NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Grippo PJ, Munshi HG, editors. Pancreatic Cancer and Tumor Microenvironment. Trivandrum (India): Transworld Research Network; 2012.

Cover of Pancreatic Cancer and Tumor Microenvironment

Pancreatic Cancer and Tumor Microenvironment.

Show details

Chapter 4Pancreatic cancer and the tumor microenvironment: Mesenchyme’s role in pancreatic carcinogenesis

.

Author Information

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 anti- metastatic 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 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 ‘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 cell- stroma 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. 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 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 TP53- inactivating 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 chemically- induced 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 deficient mice are most often used (16) but experiments have also been done with immune-competent mice (17), using tumoral material from immune- compatible 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 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 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-and- run’ 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 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 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 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 (Pdx1- Cre;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 Ptf1- Cre), 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 high- grade 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 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 tet- transactivator 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 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-to- ductal 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. cytokeratin- positive duct cells including centroacinar cells) continuously supplied the acinar compartment during physiological organ maintenance. The fifth 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 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 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.

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 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 anti- angiogenic 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 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 cell- stroma 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 cell- derived 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 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), 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. Anti- angiogenics or gemcitabine (113, 114), for example, showed very promising effects in transplanted models but turned out to be inconclusive in clinical 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/β- Catenin, 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 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-Rosa26- SmoM2-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; 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 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 tamoxifen- dependent 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; Rosa26-LSL-SmoM2-YFP mice was not sufficient to induce lesions. Neither does it potentate KrasG12D-driven pancreatic adenocarcinoma progression in Pdx1- Cre; 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;LSL- KrasG12D;TP53R270H, and PdxCre;LSL-KrasG12D;SmoM2) and further separation of tumor epithelium from the surrounding stroma revealed a 13- fold 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 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;LSL- KrasG12D;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;LSL- KrasG12D;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 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 Smad4- mediated 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-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 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 TGFβ- insensitive 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 cancer- associated 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 tumor- suppressor to an oncogene stimulated renewed interest in developing therapeutics that would specifically attempt to inhibit the oncogenic properties and preserve/restore the onco-suppressive properties of TGFβ. The ‘double-edged sword’ of TGF-β exerts both cell-specific and context- dependent effects, and TGFβ represents the archetypal factor acting in cell- autonomous 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 time-consuming.

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 anti- cancer drugs. Such out of targets and side effects result from the blockage of pathways, which has multiple effects, sometimes antagonistic, depending on 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.
Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. CA Cancer J Clin. 2003;53:5. [PubMed: 12568441]
2.
Hruban RH, Adsay NV, Albores-Saavedra J, Compton C, Garrett ES, Goodman SN, Kern SE, Klimstra DS, Kloppel G, Longnecker DS, Luttges J, Offerhaus GJ. Am J Surg Pathol. 2001;25:579. [PubMed: 11342768]
3.
Bachem MG, Zhou S, Buck K, Schneiderhan W, Siech M. Langenbecks Arch Surg. 2008;393:891. [PubMed: 18204855]
4.
Chu GC, Kimmelman AC, Hezel AF, DePinho RA. J Cell Biochem. 2007;101:887. [PubMed: 17266048]
5.
Jones S, Zhang X, Parsons DW, Lin JC, Leary RJ, Angenendt P, Mankoo P, Carter H, Kamiyama H, Jimeno A, Hong SM, Fu B, Lin MT, Calhoun ES, Kamiyama M, Walter K, Nikolskaya T, Nikolsky Y, Hartigan J, Smith DR, Hidalgo M, Leach SD, Klein AP, Jaffee EM, Goggins M, Maitra A, Iacobuzio-Donahue C, Eshleman JR, Kern SE, Hruban RH, Karchin R, Papadopoulos N, Parmigiani G, Vogelstein B, Velculescu VE, Kinzler KW. Science. 2008;321:1801. [PMC free article: PMC2848990] [PubMed: 18772397]
6.
Monti P, Marchesi F, Reni M, Mercalli A, Sordi V, Zerbi A, Balzano G, Carlo V, Allavena P, Piemonti L. Virchows Arch. 2004;445:236. [PubMed: 15258755]
7.
Froeling FE, Marshall JF, Kocher HM. J Biotechnol. 2010;148:16. [PubMed: 20083148]
8.
Dissin J, Mills LR, Mains DL, Black O Jr, Webster PD 3rd. J Natl Cancer Inst. 1975;55:857. [PubMed: 810597]
9.
