pancreatic neoplasia. Cancer Res 2007; 67: 8121-30.

60. Ijichi H, Chytil A, Gorska AE, et al. Aggressive pancreatic ductal adenocarcinoma in mice caused by pancreas-specific blockade of transforming growth factor-beta signaling in cooperation with active Kras expression. Genes Dev 2006; 20: 3147-60.

61. Adrian K, Strouch MJ, Zeng Q, et al. Tgfbr1 haploinsufficiency inhibits the development of murine mutant Kras-induced pancreatic precancer. Cancer Res 2009; 69: 9169-74.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 139-155 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

8. Desmoplasia and chemoresistance in pancreatic cancer Clifford J. Whatcott, Richard G. Posner, Daniel D. Von Hoff and Haiyong Han Clinical Translational Research Division, The Translational Genomics Research Institute, Scottsdale, Arizona, 85259, USA

Abstract. The desmoplastic reaction is a prominent pathological characteristic of pancreatic cancer. Desmoplasia is marked by a dramatic increase in the proliferation of alpha-smooth muscle actinpositive fibroblasts and is also accompanied by increased deposition of many extracellular matrix components. Changes in stromal cell proliferation and the deposition of extracellular matrix components result in dramatic changes in overall tissue heterogeneity and elasticity, as well as accompanying interstitial fluid pressure. These changes have been suggested to contribute to chemoresistance in cancer. Chemoresistance brought about by desmoplasia has both biological and physiological causes and consequences. In this chapter, we discuss some of the origins of desmoplasia in pancreatic cancer, and how it might be contributing to resistance to current chemotherapeutic interventions.

Introduction Pancreatic ductal adenocarcinoma (PDAC) stands as the fourth leading cause of cancer-related death in the United States, with a 5-year survival rate Correspondence/Reprint request: Dr. Clifford J. Whatcott, Clinical Translational Research Division, The Translational Genomics Research Institute, Scottsdale, Arizona, 85259, USA. E-mail: cwhatcott@tgen.org

140

Clifford J. Whatcott et al.

of less than 5% (1). Factors contributing to this poor prognosis include both intrinsic and acquired chemoresistance to the current first-line therapy of choice, gemcitabine. There are multiple factors which contribute to chemoresistance observed in pancreatic cancer. Among them, desmoplasia and the tumor microenvironment (TME) are increasingly seen as major contributors to chemoresistance in PDAC. Since Paget’s seminal ‘seed’ and ‘soil’ hypothesis was put forward in order to reconcile measureable differences in the frequency of distal sites of metastasis, the capacity of the ‘soil,’ or tumor microenvironment, to promote tumor cell growth has been increasingly recognized (2). The tumor microenvironment, or stroma, is home to many components, both cellular and molecular. Unfortunately, the contribution of the stroma to disease in patients is not fully understood in pancreatic cancer (3).

Desmoplasia, which is a result of the proliferation of alpha-smooth muscle actin-positive fibroblasts (also known as the myofibroblast, or activated pancreatic stellate cell (PSC)) and increased deposition of extracellular matrix (ECM) components, leads to reduced elasticity of tumor tissue with a concomitant increase in tumor interstitial fluid pressure (IFP). Increased IFP results in a decreased rate of perfusion of therapeutic agents and consequently decreased efficacy (4). This physiological chemoresistance has been shown to be a major contributor to the reduced efficacy of chemotherapeutics in multiple tumor types (4). In addition, desmoplasia can result in multiple signaling cascades that increase biological chemoresistance to therapeutic agents. Thus, targeting components of the tumor stroma that contribute to desmoplasia, in combination with agents directly targeting the tumor cells, is gaining traction as a potential approach for overcoming resistance and improving efficacy (5). In this chapter, we will review some of the general concepts of desmoplasia and chemoresistance, as well as how desmoplasia might be contributing to chemoresistance in pancreatic cancer.

Desmoplasia Desmoplasia, which is also referred to as the desmoplastic reaction, is a fundamental characteristic of PDAC. Originating from the Greek words desmos meaning “band” or “fastening” and plassein which is to “form” or “mold,” desmoplasia as we know clinically manifests itself in two ways: (1) significant overproduction of extracellular matrix proteins, and (2) extensive proliferation of myofibroblast-like cells (6). The resulting dense and fibrous connective tissue therefore is comprised of both cellular and non-cellular components (Figure 1). The cellular components of desmoplasia can include, among others, stellate cell-derived myofibroblast-like cells and an assortment of infiltrating immune cell types. Of the non-cellular components, multiple

Desmoplasia and chemoresistance in pancreatic cancer

141

Figure 1. Cell-cell interactions and ECM deposition contributing to desmoplasia in pancreatic cancer. The desmoplastic reaction involves both proliferation of the fibroblast/myofibroblast population, as well as deposition of multiple ECM components. The fibroblast population is the primary source of ECM components.

Infiltrating immune cells and tumor cells also contribute via various signaling mechanisms to the fibrogenesis mediated by the pancreatic stellate cells (PSCs).

TGFβ, PDGF, and FGF2 are primary signaling molecules that increase both the proliferation and the fibrogenesis of the PSCs. These signaling molecules are secreted by both tumor epithelial cells and infiltrating immune cells. Autocrine signaling of TGFβ mediated by the PSCs can also contribute to cell proliferation and increased ECM deposition.

extracellular matrix (ECM) proteins, including collagen types I, III, and IV, fibronectin, laminin, hyaluronan, as well as the glycoprotein osteonectin (also secreted protein, acidic and rich in cysteine, or SPARC), have been identified (7, 8). Desmoplasia results from multiple intercellular and intracellular biological signaling events (see Figure 1). Reports have demonstrated that transforming growth factor β (TGFβ), basic fibroblast growth factor (FGF2),

142

Clifford J. Whatcott et al.

connective tissue growth factor (CTGF), and interleukin-1β (IL-1β) each stimulate ECM production. Platelet-derived growth factor (PDGF), however, has been shown to stimulate proliferation of the myofibroblast-like cell population (3). Importantly, both the cellular and non-cellular components of desmoplasia contribute to the pathogenesis of pancreatic cancer.

Cellular components Among the multitude of cell types found within the tumor stroma, the pancreatic stellate cell is the most proliferative and is the primary synthetic source of many of the extracellular matrix components in the tumor microenvironment (6, 9). As this topic was also discussed in Chapter 1, the role of the PSC in desmoplasia will only be covered briefly here. In its quiescent state, the pancreatic stellate cell is positive for vitamin A-containing lipid droplets and serves as a reservoir for vitamin A in the normal pancreas. Its activation from a quiescent to an activated state, including changes to its proliferation rate, morphology, and sensitivity to mitogenic factors, are all primary features of PDAC. While the activation process is not yet fully understood, many of the signaling events leading to stellate cell activation have been identified. Following activation, the PSC loses its lipid droplets, undergoes changes to its morphology, and upregulates alpha smooth muscle actin (3). Intercellular signaling originating from multiple cell types, including tumor cells and immune cells, contribute to this increased activation and proliferation of PSCs in the desmoplastic reaction.

Tumor-associated macrophages Rudolf Virchow first observed in 1863 that leukocytes could infiltrate tumor tissues, suggesting a possible connection between inflammation and cancer (10). In his study, he suggested that chronic inflammation may in fact lead to the development of cancer. In diseased tissue, the immune response leading to inflammation can stall and become locked in a chronic state that fosters fibrotic conditions which may stimulate tumorigenesis and tumor progression, including increased genomic instability and increased secretion of growth factors. It is generally accepted that the reactive oxygen species (ROS) generated by phagocytic cells, including both neutrophils and macrophages, can contribute to multiple types of genotoxicity, including DNA strand breaks or DNA base modifications. Many of the inflammatory cells that infiltrate tumors often localize primarily to the tumor stroma, including tumor-associated macrophages (TAM), neutrophils, and regulatory T cells (Treg) (11). TAMs are among the most populous immune cell type to

Desmoplasia and chemoresistance in pancreatic cancer

143

infiltrate into the tumor microenvironment (12). TAMs are recruited to tumor sites by chemoattractants, including the CC chemokines CCL2, 3, 4, 5, and 8.

Also, TAMs can be recruited via VEGF, colony stimulating factor 1 (CSF-1), stromal-derived factor 1 (SDF-1), and thrombospondin-1 (TSP-1) (13-15).

Once guided to the tumor and activated, macrophages secrete various growth factors and cytokines into the immediate tumor microenvironment, indirectly contributing to an altered stromal environment and enhanced desmoplasia (8). For example, once activated, TAMs are known to express the most potent mediator of PSC proliferation, PDGF, as well as to release ROS (16). In addition, macrophages also have the capacity to stimulate the synthesis of collagen type I and fibronectin in PSC via the release of TGFβ (17, 18).

Ultimately, infiltration and localization of leukocytes to the stromal compartment vastly alters its properties, and leads to increased desmoplasia.

Neutrophils and regulatory T cells Two additional cell types thought to contribute to desmoplasia in pancreatic cancer include the neutrophils and regulatory T (Treg) cells. While the neutrophils, along with the TAMs can function as phagocytes at sites of inflammation, they act as the first responders in acute inflammatory responses with the release of toxic granules that are a significant source of ROS. ROS are a significant source of genomic instability and indeed have been shown to promote carcinogenesis in pancreatic ductal epithelial cells (19, 20). However, ROS can also directly result in the activation of PSCs. Furthermore, ROS can stimulate the spontaneous, non-enzymatic activation of TGFβ in vivo (21). By release of ROS, therefore, neutrophils may be contributing to the mobilization of TGFβ signaling. In addition, Nozawa et al. suggest that infiltrating neutrophils mediate the angiogenic switch and ECM remodeling through their expression of MMP-9 (22). In this way, neutrophils may contribute to an altered stromal compartment supportive of tumor progression. Treg cells play a significant role in the immune response to cancer. While Treg cells do not necessarily induce carcinogenesis or mutagenesis, they play an important role in suppressing an effective response that the immune system may attempt to mount. Reports have demonstrated that the prevalence of Treg cells is increased in patients with PDAC, suggesting that they may play a significant role in suppressing an immune response in PDAC (23, 24). In addition, Treg cells express and release multiple factors, including TGFβ, into the tumor microenvironment and have been shown to exhibit potent anti-inflammatory effects and participate in suppressing auto-immunity (25-28). The Treg cell population, which is defined as CD4+CD25+Foxp3+, expresses both active and latent TGFβ at it its

144

Clifford J. Whatcott et al.

cell surface that is absent in CD4+CD25- cells (31). Though it does not appear that TGFβ expression is entirely necessary for its immunosuppressive effects, the release of Treg-derived TGFβ may further contribute to desmoplasia in PDAC following receptor binding on the PSC (18, 29).

The cellular components of desmoplasia include the PSCs and infiltrating immune cells. In its normal quiescent state, the pancreatic stellate cell functions in the metabolism of retinoids, the development of the pancreas, and in immune cell regulation by means of its antigen presentation abilities (30). In PDAC, however, proliferating stromal cells prevent the normal function of the pancreas. Utilizing various signaling mechanisms, both tumor and immune cells contribute to the dramatic proliferation of the stromal cells.

In contrast, non-cellular components of desmoplasia arise from increased deposition of extracellular matrix components.

Non-cellular components While desmoplasia is characterized by a marked increase in the proliferation of the myofibroblasts and the infiltration of multiple immune cell types, the deposition of ECM components contributes significantly to the pathogenic potential of the desmoplastic reaction. The non-cellular components of desmoplasia consist largely of the extracellular matrix components: the fibrous proteins (e.g. collagen) and the polysaccharide chain glycosaminoglycans (GAGs, including hyaluronan) (Table 1) (31-41). In the normal pancreas, the GAGs function structurally to support compressive forces on the tissue. Conversely, the fibrous proteins serve to support any tensile forces on the tissue (33). The various ECM components also serve to allow for proper diffusion of various molecules, including nutrients and hormones, at the bloodstream:tissue interface. In the diseased pancreas, the significant over-production of ECM components can be described as the Table 1. Extracellular matrix components and their contribution to chemoresistance.

Desmoplasia and chemoresistance in pancreatic cancer

145

failed resolution of a healing wound, which leads to fibrosis. TGFβ-mediated cell signaling has been recognized as the primary signaling mechanism for fibrogenesis (42). Here we will discuss some of the signaling mechanisms contributing to the increased deposition of the ECM components of desmoplasia and how they might contribute to the pathogenesis of PDAC. Immunohistochemical analyses of PDAC have reported increased expression of collagen, including types I, III, and IV, localizing to the extracellular space (43-45). Collagen type I is composed of two alpha1 and one alpha2 protein subunits which are encoded by the COL1A1 and COL1A2 genes, respectively (46). Following translation of pro-collagen peptide chains, the peptides are processed in the endoplasmic reticulum before being secreted via the Golgi apparatus. Subsequent processing by extracellular proteases allows the collagen to coalesce into fibrils (on the order of 1μm in diameter) and collagen fibers (up to 10μm in diameter) (33). The natural arrangement of crosslinking within the collagen fibrils makes the collagen fibers highly resistant to tensile forces, and capable of dissipating significant deformation energy (47). Collagen expression is upregulated in response to TGFβ/Smad signaling and is a product of activated fibroblasts (3, 46). Importantly, collagen fibers can function in concert with the fibroblast-mediated tissue contraction to help in providing structural support in closing and healing wounds (42). In PDAC, this same function of collagen decreases tissue elasticity and increases interstitial fluid pressure, resulting in reduced drug perfusion.

Hyaluronan, a protein-free GAG, is also an important component of the ECM and contributes to tissue rigidity and decreased elasticity (34, 48).

Hyaluronan (HA) accumulation within diseased tissue is the product of increased synthesis and is secreted by activated fibroblasts in pancreatic cancer (49). Although increased expression of hyaluronan synthases in tumor cells has been observed, the overall regulation of hyaluronan synthesis is not fully understood (50). HA maintains significant interaction with water molecules, and functions to preserve tissue hydration in the normal pancreas.

However, in diseased tissue and with increased deposition of HA, interstitial edema can ensue. Consequently, interstitial fluid pressure increases, and the ability of the interstitia to conduct fluid is decreased (51). The lack of a functional lymphatic system in tumor tissue combines with this increased interstitial edema brought about by HA to significantly reduce the exchange of various molecules at the bloodstream:tissue interface, including chemotherapeutics. The significant accumulation of HA in PDAC is one characteristic feature of pancreatic cancer that contributes to the observed desmoplastic reaction, and a decrease of chemotherapeutic penetration.

The deposition of multiple ECM components in the desmoplastic reaction in PDAC is a characteristic that contributes significantly to the

146

Clifford J. Whatcott et al.

pathogenesis of tumors and ultimately to chemoresistance. Both the proliferation of the fibroblast cell compartment and the increased deposition of ECM components contribute to the desmoplastic reaction and subsequent chemoresistance. Importantly, these cellular and non-cellular components of desmoplasia contribute to biological and physiological forms of chemoresistance in PDAC.

