Pancreatic stellate cells: New target in the treatment of chronic pancreatitis


Professor Rakesh K Tandon, Department of Gastroenterology, Pushpawati Singhania Research Institute, Sheikh Sarai II, Press Enclave Marg, New Delhi 110017, India.


Chronic pancreatitis (CP) is characterized by progressive fibrosis, pain and/or loss of exocrine and endocrine functions. Recent in vitro and in vivo experiments have proven objectively the role of activated pancreatic stellate cells (PSC) in fibrogenesis in CP. Molecular mediators shown to regulate the pathogenesis include transforming growth factor beta (TGF-β), platelet-derived growth factor (PDGF), and pro-inflammatory cytokines such as IL-1, IL-6 and TNF-α. Furthermore, molecular pathways involving mitogen-activated protein kinases (MAPK), phosphatidyl inositol 3-kinase (PI3K), Ras superfamily G proteins, serine threonine protein kinase Raf-1 and peroxisome proliferator activated receptor gamma (PPAR-γ) have been elucidated. Understanding of the pathogenesis has led to identification of novel molecular targets and development of potential newer therapeutic agents. Those found to retard the progression of experimental CP and fibrosis in animal models include interferon (IFN) β and IFN-γ; a Japanese herbal medicine called Saiko-keishi-to (TJ-10); curcumin; PPAR-γ ligand (troglitazone); antioxidants (vitamin A, vitamin E, DA 9601 and epigallocatechin-3-gallate); a protease inhibitor (camostat mesilate) and hydroxymethylglutaryl-CoA inhibitor (lovastatin). This review summarizes the current literature addressing the role of different pharmacological agents aimed at reducing or preventing inflammation and the consequent fibrogenesis in CP.

Where do pancreatic stellate cells stand in chronic pancreatitis?

Chronic pancreatitis (CP) is a dynamic inflammatory process that is characterized by progressive fibrosis, pain and/or loss of exocrine and endocrine functions.1,2 Recent in vitro and in vivo studies have shown objectively the role of activated pancreatic stellate cells (PSC) in fibrogenesis in CP.3 PSC have been shown to synthesize and secrete matrix proteins,4 matrix metalloproteinases (MMP) and tissue inhibitors of matrix metalloproteinases (TIMP),5 thereby indicating that PSC can cause both synthesis and degradation of extracellular matrix. It is therefore the balance between fibrogenesis and matrix degradation that determines the fate of the pancreatic tissue, that is, maintenance of normal architecture or development of progressive fibrosis. Pancreatic stellate cells are activated by a number of factors that can be grouped primarily into inflammatory mediators, oxidant stress and toxins. The pro-inflammatory mediators that have been shown to activate PSC include interleukin (IL) 1, IL-6 and tumor necrosis factor (TNF) α. One of the major toxins that causes CP is ethanol. In vitro studies have shown that ethanol can produce CP through a number of mechanisms.6 Rat PSC have been shown to contain the ethanol metabolizing enzyme alcohol dehydrogenase (ADH). This enzyme metabolizes ethanol to produce acetaldehyde which itself causes significant intraglandular oxidant stress thus adding further to the stress caused by ethanol. Besides generating oxidant stress, ethanol can also predispose the pancreas to auto-digestion and necroinflammation. These events subsequently lead to production of pro-inflammatory cytokines namely, TNF-α, IL-1 and IL-6; they perpetuate further the pancreatic inflammation and activate PSC. Activated PSC are marked by the disappearance of fat globules and appearance of alpha smooth muscle actin (SMA). Moreover, they transform their shapes into a myofibroblast phenotype. Two other mediators that have been shown to have a significant impact on the activity of PSC are platelet-derived growth factor (PDGF) and transforming growth factor (TGF) β that are liberated by inflammatory cells. PDGF appears to be a potent mitogen and chemoattractant for PSC7,8 whereas TGF-β stimulates PSC to synthesize and secrete matrix proteins such as type 1 collagen, fibronectin and laminin, and MMP 2, 3 and 13.9–11 Besides being stimulated by mediators liberated by inflammatory cells, PSC also get activated by their own mediators. This autocrine activation is mediated by TGF-β, and a major pathway for this regulation is the extracellular signal related kinase (ERK1/2) pathway12 that is related to a mitogen-activated protein kinase (MAPK). This autocrine activation explains the progressive nature of pancreatic fibrosis in CP despite the cessation of the acute inciting event or the absence of recurrent acute insults. Other MAPK pathways that are activated by ethanol and oxidant stress include p38 kinase and c-Jun amino terminal (JNK) kinase pathways.13 Recent studies have also unveiled other molecular mediators such as the Ras superfamily G proteins and peroxisome proliferator activated receptor gamma (PPAR-γ) that have been shown to have a role in PSC activation and proliferation.14–16 The Ras superfamily G proteins undergo isoprenylation, a process that requires intermediates of cholesterol biosynthesis which is primarily regulated by the rate-limiting enzyme hydroxymethylglutaryl (HMG)-CoA reductase. In contrast, PPAR-γ is a member of the nuclear receptor family of transcription factors that control growth, differentiation and inflammation.