Lilja HS, Hyde E, Longnecker DS, Yager JD Jr. Cancer Res. 1977;37:3925. [PubMed: 908032]
10.
Pour P, Althoff J, Kruger FW, Mohr U. Cancer Lett. 1977;2:233. [PubMed: 45725]
11.
Pour P, Althoff J, Kruger FW, Mohr U. J Natl Cancer Inst. 1977;58:1449. [PubMed: 857032]
12.
Pour P, Gingell R, Langenbach R, Nagel D, Grandjean C, Lawson T, Salmasi S. Cancer Res. 1980;40:3585. [PubMed: 7438045]
13.
Naito Z, Takahashi M, Furukawa F, Toyoda K, Hasegawa R, Hayashi Y. Acta Pathol Jpn. 1986;36:1359. [PubMed: 3788562]
14.
Rivera J. A., Graeme-Cook F., Werner J.Z'Graggen, K., Rustgi, A. K., Rattner, D. W., Warshaw, A. L., Fernandez-del Castillo, CSurgery 199712282. [PubMed: 9225919]
15.
Osvaldt AB, Wendt LR, Bersch VP, Backes AN, de Cassia ASR, Edelweiss MI, Rohde L. Surgery. 2006;140:803. [PubMed: 17084724]
16.
Caldas C, Hahn SA, da Costa LT, Redston MS, Schutte M, Seymour AB, Weinstein CL, Hruban RH, Yeo CJ, Kern SE. Nat Genet. 1994;8:27. [PubMed: 7726912]
17.
Tseng WW, Winer D, Kenkel JA, Choi O, Shain AH, Pollack JR, French R, Lowy AM, Engleman EG. Clin Cancer Res. 2010;16:3684. [PMC free article: PMC3085509] [PubMed: 20534740]
18.
Yoshida T, Hanahan D. Am J Pathol. 1994;145:671. [PMC free article: PMC1890334] [PubMed: 7521578]
19.
Apte MV, Haber PS, Applegate TL, Norton ID, McCaughan GW, Korsten MA, Pirola RC, Wilson JS. Gut. 1998;43:128. [PMC free article: PMC1727174] [PubMed: 9771417]
20.
Bachem MG, Schneider E, Gross H, Weidenbach H, Schmid RM, Menke A, Siech M, Beger H, Grunert A, Adler G. Gastroenterology. 1998;115:421. [PubMed: 9679048]
21.
Bardeesy N, Sinha M, Hezel AF, Signoretti S, Hathaway NA, Sharpless NE, Loda M, Carrasco DR, DePinho RA. Nature. 2002;419:162. [PubMed: 12226664]
22.
Jhappan C, Stahle C, Harkins RN, Fausto N, Smith GH, Merlino GT. Cell. 1990;61:1137. [PubMed: 2350785]
23.
Hasty P, Ramirez-Solis R, Krumlauf R, Bradley A. Nature. 1991;350:243. [PubMed: 1672446]
24.
Johnson L, Mercer K, Greenbaum D, Bronson RT, Crowley D, Tuveson DA, Jacks T. Nature. 2001;410:1111. [PubMed: 11323676]
25.
Quaife CJ, Pinkert CA, Ornitz DM, Palmiter RD, Brinster RL. Cell. 1987;48:1023. [PubMed: 3470144]
26.
Ornitz DM, Hammer RE, Messing A, Palmiter RD, Brinster RL. Science. 1987;238:188. [PubMed: 2821617]
27.
Sandgren EP, Luetteke NC, Palmiter RD, Brinster RL, Lee DC. Cell. 1990;61:1121. [PubMed: 1693546]
28.
Sandgren EP, Quaife CJ, Paulovich AG, Palmiter RD, Brinster RL. Proc Natl Acad Sci U S A. 1991;88:93. [PMC free article: PMC50755] [PubMed: 1986386]
29.
Glasner S, Memoli V, Longnecker DS. Am J Pathol. 1992;140:1237. [PMC free article: PMC1886527] [PubMed: 1316086]
30.
Sandgren EP, Luetteke NC, Qiu TH, Palmiter RD, Brinster RL, Lee DC. Mol Cell Biol. 1993;13:320. [PMC free article: PMC358911] [PubMed: 8417334]
31.
Wagner M, Luhrs H, Kloppel G, Adler G, Schmid RM. Gastroenterology. 1998;115:1254. [PubMed: 9797382]
32.
Grippo PJ, Nowlin PS, Demeure MJ, Longnecker DS, Sandgren EP. Cancer Res. 2003;63:2016. [PubMed: 12727811]
33.