Desmoplasia induced chemoresistance Chemoresistance can occur by multiple mechanisms in cancer. It can arise from physiological barriers to drug absorption or penetration into target tissues, or from biological mechanisms within individual tumor cells which reduce the effectiveness at their intended site of action. Biological chemoresistance appears to arise as the result of three different general types of mechanisms: first, target cells can develop resistance to drug uptake; second, target cells can develop altered sensitivity to drugs at their intended targets through, for example, increased expression of anti-apoptotic proteins; and third, target cells can develop increased efflux of drugs, preventing the drugs from reaching their intended site of action (52). Physiological chemoresistance can include poor tissue vasculature or perfusion, which may result from increased interstitial fluid pressure and increased production of extracellular matrix proteins that arises from the desmoplastic reaction (53).

Chemoresistance is a problem in pancreatic cancer due to the presence of both biological and physiological mechanisms. In this section, we will review some of these mechanisms, how desmoplasia can contribute to these forms of chemoresistance, and approaches to overcoming chemoresistance therapeutically.

Biological chemoresistance Biological chemoresistance has been well documented in multiple tumor types. Reduced uptake of drugs is one mechanism by which cells resist chemotherapeutic activity. A well documented example of such an effect is seen in the resistance to such cancer chemotherapeutics as the cytotoxic folate analogs, including methotrexate. Mutations arising in either the folate binding protein, or the reduced folate transporter give way to decreased transport of methotrexate and subsequent chemoresistance (54). In pancreatic cancer, epigenetic changes in the epithelial tumor cells, resulting in reduced expression of the nucleoside transporter required for gemcitabine uptake, hENT1, has also been reported to contribute to chemoresistance (55).

Increased efflux of drugs by target cells, including the expression of the ATP-

Desmoplasia and chemoresistance in pancreatic cancer

147

binding cassette (ABC) efflux transporter family of proteins, is one way in which cells can exhibit chemoresistance to a number of therapeutics. Indeed, expression of ABC transporters such as MDR1 or MRP1-6 can evoke resistance to numerous hydrophobic natural-products, such as doxorubicin, vinblastine, and paclitaxel, and to the nucleoside analogs, respectively (56-58), by reducing intracellular drug concentrations. Both of these proteins are widely expressed in human cancers. Drug resistance has also been observed resulting from altered cellular sensitivity to drugs’ mechanisms of action. For example, increased expression of anti-apoptotic proteins such as the Bcl-2 family of proteins confers a growth advantage to tumor cells but also confers drug resistance (59). Furthermore, increased drug metabolism, increased repair of DNA damage, altered intracellular drug targets, and altered cell cycle checkpoints have all been identified as potential mechanisms of biological chemoresistance (57). The non-cellular components of desmoplasia contribute to the biological chemoresistance through multiple mechanisms, including stimulating the expression of ABC multidrug resistance transporters in tumor cells. One report demonstrated that the binding of HA to its receptor, CD44, resulted in a Stat-3-mediated increase in the expression of MDR1 in breast and ovarian tumor cell lines (60). Importantly, similar observations of induced chemoresistance by CD44 activity have also been reported in pancreatic cancer cell lines (61). In addition, HA interaction with CD44 has also been shown to activate the phosphoinositide 3-kinase/AKT signaling pathway. Activation of AKT signaling is upregulated in many cancers, resulting in phosphorylation of Bad, and downregulation of apoptosis (62-64). In this sense, desmoplasia and chemoresistance are tightly regulated and interwoven. Factors stimulating desmoplasia in fact result in chemoresistance on multiple levels, both biologically and physiologically.

Physiological chemoresistance Physiological chemoresistance has also been well documented in cancer, though studies of its clinical relevance are made difficult by the lack of good models to study their effects. Physiological chemoresistance, also referred to as “host factors� by Gottesman (57), can develop following biological changes in tumor cells but arises largely as a consequence of extracellular changes that occur in the stromal compartment. Physiological chemoresistance can arise from, among others, increased interstitial fluid pressure (as a consequence of the desmoplastic reaction), increased extracellular matrix protein deposition afforded by the expansion of the stromal compartment, or from the development of a disorganized tumor

148

Clifford J. Whatcott et al.

vasculature that gives rise to poor tumor perfusion. The mechanisms that lead to some of these effects, including increases in IFP, are not well understood.

Tumor extracellular matrix components contribute significantly to physiological chemoresistance in cancer. Collagens I, III, and IV are secreted at high levels in PDAC. Tumor collagen content, specifically, has been shown to maintain an inverse relationship to that of macromolecule penetration. In one report it was demonstrated that a U87 (human glioblastoma) xenograft tumor containing high collagen content displayed a significantly decreased diffusion coefficient (0.87 versus 1.89 Ă— 10-7 cm2 s-1) of IgG molecules compared to a LS174T (human colon adenocarcinoma) xenograft tumor containing little collagen (65). Further confirming the specific contribution of collagen, it was noted that collagenase treated U87 tumors showed diffusion coefficients similar to the untreated LS174T tumors with low collagen content. Similar effects have also been observed in pancreatic cancer. Diop-Frimpong and colleagues demonstrated that the collagen-reducing effects of the anti-hypertensive agent, losartan, resulted in a significant increase in the penetration of nanoparticles (66). Using an orthotopic xenograft mouse model, they demonstrated that the increased collagen content, and decreased penetration of nanoparticles, in their L3.6pl tumors corresponded to a significantly decreased response to doxorubicin treatment. Taken together, these results suggest that collagen content may be sufficient to predict the delivery of drugs to tumor, and to predict physiological resistance to therapy.

Hyaluronan has also been shown to contribute to physiological chemoresistance in cancer. As a component of the ECM, HA contributes to the structural rigidity of the network of ECM proteins and may act as a molecular sieve to molecules extravasating from the vasculature (67). In acting as a molecular sieve, HA is thought to reduce the penetration of chemotherapeutics in PDAC. There is, however, some debate as to the extent or role that HA may play in tissue diffusion. Indeed, it has been shown that tissue elasticity and hydraulic conductivity do not correlate with total tissue hyaluronan content (65). Further, it has been shown that depletion of tissue HA by hyaluronidase decreased the diffusion of large molecules, namely FITC-dextran (500kDa molecular weight) (34). However, additional studies have demonstrated that the molecular selectivity of the ECM is tightly regulated through variation of the ECM composition, and that this affects molecule penetration based on charge and size (68). Furthermore, it has been demonstrated that depletion of HA in a PC3 human prostate xenograft model both decreased tumor IFP and increased vascular area (69). These studies suggest that HA can indeed increase IFP in tumors, but also that its ability to

Desmoplasia and chemoresistance in pancreatic cancer

149

elicit a decrease in the penetration of chemotherapeutics will vary depending on the charge and size characteristics of the drug used.

In their paper on the potential efficacy of hedgehog inhibitors in pancreatic cancer, Olive et al outlined several careful experiments demonstrating in a transgenic mouse model for PDAC that tumors arising spontaneously within the pancreas are in fact poorly vascularized, and show poor drug perfusion relative to transplanted xenograft tumors or normal pancreatic tissue (70). Known most widely for its role in embryonic development, the hedgehog pathway has also been shown to play important roles in angiogenesis. This angiogenic potential of the hedgehog pathway has been reported in ischemic limbs and in the cornea of adult mice, or in rat models of diabetic neuropathy (71, 72). However, as shown by Olive and colleagues, the clinical efficacy of the hedgehog inhibitors in treating PDAC and in increasing overall survival is working through a mechanism that is not anti-angiogenic, but rather potently pro-perfusion. The researchers demonstrated that the hedgehog signaling inhibition increased endothelial cell proliferation and decreased fibroblast proliferation. The results presented in this study indicate that increased perfusion and decreased fibrosis are potentially important in treating PDAC. Certainly, however, more thorough studies characterizing the role of hedgehog inhibition in both pancreatic tumor and stromal cells will be necessary and will help translate the findings into the clinic.

Mechanisms of chemoresistance arise from both physiological barriers and biological adaptations of the tumor epithelial cells. Considering the number of obstacles to effective and penetrating treatment, we propose that future regimens consider a multipronged approach to targeting pancreatic cancer. By targeting the pathways involved in desmoplasia, in combination with cytotoxic agents to inhibit tumor cell growth, we would anticipate increased efficacy resulting in greater survival in pancreatic cancer patients.

Potential therapeutic approaches for reducing desmoplasia induced chemoresistance Many approaches have been pursued in hopes of increasing the efficacy of chemotherapeutics against tumor epithelial cells in pancreatic cancer.

Unfortunately, the desmoplastic reaction may abrogate much of the therapeutic activity of test agents. Indeed, we have observed significant chemoresistance to many therapeutics by primary human fibroblasts isolated from pancreatic tumors, even in the absence of a supportive microenvironment and collagen or hyaluronan network (unpublished data).

Therapeutics reducing the contribution of the desmoplastic reaction to

150

Clifford J. Whatcott et al.

chemoresistance are being actively pursued as a potential therapeutic approach (5).

Members of the TGFβ family of mediators have been implicated in the activation of the PSC. TGFβ proteins are multifunctional cytokines known to be involved in a host of cellular functions, including regulation of immune cell function, cell growth and differentiation, and in ECM production (73).

The canonical TGFβ signaling involves the binding to TGFβ receptors (TGFβRI/II) and subsequent activation of downstream Smad-mediated transcription. Since Smad4/DPC4 is reportedly mutated or deleted in approximately 55% of pancreatic tumors, it is unclear mechanistically how TGFβ antagonists may offer clinical benefit (74). Indeed, studies have demonstrated the complex nature of TGFβ signaling, demonstrating both tumor-suppressive and tumor-promoting features of TGFβ signaling (75).

Despite the complexity and multifunctional nature of the signaling pathways, recent studies have indicated that interventions aiming at TGFβ signaling can have therapeutic benefit, without the danger of the potential side-effects that may result from inhibition of its tumor suppressive activity (reviewed in 76).

Indeed, both the TGFβ antagonist, SR2F, and a TGFβ-neutralizing antibody were reportedly administered in animals with few side effects (77, 78). These studies have demonstrated that TGFβ antagonism can prevent metastases, as well as prevent cerulein-induced fibrosis and TGFβ signaling (77, 79, 80). As it appears that TGFβR1 haploinsufficiency can itself significantly inhibit the development of fibrosis and progression of precancerous lesions in a mouse model for PDAC, efforts have been made to examine the effects of TGFβ inhibition on the secretion of ECM components in fibroblast cells (80, 81).

Multiple studies taking different approaches have observed decreased ECM (fibronectin, collagen, and others) deposition by fibroblasts following inhibition of TGFβ-mediated signaling (80, 82, 83). While it is important to note that the authors commented on the potential importance of the TGFβR1/TGFβR2 ratio on downstream signaling and outcome, inhibition of TGFβ-mediated signaling may yet be a viable pathway for therapeutic intervention if, as Yingling and colleagues suggested, “patient populations in which the tumor-promoting role of TGFβ signaling predominates” are identified (76, 81). Furthermore, due to the important nature of TGFβ in the perpetuation of PSC activation, recent studies have focused on utilizing TGFβ antagonists in targeting pancreatic fibrosis as a result of chronic pancreatitis (84). The translation of these findings to the clinic could potentially make a significant impact on the treatment of patients with PDAC. As reduction in PSC load and fibrotic buildup comes as a result of TGFβ antagonism, we would hope to see significant increases in the penetration of drugs targeting tumor cells.

Desmoplasia and chemoresistance in pancreatic cancer

151

In addition to the TGFβ antagonists, the peroxisome proliferatoractivated receptor γ (PPARγ) agonists are also being investigated for its antidesmoplasia activity. PPARγ has been studied in a variety of conditions, but is most recognized in its ability to increase insulin sensitivity (85, 86).

Currently, the PPARγ ligand troglitazone (also known as Rezulin) is used clinically as an oral antihyperglycemic agent for the management of type II diabetes. Other ligands for PPARγ include lysophosphatidic acid, a few specific prostaglandins, and the thiazolidinediones (87-89). Besides insulin biology, several groups have shown that PPARγ is also involved in, for example, cell differentiation, inflammation regulation, and fibrosis in the lung (90-92). Following activation, its anti-fibrotic capabilities appear to be mediated by its ability to inhibit TGFβ signaling, as is especially noted in its ability to inhibit differentiation of fibroblasts to myofibroblasts (93). By activating PPARγ via a natural or synthetic ligand, fibrosis can be reduced and clinical benefit achieved through inhibition of the stromal contribution made by myofibroblasts such as their increased collagen production and increased matrix metalloproteinase expression. Indeed, improved organ function and decreased fibrosis has been reported in steatohepatitis patients being treated with pioglitazone, a PPARγ agonist, for diabetes (94).

The target in the aforementioned approaches is clearly the stromal compartment, specifically the myofibroblast, and would function as part of a larger therapeutic regimen. In addition to these approaches, enzymatic degradation of ECM components has also been proposed. Though the collagenases have been shown to enhance the penetration of macromolecules, sensitivity to and stability in physiological pH may be a therapeutic hindrance (34). To our knowledge, no collagenase is yet clinically available. In contrast, hyaluronidase has been used clinically in multiple trials, and has been shown to enhance tumor sensitivity to chemotherapeutics, even in tumors deemed chemoresistant (95). Importantly, hyaluronidase treatment of cultured tumor cells has been shown to enhance the penetration of drug into cultured spheroid cells (96). Early clinical trials utilizing hyaluronidase employed a bovine form of the enzyme from which allergic reactions had been reported.

Currently, a recombinant human hyaluronidase has been made available for early pilot clinical studies and has been shown not to induce significant allergic reactions (97). The availability of the human recombinant hyaluronidase is particularly exciting as many of the early clinical studies demonstrating increased chemotherapeutic efficacy of anti-tumor agents administered in combination with bovine hyaluronidase may be expanded upon. It is our hope that such agents will effect greater therapeutic outcomes for patients.

152

Clifford J. Whatcott et al.

Conclusions Desmoplasia is indeed a fundamental characteristic of pancreatic cancer that contributes significantly to its chemoresistance. Deposition of ECM proteins such as collagen and hyaluronan contributes physiologically to chemoresistance by mediating a decrease in macromolecular diffusion in pancreatic tissues. Their deposition also mediates biological signaling cascades that result in, for example, decreased apoptotic signaling by Bad. Therefore, targeting desmoplasia, which is the response of the tumor microenvironment to molecular cues from the epithelial cell compartment, constitutes a viable, combinatorial therapeutic strategy in pancreatic cancer.

Whether through enzymatic digestion of ECM components, or inhibitors of myofibroblast activation, targeting desmoplasia in pancreatic cancer will likely improve outcomes for patients with pancreatic cancer where chemoresistance remains a significant hurdle to effective systemic treatment.