This review summarizes the current literature addressing the role of different pharmacological agents aimed at reducing or preventing inflammation and the consequent fibrogenesis in CP.

Newer treatment modalities under trial

The current treatment of CP is aimed at control of pain and replacement of the lost exocrine and endocrine functions. With the discovery of PSC and some understanding of the molecular mediators and pathways involved in the pathogenesis of CP, inhibition of the activation and functions of PSC has become a potential goal directed to the treatment and prevention of pancreatic inflammation and fibrosis. Studies aimed at achieving these goals have remained limited to experimental models so far. Table 1 lists the pharmacological agents that have shown promise in experimental CP. Table 2 summarizes the animal studies that have used these agents against cultured PSC and in experimental models of CP.

Table 1.  Pharmacological agents that have been shown to prevent or improve chronic pancreatitis in animal models
  1. IFN, interferon; MAPK, mitogen-activated protein kinase; PDGF, platelet-derived growth factor; PPAR, peroxisome proliferator activated receptor.

Group 1: Anti-inflammatory and immunomodulatory compounds
 Japanese herbal derivative
  Saiko-keishi-to (TJ-10)
 Herbal polyphenol compound
 PPAR-γ ligand
Group 2: Antioxidant compounds
 Vitamin A
 Vitamin E
  DA 9601
 Polyphenol compound
Group 3: Protease inhibitor
 Camostat mesilate
Group 4: HMG-CoA reductase inhibitor
Group 5: Others
 PDGF inhibitors
 p38 MAPK blocker
Table 2.  Summary of animal studies that have used newer therapeutic agents to inhibit fibrosis in chronic pancreatitis
A. In vivo studies
Author, yearDrug and time of administrationAnimal modelResults
Su et al. 199918Saiko-keishi-to (TJ-10); Administered along with induction of CPSpontaneous CP in WBN/Kob and Wistar ratsIncreased pancreatic weight at 16 weeks (< 0.05)
Reduced serum amylase at all ages
Normal histology at 12 weeks and slight interstitial edema and inflammatory cell infiltrate at 16 week (< 0.05)
CP changes evident at 20 weeks in treated rats compared with 12 weeks in untreated ones
Slight expression of PAP mRNA in inflamed tissue at 16 weeks in treated rats compared with full expression at 8 weeks in untreated ones
Su et al. 200143Saiko-keishi-to (TJ-10); Administered along with induction of CPSpontaneous CP in WBN/Kob ratsIncrease in pancreatic wet weight at 16 week (< 0.001) and 24 week (< 0.01)
No histologic changes of CP at 12 weeks and significant reduction in CP histologic scores at 16 week (< 0.001)
Significant reduction in acinar degeneration at 20 weeks (< 0.05)
Detection of TGF-β mRNA from 12 week in treated rats compared to 4 week in untreated ones
Van Westerloo et al. 200552Troglitazone; Administered along with induction of CPCerulein-induced CP in female C57BL/6 miceSignificant attenuation of markers of CP (< 0.05)
Significant reduction of myeloperoxidase concentration (< 0.05)
Partial prevention of reduction in intrapancreatic acinar cell/HPF in pancreatic tissue (< 0.05)
Complete prevention of rise in active TGF-β levels (< 0.05) in pancreatic tissue
Hisada et al. 200553Troglitazone; Administered along with induction of CPSpontaneous CP model in male WBN/Kob ratsInhibition of extracellular matrix synthesis
Inhibition of expression of α-SMA, procollagen I and III, fibronectin
Inhibition of NFκB binding activity
Gomez et al. 200456Vitamin E; Administered from 4 days prior to induction of CPCerulein-induced pancreatitis in male Wistar ratsReduced TGF-β levels
Reduced pancreatic weight
Reduced pancreatic hydroxyproline and plasma hyaluronic acid
Reduction in myofibroblast number
Improvement in fibrosis score.
Yoo et al. 200558DA 9601; Administered along with induction of CPCerulein-induced CP in miceReduction in pancreatic fibrosis
Reduction in NFκB binding activity
Reduction in myeloperoxidase and iNOS
Reduced expression of α-SMA and type I collagen
B. In vitro studies
Author, YearDrugPSC typeResults
  1. CP, chronic pancreatitis; DDC, diethyldithiocarbamate; ERK, extracellular signal related kinase; IL, interleukin; MAPK, mitogen-activated protein kinase; MCP, monocyte chemoattractant protein; iNOS, inducible nitric oxide synthetase; NFκB, nuclear factor κB; PAP, pancreatitis-associated protein; PDGF, platelet-derived growth factor; PSC, pancreatic stellate cells; SMA, smooth muscle actin; STAT, signal transduction and activation of transcription; TGF, transforming growth factor; TNF, tumor necrosis factor; TUNEL, TdT mediated X-dUTP nick end labeling.