Wagner M. Genes Dev. 2001;15:286. [PMC free article: PMC312620] [PubMed: 11159909]
34.
Jacks T, Remington L, Williams BO, Schmitt EM, Halachmi S, Bronson RT, Weinberg RA. Curr Biol. 1994;4:1. [PubMed: 7922305]
35.
Bardeesy N, Morgan J, Sinha M, Signoretti S, Srivastava S, Loda M, Merlino G, DePinho RA. Mol Cell Biol. 2002;22:635. [PMC free article: PMC139752] [PubMed: 11756558]
36.
Tuveson DA. Cancer Res. 2006;66:242. [PubMed: 16397237]
37.
Gu G, Dubauskaite J, Melton DA. Development. 2002;129:2447. [PubMed: 11973276]
38.
Kawaguchi Y, Cooper B, Gannon M, Ray M, MacDonald RJ, Wright CVE. Nat Genet. 2002;32:128. [PubMed: 12185368]
39.
Wu KL, Gannon M, Peshavaria M, Offield MF, Henderson E, Ray M, Marks A, Gamer LW, Wright CV, Stein R. Mol Cell Biol. 1997;17:6002. [PMC free article: PMC232449] [PubMed: 9315659]
40.
Jackson EL, Willis N, Mercer K, Bronson RT, Crowley D, Montoya R, Jacks T, Tuveson DA. Genes Dev. 2001;15:3243. [PMC free article: PMC312845] [PubMed: 11751630]
41.
Hingorani SR, Petricoin EF, Maitra A, Rajapakse V, King C, Jacobetz MA, Ross S, Conrads TP, Veenstra TD, Hitt BA, Kawaguchi Y, Johann D, Liotta LA, Crawford HC, Putt ME, Jacks T, Wright CV, Hruban RH, Lowy AM, Tuveson DA. Cancer Cell. 2003;4:437. [PubMed: 14706336]
42.
Aguirre AJ. Genes Dev. 2003;17:3112. [PMC free article: PMC305262] [PubMed: 14681207]
43.
Bardeesy N, Aguirre AJ, Chu GC, Cheng KH, Lopez LV, Hezel AF, Feng B, Brennan C, Weissleder R, Mahmood U, Hanahan D, Redston MS, Chin L, Depinho RA. Proc Natl Acad Sci U S A. 2006;103:5947. [PMC free article: PMC1458678] [PubMed: 16585505]
44.
Hingorani SR, Wang L, Multani AS, Combs C, Deramaudt TB, Hruban RH, Rustgi AK, Chang S, Tuveson DA. Cancer Cell. 2005;7:469. [PubMed: 15894267]
45.
Bardeesy N, Cheng KH, Berger JH, Chu GC, Pahler J, Olson P, Hezel AF, Horner J, Lauwers GY, Hanahan D, DePinho RA. Genes Dev. 2006;20:3130. [PMC free article: PMC1635148] [PubMed: 17114584]
46.
Izeradjene K, Combs C, Best M, Gopinathan A, Wagner A, Grady WM, Deng CX, Hruban RH, Adsay NV, Tuveson DA, Hingorani SR. Cancer Cell. 2007;11:229. [PubMed: 17349581]
47.
Kojima K, Vickers SM, Adsay NV, Jhala NC, Kim HG, Schoeb TR, Grizzle WE, Klug CA. Cancer Res. 2007;67:8121. [PubMed: 17804724]
48.
Morton JP, Jamieson NB, Karim SA, Athineos D, Ridgway RA, Nixon C, McKay CJ, Carter R, Brunton VG, Frame MC. Gastroenterology. 2010;139:586. [PMC free article: PMC3770904] [PubMed: 20452353]
49.
Ijichi H, Chytil A, Gorska AE, Aakre ME, Fujitani Y, Fujitani S, Wright CV, Moses HL. Genes Dev. 2006;20:3147. [PMC free article: PMC1635149] [PubMed: 17114585]
50.
Vincent DF, Yan KP, Treilleux I, Gay F, Arfi V, Kaniewski B, Marie JC, Lepinasse F, Martel S, Goddard-Leon S, Iovanna JL, Dubus P, Garcia S, Puisieux A, Rimokh R, Bardeesy N, Scoazec JY, Losson R, Bartholin L. PLoS Genet. 2009;5:e1000575. [PMC free article: PMC2706992] [PubMed: 19629168]
51.
Nolan-Stevaux O., Lau J., Truitt M. L., Chu G. C., Hebrok M.Fernandez- Zapico, M. E., Hanahan, DGenes Dev 20092324. [PMC free article: PMC2632164] [PubMed: 19136624]
52.