Acknowledgements Research in the authors’ laboratories is supported by grants from the NIH/NCI (CA109552, CA140924, and CA095031), Stand Up to Cancer (SU2C), and the Katz Family Foundation.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Jemal, A., Siegel, R., Xu, J., Ward, E. 2010, CA Cancer J Clin, 60, 277.

Paget, S. 1889, The Lancet, 133, 571.

Shimizu, K. 2008, J Gastroenterol, 43, 823.

Heldin, C. H., Rubin, K., Pietras, K., Ostman, A. 2004, Nat Rev Cancer, 4, 806.

Garber, K. 2010, J. Natl. Cancer Inst., 102, 448.

Yen, T. W., Aardal, N. P., Bronner, M. P., et al. 2002, Surgery, 131, 129.

Bachem, M. G., Schneider, E., Gross, H., et al. 1998, Gastroenterology, 115, 421.

Apte, M. V., Haber, P. S., Darby, S. J., et al. 1999, Gut, 44, 534.

Faouzi, S., Le Bail, B., Neaud, V., et al. 1999, J Hepatol, 30, 275.

Balkwill, F., Mantovani, A. 2001, Lancet, 357, 539.

Negus, R. P., Stamp, G. W., Hadley, J., Balkwill, F. R. 1997, Am J Pathol, 150, 1723.

Solinas, G., Germano, G., Mantovani, A., Allavena, P. 2009, J Leukoc Biol, 86, 1065.

Murdoch, C., Giannoudis, A., Lewis, C. E. 2004, Blood, 104, 2224.

Coffelt, S. B., Hughes, R., Lewis, C. E. 2009, Biochim Biophys Acta, 1796, 11.

Martin-Manso, G., Galli, S., Ridnour, L. A., Tsokos, M., Wink, D. A., Roberts, D. D. 2008, Cancer Res, 68, 7090.

Desmoplasia and chemoresistance in pancreatic cancer

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

153

Ross, R. 1989, Lancet, 1, 1179.

Schmid-Kotsas, A., Gross, H. J., Menke, A., et al. 1999, Am J Pathol, 155, 1749.

Aoyagi, Y., Oda, T., Kinoshita, T., et al. 2004, Br J Cancer, 91, 1316.

Toyokuni, S., Okamoto, K., Yodoi, J., Hiai, H. 1995, FEBS Lett, 358, 1.

Vaquero, E. C., Edderkaoui, M., Pandol, S. J., Gukovsky, I., Gukovskaya, A. S. 2004, J Biol Chem, 279, 34643.

Koli, K., Myllarniemi, M., Keski-Oja, J., Kinnula, V. L. 2008, Antioxid Redox Signal, 10, 333.

Nozawa, H., Chiu, C., Hanahan, D. 2006, Proc Natl Acad Sci U S A, 103, 12493.

Liyanage, U. K., Moore, T. T., Joo, H. G., et al. 2002, J Immunol, 169, 2756.

Liyanage, U. K., Goedegebuure, P. S., Moore, T. T., et al. 2006, J Immunother, 29, 416.

Li, M. O., Sanjabi, S., Flavell, R. A. 2006, Immunity, 25, 455.

Li, M. O., Wan, Y. Y., Sanjabi, S., Robertson, A. K., Flavell, R. A. 2006, Annu Rev Immunol, 24, 99.

Moore, K. W., de Waal Malefyt, R., Coffman, R. L., O'Garra, A. 2001, Annu Rev Immunol, 19, 683.

Miyara, M., Sakaguchi, S. 2007, Trends Mol Med, 13, 108.

Kullberg, M. C., Hay, V., Cheever, A. W., et al. 2005, Eur J Immunol, 35, 2886.

Friedman, S. L. 2008, Physiol Rev, 88, 125.

Miyamoto, H., Murakami, T., Tsuchida, K., Sugino, H., Miyake, H., Tashiro, S. 2004, Pancreas, 28, 38.

McKee, T. D., Grandi, P., Mok, W., et al. 2006, Cancer Res, 66, 2509.

Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., Watson, J. Extracellular matrix of animals. Molecular Biology of the Cell. 3rd Ed. ed. New York: Garland Publishing; 1994. p. 971.

Magzoub, M., Jin, S., Verkman, A. S. 2008, FASEB J, 22, 276.

Stern, R. 2008, Semin Cancer Biol, 18, 275.

Ricciardelli, C., Sakko, A. J., Ween, M. P., Russell, D. L., Horsfall, D. J. 2009, Cancer Metastasis Rev, 28, 233.

Marastoni, S., Ligresti, G., Lorenzon, E., Colombatti, A., Mongiat, M. 2008, Connect Tissue Res, 49, 203.

Kaspar, M., Zardi, L., Neri, D. 2006, Int J Cancer, 118, 1331.

Sethi, T., Rintoul, R. C., Moore, S. M., et al. 1999, Nat Med, 5, 662.

Mahadevan, D., Von Hoff, D. D. 2007, Mol Cancer Ther, 6, 1186.

Desai, N., Trieu, V., Damascelli, B., Soon-Shiong, P. 2009, Transl Oncol, 2, 59.

Raghow, R. 1991, Chest, 99, 61S. Imamura, T., Iguchi, H., Manabe, T., et al. 1995, Pancreas, 11, 357.

Mollenhauer, J., Roether, I., Kern, H. F. 1987, Pancreas, 2, 14.

Linder, S., Castanos-Velez, E., von Rosen, A., Biberfeld, P. 2001, Hepatogastroenterology, 48, 1321.

Verrecchia, F., Mauviel, A. 2004, Cell Signal, 16, 873.

Buehler, M. J. 2006, Proc Natl Acad Sci U S A, 103, 12285.

Laurent, T. C., Fraser, J. R. 1992, FASEB J, 6, 2397.

Hiltunen, E. L., Anttila, M., Kultti, A., et al. 2002, Cancer Res, 62, 6410.

154

Clifford J. Whatcott et al.

50. Itano, N., Sawai, T., Atsumi, F., et al. 2004, J Biol Chem, 279, 18679.

51. Levick, J. R. 1987, Q J Exp Physiol, 72, 409.

52. Szakacs, G., Paterson, J. K., Ludwig, J. A., Booth-Genthe, C., Gottesman, M. M. 2006, Nat Rev Drug Discov, 5, 219.

53. Minchinton, A. I., Tannock, I. F. 2006, Nat Rev Cancer, 6, 583.

54. Longo-Sorbello, G. S., Bertino, J. R. 2001, Haematologica, 86, 121.

55. Giovannetti, E., Del Tacca, M., Mey, V., et al. 2006, Cancer Res, 66, 3928.

56. Borst, P., Evers, R., Kool, M., Wijnholds, J. 1999, Biochim Biophys Acta, 1461, 347.

57. Gottesman, M. M. 2002, Annu. Rev. Med., 53, 615.

58. Higgins, C. F. 1992, Annu Rev Cell Biol, 8, 67.

59. Hamacher, R., Schmid, R. M., Saur, D., Schneider, G. 2008, Mol Cancer, 7, 64.

60. Bourguignon, L. Y., Peyrollier, K., Xia, W., Gilad, E. 2008, J Biol Chem, 283, 17635.

61. Hong, S. P., Wen, J., Bang, S., Park, S., Song, S. Y. 2009, Int J Cancer, 125, 2323.

62. Fujita, Y., Kitagawa, M., Nakamura, S., et al. 2002, FEBS Lett, 528, 101.

63. Benitez, A., Yates, T. J., Lopez, L. E., Cerwinka, W. H., Bakkar, A. A., Lokeshwar, V. B. 2011, Cancer Res, 71, 4085.

64. Mitsiades, C. S., Mitsiades, N., Koutsilieris, M. 2004, Curr Cancer Drug Targets, 4, 235.

65. Netti, P. A., Berk, D. A., Swartz, M. A., Grodzinsky, A. J., Jain, R. K. 2000, Cancer Res, 60, 2497.

66. Diop-Frimpong, B., Chauhan, V. P., Krane, S., Boucher, Y., Jain, R. K. 2011, Proc Natl Acad Sci U S A, 108, 2909.

67. Stern, R. 2004, Eur J Cell Biol, 83, 317.

68. Lieleg, O., Baumgartel, R. M., Bausch, A. R. 2009, Biophys J, 97, 1569.

69. Thompson, C. B., Shepard, H. M., O'Connor, P. M., et al. 2010, Mol Cancer Ther, 9, 3052.

70. Olive, K. P., Jacobetz, M. A., Davidson, C. J., et al. 2009, Science, 324, 1457.

71. Pola, R., Ling, L. E., Silver, M., et al. 2001, Nat Med, 7, 706.

72. Kusano, K. F., Allendoerfer, K. L., Munger, W., et al. 2004, Arterioscler Thromb Vasc Biol, 24, 2102.

73. Naber, H. P., ten Dijke, P., Pardali, E. 2008, Curr Cancer Drug Targets, 8, 466.

74. Hahn, S. A., Schutte, M., Hoque, A. T., et al. 1996, Science, 271, 350.

75. Wakefield, L. M., Roberts, A. B. 2002, Curr Opin Genet Dev, 12, 22.

76. Yingling, J. M., Blanchard, K. L., Sawyer, J. S. 2004, Nat Rev Drug Discov, 3, 1011.

77. Yang, Y. A., Dukhanina, O., Tang, B., et al. 2002, J Clin Invest, 109, 1607.

78. Ruzek, M. C., Hawes, M., Pratt, B., et al. 2003, Immunopharmacol Immunotoxicol, 25, 235.

79. Muraoka, R. S., Dumont, N., Ritter, C. A., et al. 2002, J Clin Invest, 109, 1551.

80. Zion, O., Genin, O., Kawada, N., et al. 2009, Pancreas, 38, 427.

81. Adrian, K., Strouch, M. J., Zeng, Q., et al. 2009, Cancer Res, 69, 9169.

Desmoplasia and chemoresistance in pancreatic cancer

155

82. Mulsow, J. J., Watson, R. W., Fitzpatrick, J. M., O'Connell, P. R. 2005, Ann Surg, 242, 880.

83. Fu, K., Corbley, M. J., Sun, L., et al. 2008, Arterioscler Thromb Vasc Biol, 28, 665.

84. Nagashio, Y., Ueno, H., Imamura, M., et al. 2004, Lab Invest, 84, 1610.

85. Kobayashi, M., Iwanishi, M., Egawa, K., Shigeta, Y. 1992, Diabetes, 41, 476.

86. Kellerer, M., Kroder, G., Tippmer, S., et al. 1994, Diabetes, 43, 447.

87. McIntyre, T. M., Pontsler, A. V., Silva, A. R., et al. 2003, Proc Natl Acad Sci U S A, 100, 131.

88. Forman, B. M., Tontonoz, P., Chen, J., Brun, R. P., Spiegelman, B. M., Evans, R. M. 1995, Cell, 83, 803.

89. Shimizu, K., Shiratori, K., Hayashi, N., Kobayashi, M., Fujiwara, T., Horikoshi, H. 2002, Pancreas, 24, 184.

90. Burgess, H. A., Daugherty, L. E., Thatcher, T. H., et al. 2005, Am J Physiol Lung Cell Mol Physiol, 288, L1146.

91. Buckingham, R. E. 2005, Hepatol Res, 33, 167.

92. Hummasti, S., Tontonoz, P. 2006, Mol Endocrinol, 20, 1261.

93. Sime, P. J. 2008, J Investig Med, 56, 534.

94. Belfort, R., Harrison, S. A., Brown, K., et al. 2006, N Engl J Med, 355, 2297.

95. Baumgartner, G., Gomar-Hoss, C., Sakr, L., Ulsperger, E., Wogritsch, C. 1998, Cancer Lett, 131, 85.

96. Kohno, N., Ohnuma, T., Truog, P. 1994, J Cancer Res Clin Oncol, 120, 293.

97. Yocum, R. C., Kennard, D., Heiner, L. S. 2007, J Infus Nurs, 30, 293.

Transworld Research Network 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India

Pancreatic Cancer and Tumor Microenvironment, 2012: 157-184 ISBN: 978-81-7895-548-3 Editors: Paul J. Grippo and Hidayatullah G. Munshi

9. Therapeutic targeting of pancreatic stroma Andrew S. Liss and Sarah P. Thayer

Department of Surgery and the Andrew L. Warshaw, MD, Institute for Pancreatic Cancer Research, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02114, USA

Abstract. Pancreatic ductal adenocarcinoma (PDAC) is the fourth highest cause of cancer-related deaths in the United States, and for the vast majority of patients chemotherapy is the only course of treatment. The recent use FOLFIRINOX has provided the most significant survival benefit (~4 months) of any treatment brought to the clinic in nearly two decades. Despite recent advances, chemotherapeutics have proven woefully inadequate to treat this cancer. One of the hallmarks of PDAC is the presence of an extensive desmoplastic reaction consisting of pancreatic stellate cells, fibroblasts, immune cells, vasculature, and extracellular matrix. The complex interaction between cancer cells and the tumor microenvironment in PDAC is beginning to be understood, and the disruption of these interactions is a promising new avenue for therapeutic targeting of PDAC. In this chapter we will discuss a variety of therapies now entering the clinic that specifically target the tumor stroma. Among these are therapies that reduce the tumor stroma, allowing for a more effective delivery of conventional therapeutics to the cancer cells; antibodies that activate a macrophage-mediated immune response to the tumor and overcome its immunosuppressive environment; and inhibitors of signaling pathways that promote the vascularization of the tumor.

Correspondence/Reprint request: Dr. Sarah P. Thayer, Department of Surgery and the Andrew L. Warshaw MD, Institute for Pancreatic Cancer Research, Massachusetts General Hospital and Harvard Medical School Boston, Massachusetts 02114, USA. E-mail: sthayer@partners.org

158

Andrew S. Liss & Sarah P. Thayer

Introduction Pancreatic ductal adenocarcinoma (PDAC) is only the 14th most common cancer, but it is the fourth highest cause of cancer related deaths in the United States. Despite a dismally low five year survival rate (less than 5%), research on this disease has been limited due to the small number of people affected and an elderly patient population. Currently the only hope for patients diagnosed with PDAC is surgical resection. However, 80% of patients present with locally advanced or metastatic disease at the time of diagnosis, making them ineligible for surgical intervention. The standard of care for these patients for the past two decades has been chemotherapy with the nucleoside analog gemcitabine. However, the use of gemcitabine to treat these patients has resulted in only a five-week increase in median survival [1]. Combining gemcitabine treatment with platin-based agents (oxaliplatin and cisplatin) or a thymadilate synthetase inhibitor (capecitabine) has failed to enhance its therapeutic potential [2-4]. A phase III study has demonstrated that relative to treatment with gemcitabine alone, combinatorial treatment with erlotinib, an EGFR inhibitor, and gemcitabine results in a modest increase in median survival (6.24 vs. 5.91 months) for patients with unresectable PDAC [5]. More recently the use of FOLFIRINOX (oxaliplatin, irinotecan, fluorouracil, and leucovorin) in the treatment of metastatic PDAC has shown dramatic results. A phase II/III trial with 342 patients demonstrated that FOLFIRINOX improved median survival of patients by more than four months relative to gemcitabine (11.1 months vs. 6.8 months) [6].