Baumert et al. 200634Interferon β and γCultured PSC of inbred male LEW.1 W ratsInhibition of PSC proliferation and collagen synthesis
Dose-dependent inhibition of DNA synthesis
Inhibition of α-SMA expression
Induction of STAT 1 and STAT 3 expression
Masamune et al. 200646CurcuminCultured rat PSCInhibition of PDGF-induced PSC proliferation
Inhibition of expression of α-SMA gene
Inhibition of IL-1β and TNF-α induced MCP-1 production
Inhibition of transition of PSC to myofibroblast
Inhibition of collagen I and III gene expression
McCarroll et al. 200655Vitamin ACultured rat PSCInhibition of PSC proliferation
Inhibition of expression of α-SMA, collagen I, fibronectin and laminin
Inhibition of activation of MAPK

Anti-inflammatory and immunomodulatory agents


Interferons (IFN) are multifunctional cytokines that have anti-inflammatory, antiproliferative, immunomodulatory and antiviral activity.29 Interferons act by binding to and thereby activating specific cell surface receptors on target cells.30 This activation subsequently triggers transduction of signals to the nuclei via tyrosine kinases of the Janus family (JAK) and signal transduction and activation of transcription (STAT) factors.31 The use of IFN as anti-inflammatory mediators in animal models of CP was triggered by studies that showed an inhibitory role of these cytokines on activated hepatic stellate cells (HSC).23,32,33

In a recent in vitro experiment on cultured PSC of inbred male LEW1W rats,34 IFN-β and IFN-γ were shown to inhibit PSC proliferation and collagen synthesis. There was a dose-dependent inhibition of DNA synthesis by both IFN-β and IFN-γ; and IFN-γ was specifically found to inhibit alpha-SMA expression, meaning that IFN-γ inhibits activation of PSC. Both IFN-β and IFN-γ were also found to strongly induce STAT-1 and STAT-3 tyrosine phosphorylation, implying effective inhibitory signal transduction by IFN in PSC.

Saiko-keishi-to (TJ-10)

TJ-10 is a mixture of extracts of nine Japanese herbs with different pharmacological actions (Table 3).35–41 Overall it has been shown to have anti-inflammatory, analgesic and immunomodulatory properties.42 TJ-10 has been in clinical use in Japan for human CP for several years, although the mechanism was not clearly understood.17 Recent in vivo animal studies by Su et al.18,43 and Motoo et al.44 have elucidated the phenotypic effects and mechanisms of action of TJ-10 in CP.

