SEARCH

SEARCH BY CITATION

Keywords:

  • pancreatic cancer;
  • pancreatic fibrosis;
  • pancreatic stellate cell;
  • pancreatitis;
  • stromal–epithelial interaction

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References

Pancreatitis and pancreatic cancer represent two major diseases of the exocrine pancreas. Pancreatitis exhibits both acute and chronic manifestations. The commonest causes of acute pancreatitis are gallstones and alcohol abuse; the latter is also the predominant cause of chronic pancreatitis. Recent evidence indicates that endotoxinemia, which occurs in alcoholics due to increased gut permeability, may trigger overt necroinflammation of the pancreas in alcoholics and one that may also play a critical role in progression to chronic pancreatitis (acinar atrophy and fibrosis) via activation of pancreatic stellate cells (PSCs). Chronic pancreatitis is a major risk factor for the development of pancreatic cancer, which is the fourth leading cause of cancer-related deaths in humans. Increasing attention has been paid in recent years to the role of the stroma in pancreatic cancer progression. It is now well established that PSCs play a key role in the production of cancer stroma and that they interact closely with cancer cells to create a tumor facilitatory environment that stimulates local tumor growth and distant metastasis. This review summarizes recent advances in our understanding of the pathogenesis of alcoholic pancreatitis and pancreatic cancer, with particular reference to the central role played by PSCs in both diseases. An improved knowledge of PSC biology has the potential to provide an insight into pathways that may be therapeutically targeted to inhibit PSC activation, thereby inhibiting the development of fibrosis in chronic pancreatitis and interrupting stellate cell–cancer cell interactions so as to retard cancer progression.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References

Inflammation of the pancreas (pancreatitis) and pancreatic cancer represent two major diseases of the exocrine pancreas, both of which can be associated with significant morbidity and mortality. Excessive alcohol consumption is an important causative factor for both acute and chronic pancreatitis,1 and considerable research effort has been invested in elucidating the molecular mechanisms responsible for alcohol-induced pancreatic injury. Investigators in this area have always been mindful of an apparent clinical paradox with respect to alcoholic pancreatitis—i.e. while it is established that alcohol exerts dose-related toxic effects on the pancreas, it is also evident that only a minority of heavy drinkers develop pancreatitis. Thus, in parallel with studies examining constant effects of alcohol on the pancreas, researchers have worked on identifying the factors that may increase individual susceptibility to alcoholic pancreatitis.

With regard to the constant effects of alcohol on the pancreas, it is now well established that the pancreas can metabolize alcohol to its toxic metabolites, acetaldehyde, and fatty acid ethyl esters (FAEEs); by-products of ethanol metabolism are reactive oxygen species (ROS) which can induce oxidant stress within the gland.2 There are also several lines of evidence indicating that the deleterious effects of alcohol and its metabolites on acinar cell organelles and signaling pathways predispose the cell to premature intracellular activation of digestive enzymes by lysosomal enzymes. Thus alcohol (and its metabolites) have been shown to (i) destabilize lysosomes (containing lysosomal enzymes) and zymogen granules (containing digestive enzymes),3,4 an effect mediated by oxidant stress, cholesteryl esters,4 and FAEE;5 (ii) increase digestive and lysosomal enzyme content (due to increased synthesis and impaired secretion of these enzymes);6 (iii) increase the activation of transcription factors (NF-κB and AP-1) which regulate cytokine expression;7 and (iv) induce a sustained increase in cytoplasmic ionic calcium (Ca++) causing mitochondrial calcium overload that can lead to mitochondrial depolarization and cell necrosis.8

The above changes are thought to sensitize the pancreatic exocrine cell such that, in the presence of an appropriate trigger/co-factor, overt necroinflammation is initiated. It is also recognized that repeated episodes of acute alcohol-induced pancreatic injury lead to increasing residual damage to the gland, eventually resulting in the changes of chronic pancreatitis, that is, atrophy and fibrosis (the necrosis–fibrosis sequence9).