Most, if not all, of the chemotherapeutic regimens for PDAC are directed at the cancer and not the microenvironment. The stroma of PDAC tumors is composed of abundant extracellular matrix and a complex set of cells including pancreatic stellate cells, fibroblasts, immune cells, and vasculature, each of which contributes to the survival and spread of the disease. There is a complex interplay between the cancer cells and the various components of the tumor microenvironment. Recently there has been a greater focus on defining how the stroma contributes to tumorigenesis; the results of these studies have begun to reveal novel therapeutic interventions for this cancer (Figure 1).

In this chapter we will review the efforts that have been made to develop therapeutics that specifically target the various components of the stroma. We will discuss classical approaches that target the tumor vasculature and early events in epithelial-to-mesenchymal transitions that promote the invasiveness of cancer cells. In addition, we will describe exciting new therapies that target paracrine pathways that promote angiogenesis, desmoplastic response and chemoresistance, as well as the host immune response.

159

Targeting the pancreatic stroma

Figure 1. Targeting the tumor microenvironment in PDAC. Cancer cells and the various components of the desmoplastic stroma produce soluble factors that are beneficial to the growth of both the stroma and the cancer. Therapeutic targets are identified that have been exploited to develop therapeutics directed at various components of the tumor microenvironment.

Targeting the stroma chemotherapuetics

to

enhance

conventional

PDAC is characterized by an extensive desmoplastic reaction which has recently been implicated in the chemoresistance of PDAC (Chapter 3).

These tumors contain an abundant stroma and poor vascular perfusion, resulting in a tumor vasculature that rarely comes in close proximity to the adenomatous component of the tumor [7]. It has been proposed that the failure of chemotherapeutics to effectively treat this disease is not due to the inherent resistance of this cancer to the drugs, but rather to an inability of the drugs to reach the cancer cells. Recent work targeting the Sonic hedgehog (Shh) signaling pathway suggests that reducing desmoplasia in PDAC allows for a more effective delivery of chemotherapeutic agents to the tumor.

Sonic Hedgehog pathway Shh is a lipid-modified protein that is secreted by cells and is involved in the regulation of developmental programs during embryogenesis; it is

160

Andrew S. Liss & Sarah P. Thayer

described in detail in Chapter 6. Briefly, in the absence of Shh, the signaling pathway is kept inactive by Patch (Ptch) proteins, Ptch1 and Ptch2, which are localized at the base of primary cilia (Figure 2). These twelve-pass membrane-spanning proteins prevent the localization of the seven-pass membrane-spanning protein Smoothened (Smo) to the primary cilia of cells.

The binding of the secreted Hh ligands to Ptch relieves this inhibition, allowing for the relocalization of Smo from vesicles to the primary cilia, a key step in the activation of downstream signaling. The activation of Smo results in the activation of the Gli family of transcription factors (Gli1, Gli2, and Gli3). Gli1 is believed to function as a transcriptional activator while Gli2/3 are transcriptional repressors. In the absence of Hh signaling Gli1 is transcriptionally silent and Gli2/3 undergoes proteolytic cleavage and functions to repress Hh-specific genes. Upon activation of the pathway, processing of Gli2/3 is reduced and there is a strong expression of Gli1, making

Figure 2. Model of the Shh signaling pathway. A. In the absence of Shh, the Ptch receptor inhibits Smo-containing vesicles from fusing to the cell membrane. Gli proteins are localized to the primary cilium by association with SuFu. Proteolytic processing of Gli creates a transcriptionally inactive form (Glir) which enters the nucleus to silence Shh target genes. B. The binding of Shh to Ptch allows for Smo localization to the primary cilium, resulting in the release of transcriptionally active Gli (Glia) from SuFu. Active Gli binds to the promoters of Shh responsive genes to activate transcription. Inhibitors that block the Shh signaling pathway are shown.

Targeting the pancreatic stroma

161

Gli1 one of the best characterized Hh responsive genes and one of the strongest predictors of Hh pathway activation.

Recent studies have demonstrated that Shh exerts its effect principally through a paracrine pathway in PDAC. These experiments have shown that Shh produced by tumor cells plays a key role in the recruitment and maintenance of the tumor mesenchyme. Mouse models employing orthotopic tumors of PDAC demonstrates a reduced desmoplasia in tumors from mice treated with neutralizing antibodies against Shh [8]. Further, tumors formed by transformed pancreatic ductal epithelial cells that overexpress Shh exhibit a greater desmoplastic reaction than tumors derived from control cells. In addition to their role in promoting the desmoplastic reaction observed in PDAC, the stromal cells of these tumors uniquely exhibit evidence of Hh pathway activation. Microarray analysis of microdissected human tumors reveals that cancer-associated fibroblasts express high levels of Smo [9].

Additional studies have demonstrated that Gli activity is restricted to the stroma [10] and that Smo is localized on the primary cilia of stromal fibroblasts and vessels [11].

Preclinical modeling with a genetically engineered mouse model of PDAC has provided evidence for the potential clinical benefit of targeting Shh in the treatment of PDAC [7]. In this study, mice treated with gemcitabine and the Shh inhibitor IPI-926 exhibited a greater than two-fold increase in survival relative to gemcitabine alone. Of particular interest was the mechanism for this enhanced survival. Shh inhibitor-treated mice (+/- gemcitabine) had tumors that were depleted of the stromal compartment after one week of treatment. Shh inhibitor-treated tumors had a higher mean vascular density than control tumors. Although this effect was transient, it correlated with better delivery of gemcitabine to the tumor cells. These results suggest that one of the mechanisms of the poor therapeutic response of PDAC to gemcitabine is the inefficient delivery of this drug to the cancer cells. Disruption of the tumor microenvironment with Shh inhibitors to allow for better drug delivery affords a novel treatment option.

The results of this preclinical study provide exciting evidence for the use of Shh inhibitors in the therapeutic treatment of PDAC. Due to the role of Hh pathway activation in the development of a wide variety of cancers, there have been extensive efforts to develop pharmacological agents to inhibit this pathway. Direct inhibition of Shh activity has been employed in preclinical animal testing and cell culture assays with neutralizing antibodies, such as 5E1 [11-13]. A recent report identified a small molecule antagonist of Shh, robotnikinin [14]. However, there are presently no therapies in the clinic that target Shh.

162

Andrew S. Liss & Sarah P. Thayer

Drugs have been developed to inhibit Shh signaling at various points along the signaling pathway. Most of these therapeutics are directed at Smo and have been studied in several solid tumors, most recently PDAC. The first Hh inhibitor identified was the naturally occurring compound cyclopamine [15]. This teratogenic compound was discovered, in part, based on the induction of holoprosencephaly in lambs born to ewes that consume Veratrum californicum during pregnancy [16, 17]. Cyclopamine directly binds to the heptahelical bundle of Smo, altering its conformation to one that is incapable of downstream signaling [18]. The first synthetic molecule directed at Smo was developed by Curis, Cur61414 [19]. Since that time a number of chemical inhibitors targeting Smo have been developed and brought to the clinic, including GDC-0449 (Genentech/Curis), IPI-926 (Infinity Pharmaceuticals), LDE225 (Novartis), BMS-833923 (Exelixis/Bristol-Myers Squibb), and PF-04449913 (Pfizer) [20]. For all of these compounds there are no fewer than 49 clinical trials currently in progress in the United States. GDC-0449 and IPI-926 are the best studied among them; to date there are six clinical trials utilizing these two compounds in the treatment of PDAC (Table 1). Preliminary results from a Phase Ib/II trial using IPI-926 in combination with gemcitabine reveal that the drug is well tolerated in patients with metastatic pancreatic cancer; radiographic partial responses were seen in 3/9 patients [21].

In addition to targeting the upstream Shh signaling pathway, efforts have been made to target the final step of this pathway, the Gli family of transcription factors. This has become increasingly important as recent studies have demonstrated a growing number of signaling pathways that modulate the expression and activity of this family of zinc finger transcription factors. Due the complex regulation of Gli proteins, the identification of inhibitors that directly target these transcription factors has been challenging. The activity of Gli proteins is also affected by a variety of kinases including MAPK, PI3K, and PKA [22-24]. Many of the compounds identified as inhibitors of Gli may actually modulate one of these regulatory proteins and therefore likely have broader effects on the cell than the Hh pathway. A classic example of this can be found in forskolin, which through the activation of adenylate cyclase upregulates PKA, resulting in the maintenance of repressive Gli isoforms [24]. More recently identified inhibitors of Gli, acrylaflavin C and physalin F, indirectly antagonize Gli activity through the PKC/MAPK pathways [25]. Compounds known to directly affect Gli activity have now been identified. Two such molecules are GANT58 and GANT61. These compounds were identified in a cell-based screen for their ability to inhibit Gli1 transcriptional activity [26]. They appear to affect Gli activity in the nuclei of cells, as they are effective inhibitors

163

Targeting the pancreatic stroma

Table 1. Clinical Trials Employing Shh Pathway Inhibitors in the Treatment of Pancreatic Ductal Adenocarcinoma.

Hh Inhibitor

Other Interventions

GDC-0449

Gemcitabine and nab-Paclitaxel

GDC-0449

none

GDC-0449

Gemcitabine

GDC-0449

Gemcitabine

GDC-0449

Erlotinib and Gemcitabine

IPI-926

none

IPI-926

Gemcitabine

IPI-926

FOLFIRINOX

Study Title A Phase II Study of Gemcitabine and NabPaclitaxel in Combination With GDC-0449 (Hedgehog Inhibitor) in Patients With Previously Untreated Metastatic Adenocarcinoma of the Pancreas Proof of Mechanism Study of an Oral Hedgehog Inhibitor (GDC-0449) in Patients With Resectable Pancreatic Ductal Adenocarcinoma in the Pre-operative Window Period Cancer Stem Cells and Inhibition of Hedgehog Pathway Signaling in Advanced Pancreas Cancer: A Pilot Study of GDC-0449 in Combination With Gemcitabine A Multi-Center, Double Blind, PlaceboControlled, Randomized Phase II of Gemcitabine Plus GDC-0449 (NSC 747691), a Hh Pathway Inhibitor, in Patients With Metastatic Pancreatic Cancer (10052747) Phase I Trial of the Combination of GDC-0449 and Erlotinib +/- Gemcitabine A Phase 1 Study of IPI-926 in Patients With Advanced and/or Metastatic Solid Tumor Malignancies A Phase 1b/2 Study Evaluating IPI-926 in Combination With Gemcitabine in Patients With Metastatic Pancreatic Cancer A Phase I Study of FOLFIRINOX Plus IPI-926 for Advanced Pancreatic Adenocarcinoma

Phase

Trial Identifier

Phase II

NCT01088815

Phase II

NCT01096732

Phase II

NCT01195415

Phase II

NCT01064622

Phase I

NCT00878163

Phase I

NCT00761696

Phase I/II

NCT01130142

Phase I

NCT01383538

Clincaltrials.gov August 2011

of a Gli protein with a mutant nuclear export signal. Targeted pathway analysis and microarray experiments did not detect additional pathways that were altered by these compounds, and cells treated with GANT61 exhibited reduced Gli DNA binding. A recent report has described the identification of four novel small molecule inhibitors of Gli, HPI-1, HPI-2, HPI-3, and HPI-4 [27]. Each of these compounds inhibits Hh signaling downstream of Smo, but employs unique strategies. HPI-1 is able to block the activity of both Gli1 and Gli2. The inhibition of Gli2 activation is likely accomplished by an indirect mechanism that prevents the conversion of Gli2 repressor to activator. In contrast, inhibition of Gli1 by HPI-1 is more likely direct, as it results in increased levels of Gli1 in cells. HPI-2 and HPI-3 both block Gli2 activator formation. Although their mechanism of action is not yet defined, these compounds result in Gli2 accumulation in the cilia, suggesting that they function through a cilia-dependent step. HPI-4 nonspecifically blocked Gli activation through the perturbation of cilia formation in cell lines. While the development of therapeutic compounds that inhibit Gli activity is still in its early stages, there appears to be a variety of druggable mechanisms to block the activity of these transcription factors.

164

Andrew S. Liss & Sarah P. Thayer

Since PDAC cells require Smo-independent Gli activity for their survival, such compounds may have the added benefit of targeting both stromal and cancer cells [28].

The characterization of Shh in PDAC in recent years has focused on paracrine signaling to the cells of the stroma. This represents a change of experimental emphasis from earlier work which focused on an autocrine role for Shh. This shift resulted from experiments utilizing cell lines treated with the Smo inhibitor cyclopamine, which demonstrated altered biological effects on PDAC cell lines that did not correlate with Hh pathway inhibition.

However, these results were not always clear, as observed in cyclopaminetreated cells that exhibited reduced proliferation and high levels of apoptosis [29]. Proliferation of these cells could be restored with an activator of Smo, but there was no protective effect observed for apoptosis, suggesting that only some of the biological changes were Hh-independent. Recent studies have demonstrated a number of signaling pathways that can activate Gli independently of Shh, including those found dysregulated in PDAC such as KRAS and TGF-β [28]. Additionally, recent work characterizing the cancer stem cell (or tumor initiating cell) population of PDAC has demonstrated that these cells express high levels of Shh [30]. Experiments with cells from PDAC and other cancers have shown that the cancer stem cell population is responsive to Shh inhibitors and that Shh makes these cells more susceptible to traditional chemotherapeutic drugs [31]. These studies demonstrate that in addition to playing a role in targeting the stroma, Shh inhibitors may exhibit direct effects on the cancer cells.

Transforming growth factor-β pathway The observation that reducing the stromal component of the tumor by inhibiting Shh allows for a more effective delivery of conventional therapeutics opens the door for similar strategies. The transforming growth factor (TGF)-β pathway appears to have tremendous promise, as it has been implicated in the regulation of both the growth of cancer cells and the formation of the tumor microenvironment. This pathway is composed of three TGF- β ligands (TGF-β1, -β2, and -β3) and three TGF-β receptors [32].

The binding of ligand to the type II receptor allows the recruitment and phosphorylation of the Type I receptor. The TGF-β type I receptors contain intracellular kinase domains which phosphorylate and activate receptorassociated Smad proteins. These activated Smad proteins dimerize with Smad4 and translocate to the nucleus, where they act to regulate gene expression.