Table 3.  Various components of Saiko-keishi-to (TJ-10)
Bupleurium root
Scutellaria root
Glycyrrhiza root
Cinnamon bark
Ginseng root

In an in vivo experiment by Su et al.18 on a spontaneous CP model in WBN/Kob and male Wistar rats, it was shown that TJ-10 reduced serum amylase levels at all ages and maintained normal pancreatic histology for up to 12 weeks after induction of pancreatitis. There was only mild interstitial edema and inflammatory cell exudation at 16 weeks in the treated rats and changes of CP were evident after 20 weeks compared with 12 weeks in the untreated rats (< 0.05). Moreover, the expression of pancreatitis-associated protein (PAP) mRNA in the inflamed pancreatic tissue was minimal even at 16 weeks in the treated rats as compared with full expression as early as 8 weeks in the untreated ones. PAP is a protein that is expressed in CP tissue and acts as an antiapoptotic factor in acinar cell injury.44 It was shown that the PAP mRNA can be expressed in pancreatic tissue prior to the appearance of histological changes of CP.18 Similar results were obtained in another subsequent study by the same group.43 In addition, there was significant reduction in acinar degeneration at 20 weeks in the treated rats and the appearance of TGF-β mRNA was delayed until 12 weeks in the treated rats compared with 4 weeks in the untreated ones.

These studies clearly show that TJ-10 delays the development of CP in experimental models and also reduces the severity of CP.

Curcumin (diferuloylmethane)

Curcumin (Fig. 1) is a yellow crystalline powder that is a major active component of turmeric (Curcuma longa Linn.) that is used as a condiment in cooking in India.45 It belongs to the polyphenol family of compounds and has been shown to have anti-inflammatory, antifibrotic and anticancer properties.24 In a in vitro study by Masamune et al.46 that used cultured rat PSC, it was shown that curcumin could inhibit PDGF-induced PSC proliferation and expression of the alpha-SMA gene. It also inhibited IL-1β and TNF-α induced monocyte chemoattractant protein (MCP) 1 production and activation of activator protein (AP) 147 and MAP kinases (namely ERK, JAK and p38 kinase). AP-1, like nuclear factor κB (NFκB), is a redox-sensitive transcriptional factor that is composed primarily of heterodimers of the various proteins of the Fos and Jun family.47,48 Besides its role in the oxidant injury and inflammatory cascade, it has also been implicated in carcinogenesis.49 Curcumin also inhibited the activation of quiescent PSC to the myofibroblast phenotype and gene expression of collagens I and III. This in turn inhibited production of collagen.

Figure 1.

Chemical structure of curcumin.

PPAR-γ ligand


The thiazolidinedione derivative troglitazone is the most well studied PPAR-γ ligand against CP. Its use in experimental CP was triggered by experiments that showed PPAR-γ ligand induced effective reduction of inflammation50 and reduced TGF-β production in hepatic stellate cells, aortic smooth muscles and kidneys.19,51 van Westerloo et al.52 had shown that in a cerulein-induced CP model in female C57BL/6 mice, troglitazone reduced the number of PSC in pancreatic tissue. There was complete prevention in the rise of the level of active TGF-β and significant reduction in the concentration of myeloperoxidase. There was also prevention in reduction of acinar cell number within the pancreas. Shimuzu et al.20 found that troglitazone could also block progression of the cell cycle beyond the G1 phase by a PPAR-γ independent mechanism, thereby inhibiting PSC proliferation.

In another experiment53 in a spontaneous CP model in male WBN/Kob rats, troglitazone was found to inhibit extracellular matrix synthesis by reducing the expression of α-SMA, procollagen I, procollagen III and fibronectin. Besides this, it also inhibited NFκB binding activity, thereby inhibiting the downstream inflammatory cascade.


Vitamin A

In the quiescent state, PSC contain vitamin A in the cytoplasm that disappears during their activation.54 Pancreatic stellate cells are equipped with identical metabolic pathways that metabolize ethanol and vitamin A.25 The relationship between retinol and its metabolites on ethanol-induced PSC activation was, however, not clear until recently. In a study by McCarroll et al.,55 retinol, all-trans retinoic acid (ATRA) and 9-cis retinoic acid were found to inhibit significantly ethanol-induced PSC proliferation, expression of alpha-SMA, collagen I, fibronectin and laminin. It also inhibited the activation of all the three classes of MAPK.