Skoulidis F, Cassidy LD, Pisupati V, Jonasson JG, Bjarnason H, Eyfjord JE, Karreth FA, Lim M, Barber LM, Clatworthy SA, Davies SE, Olive KP, Tuveson DA, Venkitaraman AR. Cancer Cell. 2010;18:499. [PubMed: 21056012]
53.
De La O.J.-P., Emerson, L. L., Goodman, J. L., Froebe, S. C., Illum, B. E., Curtis, A. B., Murtaugh, L. CProc Natl Acad Sci U S A 200810518907. [PMC free article: PMC2585942] [PubMed: 19028876]
54.
Heiser PW, Cano DA, Landsman L, Kim GE, Kench JG, Klimstra DS, Taketo MM, Biankin AV, Hebrok M. Gastroenterology. 2008;135:1288. [PMC free article: PMC2613004] [PubMed: 18725219]
55.
Krantz SB, Shields MA, Dangi-Garimella S, Cheon E, Barron MR, Hwang RF, Rao MS, Grippo PJ, Bentrem DJ, Munshi HG. Mol Cancer Res. 2011 [PMC free article: PMC3196812] [PubMed: 21856775]
56.
Lang GA, Iwakuma T, Suh YA, Liu G, Rao VA, Parant JM, Valentin-Vega YA, Terzian T, Caldwell LC, Strong LC, El-Naggar AK, Lozano G. Cell. 2004;119:861. [PubMed: 15607981]
57.
Olive KP, Tuveson DA, Ruhe ZC, Yin B, Willis NA, Bronson RT, Crowley D, Jacks T. Cell. 2004;119:847. [PubMed: 15607980]
58.
Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery CA Jr, Butel JS, Bradley A. Nature. 1992;356:215. [PubMed: 1552940]
59.
Purdie CA, Harrison DJ, Peter A, Dobbie L, White S, Howie SE, Salter DM, Bird CC, Wyllie AH, Hooper ML, et al. Oncogene. 1994;9:603. [PubMed: 8290271]
60.
Hezel AF, Gurumurthy S, Granot Z, Swisa A, Chu GC, Bailey G, Dor Y, Bardeesy N, Depinho RA. Mol Cell Biol. 2008;28:2414. [PMC free article: PMC2268441] [PubMed: 18227155]
61.
Stanger BZ, Stiles B, Lauwers GY, Bardeesy N, Mendoza M, Wang Y, Greenwood A, Cheng K-h, McLaughlin M, Brown D. Cancer Cell. 2005;8:185. [PubMed: 16169464]
62.
Jorgensen MC, Ahnfelt-Ronne J, Hald J, Madsen OD, Serup P, Hecksher-Sorensen J. Endocr Rev. 2007;28:685. [PubMed: 17881611]
63.
Guerra C, Schuhmacher AJ, Canamero M, Grippo PJ, Verdaguer L, Perez-Gallego L, Dubus P, Sandgren EP, Barbacid M. Cancer Cell. 2007;11:291. [PubMed: 17349585]
64.
Dor Y, Brown J, Martinez OI, Melton DA. Nature. 2004;429:41. [PubMed: 15129273]
65.
Zhou Q, Brown J, Kanarek A, Rajagopal J, Melton DA. Nature. 2008;455:627. [PubMed: 18754011]
66.
Habbe N, Shi G, Meguid RA, Fendrich V, Esni F, Chen H, Feldmann G, Stoffers DA, Konieczny SF, Leach SD, Maitra A. Proc Natl Acad Sci U S A. 2008;105:18913. [PMC free article: PMC2596215] [PubMed: 19028870]
67.
Friedlander SYG, Chu GC, Snyder EL, Girnius N, Dibelius G, Crowley D, Vasile E, DePinho RA, Jacks T. Cancer Cell. 2009;16:379. [PMC free article: PMC3048064] [PubMed: 19878870]
68.
Premsrirut PK, Dow LE, Kim SY, Camiolo M, Malone CD, Miething C, Scuoppo C, Zuber J, Dickins RA, Kogan SC, Shroyer KR, Sordella R, Hannon GJ, Lowe SW. Cell. 2011;145:145. [PMC free article: PMC3244080] [PubMed: 21458673]
69.
Brembeck FH, Schreiber FS, Deramaudt TB, Craig L, Rhoades B, Swain G, Grippo P, Stoffers DA, Silberg DG, Rustgi AK. Cancer Res. 2003;63:2005. [PubMed: 12727809]
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 JM, Behringer RR, Nakamura T, Uemoto S. Nat Genet. 2011;43:34. [PubMed: 21113154]
71.