Targeting the pancreatic stroma

165

Elevated levels of TGF-β have been observed in PDAC and correlate with a reduced survival [33]. The mechanism by which TGF-β contributes to PDAC is likely multifaceted. TGF-β has been implicated in angiogenesis, immunosuppression, activation of pancreatic stellate cells, and promoting epithelial-to-mesenchymal transition and proliferation of PDAC cells. A role for TGF-β in promoting desmoplasia in PDAC is illustrated in studies demonstrating that primary cultures of pancreatic stellate cells exhibit an increased proliferation in response to media conditioned by PDAC cell lines [34]. These stimulated stellate cells exhibit an increase in extracellular matrix production that is dependent on TGF-β activity. The co-injection of PDAC cell lines and cultured pancreatic stellate cells into nude mice results in more rapid tumor growth and the formation of tumors with a greater desmoplastic response than observed with tumors formed from PDAC cell lines alone.

Recent experiments characterizing fibrosis in a mouse model of pancreatitis implicate TGF-β in the activation of endogenous pancreatic stellate cells. A genetically engineered mouse that over-expressed Smad7 (an intracellular inhibitor of TGF-β signaling) in the pancreas exhibited reduced fibrosis and extracellular matrix production after induction of chronic pancreatitis [35].

These experiments further demonstrate that TGF-β signaling is required for the activation of pancreatic stellate cells, which have previously been implicated as the source of pancreatitis-induced fibrosis.

The TGF-β signaling pathway allows a variety of therapeutic mechanisms to be employed for its inhibition; examples of these may be found in clinical trials (Table 2). A number of approaches have been developed to prevent the binding of TGF-β to its cellular receptors, including antibodies that prevent signaling by binding to TGF-β or its receptors. A neutralizing antibody, fresolimumab (GC1008), which targets all three TGF-β molecules, is now in phase II trials for the treatment of metastatic breast cancer and malignant pleural mesothelioma [36]. Additionally, ligand trap approaches have been employed in preclinical models to inhibit TGF-β activity. Soluble forms of the TGF-β type II and III receptors have been employed to sequester TGF-β from the cellular receptors [37, 38]. Soluble forms of these receptors have been demonstrated to inhibit the invasiveness of PDAC cell lines and the formation of tumors in a xenograft mouse model.

The TGF-β type I and II receptors contain intracellular serine/threonine kinase domains which are responsible for downstream activation of Smad transcription factors. A number of inhibitors of these kinases have been developed, including LY2109761, which has shown efficacy in a preclinical model of PDAC [39]. Although an extensive analysis of the mechanism by

166

Andrew S. Liss & Sarah P. Thayer

Table 2. Clinical Trials Investigating Inhibitors of the TGF-β Pathway.

Drug

Target

GC1008 (Fresolimumab)

TGF-β

LY215729

TGF-β Type I Receptor

LY573636

TGF-β Type I Receptor

AP 12009 TGF-β2 (Trabedersen) Clincaltrials.gov August 2011

Condition Malignant Pleural Mesothelioma Metastatic Breast Cancer Myelofibrosis Pancreatic Cancer Hepatocellular Carcinoma Glioma Acute Myeloid Leukemia Renal Cell Cancer Metastatic Melanoma Advanced Solid Tumors Solid Tumors Ovarian Cancer Advanced Solid Tumors Pancreatic and Colorectal Neoplasms and Melanoma Anaplastic Astrocytoma Glioblastoma

Phase Phase II Phase I Phase I Phase I/II Phase II Phase I/II Phase I Phase I Phase II Phase I Phase I Phase II Phase I Phase I Phase III

Trial Identifier NCT01112293 NCT01401062 NCT01291784 NCT01373164 NCT01246986 NCT01220271 NCT00718159 NCT01258348 NCT00383292 NCT01214668 NCT01215916 NCT00428610 NCT01284335 NCT00844064 NCT00761280

which these inhibitors affect tumorigenesis has not been performed, treatment of mice with orthotopic PDAC tumors with both LY2109761 and gemcitabine reduced both tumor burden and metastatic lesions relative to control mice. While this inhibitor has not been employed in patients, at least two kinase inhibitors (LY2157299 and LY573636) that target the TGF-β pathway have been developed by Eli Lilly and Company and are now in clinical trials for the treatment of cancers. LY2157299 is in phase 1b/II trials for the treatment of pancreatic cancer.

An additional approach that reduces the expression of TGF-β has been developed. Phosphorothioate antisense oligonucleotides that target TGF-β1 (AP 11014) and TGF-β2 (AP 12009) effectively reduce the secretion of TGF-β from tumor cells [40, 41]. In phase IIb clinical trials of patients with glioma, AP 12009 (trabedersen) provided a survival advantage relative to conventional treatments [42, 43]. Preclinical testing of AP 12009 in a mouse model of PDAC has demonstrated its therapeutic potential. AP 12009 dramatically reduced the secretion of TGF-β from PDAC cell lines and reduced the growth of orthotopically implanted xenograft tumors. Consistent with the varied roles of TGF-β in PDAC, AP 12009 diminished the immunosuppressive effects of PDAC cells and inhibited tumor angiogenesis in mice. Phase I/II clinical trials are now under way for adult patients with advanced melanoma, colorectal or pancreatic carcinoma [44]. An interim report reveals that 23 of the 33 patients treated to date had pancreatic adenocarcinoma. Median overall survival for pancreatic cancer patients in the 2nd schedule is 13.2 months. One of the patients in schedule 1 experienced a complete remission and is alive after 38 months, and two patients in schedule 2 have stable disease and are alive at 14.8 months after beginning AP 12009.

Although a randomized phase II trial comparing this to standard treatment

Targeting the pancreatic stroma

167

remains to be done, these early reports suggest that AP 12009 has a good safety profile. Targeting the mesenchyme using TGF-β inhibitors may thus be a rational approach in the treatment of pancreatic cancer.

Extracellular matrix A key feature of the microenvironment of PDAC is the extensive amount of extracellular matrix (ECM) found in the tumor. Pancreatic stellate cells are likely responsible for much of this extracellular matrix, as they have been shown to produce collagen type I and c-fibronectin in response to pancreatic cancer cells [34]. The breakdown of the ECM is critical for localized growth of cancer cells as well as their systemic dissemination. The matrix metalloproteinase family plays a key role in the regulation of the ECM and is a therapeutic target for the treatment of cancer.

Matrix metalloproteinases The epithelial to mesenchymal (EMT) transition is one of the defining steps a cancer cell undergoes to become invasive and eventually metastatic [45]. One of the hallmarks of EMT is the upregulation of the matrix metalloproteinase (MMP) family of enzymes [46]. These enzymes play a key role in the breakdown of the basement membrane surrounding the cancer cells. The MMP family is composed of at least 24 structurally related proteins that have a diverse set of substrates including collagen, lamin, fibronectin, and gelatin [47, 48]. The enzymatic activity of these proteases can be regulated by a second family of proteins, tissue inhibitors of matrix metalloproteinases (TIMP 1, 2, 3, and 4) [49]. It is thought that the balance of MMP and TIMP proteins controls the integrity of the ECM and that alterations in the relative levels among these proteins potentiate malignant disease.

Elevated levels and/or activities of MMP1, 2, 3, 7, 9, and 11 have been observed in clinical samples of PDAC [50-55]. Likewise, elevated levels of TIMP 1 and 2 have been reported [51-53]. Recently the molecular mechanisms by which MMPs are activated in PDAC have begun to be elucidated. PDAC cell lines induce the expression of a membrane-bound MMP, MMP14, in response to Type I collagen [56]. The expression of MMP14 enhances the invasiveness and migration of these cells. Abundant levels of collagen are found in the stroma of PDAC, suggesting that the interaction of the cancer cells with this collagen initiates an invasive phenotype [57]. MMP2 is associated with the invasiveness of PDAC cells and the formation of the desmoplastic stroma [58]. Pancreatic stellate cells

168

Andrew S. Liss & Sarah P. Thayer

are an additional source of MMP2 in PDAC tumors [59]. MMP9 has also been shown to enhance the invasiveness of PDAC cells as well as to play an important role in tumor angiogenesis [60]. Preclinical testing of inhibitors that target MMP family members has demonstrated potential efficacy of these drugs. In a chemically induced model of PDAC it was demonstrated that RO 28-2653, a specific inhibitor of MMP2 and MMP9, reduced the incidence of liver metastasis [61]. The use of a broad spectrum MMP inhibitor, batimastat, in an orthotopic mouse model of PDAC reduced the tumor volume and frequency of metastatic lesions relative to control mice [62, 63].

Because of these preclinical studies, two MMP inhibitors have been brought to clinical trials for the treatment of PDAC. A phase III trial enrolling 277 patients studied the efficacy of tanomastat (Bay 12-9566), an inhibitor of MMP-2, MMP-3, MMP-9, and MMP-13. A comparison of patients treated with tanomastat versus gemcitabine revealed that patients treated with gemcitabine exhibited a longer median survival (6.59 months vs.

3.74 months) as well as disease-free progression (3.5 months vs. 1.68 months) than those treated with tanomastat [64]. A second trial investigated the efficacy of marimistat, an inhibitor of MMP-1, MMP-2, MMP-7, MMP-9, and MMP-14 [65]. This study examined the combinatorial treatment of gemcitabine and marimistat compared with gemcitabine alone in a group of 239 patients. However, there was no difference in median survival, one-year survival, or disease-free progression. Despite the positive results of preclinical studies, there was no demonstrable benefit for the treatment of PDAC from the use of MMP inhibitors alone or in combination with traditional chemotherapy.

The failure of these MMP inhibitors to show clinical benefit was not limited to patients with PDAC. Despite the exciting promise of this class of drugs, there has been no benefit observed from these inhibitors in any cancer.

There are likely many reasons for the failure of these drugs to elicit the desired biological change. The concept of targeting MMPs was based on the importance of these enzymes in the breakdown of the basement membrane allowing for invasion of the cancer cells. While this is an early event in tumorigenesis, the patients in these studies have advanced disease; perhaps these compounds would prove more effective in patients with earlier stage disease. Additionally, the lack of specificity in MMP inhibition by these drugs likely contributes to their ineffectiveness. Since their initial characterization in tumor progression, the large number of additional roles MMPs play has become more apparent. Not only have some MMPs (MMP-8 and MMP-12) been demonstrated to have anti-tumor effects, but there is a growing list of non-extracellular matrix proteins that are substrates for the

Targeting the pancreatic stroma

169

proteases [66]. Highly specific MMP inhibitors may prove to be a more effective approach to targeting this class of enzymes. MMP inhibitors that selectively target MMP-2, MMP-9, and MMP-14, alone or in combination, have demonstrated promising results in preclinical models [67-70]. It is hoped that selectively targeting those MMPs that play key roles in tumorigenesis may yield a more dramatic response in clinical trials.

Targeting the tumor vasculature As a tumor grows beyond microscopic size it requires the formation of new blood vessels to provide nutrients and oxygen for growth and for the removal of metabolic byproducts. Tumors initiate the formation of new blood vessels through the process of angiogenesis. Unlike vasculogenesis, which forms new vessels de novo, angiogenesis forms new blood vessels through the remodeling of existing vasculature. With the exception of specialized conditions such as wound repair, this process is largely held in check after embryogenesis by the proper balance of pro- and anti-angiogenic factors [71].

In the tumor, the “angiogenic switch� is initiated by an increase in proangiogenic factors and/or a decrease in anti-angiogenic factors, allowing for the neovascularization of the tumor. While solid tumors efficiently promote angiogenesis, this event is not as orderly a process as occurs under normal conditions. Generally tumor vessels are composed of a single layer of both endothelial cells and basement membrane, and often lack surrounding smooth muscle [72, 73]. It is likely that these features of the tumor vasculature potentiate the migration of metastatic cells. The process of angiogenesis is complex and involves a large number of soluble and cellular factors [74].

Current efforts in targeting these pathways for drug development have focused on a few major pathways.

VEGF pathway Some of the earliest factors studied in promoting angiogenesis are members of the vascular endothelial growth factor (VEGF) family. This family of secreted factors is composed of six members: VEGF-A, VEGF-B, VEGF-C, VEGF-D, VEGF-E, and placenta growth factor (PIGF) [75]. Of these factors VEGF-A is the best studied and plays a key role in promoting the proliferation and migration of endothelial cells. Members of the VEGF family form dimers and bind to the VEGF receptor (VEGFR) family of receptor tyrosine kinases. This family is composed of VEGFR1, VEGFR2, and VEGFR3, which exhibit specificity with the VEGF factors they bind.

VEGFR1 and VEGFR2 efficiently bind VEGF-A to promote the angiogenic

170

Andrew S. Liss & Sarah P. Thayer

pathway, while VEGFR3 binds VEGF-C and VEGF-D and is mainly involved in lymphangiogenesis [76, 77]. The binding of VEGF to its cognate VEGFR initiates receptor dimerization, resulting in phosphorylation and activation of the intracellular kinase domain. Activated VEGFR2 exerts its angiogenic effects by stimulating phospholipase C gamma (PLCÎł), resulting in higher levels of protein kinase C (PKC) activity [78-80]. While VEGF-A binds to both VEGFR1 and VEGFR2, the intracellular kinase activity of VEGFR1 is dispensable for the development of animals as well as tumor angiogenesis [81]. It appears that the role of VEGFR1 is to bind and recruit VEGFA from the microenvironment, facilitating its interaction with VEGFR2.

The elevated expression of VEGF has been observed in a wide range of solid tumors, and preclinical studies have demonstrated an important role for the VEGF pathway in tumorigenesis. Furthermore, there is evidence to suggest an important role for VEGF in angiogenesis of PDAC. PDAC cell lines secrete active VEGF-A [82]. Studies have also shown a correlation between levels of VEGF-A and the density of blood vessels in tumors [83, 84]. Elevated levels of VEGF-A in patients also correlate with greater disease progression. In addition to these expression studies, VEGF-A activity has been demonstrated to be important in the development of tumors in a mouse model [85]. Given the potential for VEGF inhibitors in a broad spectrum of cancers, there has been tremendous interest in developing inhibitors of this pathway, a number of which have been employed in clinical trial of PDAC. Two strategies have been used to disrupt this pathway, inhibiting the binding of VEGF-A to its receptor and blocking the kinase activity of VEGFRs (Table 3). A number of monoclonal antibodies have been developed that specifically bind to VEGF-A, preventing its binding to VEGFR1 and VEGFR2. One such therapy is bevacizumab (Avastin), which was developed by Genentech and Roche [86]. Bevacizumab was the first clinically available therapy in the United States to block angiogenesis. More recently, ranibizumab, an Fab fragment of this antibody, was developed and approved for the clinic. Monoclonal antibodies that bind to VEGFR and prevent VEGF binding have also been developed. IMC-1121b specifically recognizes VEGFR2, preventing VEGF-A binding and subsequent VEGF signaling. Alternative methods to physically prevent VEGF-A function are also utilized. Aflibercept was developed by joining the extracellular domains of VEGFR1 and VEGFR2 with an IgG1 Fc region, creating VEGF-Trap [87].