In another earlier study,21 ATRA was found to inhibit PSC proliferation and collagen synthesis, which was most likely mediated through transrepression of AP-1 signaling. In contrast to the study by McCarroll et al. ATRA in this study did not inhibit ERK activation and did not block the phenotypic transition of PSC towards myofibroblast.

Vitamin E

In a cerulein-induced pancreatitis model in male Wistar rats,56 vitamin E was found to reduce the number of myofibroblasts and improve the fibrosis score. Moreover, there was also a reduction in the level of TGF-β, pancreatic hydroxyproline and plasma hyaluronic acid, in addition to reduction in oxidative stress.

DA 9601

DA 9601 is a novel agent that has been extracted from the Asiatic traditional medicine Artemisia asiatica Nakai (Asteraceae). It contains a pharmacologically active flavonoid called eupatilin,57 which has been shown to have anti-inflammatory and antioxidant effects on experimental models of gastric, hepatic and pancreatic injury.22 It also induces apoptosis of human promyelocytic leukemia cells.57

In a cerulein-induced model of CP in mice,58 DA 9601 reduced the degree of pancreatic fibrosis and NFκB levels. In addition, it also reduced the enzymes myeloperoxidase and inducible nitric oxide synthetase (iNOS). There was also an increase in the levels of heat shock protein (HSP) 70 and metallothionein, molecules that have cytoprotective effects.


Epigallocatechin-3-gallate (EGCG; Fig. 2) is an antioxidant polyphenol purified from green tea and has the most potent antioxidant activity among the purified green tea polyphenols.59 Its antioxidant activity has been found to be more than that of vitamins E and C.60 Studies on isolated macrophages have shown that EGCG decreased the activation of NFκB that resulted in inhibition of lipopolysaccharide-induced nitric oxide production (anti-inflammatory action).61 EGCG also inhibited PDGF-induced proliferation of vascular smooth muscle cells (antiproliferative action)62 and induced rapid apoptosis in cancer cells (anticancer activity).26

Figure 2.

Chemical structure of epigallocatechin-3-gallate (EGCG).

In one of the foremost studies on isolated PSC from male Wistar rats,27 it was shown that EGCG could abolish ethanol-induced lipid peroxidation of cell membrane. Moreover, there was a loss of superoxide dismutase (SOD) activity and suppression of gene expression of various SOD enzymes. EGCG also suppressed ethanol-induced MAPK activation and synthesis and secretion of procollagen I and collagen.

In another in vitro experiment from Japan,63 EGCG was found to inhibit PDGF-induced proliferation of PSC in a dose-dependent manner. There was an inhibition of cell cycle progression of PSC beyond the G1 phase, as evidenced by reduction in the number of PDGF-treated S-phase PSC that had been pre-exposed to EGCG. EGCG was found to reduce PDGF-induced cyclin D1 expression while there was an increase in the level of p27Kip1 expression. Cyclin D proteins are essential molecules of the cell cycle that mediate progression of the cell cycle through G1 phase and entry into the S phase,64 while p27Kip1 is a member of the Cip/Kip family of cyclin dependent kinase (cdk) inhibitors that inhibits G1/S transition. As demonstrated in the experiment, the mechanism of action of EGCG appeared to be a dose-dependent inhibition of tyrosine phosphorylation of the PDGF-β receptor and thereby the resulting downstream activation of the ERK and PI3 kinase/Akt pathways. Tyrosine phosphorylation of the PDGF-β receptor is a key mechanism that triggers intracellular signaling after extracellular PDGF stimulation.65,66 The inhibition was most likely caused by the galloyl group in the third position of the EGCG. The galloyl group was found in a previous study to interfere with PDGF-BB induced mitogenic signaling in vascular smooth muscle cells.62 There was, however, no change in the PDGF-β receptor expression, which was in contrast to that in hepatic stellate cells (HSC).67

These experiments thus demonstrate the antioxidant activity of EGCG on PSC.