With respect to individual susceptibility to alcoholic pancreatitis, there have been numerous studies examining diet,10 type and pattern of drinking,10,11 smoking,12 hyperlipidemia,13 and inherited factors14 in alcoholics with pancreatitis compared with alcoholics without pancreatitis and/or healthy controls. Despite concerted efforts, no factor has been unequivocally identified as being associated with alcoholic pancreatitis. However, the findings of two recent studies are of some interest. Miyasaka et al.15 have reported the presence of a polymorphism of the gene that encodes for carboxyl ester lipase (CEL) in alcoholics with pancreatitis compared with those without pancreatitis. As CEL is a FAEE synthase enzyme, mutations in the CEL gene may influence FAEE formation in these patients. However, the functional significance of the CEL polymorphism is yet to be determined.

More recently, the protective PRSS2 (anionic trypsinogen) variant, G191R, was reported to be significantly less common in patients with alcoholic pancreatitis compared with healthy controls.16 This variant has been shown to mitigate intrapancreatic trypsin activity, thereby playing a protective role against chronic pancreatitis.

Candidate susceptibility factors that have not yet been examined fully include gene polymorphisms of proteins relevant to cellular antioxidant defenses, minor cystic fibrosis mutations, and environmental factors such as bacterial endotoxin. The basis for examining bacterial endotoxin as a putative trigger/susceptibility factor is the well-reported occurrence of endotoxinemia in alcoholics secondary to an alcohol-related increase in gut permeability; this allows gut bacteria or bacterial products to enter the portal circulation.17 Interestingly, a positive correlation has been reported between higher circulating lipopolysaccharide (LPS, an endotoxin found in the cell wall of Gram negative bacteria such as Escherichia coli) levels and increasing severity of acute pancreatitis.18

Recent experimental studies have provided new insights into the role of endotoxin as a trigger factor initiating overt pancreatic injury in the alcohol-exposed pancreas (Fig. 1). In this regard, Vonlaufen and colleagues19 have reported a study in which rats were fed an alcohol diet for 8 weeks and then challenged with either single or three repeated doses of intravenous LPS. While alcohol alone failed to produce any demonstrable pancreatic injury (in keeping with the literature), exposure of alcohol-fed rats to a single dose of LPS resulted in significant pancreatitis characterized by acinar cell vacuolization and necrosis, inflammatory cell infiltration, and hemorrhage of the gland. This study was the first to identify a physiologically relevant factor that could initiate pancreatic injury after chronic alcohol exposure. Importantly, repeated challenge with LPS in alcohol-fed animals resulted in the characteristic changes of chronic pancreatitis (acinar atrophy and fibrosis).

image

Figure 1. Pathogenesis of alcoholic pancreatitis. The figure depicts the current hypothesis for the pathogenesis of alcoholic pancreatitis. Ethanol, via its toxic metabolites and the generation of oxidant stress, exerts direct effects on pancreatic acinar cells which predispose the gland to autodigestive injury. The presence of an appropriate trigger factor, such as endotoxin, initiates overt pancreatic necroinflammation. Pancreatic stellate cells are activated by cytokines released during the necroinflammatory process as well as by direct effects of endotoxin and oxidant stress. Activated PSCs secrete excessive amounts of extracellular matrix proteins leading to the development of pancreatic fibrosis.

Download figure to PowerPoint

Prior to the study by Vonlaufen et al.19 researchers had used other types of second ‘hits’ to produce overt injury in alcohol-fed animals, such as supraphysiological doses of the secretagogue cholecystokinin,20 toxins (trinitrobenzene sulfonic acid),21 and surgical interventions (bile duct ligation22). However, these additional factors are unlikely to be ever encountered in the clinical situation, making the clinical relevance of these models doubtful. In contrast, the alcohol feeding/LPS challenge model exhibiting both acute and chronic features of pancreatitis, provides, for the first time, a clinically relevant tool to study the chronology of alcohol-induced pancreatic injury, as well as the possible reversibility of established pancreatic damage after alcohol withdrawal. With regard to the latter, preliminary studies have demonstrated that alcohol withdrawal after established pancreatitis results in reversal of pancreatic injury as early as at 7 days of withdrawal, while continuation of alcohol leads to persistence of pancreatitis.23