Aflibercept sequesters VEGF-A, preventing its signaling through VEGFR. There are also a number of broad-spectrum RTK inhibitors that target the

Table 3. Clinical Trials Investigating VEGF/VEGFR Inhibitors in the Treatment of Pancreatic Ductal Adenocarcinoma.

Targeting the pancreatic stroma 171

172

Andrew S. Liss & Sarah P. Thayer

tyrosine kinase activity of the VEGFRs. Sunitiab targets all three VEGFRs as well as other RTKs such as PDGFRs, FLT3, RET and c-Kit [88]. Sorafonib inhibits the activity of VEGFR2 and VEGFR3 as well as PDGFRs, FGFR1, FLT3, c-Kit, and MAPK [89-91]. Axitinib inhibits the activity of VEGFR1, VEGFR2, and VEGFR3 as well as PDGFRs and c-Kit [92].

The use of inhibitors of the VEGF pathway for therapeutic treatment of PDAC has not yielded promising results. A number of phase II and III studies have recently been published that evaluated the efficacy of anti-angiogenic therapies alone or in combination with conventional chemotherapy. A phase III trial was performed to evaluate the efficacy of axitinib treatment in combination with gemcitabine in patients with metastatic PDAC [93]. Despite promising results of earlier studies with this compound, no significant survival was observed in patients treated with axitinib with gemcitabine relative to those treated with gemcitabine plus placebo (mean survival 8.5 months vs. 8.3 months) [94, 95]. Similarly, a phase III trial of bevacizumab in combination with gemcitabine did not demonstrate any significant advantage over treatment with gemcitabine alone [96]. Preclinical studies that demonstrated a synergistic effect of anti-angiogenic agents with drugs that target the EGFR pathway have led to clinical trials that target both of these pathways [97, 98]. However, a phase II trial testing the EGFR inhibitor cetuximab with bevacizumab in 61 patients showed no clinical benefit with or without combinatorial treatment of gemcitabine [99]. A recently published phase III study evaluated the use of bevacizumab and erlotinib (an EGFR inhibitor) in 301 patients with advanced cases of PDAC [100]. No meaningful change in overall patient survival was observed in this study. The cumulative results of these clinical trials support the idea that the VEGF pathway in PDAC, although an attractive therapeutic target, has failed to show clinical benefit. The mechanism of the failure remains unknown, as there are no reports that anti-VEGF therapy changes the tumor vasculature in patients from these trials.

Angiopoietin pathway A large number of mitogenic factors which can contribute to angiogenesis are overexpressed in PDAC, including members of the FGF family, EGF, PDGF-beta, and TGF-Îą [101]. In addition, angiopoietins are now thought to play a role in PDAC angiogenesis [102, 103]. There are four secreted angiopoietins, Ang1, Ang2, Ang3, and Ang4, which can bind to the receptor tyrosine kinase Tie2 to activate downstream signaling. There is an additional Tie receptor, Tie1, which does not bind Ang proteins but rather heterodimerizes with activated Tie2 to regulate signaling. While this pathway was initially described as inhibitory towards angiogenesis, recent studies

Targeting the pancreatic stroma

173

have revealed a context-dependent role for a promoting angiogenesis.

Experiments demonstrated that there are unique Ang1/Tie2 complexes formed depending on the presence of cell-cell contact or extracellular matrix [104, 105]. Tie2 is localized to the site of cell-cell contact through Tie2Ang1-Tie2 interactions that bridge the two cells. Activation of Tie2 signaling by this mechanism results in the activation of the AKT pathway, promoting endothelial cell quiescence. However, Tie2 activation in the absence of cellcell contact occurs by the binding of ECM-bound Ang1. Tie2 activated by this mechanism promotes at least two downstream pathways. The adaptor protein Dok-R is phosphorylated by Tie2, promoting endothelial cell migration [106]. In addition, ECM-bound Ang-1 activation of Tie2 results in stimulation of the ERK pathway, which promotes angiogenesis by inducing endothelial cell migration [107].

Drugs developed to inhibit signaling of the Ang-Tie pathway are now undergoing clinical testing. Therapeutic antibodies that block Ang1 and Ang2 (AMG 386, Amgen) or specifically block Ang2 interaction with Tie2 (CVX241 and CVX-060) are being evaluated for a variety of advanced solid tumors including ovarian, lung, and colorectal cancer as well as glioblastomas [108, 109]. While these clinical studies are not yet mature, these drugs have clinical activity and are well tolerated, making them appropriate for future studies in patients with PDAC.

Notch pathway The Notch signaling pathway is a key developmental pathway that is involved in the regulation of cell proliferation and apoptosis. Studies characterizing the role of Notch during embryogenesis reveal a role for this pathway in the formation of blood vessels [110-113]. The mammalian Notch signaling pathway is composed of four single-pass integral membrane Notch receptors (Notch 1-4) and five Notch ligands (DLL1, DLL3, DLL4, Jagged 1 and Jagged 2). Notch receptors bind to membrane-bound Notch ligands on neighboring cells, resulting in a two-step proteolytic activation of the receptor. Notch receptors are initially cleaved by members of the ADAM family of metalloproteinases, allowing for further processing by Îłâˆ’secretase.

The final cleavage event releases the intracellular domain of the Notch receptor, allowing for its nuclear translocation, where it cooperates with additional transcription factors (such as CSL and Mastermind) to regulate gene expression.

Like other developmental pathways, the activation of Notch has been linked to tumorigenesis, including pancreatic cancer [114]. Preventing Notch activation with an inhibitor of Îł-secretase inhibited the growth of 13/26

174

Andrew S. Liss & Sarah P. Thayer

PDAC cell lines in vitro [115]. Furthermore, in a genetically engineered model of PDAC, inhibition of Îł-secretase activity prevented the progression of premalignant PanIN lesions to PDAC in 100% of treated animals.

Emerging evidence suggests a role for this pathway in PDAC tumor angiogenesis. Downregulation of Notch-1 in BxPC3 cells resulted in a decrease in VEGF mRNA levels and a corresponding decrease in secreted VEGF [116]. High levels of the Notch ligand Delta-like ligand-4 (DLL4) have been correlated with poor prognosis in patients who underwent surgical resection for PDAC [117]. A xenograft mouse model employing PK-1 cells demonstrated that employing a neutralizing antibody to disrupt signaling of DLL4 greatly reduced vascular density of tumors. Likewise, a role for DLL4 in tumor angiogenesis has been observed in other systems [118, 119].

Pharmaceutical inhibition of Notch signaling has focused on the proteolytic activation of Notch receptors. A number of Îł-secretase inhibitors (GSIs) have been developed that prevent the final cleavage event required for Notch activation. The potential clinical effectiveness of this class of compound in the treatment of PDAC was first demonstrated in a mouse model of PDAC [115]. Two GSIs, R04929097 and MK0752, are now in clinical trials for the treatment of PDAC (Table 4). RO4929097 was able to inhibit tumor growth by 50-80% in preclinical xenograft mouse models of PDAC [120]. While clinical data on the efficacy of this drug in the treatment of PDAC is not yet available, early phase I trials have demonstrated that it is well tolerated in patients [121].

Table 4. Clinical Trials Employing Notch Inhibitors in the Treatment of Pancreatic Ductal Adenocarcinoma. p y g Hh Inhibitor

Other Interventions

MK0752

Gemcitabine

RO4929097

none

RO4929097

none

RO4929097

Cediranib maleate

RO4929097

Gemcitabine

Clincaltrials.gov August 2011

Study Title A Cancer Research UK Phase I/IIa Trial of an Oral Notch Inhibitor (MK-0752) in Combination With Gemcitabine in Patients With Stage IV Pancreatic Cancer A Neoadjuvant Pharmacodynamic Study Of RO4929097 (RO) in Pancreas Cancer A Phase II Study of the Gamma Secretase Inhibitor RO4929097 in Previously Treated Metastatic Pancreas Cancer A Phase 1, Pharmacokinetic and Pharmacodynamic Study of the Combination of RO4929097 and Cediranib in Patients With Advanced Solid Tumors A Phase I Study of R04929097 in Combination With Gemcitabine in Patients With Advanced Solid Tumors

Phase

Trial Identifier

Phase I/II

NCT01098344

Phase I

NCT01192763

Phase II

NCT01232829

Phase I

NCT01131234

Phase I

NCT01145456

Targeting the pancreatic stroma

175

Activation of the immune response A dense inflammatory infiltrate is found throughtout the desmoplastic stroma of PDAC (Chapter 9). It is believed that most cancer cells are identified and cleared by the immune system at a very early stage in their development. For many years it was thought that the tumor growth was kept in check by the immune system, that advanced disease occurred as a breakdown in the immune response, and that the successful formation of a tumor involved the establishment of a localized immunosuppressed environment [122]. Paradoxically, the immunoprivileged tumor is not characterized by the absence of immune cells, but rather exhibits an extensive immune infiltrate.

This infiltrate is largely composed of immune cells that under normal conditions the body uses to dampen the immune response. These cells include myeloid-derived suppressor cells (MDSCs), regulatory T-cells (Treg cells), and macrophages [123]. The active suppression of the immune system at the site of the tumor is a common event in tumor development [124].

Recently it was demonstrated that alterations in the immune response to PDAC occur at a very early stage in the development of the disease [125].

Experiments utilizing a genetically engineered mouse model of PDAC expressing an activated KRAS allele (KrasG12D) indicated that an immunosuppressive environment is established as early as preinvasive PanIN lesions. PanINs, as well as invasive PDAC, were characterized by the infiltration of immune cells that dampen the immune response, including macrophages and MDSCs. There was an inverse correlation between the number of effector T-cells and MDSCs in tumors. Suppression of T-cell function by MDSCs has been attributed to a number of mechanisms. MDSCs exhibit high levels of Arg1 and iNOS [126, 127]. These enzymes both utilize L-arginine as a substrate in their metabolic cycle, resulting in the depletion of L-arginine from the microenvironment. Interestingly, the lack of L-arginine results in the downregulation of the CD3zeta chain on T-cells and subsequent inhibition of T-cell proliferation [128, 129]. In addition, elevated iNOS activity results in increased NO production, which can inhibit signaling downstream of the IL2 receptor [130]. In addition to elevated levels of MDSCs, these pancreata also had significant numbers of Treg cells. A major function of Treg cells is to end the T-cell-mediated immune response, and these cells have been implicated in limiting the immune response in tumors [131]. Due to the early establishment of an immunosuppressive state, therapeutics directed at activating the immune response against PDAC are an attractive target of research.

176

Andrew S. Liss & Sarah P. Thayer

The development of a T-cell-dependent anti-tumor response is largely dependent on the activation of the cell surface protein CD40 [132]. During the normal immune response, na誰ve CD8+ cells are stimulated to become cytotoxic T-lymphocytes (CTLs) upon binding to antigen-presenting cells (APCs). However, in order for APCs to efficiently promote CTL formation, they must first be primed by the binding of CD4+ T-helper cells. This priming occurs through the binding of the CD40 molecule on the surface of the APC by CD154 on the surface of the CD4+ T-helper cell [133]. It has been thought that a limiting step in the development of an effective CTL response against tumors was the effective priming of APCs. Key experiments demonstrated that the binding of T-helper cells could be efficiently substituted by the addition of an antibody specific to CD40 [134-136]. The activation of CD40 licenses antigen-presenting cells for tumor-specific T-cell priming and activation. Based on these experiments, a number of antibodies have been developed for the activation of CD40 [137].

A clinical trial was recently performed to determine whether activation of CD40 could overcome the immunosuppressive environment in PDAC and induce T-cell-mediated antitumor activity [138]. This trial enrolled 21 patients who were given a combinatorial therapy of a CD40 monoclonal antibody, CP-870,893 and gemcitabine. These patients had an overall survival of 7.4 months with a median disease-free survival of 5.6 months.

While this study size was small, the observed results are an improvement over the use of gemcitabine alone, which had an overall survival of 5.7 months with a median disease-free survival of 2.3 months.

The mechanism of the antitumor effects of this therapy is predominantly mediated by macrophages. Despite eliciting a strong T-cell response in a mouse model, these cells were not responsible for the observed anti-tumor activity. Interestingly, the authors observed that the CD40-specific antibody bound to macrophages in the peripheral blood and that these macrophages specifically migrated to the tumor. Antibodies to CD40 did not activate macrophages already residing in the tumor, but rather recruited macrophages from the peripheral blood to tumor. The mechanism by which the activation of CD40 results in the migration of macrophages to PDAC tumors has not been defined.

Conclusion Therapeutic targeting of PDAC has historically focused on the cancer cells. Scientific investigation over the past decade has focused on the cancer genome, and has failed or been slow to identify more effective targets for this cancer. Over the past few years, however, it has become increasingly evident

Targeting the pancreatic stroma

177

that the tumor and its stromal microenvironment have a dynamic and reciprocal interaction that plays a critical role in tumor initiation, progression, metastasis and chemoresistance, and that these interactions can be exploited for novel therapeutic targets. We are just beginning to understand these complex interactions and to discover that the stroma not only is composed of cancer-associated fibroblasts with an extensive ECM which contributes to chemoresistance, but is also composed of inflammatory cells and mediators which may cause a local immunosuppressive environment, as well as blood and lymphatic vessels important in the initiation, growth and metastatic spread of this cancer. Much more work remains to be done before the promise of preclinical models is realized in the clinic. The picture that is clearly evolving, however, suggests that future strategies targeting both pancreatic ductal adenocarcinoma and its microenvironment will be needed in order to effectively treat this cancer.

References 1.

2.

3.

4.

5.

6.

Burris, H.A., 3rd, Moore, M.J., Andersen, J., Green, M.R., Rothenberg, M.L., Modiano, M.R., Cripps, M.C., Portenoy, R.K., Storniolo, A.M., Tarassoff, P., Nelson, R., Dorr, F.A., Stephens, C.D. and Von Hoff, D.D. 1997, J. Clin. Oncol., 15, 2403.

Bernhard, J., Dietrich, D., Scheithauer, W., Gerber, D., Bodoky, G., Ruhstaller, T., Glimelius, B., Bajetta, E., Schuller, J., Saletti, P., Bauer, J., Figer, A., Pestalozzi, B.C., Kohne, C.H., Mingrone, W., Stemmer, S.M., Tamas, K., Kornek, G.V., Koeberle, D. and Herrmann, R. 2008, J. Clin. Oncol., 26, 3695.

Heinemann, V., Quietzsch, D., Gieseler, F., Gonnermann, M., Schonekas, H., Rost, A., Neuhaus, H., Haag, C., Clemens, M., Heinrich, B., Vehling-Kaiser, U., Fuchs, M., Fleckenstein, D., Gesierich, W., Uthgenannt, D., Einsele, H., Holstege, A., Hinke, A., Schalhorn, A. and Wilkowski, R. 2006, J. Clin. Oncol., 24, 3946.