Direkze NC, Hodivala-Dilke K, Jeffery R, Hunt T, Poulsom R, Oukrif D, Alison MR, Wright NA. Cancer Res. 2004;64:8492. [PubMed: 15574751]
72.
Yauch RL, Gould SE, Scales SJ, Tang T, Tian H, Ahn CP, Marshall D, Fu L, Januario T, Kallop D, Nannini-Pepe M, Kotkow K, Marsters JC, Rubin LL, de Sauvage FJ. Nature. 2008;455:406. [PubMed: 18754008]
73.
Paget S. Cancer Metastasis Rev. 1989;8:98. [PubMed: 2673568]
74.
Raimondi S, Lowenfels AB, Morselli-Labate AM, Maisonneuve P, Pezzilli R. Best Pract Res Clin Gastroenterol. 2010;24:349. [PubMed: 20510834]
75.
Lampel M, Kern HF. Virchows Arch A Pathol Anat Histol. 1977;373:97. [PubMed: 139754]
76.
Adler G, Hupp T, Kern HF. Virchows Arch A Pathol Anat Histol. 1979;382:31. [PubMed: 157597]
77.
Steinle AU, Weidenbach H, Wagner M, Adler G, Schmid RM. Gastroenterology. 1999;116:420. [PubMed: 9922324]
78.
Mizunuma T, Kawamura S, Kishino Y. J Nutr. 1984;114:467. [PubMed: 6199486]
79.
Su KH, Cuthbertson C, Christophi C. HPB Oxf. 2006;8:264. [PMC free article: PMC2023897] [PubMed: 18333137]
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. Gastroenterology. 2005;129:696. [PubMed: 16083722]
81.
Ohmuraya M, Hirota M, Araki K, Baba H, Yamamura K. Pancreas. 2006;33:104. [PubMed: 16804421]
82.
Selig L, Sack U, Gaiser S, Kloppel G, Savkovic V, Mossner J, Keim V, Bodeker H. BMC Gastroenterol. 2006;6:30. [PMC free article: PMC1637108] [PubMed: 17069643]
83.
Archer H, Jura N, Keller J, Jacobson M, Barsagi D. Gastroenterology. 2006;131:1844. [PubMed: 17087933]
84.
Marrache F, Tu SP, Bhagat G, Pendyala S, Osterreicher CH, Gordon S, Ramanathan V, Penz-Osterreicher M, Betz KS, Song Z, Wang TC. Gastroenterology. 2008;135:1277. [PMC free article: PMC2707078] [PubMed: 18789941]
85.
Ji B, Tsou L, Wang H, Gaiser S, Chang DZ, Daniluk J, Bi Y, Grote T, Longnecker DS, Logsdon CD. Gastroenterology. 2009;137:1072. [PMC free article: PMC2789008] [PubMed: 19501586]
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 OJ, Murtaugh LC. Cell Cycle. 2009;8:1860. [PMC free article: PMC2719432] [PubMed: 19440048]
88.
Carriere C, Young AL, Gunn JR, Longnecker DS, Korc M. Biochem Biophys Res Commun. 2009;382:561. [PMC free article: PMC2927854] [PubMed: 19292977]
89.
Morris JPt, Cano DA, Sekine S, Wang SC, Hebrok M. J Clin Invest. 2010;120:508. [PMC free article: PMC2810083] [PubMed: 20071774]
90.
Lesina M, Kurkowski MU, Ludes K, Rose-John S, Treiber M, Kloppel G, Yoshimura A, Reindl W, Sipos B, Akira S, Schmid RM, Algul H. Cancer Cell. 2011;19:456. [PubMed: 21481788]
91.
Fukuda A, Wang SC, Morris JPt, Folias AE, Liou A, Kim GE, Akira S, Boucher KM, Firpo MA, Mulvihill SJ, Hebrok M. Cancer Cell. 2011;19:441. [PMC free article: PMC3075548] [PubMed: 21481787]
92.
Bachem MG, Schunemann M, Ramadani M, Siech M, Beger H, Buck A, Zhou S, Schmid-Kotsas A, Adler G. Gastroenterology. 2005;128:907. [PubMed: 15825074]
93.
Neesse A, Wagner M, Ellenrieder V, Bachem M, Gress TM, Buchholz M. Pancreatology. 2007;7:380. [PubMed: 17703087]
94.