Protease inhibitor

Camostat mesilate

Camostat mesilate (CM) is a synthetic low molecular weight serine protease inhibitor that can inhibit proteases such as trypsin, kallikrein, thrombin, plasmin and C1 esterase.68 It has also been shown that CM can inhibit plasmin-dependent activation of latent TGF-β thereby preventing hepatic fibrosis in rat hepatocytes.69 It has been used for CP in Japan even though the exact mechanisms were not elucidated. Recent experimental studies have now clarified the mechanism thus justifying their clinical use.

Earlier studies had shown that CM could increase pancreatic secretion and pancreatic weight in rats,70 and prevent the progression of cerulein-induced CP in rats.71 In a study by Su et al.72 in a spontaneous CP model in WBN/Kob rats, there was significant reduction in expression of mRNA of PAP, P8, IL-6 and TGF-β at 12 weeks in CM-treated rats compared with the untreated ones. In another diethyldithiocarbamate (DDC) induced CP model in male Wistar rats,73 there was a significant reduction in the area of pancreatic fibrosis in the treated rats. Moreover, there was a reduction in the ratio of α-SMA positive cells to desmin-positive cells in the CM-treated rats; as was the reduction in the level of prolyl hydroxylase, a marker of collagen synthesis. In the most recently published study74 that used a dibutyltin dichloride (DBTC) induced CP model in male Lewis rats, it was found that in vivo there was inhibition of inflammation, cytokine expression and fibrosis in the pancreas after exposure to CM. In vitro, it was found that there was an inhibition of the production of MCP-1 by monocytes and PSC. There was also inhibition of the production of TNF-α by monocytes and proliferation of PSC; however, in this study, expression of procollagen α1 was not influenced by CM.

HMG-coenzyme A reductase inhibitor


Lovastatin is a member of the HMG-CoA reductase inhibitor family that has cholesterol-lowering action. Recent experiments have shown that it also has pro-apoptotic, antiproliferative and antimetastatic action against various cancer cell lines including human pancreatic cancer.28,75 In the context of CP,76in vitro experiments using cultured rat PSC have shown that lovastatin can lead to a significant reduction in DNA synthesis. TdT mediated X-dUTP nick end labeling (TUNEL) assay showed that it could produce a dose-dependent increase in apoptotic cells. The fundamental mechanism of action was found to be inhibition of mevalonic acid, which is a precursor of lipid moieties required for the isoprenylation of the Ras superfamily G proteins. This in turn inhibited the PDGF-activated Raf-Ras ERK pathway,77 thereby inhibiting PSC proliferation. In addition, there was an interference in the PDGF-induced membrane translocation of RhoA,78 a protein that is required for regulation of actin cytoskeleton, cell adhesion and motility. This led to impairment in PSC migration.


Besides the above-mentioned pharmacological agents, another molecule that was studied was a PDGF inhibitor trapidil. It was found that it could efficiently inhibit PDGF-induced ERK activation and PSC proliferation.79 Two p38-induced MAP kinase inhibitors, 4-(4-flurophenyl)-2-(4-methylsulfinphenyl)-5-(4-pyridyl)imidazole and 4-(4-flurophenyl)-2-(4-hydroxyphenyl)-5-(4-pyridyl)1H-imidazole, were also found to inhibit PDGF-induced PSC proliferation.80


Results from the various studies have clearly demonstrated that functions of PSC can be inhibited in CP. Though a few experiments with the same compound had divergent results in certain parameters, for example the effect of CM in procollagen α1 expression, this could be due to differences in experimental designs. Overall, all of the agents mentioned above inhibited PSC activation and the resulting fibrosis, the hallmark of CP. These experiments have also elucidated the molecular mechanisms of action of the different agents. Thus, treatment, prevention and cessation of progression of CP appear feasible; however, it is worthwhile mentioning that the therapeutic modalities in these experimental models of pancreatitis were administered along with or prior to the induction of CP (Table 2), which does not simulate a clinical situation. Therefore, the next big step now is to design phase I and subsequent trials for human CP so that the experimental results, especially for compounds such as TJ-10 and CM (molecules that have already been in use for CP in China and Japan), can be translated from the bench to the bedside.