Louvet, C., Labianca, R., Hammel, P., Lledo, G., Zampino, M.G., Andre, T., Zaniboni, A., Ducreux, M., Aitini, E., Taieb, J., Faroux, R., Lepere, C. and de Gramont, A. 2005, J. Clin. Oncol., 23, 3509.

Moore, M.J., Goldstein, D., Hamm, J., Figer, A., Hecht, J.R., Gallinger, S., Au, H.J., Murawa, P., Walde, D., Wolff, R.A., Campos, D., Lim, R., Ding, K., Clark, G., Voskoglou-Nomikos, T., Ptasynski, M. and Parulekar, W. 2007, J. Clin.

Oncol., 25, 1960.

Conroy, T., Desseigne, F., Ychou, M., Bouche, O., Guimbaud, R., Becouarn, Y., Adenis, A., Raoul, J.L., Gourgou-Bourgade, S., de la Fouchardiere, C., Bennouna, J., Bachet, J.B., Khemissa-Akouz, F., Pere-Verge, D., Delbaldo, C., Assenat, E., Chauffert, B., Michel, P., Montoto-Grillot, C. and Ducreux, M. 2011, N. Engl. J. Med., 364, 1817.

178

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Andrew S. Liss & Sarah P. Thayer

Olive, K.P., Jacobetz, M.A., Davidson, C.J., Gopinathan, A., McIntyre, D., Honess, D., Madhu, B., Goldgraben, M.A., Caldwell, M.E., Allard, D., Frese, K.K., Denicola, G., Feig, C., Combs, C., Winter, S.P., Ireland-Zecchini, H., Reichelt, S., Howat, W.J., Chang, A., Dhara, M., Wang, L., Ruckert, F., Grutzmann, R., Pilarsky, C., Izeradjene, K., Hingorani, S.R., Huang, P., Davies, S.E., Plunkett, W., Egorin, M., Hruban, R.H., Whitebread, N., McGovern, K., Adams, J., Iacobuzio-Donahue, C., Griffiths, J. and Tuveson, D.A. 2009, Science, 324, 1457.

Bailey, J.M., Swanson, B.J., Hamada, T., Eggers, J.P., Singh, P.K., Caffery, T., Ouellette, M.M. and Hollingsworth, M.A. 2008, Clin. Cancer Res., 14, 5995.

Walter, K., Omura, N., Hong, S.M., Griffith, M., Vincent, A., Borges, M. and Goggins, M. 2010, Clin. Cancer Res., 16, 1781.

Tian, H., Callahan, C.A., DuPree, K.J., Darbonne, W.C., Ahn, C.P., Scales, S.J. and de Sauvage, F.J. 2009, Proc. Natl. Acad. Sci. U.S.A., 106, 4254.

Bailey, J.M., Mohr, A.M. and Hollingsworth, M.A. 2009, Oncogene, 28, 3513.

Ericson, J., Morton, S., Kawakami, A., Roelink, H. and Jessell, T.M. 1996, Cell, 87, 661.

O'Toole, S.A., Machalek, D.A., Shearer, R.F., Millar, E.K., Nair, R., Schofield, P., McLeod, D., Cooper, C.L., McNeil, C.M., McFarland, A., Nguyen, A., Ormandy, C.J., Qiu, M.R., Rabinovich, B., Martelotto, L.G., Vu, D., Hannigan, G.E., Musgrove, E.A., Christ, D., Sutherland, R.L., Watkins, D.N. and Swarbrick, A. 2011, Cancer Res., 71, 4002.

Stanton, B.Z., Peng, L.F., Maloof, N., Nakai, K., Wang, X., Duffner, J.L., Taveras, K.M., Hyman, J.M., Lee, S.W., Koehler, A.N., Chen, J.K., Fox, J.L., Mandinova, A. and Schreiber, S.L. 2009, Nat. Chem. Biol., 5, 154.

Cooper, M.K., Porter, J.A., Young, K.E. and Beachy, P.A. 1998, Science, 280, 1603.

Binns, W., James, L.F., Shupe, J.L. and Everett, G. 1963, Am. J. Vet. Res., 24, 1164.

Keeler, R.F. and Binns, W. 1968, Teratology, 1, 5.

Chen, J.K., Taipale, J., Cooper, M.K. and Beachy, P.A. 2002, Genes Dev., 16, 2743.

Williams, J.A., Guicherit, O.M., Zaharian, B.I., Xu, Y., Chai, L., Wichterle, H., Kon, C., Gatchalian, C., Porter, J.A., Rubin, L.L. and Wang, F.Y. 2003, Proc.

Natl. Acad. Sci. U.S.A., 100, 4616.

Merchant, A.A. and Matsui, W. 2010, Clin. Cancer. Res., 16, 3130.

Stephenson, J., Richards, D.A., Wolpin, B.M., Becerra, C., Hamm, J.T., Messersmith, W.A., Devens, S., Cushing, J., Goddard, J., Schmalbach, T. and Fuchs, C.S. 2011, J. Clin. Oncol., 29, suppl; abstr 4114.

Seto, M., Ohta, M., Asaoka, Y., Ikenoue, T., Tada, M., Miyabayashi, K., Mohri, D., Tanaka, Y., Ijichi, H., Tateishi, K., Kanai, F., Kawabe, T. and Omata, M. 2009, Mol. Carcinog., 48, 703.

Mizuarai, S., Kawagishi, A. and Kotani, H. 2009, Mol. Cancer, 8, 44.

Kaesler, S., Luscher, B. and Ruther, U. 2000, Biol. Chem., 381, 545.

Targeting the pancreatic stroma

179

25. Hosoya, T., Arai, M.A., Koyano, T., Kowithayakorn, T. and Ishibashi, M. 2008, Chembiochem, 9, 1082.

26. Lauth, M., Bergstrom, A., Shimokawa, T. and Toftgard, R. 2007, Proc. Natl.

Acad. Sci. U.S.A., 104, 8455.

27. Hyman, J.M., Firestone, A.J., Heine, V.M., Zhao, Y., Ocasio, C.A., Han, K., Sun, M., Rack, P.G., Sinha, S., Wu, J.J., Solow-Cordero, D.E., Jiang, J., Rowitch, D.H. and Chen, J.K. 2009, Proc. Natl. Acad. Sci. U.S.A., 106, 14132.

28. Nolan-Stevaux, O., Lau, J., Truitt, M.L., Chu, G.C., Hebrok, M., FernandezZapico, M.E. and Hanahan, D. 2009, Genes Dev., 23, 24.

29. Xu, X., Guo, C., Liu, J., Yang, W., Xia, Y., Xu, L., Yu, Y. and Wang, X. 2009, J. Carcinog., 8, 1.

30. Li, C., Heidt, D.G., Dalerba, P., Burant, C.F., Zhang, L., Adsay, V., Wicha, M., Clarke, M.F. and Simeone, D.M. 2007, Cancer Res., 67, 1030.

31. Song, Z., Yue, W., Wei, B., Wang, N., Li, T., Guan, L., Shi, S., Zeng, Q., Pei, X. and Chen, L. 2011, PLoS One, 6, e17687.

32. ten Dijke, P. and Arthur, H.M. 2007, Nat. Rev. Mol. Cell Biol., 8, 857.

33. Friess, H., Yamanaka, Y., Buchler, M., Ebert, M., Beger, H.G., Gold, L.I. and Korc, M. 1993, Gastroenterology, 105, 1846.

34. Bachem, M.G., Schunemann, M., Ramadani, M., Siech, M., Beger, H., Buck, A., Zhou, S., Schmid-Kotsas, A. and Adler, G. 2005, Gastroenterology, 128, 907.

35. He, J., Sun, X., Qian, K.Q., Liu, X., Wang, Z. and Chen, Y. 2009, Biochim.

Biophys. Acta., 1792, 56.

36. Trachtman, H., Fervenza, F.C., Gipson, D.S., Heering, P., Jayne, D.R., Peters, H., Rota, S., Remuzzi, G., Rump, L.C., Sellin, L.K., Heaton, J.P., Streisand, J.B., Hard, M.L., Ledbetter, S.R. and Vincenti, F. 2011, Kidney Int., 79, 1236.

37. Rowland-Goldsmith, M.A., Maruyama, H., Kusama, T., Ralli, S. and Korc, M. 2001, Clin. Cancer Res., 7, 2931.

38. Gordon, K.J., Dong, M., Chislock, E.M., Fields, T.A. and Blobe, G.C. 2008, Carcinogenesis, 29, 252.

39. Melisi, D., Ishiyama, S., Sclabas, G.M., Fleming, J.B., Xia, Q., Tortora, G., Abbruzzese, J.L. and Chiao, P.J. 2008, Mol. Cancer. Ther., 7, 829.

40. Schlingensiepen, K.-H., Bischof, A., Egger, T., Hafner, M., Herrmuth, H., Jachimczak, P., Kielmanowicz, M., Niewel, M., Zavadova, E. and Stauder, G. 2004, J. Clin. Oncol., 22, 3132.

41. Schlingensiepen, K.H., Schlingensiepen, R., Steinbrecher, A., Hau, P., Bogdahn, U., Fischer-Blass, B. and Jachimczak, P. 2006, Cytokine Growth Factor Rev., 17, 129.

42. Bogdahn, U., Hau, P., Stockhammer, G., Venkataramana, N.K., Mahapatra, A.K., Suri, A., Balasubramaniam, A., Nair, S., Oliushine, V., Parfenov, V., Poverennova, I., Zaaroor, M., Jachimczak, P., Ludwig, S., Schmaus, S., Heinrichs, H. and Schlingensiepen, K.H. 2011, Neuro Oncol., 13, 132.

43. Hau, P., Jachimczak, P., Schlingensiepen, R., Schulmeyer, F., Jauch, T., Steinbrecher, A., Brawanski, A., Proescholdt, M., Schlaier, J., Buchroithner, J., Pichler, J., Wurm, G., Mehdorn, M., Strege, R., Schuierer, G., Villarrubia, V., Fellner, F., Jansen, O., Straube, T., Nohria, V., Goldbrunner, M., Kunst, M.,

180

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

Andrew S. Liss & Sarah P. Thayer

Schmaus, S., Stauder, G., Bogdahn, U. and Schlingensiepen, K.H. 2007, Oligonucleotides, 17, 201.

Oettle, H., Hilbig, A., Seufferlein, T., Schmid, R.M., Luger, T., von Wichert, G., Schmaus, S., Heinrichs, H. and Schlingensiepen, K. 2009, J. Clin. Oncol., 27, (suppl; abstr 4619) Kalluri, R. and Weinberg, R.A. 2009, J. Clin. Invest., 119, 1420.

Orlichenko, L.S. and Radisky, D.C. 2008, Clin. Exp. Metastasis, 25, 593.

Ra, H.J. and Parks, W.C. 2007, Matrix Biol., 26, 587.

Jones, L., Ghaneh, P., Humphreys, M. and Neoptolemos, J.P. 1999, Ann. N. Y. Acad. Sci., 880, 288.

Nagase, H., Visse, R. and Murphy, G. 2006, Cardiovasc. Res., 69, 562.

Satoh, K., Ohtani, H., Shimosegawa, T., Koizumi, M., Sawai, T. and Toyota, T. 1994, Gastroenterology, 107, 1488.

Gress, T.M., Muller-Pillasch, F., Lerch, M.M., Friess, H., Buchler, M. and Adler, G. 1995, Int. J. Cancer, 62, 407.

Bramhall, S.R., Stamp, G.W., Dunn, J., Lemoine, N.R. and Neoptolemos, J.P. 1996, Br. J. Cancer, 73, 972.

Bramhall, S.R., Neoptolemos, J.P., Stamp, G.W. and Lemoine, N.R. 1997, J. Pathol., 182, 347.

Koshiba, T., Hosotani, R., Wada, M., Miyamoto, Y., Fujimoto, K., Lee, J.U., Doi, R., Arii, S. and Imamura, M. 1998, Cancer, 82, 642.

Imamura, T., Ohshio, G., Mise, M., Harada, T., Suwa, H., Okada, N., Wang, Z., Yoshitomi, S., Tanaka, T., Sato, H., Arii, S., Seiki, M. and Imamura, M. 1998, J. Cancer Res. Clin. Oncol, 124, 65.

Shields, M.A., Dangi-Garimella, S., Krantz, S.B., Bentrem, D.J. and Munshi, H.G. 2011, J. Biol. Chem., 286, 10495.

Hezel, A.F., Kimmelman, A.C., Stanger, B.Z., Bardeesy, N. and Depinho, R.A. 2006, Genes Dev., 20, 1218.

Ellenrieder, V., Alber, B., Lacher, U., Hendler, S.F., Menke, A., Boeck, W., Wagner, M., Wilda, M., Friess, H., Buchler, M., Adler, G. and Gress, T.M. 2000, Int. J. Cancer, 85, 14.

Schneiderhan, W., Diaz, F., Fundel, M., Zhou, S., Siech, M., Hasel, C., Moller, P., Gschwend, J.E., Seufferlein, T., Gress, T., Adler, G. and Bachem, M.G. 2007, J. Cell Sci., 120, 512.

Nakamura, T., Kuwai, T., Kim, J.S., Fan, D., Kim, S.J. and Fidler, I.J. 2007, Neoplasia, 9, 979.

Kilian, M., Gregor, J.I., Heukamp, I., Hanel, M., Ahlgrimm, M., Schimke, I., Kristiansen, G., Ommer, A., Walz, M.K., Jacobi, C.A. and Wenger, F.A. 2006, Prostaglandins Leukot. Essent. Fatty Acids, 75, 429.

Zervos, E.E., Norman, J.G., Gower, W.R., Franz, M.G. and Rosemurgy, A.S. 1997, J. Surg. Res., 69, 367.

Zervox, E.E., Franz, M.G., Salhab, K.F., Shafii, A.E., Menendez, J., Gower, W.R. and Rosemurgy, A.S. 2000, J. Gastrointest. Surg., 4, 614.

Targeting the pancreatic stroma

181

64. Moore, M.J., Hamm, J., Dancey, J., Eisenberg, P.D., Dagenais, M., Fields, A., Hagan, K., Greenberg, B., Colwell, B., Zee, B., Tu, D., Ottaway, J., Humphrey, R. and Seymour, L. 2003, J. Clin. Oncol., 21, 3296.

65. Bramhall, S.R., Schulz, J., Nemunaitis, J., Brown, P.D., Baillet, M. and Buckels, J.A. 2002, Br. J. Cancer, 87, 161.

66. Rodriguez, D., Morrison, C.J. and Overall, C.M. 2010, Biochim. Biophys. Acta., 1803, 39.

67. Suojanen, J., Salo, T., Koivunen, E., Sorsa, T. and Pirila, E. 2009, Cancer Biol.

Ther., 8, 2362.