Schneiderhan W, Diaz F, Fundel M, Zhou S, Siech M, Hasel C, Moller P, Gschwend JE, Seufferlein T, Gress T, Adler G, Bachem MG. J Cell Sci. 2007;120:512. [PubMed: 17227797]
95.
Hwang RF, Moore T, Arumugam T, Ramachandran V, Amos KD, Rivera A, Ji B, Evans DB, Logsdon CD. Cancer Res. 2008;68:918. [PMC free article: PMC2519173] [PubMed: 18245495]
96.
Vonlaufen A, Joshi S, Qu C, Phillips PA, Xu Z, Parker NR, Toi CS, Pirola RC, Wilson JS, Goldstein D, Apte MV. Cancer Res. 2008;68:2085. [PubMed: 18381413]
97.
Kraman M, Bambrough PJ, Arnold JN, Roberts EW, Magiera L, Jones JO, Gopinathan A, Tuveson DA, Fearon DT. Science. 2010;330:827. [PubMed: 21051638]
98.
Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, Cordon-Cardo C, Guise TA, Massague J. Cancer Cell. 2003;3:537. [PubMed: 12842083]
99.
Zhang XH, Wang Q, Gerald W, Hudis CA, Norton L, Smid M, Foekens JA, Massague J. Cancer Cell. 2009;16:67. [PMC free article: PMC2749247] [PubMed: 19573813]
100.
Lu X, Wang Q, Hu G, Van Poznak C, Fleisher M, Reiss M, Massague J, Kang Y. Genes Dev. 2009;23:1882. [PMC free article: PMC2725946] [PubMed: 19608765]
101.
Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, Giri DD, Viale A, Olshen AB, Gerald WL, Massague J. Nature. 2005;436:518. [PMC free article: PMC1283098] [PubMed: 16049480]
102.
Gupta GP, Nguyen DX, Chiang AC, Bos PD, Kim JY, Nadal C, Gomis RR, Manova-Todorova K, Massague J. Nature. 2007;446:765. [PubMed: 17429393]
103.
Padua D, Zhang XH, Wang Q, Nadal C, Gerald WL, Gomis RR, Massague J. Cell. 2008;133:66. [PMC free article: PMC2390892] [PubMed: 18394990]
104.
Nguyen DX, Chiang AC, Zhang XH, Kim JY, Kris MG, Ladanyi M, Gerald WL, Massague J. Cell. 2009;138:51. [PMC free article: PMC2742946] [PubMed: 19576624]
105.
Bos PD, Zhang XHF, Nadal C, Shu W, Gomis RR, Nguyen DX, Minn AJ, van de Vijver MJ, Gerald WL, Foekens JA, Massagué J. Nature. 2009;459:1005. [PMC free article: PMC2698953] [PubMed: 19421193]
106.
Landi S. Mutat Res Rev Mutat Res. 2009;681:299. [PubMed: 19150414]
107.
Cheon EC, Strouch MJ, Barron MR, Ding Y, Melstrom LG, Krantz SB, Mullapudi B, Adrian K, Rao S, Adrian TE, Bentrem DJ, Grippo PJ. Int J Cancer. 2011;128:2783. [PMC free article: PMC3061970] [PubMed: 20725998]
108.
Lankisch PG. Pancreas. 2009;38:494. [PubMed: 19550274]
109.
Wendt LR, Osvaldt AB, Bersch VP, Schumacher Rde C, Edelweiss MI, Rohde L. Acta Cir Bras. 2007;22:202. [PubMed: 17546293]
110.
Bersch VP, Osvaldt AB, Edelweiss MI, Schumacher Rde C, Wendt LR, Abreu LP, Blom CB, Abreu GP, Costa L, Piccinini P, Rohde L. Pancreas. 2009;38:65. [PubMed: 18824948]
111.
Song Z, Bhagat G, Quante M, Baik GH, Marrache F, Tu SP, Zhao CM, Chen D, Dannenberg AJ, Wang TC. Lab Invest. 2010;90:426. [PubMed: 20065943]
112.
Hruban RH, Canto MI, Goggins M, Schulick R, Klein AP. Adv Surg. 2010;44:293. [PMC free article: PMC2966038] [PubMed: 20919528]
113.
Bruns CJ, Shrader M, Harbison MT, Portera C, Solorzano CC, Jauch KW, Hicklin DJ, Radinsky R, Ellis LM. Int J Cancer. 2002;102:101. [PubMed: 12385004]
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. Eur J Pharmacol. 2004;498:9. [PubMed: 15363970]
115.