Role of pancreatic stellate cells in pancreatic fibrosis

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References

Pancreatic stellate cells in health and injury

The most rapid advances in recent years have been those related to our understanding of the pathogenesis of pancreatic fibrosis, whereby the old concept of fibrosis as a mere end-product of chronic injury has been replaced by the view that fibrogenesis is a dynamic process, and that fibrosis is potentially reversible, at least in the early stages. Elucidation of the molecular mechanisms mediating pancreatic fibrosis over the past decade has largely been made possible by the identification, isolation, and characterization of the key cells responsible for pancreatic fibrosis, namely, pancreatic stellate cells (PSCs)24,25 (and review26); Australian researchers played a key role in these discoveries. As shown in Fig. 2, PSCs are resident cells of the pancreas, located in close proximity to the basolateral aspect of pancreatic acinar cells.

image

Figure 2. Desmin positive pancreatic stellate cells—the left panel depicts a frozen section of the pancreas immunostained for the pancreatic stellate cell (PSC) selective marker, desmin, while the right panel shows a corresponding line diagram. An acinus (A) made up of individual acinar cells is surrounded by brown stained, desmin positive stellate cells (PSCs) with a central cell body and long cytoplasmic projections extending along the basolateral aspect of acinar cells.

Download figure to PowerPoint

In the healthy pancreas, PSCs exhibit a quiescent (non-activated) phenotype with abundant vitamin A containing lipid droplets in the cytoplasm. PSCs express selective markers which differentiate them from fibroblasts; these markers include desmin, glial fibrillary acidic protein, nestin, nerve growth factor, and neural cell adhesion molecule. The stellate cells play an important role in regulating normal extracellular matrix (ECM) within the pancreas via the maintenance of a delicate balance between ECM synthesis and degradation.

During pancreatic injury, PSCs can be activated to a myofibroblast-like phenotype by several exogenous factors that include alcohol and its metabolites, oxidant stress, growth factors, and proinflammatory cytokines.27 Activated PSCs can also secrete endogenous cytokines resulting in an autocrine loop of cell activation. As a consequence, the cells synthesize and secrete excessive amounts of the ECM proteins (including collagen, laminin, fibronectin) which comprise fibrous tissue. Activated PSCs also exhibit increased proliferation and migratory ability. Recently, activated PSCs have also been shown to have a capacity for phagocytosis,28,29 a process that often represents the initial step in processing and presentation of antigens.29–31 This finding suggests that PSCs may also play a role in the immune response.

The predominant characteristics of quiescent and activated PSCs are summarized in Table 1. Transformation of PSCs from a quiescent to an activated phenotype has been the subject of intense study in recent years. Several signaling pathways/molecules that mediate this process have been identified (see reviews30,31). These include mitogen-activated protein kinases (MAPK), phosphatidylinositol kinase, protein kinase C, peroxisome proliferator activated receptor gamma, the Janus kinase – signal transducer and activator of transcription (JAK-STAT) pathway and the transcription factors NF-κB and AP-1.

Table 1.  Characteristics of quiescent and activated PSC phenotypes
 Quiescent PSCsActivated PSCs
  1. CTGF, connective tissue growth factor; ECM, extracellular matrix; IL, interleukin; MMP, matrix metalloproteinase; PDGF, platelet-derived growth factor; PSC, pancreatic stellate cell; TGFβ, transforming growth factor β; TIMP, tissue inhibitors of matrix proteinase.

Vitamin A lipid dropletsPresentAbsent
α Smooth muscle actinAbsentPresent
ProliferationLimitedIncreased
MigrationLimitedIncreased
Extracellular matrix productionLimitedIncreased
MMPs and TIMPsComplement of MMPs and TIMPs to maintain normal ECM turnoverChange in types of MMPs and TIMPs to facilitate ECM deposition
Production of cytokinesLimitedIncreased (PDGF, TGFβ, CTGF, IL1, IL6, IL15)
Capacity for phagocytosisAbsentPresent

Role of PSCs in chronic pancreatitis

Evidence of the central role played by PSCs in the fibrosis of chronic pancreatitis comes from both human studies (histological assessment of human chronic pancreatitis sections) and animal models.27 These studies have unquivocally demonstrated the presence of activated PSCs in areas of pancreatic fibrosis, and have also shown that PSCs are the primary source of collagen in the fibrotic tissue. Activation of PSCs during chronic pancreatitis is likely mediated by proliferative and profibrogenic growth factors such as platelet-derived growth factor (PDGF), transforming growth factor β (TGFβ) and connective tissue growth factor, all known to be upregulated during pancreatic necroinflammation. PSCs may also be activated by ROS (generated within the pancreas by oxidative metabolism of alcohol and/or necroinflammation).