68. Koivunen, E., Arap, W., Valtanen, H., Rainisalo, A., Medina, O.P., Heikkila, P., Kantor, C., Gahmberg, C.G., Salo, T., Konttinen, Y.T., Sorsa, T., Ruoslahti, E. and Pasqualini, R. 1999, Nat. Biotechnol., 17, 768.

69. Heikkila, P., Suojanen, J., Pirila, E., Vaananen, A., Koivunen, E., Sorsa, T. and Salo, T. 2006, Int. J. Cancer, 118, 2202.

70. Salo, T., Sorsa, T. and Lindqvist, C. 2006, Oral Oncol., 42, 955.

71. Hanahan, D. and Folkman, J. 1996, Cell, 86, 353.

72. Kawaguchi, T. and Nakamura, K. 1986, Cancer Metastasis Rev., 5, 77.

73. Warren, B.A., Shubik, P. and Feldman, R. 1978, Cancer Lett., 4, 245.

74. Carmeliet, P. and Jain, R.K. 2011, Nature, 473, 298.

75. Hiratsuka, S. 2011, Front. Biosci., 16, 1413.

76. Alitalo, K. and Carmeliet, P. 2002, Cancer Cell, 1, 219.

77. Veikkola, T., Jussila, L., Makinen, T., Karpanen, T., Jeltsch, M., Petrova, T.V., Kubo, H., Thurston, G., McDonald, D.M., Achen, M.G., Stacker, S.A. and Alitalo, K. 2001, EMBO J., 20, 1223.

78. Takahashi, T., Ueno, H. and Shibuya, M. 1999, Oncogene, 18, 2221.

79. Xia, P., Aiello, L.P., Ishii, H., Jiang, Z.Y., Park, D.J., Robinson, G.S., Takagi, H., Newsome, W.P., Jirousek, M.R. and King, G.L. 1996, J. Clin. Invest., 98, 2018.

80. Takahashi, T., Yamaguchi, S., Chida, K. and Shibuya, M. 2001, Embo J., 20, 2768.

81. Hiratsuka, S., Minowa, O., Kuno, J., Noda, T. and Shibuya, M. 1998, Proc. Natl.

Acad. Sci. U.S.A., 95, 9349.

82. Luo, J., Guo, P., Matsuda, K., Truong, N., Lee, A., Chun, C., Cheng, S.Y. and Korc, M. 2001, Int .J. Cancer, 92, 361.

83. Seo, Y., Baba, H., Fukuda, T., Takashima, M. and Sugimachi, K. 2000, Cancer, 88, 2239.

84. Itakura, J., Ishiwata, T., Friess, H., Fujii, H., Matsumoto, Y., Buchler, M.W. and Korc, M. 1997, Clin. Cancer Res., 3, 1309.

85. Schwarz, R.E., Awasthi, N., Konduri, S., Cafasso, D. and Schwarz, M.A. 2010, Ann. Surg. Oncol., 17, 1442.

86. Presta, L.G., Chen, H., O'Connor, S.J., Chisholm, V., Meng, Y.G., Krummen, L., Winkler, M. and Ferrara, N. 1997, Cancer Res., 57, 4593.

87. Holash, J., Davis, S., Papadopoulos, N., Croll, S.D., Ho, L., Russell, M., Boland, P., Leidich, R., Hylton, D., Burova, E., Ioffe, E., Huang, T., Radziejewski, C., Bailey, K., Fandl, J.P., Daly, T., Wiegand, S.J., Yancopoulos, G.D. and Rudge, J.S. 2002, Proc. Natl. Acad. Sci. U.S.A., 99, 11393.

182

Andrew S. Liss & Sarah P. Thayer

88. Chow, L.Q. and Eckhardt, S.G. 2007, J. Clin. Oncol., 25, 884.

89. Plaza-Menacho, I., Mologni, L., Sala, E., Gambacorti-Passerini, C., Magee, A.I., Links, T.P., Hofstra, R.M., Barford, D. and Isacke, C.M. 2007, J. Biol. Chem., 282, 29230.

90. von Bubnoff, N., Engh, R.A., Aberg, E., Sanger, J., Peschel, C. and Duyster, J. 2009, Cancer Res., 69, 3032.

91. Ulivi, P., Arienti, C., Amadori, D., Fabbri, F., Carloni, S., Tesei, A., Vannini, I., Silvestrini, R. and Zoli, W. 2009, J. Cell. Physiol., 220, 214.

92. Rugo, H.S., Herbst, R.S., Liu, G., Park, J.W., Kies, M.S., Steinfeldt, H.M., Pithavala, Y.K., Reich, S.D., Freddo, J.L. and Wilding, G. 2005, J. Clin. Oncol., 23, 5474.

93. Kindler, H.L., Ioka, T., Richel, D.J., Bennouna, J., Letourneau, R., Okusaka, T., Funakoshi, A., Furuse, J., Park, Y.S., Ohkawa, S., Springett, G.M., Wasan, H.S., Trask, P.C., Bycott, P., Ricart, A.D., Kim, S. and Van Cutsem, E. 2011, Lancet Oncol., 12, 256.

94. Spano, J.P., Chodkiewicz, C., Maurel, J., Wong, R., Wasan, H., Barone, C., Letourneau, R., Bajetta, E., Pithavala, Y., Bycott, P., Trask, P., Liau, K., Ricart, A.D., Kim, S. and Rixe, O. 2008, Lancet, 371, 2101.

95. Spano, J.P., Moore, M.J., Pithavala, Y.K., Ricart, A.D., Kim, S. and Rixe, O. 2011, Invest. New Drugs, 96. Kindler, H.L., Niedzwiecki, D., Hollis, D., Sutherland, S., Schrag, D., Hurwitz, H., Innocenti, F., Mulcahy, M.F., O'Reilly, E., Wozniak, T.F., Picus, J., Bhargava, P., Mayer, R.J., Schilsky, R.L. and Goldberg, R.M. 2010, J. Clin.

Oncol., 28, 3617.

97. Ciardiello, F., Bianco, R., Damiano, V., Fontanini, G., Caputo, R., Pomatico, G., De Placido, S., Bianco, A.R., Mendelsohn, J. and Tortora, G. 2000, Clin. Cancer Res., 6, 3739.

98. Jung, Y.D., Mansfield, P.F., Akagi, M., Takeda, A., Liu, W., Bucana, C.D., Hicklin, D.J. and Ellis, L.M. 2002, Eur. J. Cancer, 38, 1133.

99. Ko, A.H., Youssoufian, H., Gurtler, J., Dicke, K., Kayaleh, O., Lenz, H.J., Keaton, M., Katz, T., Ballal, S. and Rowinsky, E.K. 2011, Invest. New Drugs, 100. Van Cutsem, E., Vervenne, W.L., Bennouna, J., Humblet, Y., Gill, S., Van Laethem, J.L., Verslype, C., Scheithauer, W., Shang, A., Cosaert, J. and Moore, M.J. 2009, J. Clin. Oncol., 27, 2231.

101. Korc, M. 1998, Surg. Oncol. Clin. N. Am., 7, 25.

102. Yamazaki, M., Nakamura, K., Mizukami, Y., Ii, M., Sasajima, J., Sugiyama, Y., Nishikawa, T., Nakano, Y., Yanagawa, N., Sato, K., Maemoto, A., Tanno, S., Okumura, T., Karasaki, H., Kono, T., Fujiya, M., Ashida, T., Chung, D.C. and Kohgo, Y. 2008, Cancer Sci., 99, 1131.

103. Nakamura, K., Sasajima, J., Mizukami, Y., Sugiyama, Y., Yamazaki, M., Fujii, R., Kawamoto, T., Koizumi, K., Sato, K., Fujiya, M., Sasaki, K., Tanno, S., Okumura, T., Shimizu, N., Kawabe, J., Karasaki, H., Kono, T., Ii, M., Bardeesy, N., Chung, D.C. and Kohgo, Y. 2010, PLoS One, 5, e8824.

104. Fukuhara, S., Sako, K., Minami, T., Noda, K., Kim, H.Z., Kodama, T., Shibuya, M., Takakura, N., Koh, G.Y. and Mochizuki, N. 2008, Nat. Cell Biol., 10, 513.

Targeting the pancreatic stroma

183

105. Saharinen, P., Eklund, L., Miettinen, J., Wirkkala, R., Anisimov, A., Winderlich, M., Nottebaum, A., Vestweber, D., Deutsch, U., Koh, G.Y., Olsen, B.R. and Alitalo, K. 2008, Nat. Cell Biol., 10, 527.

106. Master, Z., Jones, N., Tran, J., Jones, J., Kerbel, R.S. and Dumont, D.J. 2001, Embo J., 20, 5919.

107. Eliceiri, B.P., Klemke, R., Stromblad, S. and Cheresh, D.A. 1998, J. Cell. Biol., 140, 1255.

108. Neal, J. and Wakelee, H. 2010, Curr. Opin. Mol. Ther., 12, 487.

109. Arnett, S.O., Teillaud, J.L., Wurch, T., Reichert, J.M., Dunlop, C. and Huber, M. 2011, MAbs, 3, 133.

110. Krebs, L.T., Xue, Y., Norton, C.R., Shutter, J.R., Maguire, M., Sundberg, J.P., Gallahan, D., Closson, V., Kitajewski, J., Callahan, R., Smith, G.H., Stark, K.L. and Gridley, T. 2000, Genes Dev., 14, 1343.

111. Gale, N.W., Dominguez, M.G., Noguera, I., Pan, L., Hughes, V., Valenzuela, D.M., Murphy, A.J., Adams, N.C., Lin, H.C., Holash, J., Thurston, G. and Yancopoulos, G.D. 2004, Proc. Natl. Acad. Sci. U.S.A., 101, 15949.

112. Krebs, L.T., Shutter, J.R., Tanigaki, K., Honjo, T., Stark, K.L. and Gridley, T. 2004, Genes Dev., 18, 2469.

113. Fischer, A., Schumacher, N., Maier, M., Sendtner, M. and Gessler, M. 2004, Genes Dev., 18, 901.

114. Mysliwiec, P. and Boucher, M.J. 2009, Adv. Med. Sci., 54, 136.

115. Plentz, R., Park, J.S., Rhim, A.D., Abravanel, D., Hezel, A.F., Sharma, S.V., Gurumurthy, S., Deshpande, V., Kenific, C., Settleman, J., Majumder, P.K., Stanger, B.Z. and Bardeesy, N. 2009, Gastroenterology, 136, 1741.

116. Wang, Z., Banerjee, S., Li, Y., Rahman, K.M., Zhang, Y. and Sarkar, F.H. 2006, Cancer Res., 66, 2778.

117. Chen, H.T., Cai, Q.C., Zheng, J.M., Man, X.H., Jiang, H., Song, B., Jin, G., Zhu, W. and Li, Z.S. 2011, Ann. Surg. Oncol., 118. Noguera-Troise, I., Daly, C., Papadopoulos, N.J., Coetzee, S., Boland, P., Gale, N.W., Lin, H.C., Yancopoulos, G.D. and Thurston, G. 2006, Nature, 444, 1032.

119. Li, J.L., Sainson, R.C., Shi, W., Leek, R., Harrington, L.S., Preusser, M., Biswas, S., Turley, H., Heikamp, E., Hainfellner, J.A. and Harris, A.L. 2007, Cancer Res., 67, 11244.

120. Luistro, L., He, W., Smith, M., Packman, K., Vilenchik, M., Carvajal, D., Roberts, J., Cai, J., Berkofsky-Fessler, W., Hilton, H., Linn, M., Flohr, A., JakobRotne, R., Jacobsen, H., Glenn, K., Heimbrook, D. and Boylan, J.F. 2009, Cancer Res., 69, 7672.

121. Tolcher, A.W., Mikulski, S.M., Messersmith, W.A., Kwak, E.L., Gibbon, D., Boylan, J., Xu, Z.X., DeMario, M. and Wheler, J.J. 2010, J. Clin. Oncol., 28 (suppl; abstr 2502) 122. Vesely, M.D., Kershaw, M.H., Schreiber, R.D. and Smyth, M.J. 2011, Annu.

Rev. Immunol., 29, 235.

123. Mellor, A.L. and Munn, D.H. 2008, Nat. Rev. Immunol., 8, 74.

184

Andrew S. Liss & Sarah P. Thayer

124. Schreiber, R.D., Old, L.J. and Smyth, M.J. 2011, Science, 331, 1565.

125. Clark, C.E., Hingorani, S.R., Mick, R., Combs, C., Tuveson, D.A. and Vonderheide, R.H. 2007, Cancer Res., 67, 9518.

126. Bronte, V. and Zanovello, P. 2005, Nat. Rev. Immunol., 5, 641.

127. Rodriguez, P.C. and Ochoa, A.C. 2008, Immunol. Rev., 222, 180.

128. Rodriguez, P.C., Quiceno, D.G., Zabaleta, J., Ortiz, B., Zea, A.H., Piazuelo, M.B., Delgado, A., Correa, P., Brayer, J., Sotomayor, E.M., Antonia, S., Ochoa, J.B. and Ochoa, A.C. 2004, Cancer Res., 64, 5839.

129. Rodriguez, P.C., Quiceno, D.G. and Ochoa, A.C. 2007, Blood, 109, 1568.

130. Bingisser, R.M., Tilbrook, P.A., Holt, P.G. and Kees, U.R. 1998, J. Immunol., 160, 5729.

131. Sakaguchi, S., Sakaguchi, N., Shimizu, J., Yamazaki, S., Sakihama, T., Itoh, M., Kuniyasu, Y., Nomura, T., Toda, M. and Takahashi, T. 2001, Immunol. Rev., 182, 18.

132. Noelle, R.J. 1998, Agents Actions Suppl., 49, 17.

133. Grewal, I.S. and Flavell, R.A. 1998, Annu. Rev. Immunol., 16, 111.

134. Diehl, L., den Boer, A.T., Schoenberger, S.P., van der Voort, E.I., Schumacher, T.N., Melief, C.J., Offringa, R. and Toes, R.E. 1999, Nat. Med., 5, 774.

135. French, R.R., Chan, H.T., Tutt, A.L. and Glennie, M.J. 1999, Nat. Med., 5, 548.

136. Sotomayor, E.M., Borrello, I., Tubb, E., Rattis, F.M., Bien, H., Lu, Z., Fein, S., Schoenberger, S. and Levitsky, H.I. 1999, Nat. Med., 5, 780.

137. Khalil, M. and Vonderheide, R.H. 2007, Update Cancer Ther., 2, 61.

138. Beatty, G.L., Chiorean, E.G., Fishman, M.P., Saboury, B., Teitelbaum, U.R., Sun, W., Huhn, R.D., Song, W., Li, D., Sharp, L.L., Torigian, D.A., O'Dwyer, P.J. and Vonderheide, R.H. 2011, Science, 331, 1612.