Kindler HL, Friberg G, Singh DA, Locker G, Nattam S, Kozloff M, Taber DA, Karrison T, Dachman A, Stadler WM, Vokes EE. J Clin Oncol. 2005;23:8033. [PubMed: 16258101]
116.
Kindler HL, Friberg G, Skoog L, Wade-Oliver K, Vokes EE. Am J Clin Oncol. 2005;28:340. [PubMed: 16062074]
117.
Singh M, Lima A, Molina R, Hamilton P, Clermont AC, Devasthali V, Thompson JD, Cheng JH, Bou Reslan H, Ho CC, Cao TC, Lee CV, Nannini MA, Fuh G, Carano RA, Koeppen H, Yu RX, Forrest WF, Plowman GD, Johnson L. Nat Biotechnol. 2010;28:585. [PubMed: 20495549]
118.
Beatty GL, Chiorean EG, Fishman MP, Saboury B, Teitelbaum UR, Sun W, Huhn RD, Song W, Li D, Sharp LL, Torigian DA, O'Dwyer PJ, Vonderheide RH. Science. 2011;331:1612. [PMC free article: PMC3406187] [PubMed: 21436454]
119.
Dosch JS, Pasca di Magliano M, Simeone DM. Pancreatology. 2010;10:151. [PubMed: 20453550]
120.
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, Hebrok M. Nature. 2003;425:851. [PMC free article: PMC3688051] [PubMed: 14520413]
121.
Mao J, Ligon KL, Rakhlin EY, Thayer SP, Bronson RT, Rowitch D, McMahon AP. Cancer Res. 2006;66:10171. [PMC free article: PMC3806052] [PubMed: 17047082]
122.
Morton JP, Mongeau ME, Klimstra DS, Morris JP, Lee YC, Kawaguchi Y, Wright CV, Hebrok M, Lewis BC. Proc Natl Acad Sci U S A. 2007;104:5103. [PMC free article: PMC1828712] [PubMed: 17372229]
123.
Pasca di Magliano M, Sekine S, Ermilov A, Ferris J, Dlugosz AA, Hebrok M. Genes Dev. 2006;20:3161. [PMC free article: PMC1635150] [PubMed: 17114586]
124.
Bailey JM, Swanson BJ, Hamada T, Eggers JP, Singh PK, Caffery T, Ouellette MM, Hollingsworth MA. Clin Cancer Res. 2008;14:5995. [PMC free article: PMC2782957] [PubMed: 18829478]
125.
Tian H, Callahan CA, DuPree KJ, Darbonne WC, Ahn CP, Scales SJ, de Sauvage FJ. Proc Natl Acad Sci U S A. 2009;106:4254. [PMC free article: PMC2647977] [PubMed: 19246386]
126.
Bailey JM, Mohr AM, Hollingsworth MA. Oncogene. 2009;28:3513. [PMC free article: PMC2910592] [PubMed: 19633682]
127.
Berman DM, Karhadkar SS, Maitra A, Montes De Oca R, Gerstenblith MR, Briggs K, Parker AR, Shimada Y, Eshleman JR, Watkins DN, Beachy PA. Nature. 2003;425:846. [PubMed: 14520411]
128.
Feldmann G, Dhara S, Fendrich V, Bedja D, Beaty R, Mullendore M, Karikari C, Alvarez H, Iacobuzio-Donahue C, Jimeno A, Gabrielson KL, Matsui W, Maitra A. Cancer Res. 2007;67:2187. [PMC free article: PMC3073370] [PubMed: 17332349]
129.
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. Mol Cancer Ther. 2008;7:2725. [PMC free article: PMC2605523] [PubMed: 18790753]
130.
Feldmann G, Habbe N, Dhara S, Bisht S, Alvarez H, Fendrich V, Beaty R, Mullendore M, Karikari C, Bardeesy N, Ouellette MM, Yu W, Maitra A. Gut. 2008;57:1420. [PMC free article: PMC2707354] [PubMed: 18515410]
131.
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. Science. 2009;324:1457. [PMC free article: PMC2998180] [PubMed: 19460966]
132.
Rifai Y, Arai MA, Koyano T, Kowithayakorn T, Ishibashi M. J Nat Prod. 2010;73:995. [PubMed: 20450170]
133.
Hahn SA, Schutte M, Hoque AT, Moskaluk CA, da Costa LT, Rozenblum E, Weinstein CL, Fischer A, Yeo CJ, Hruban RH, Kern SE. Science. 1996;271:350. [PubMed: 8553070]
134.
Hansel DE, Kern SE, Hruban RH. Annu Rev Genomics Hum Genet. 2003;4:237. [PubMed: 14527303]
135.