Observations in in vivo models are supported by the findings of in vitro studies using cultured PSCs. Researchers have confirmed that cultured PSCs are activated by several proinflammatory cytokines including PDGF, TGFβ1, interleukins (IL)1 and 6, and tumor necrosis factor-α.32,33 Furthermore, they have shown that PSCs can secrete cytokines such as TGFβ1,34 PDGF,35 and IL15.36 Relevant to alcoholic pancreatic fibrosis is the finding that exposure of cultured PSCs to clinically relevant concentrations of alcohol (10 mM (seen with social drinking) or 50 mM (associated with heavy drinking)) activates the cells.37 This activation is mediated by the conversion of alcohol to its oxidative metabolite acetaldehyde and the generation of oxidant stress within the cells.

Exposure of PSCs to LPS also induces PSC activation as evidenced by α smooth muscle actin expression, increased fibronectin (an ECM protein) synthesis and increased migration.19 These findings imply that PSCs respond to LPS via its known receptor toll-like receptor 4 (TLR4). Indeed, it has now been shown that PSCs express TLR4; moreover, this expression is induced by prior LPS exposure.19

Of particular interest are the findings that alcohol and LPS exert synergistic effects on PSC activation;19 these results support the in vivo observations described above in the alcohol+LPS model of alcoholic pancreatitis. Furthermore, both compounds inhibit PSC apoptosis in vitro,23 suggesting that the persistence of pancreatic fibrosis in vivo upon continuation of alcohol (noted earlier) is likely due to the prevention of the loss of activated PSCs, thereby allowing continued synthesis of excessive ECM proteins.

Signaling pathways mediating PSC activation

Clues to the possible pathways mediating the observed synergistic effects of alcohol and LPS on PSCs are available from studies in the liver. It is well established that once LPS binds to its receptor TLR4, a signaling cascade is initiated via two pathways—the myeloid differentiation factor 88 (MyD88) pathway and a MyD88-independent pathway.38 MAPK signaling is also known to be induced by ligand binding to TLR4.39 The end result is activation of the transcription factor NF-κB which induces the synthesis of proinflammatory cytokines such as tumor necrosis factor-α. ROS are known to stimulate TLR4 activity40 and as alcohol metabolism in the pancreas leads to ROS formation, it is possible that alcohol-induced oxidant stress sensitizes pancreatic cells to the effect of LPS. Even more intriguing is the finding that pancreatic digestive enzymes, particularly pancreatic elastase, induces transcription factors in mononuclear cells via TLR4 activation.41 Thus, given that premature activation of digestive enzymes is a critical event in alcoholic pancreatitis, it may be speculated that prematurely activated elastase may predispose cells to LPS-induced injury via TLR4 activation. Obviously, further studies using specific inhibitors of relevant receptors and signaling molecules are needed to clarify this issue.

Role of PSCs in the desmoplastic reaction of pancreatic cancer

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References

As noted earlier, chronic pancreatitis is a major risk factor for pancreatic cancer.42 The two conditions have in common a striking histopathological feature, abundant fibrosis. Interestingly, the stromal compartments of chronic pancreatitis and pancreatic cancer have been reported to have at least 107 genes in common as assessed by microarray analysis.43 Researchers investigating the pathogenesis of pancreatic cancer have traditionally concentrated on the biology of the cancer cells themselves. It is only in the past 5 years that investigators have turned their attention to the possible role of the stromal/desmoplastic reaction in the progression of pancreatic cancer.

Analysis of human pancreatic cancer sections by immunohistochemistry for PSC-selective markers, special histological stains (e.g. Sirius Red for collagen, and in situ hybridization) has established that the cells responsible for producing the desmoplastic reaction in pancreatic cancer are PSCs.44 The possible interactions between PSCs and pancreatic cancer cells have now been examined using both in vitro and in vivo approaches (see review45). For in vitro studies, PSCs and pancreatic cancer cells have been either co-cultured or exposed to conditioned medium from cancer cells or PSCs, respectively. Pancreatic cancer cells have been shown to induce PSC activation, as indicated by increased proliferation, ECM synthesis, and migration. In turn, PSCs induce cancer cell migration and proliferation but inhibit cancer cell apoptosis thereby effectively enhancing the survival of cancer cells (Fig. 3). These findings suggest that pancreatic cancer cells are able to recruit host PSCs to their immediate vicinity, while PSCs reciprocate by facilitating cancer cell growth as well as local invasion. Factors mediating cancer cell-induced PSC activation include fibroblast growth factor and PDGF,44,46 while PSC-induced cancer cell proliferation is mediated, at least in part, by PDGF.47

image

Figure 3. Stromal–epithelial interactions. This figure depicts the known interactions between pancreatic cancer cells and pancreatic stellate cells. Pancreatic cancer cells increase pancreatic stellate cell (PSC) proliferation, extracellular matrix (ECM) synthesis and migration. In turn, PSCs increase cancer cell proliferation and migration but decrease cancer cell apoptosis.

Download figure to PowerPoint

The in vivo approach to assess tumor cell–stromal cell interactions has involved subcutaneous46 and orthotopic47,48 xenograft models of pancreatic cancer in immunocompromised mice. The orthoptopic model is preferred because tumors are produced within the pancreas itself, thereby simulating the known microenvironment of the cancer. BALB/c nude mice injected with a mixture of pancreatic cancer cells (MiaPaCa-2 or AsPC-1) and PSCs directly into the pancreas exhibit accelerated pancreatic tumor growth and, importantly, increased regional and distant metastasis.47,48 Tumors produced by the mixture of cancer cells and PSCs are characterized by the presence of activated PSCs and bands of fibrosis (resembling desmoplasia) as well as increased proliferation and decreased apoptosis of cancer cells. These data strongly support the concept that PSCs play an active role in facilitating the progression of pancreatic cancer in terms of both local growth and distant spread.

More recent studies indicate that PSCs stimulate tumor angiogenesis and are capable of transendothelial migration, implying that PSCs can intravasate/extravasate to and from blood vessels.49 Co-culture with cancer cells further stimulates transendothelial migration of PSCs. In keeping with this theme of PSCs migrating into blood vessels is an intriguing and novel finding in the orthotopic model, obtained by using a gender mismatch approach. This involved injecting the pancreas of female mice with female cancer cells + male human PSCs; fluorescent in situ hybridization experiments have demonstrated the presence of y chromosomes (male human PSCs) in metastatic nodules.49 These results strongly indicate that PSCs accompany cancer cells to distant metastatic sites,49 and represent the first evidence that stromal cells from the primary tumor may travel to distant metastatic sites where they might facilitate the seeding, survival, and growth of cancer cells.

Pancreatic stellate cells from the primary site most likely facilitate the recruitment of stellate cells/native stromal cells within the metastatic organs/tissues, thereby enhancing the tumor facilitatory microenvironment within the distant sites. Identification and characterization of the factors mediating the observed interactions between stromal cells and cancer cells is the next important step in this area of research. The ultimate goal is to develop effective therapies that target specific molecules/pathways so as to interrupt the stromal–epithelial interaction and inhibit cancer progression.

Concluding remarks

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References

In summary, there have been important advances in the areas of pancreatitis and pancreatic cancer in recent years, with a number of these coming from work performed by Australian researchers. The development of a clinically relevant model of alcoholic pancreatitis will now enable the determination of the pathways responsible for progression and reversal of pancreatitis in alcoholics. The progress in recent years in our understanding of the central role of PSCs in chronic pancreatitis as well as pancreatic cancer has opened up several new avenues of research related to elucidation of the molecular mechanisms of fibrogenesis and development of therapeutic strategies to prevent /reverse PSC activation thereby effectively preventing/retarding the fibrotic process.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References

Work by the Pancreatic Research Group described in this review is supported by grant funding from the National Health and Medical Research Council of Australia and the Cancer Council of New South Wales. M Apte is a member of the New South Wales Pancreatic Cancer Network established with the aid of a STREP grant (on which she is a Chief Investigator) funded by the NSW Cancer Council.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Role of pancreatic stellate cells in pancreatic fibrosis
  5. Role of PSCs in the desmoplastic reaction of pancreatic cancer
  6. Concluding remarks
  7. Acknowledgment
  8. References