Kuang C, Xiao Y, Liu X, Stringfield TM, Zhang S, Wang Z, Chen Y. Proc Natl Acad Sci U S A. 2006;103:1858. [PMC free article: PMC1413640] [PubMed: 16443684]
136.
Dupont S, Zacchigna L, Cordenonsi M, Soligo S, Adorno M, Rugge M, Piccolo S. Cell. 2005;121:87. [PubMed: 15820681]
137.
He W, Dorn DC, Erdjument-Bromage H, Tempst P, Moore MA, Massague J. Cell. 2006;125:929. [PubMed: 16751102]
138.
Dupont S, Mamidi A, Cordenonsi M, Montagner M, Zacchigna L, Adorno M, Martello G, Stinchfield MJ, Soligo S, Morsut L, Inui M, Moro S, Modena N, Argenton F, Newfeld SJ, Piccolo S. Cell. 2009;136:123. [PubMed: 19135894]
139.
Doisne JM, Bartholin L, Yan KP, Garcia CN, Duarte N, Le Luduec JB, Vincent D, Cyprian F, Horvat B, Martel S, Rimokh R, Losson R, Benlagha K, Marie JC. J Exp Med. 2009;206:1365. [PMC free article: PMC2715067] [PubMed: 19451264]
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, [PMC free article: PMC3378851] [PubMed: 22469842]
141.
Bartholin L, Cyprian FS, Vincent D, Garcia CN, Martel S, Horvat B, Berthet C, Goddard-Léon S, Treilleux I, Rimokh R, Marie JC. Genesis. 2008;46:724. [PubMed: 18821589]
142.
Vincent DF, Kaniewski B, Powers SE, Havenar-Daughton C, Marie JC, Wotton D, Bartholin L. Genesis. 2010;48:559. [PMC free article: PMC2944915] [PubMed: 20645310]
143.
Vincent DF, 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 CS, Sahai E. Nat Cell Biol. 2009;11:1287. [PMC free article: PMC2773241] [PubMed: 19838175]
145.
Meulmeester E, Ten Dijke P. J Pathol. 2011;223:205. [PubMed: 20957627]
146.
Muraoka RS, Koh Y, Roebuck LR, Sanders ME, Brantley-Sieders D, Gorska AE, Moses HL, Arteaga CL. Mol Cell Biol. 2003;23:8691. [PMC free article: PMC262670] [PubMed: 14612410]
147.
Muraoka-Cook RS, Dumont N, Arteaga CL. Clin Cancer Res. 2005;11:937s. [PubMed: 15701890]
148.
Muraoka-Cook RS, Shin I, Yi JY, Easterly E, Barcellos-Hoff MH, Yingling JM, Zent R, Arteaga CL. Oncogene. 2006;25:3408. [PubMed: 16186809]
149.
Siegel PM. Proc Natl Acad Sci U S A. 2003;100:8430. [PMC free article: PMC166246] [PubMed: 12808151]
150.
Rowland-Goldsmith MA, Maruyama H, Matsuda K, Idezawa T, Ralli M, Ralli S, Korc M. Mol Cancer Ther. 2002;1:161. [PubMed: 12467210]
151.
Gaspar NJ, Li L, Kapoun AM, Medicherla S, Reddy M, Li G, O'Young G, Quon D, Henson M, Damm DL, Muiru GT, Murphy A, Higgins LS, Chakravarty S, Wong DH. Mol Pharmacol. 2007;72:152. [PubMed: 17400764]
152.
Melisi D, Ishiyama S, Sclabas GM, Fleming JB, Xia Q, Tortora G, Abbruzzese JL, Chiao PJ. Mol Cancer Ther. 2008;7:829. [PMC free article: PMC3088432] [PubMed: 18413796]
153.
Kano MR, 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. Proc Natl Acad Sci U S A. 2007;104:3460. [PMC free article: PMC1800736] [PubMed: 17307870]
154.
Kano MR, Komuta Y, Iwata C, Oka M, Shirai YT, Morishita Y, Ouchi Y, Kataoka K, Miyazono K. Cancer Sci. 2009;100:173. [PubMed: 19037999]
155.
Schlingensiepen KH, Jaschinski F, Lang SA, Moser C, Geissler EK, Schlitt HJ, Kielmanowicz M, Schneider A. Cancer Sci. 2011 [PubMed: 21366804]
Copyright © 2012, Transworld Research Network.
Bookshelf ID: NBK98929PMID: 22876381

Views

  • PubReader
  • Print View
  • Cite this Page

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...