Article Text


Pancreatic stellate cells: new kids become mature
  1. M Pinzani
  1. Correspondence to:
    Dr M Pinzani
    Dipartimento di Medicina Interna, Università degli Studi di Firenze, Viale GB Morgagni, 85, 50134 Firenze, Italy; m.pinzani{at}

Statistics from

Vitamin A and its metabolites can reverse activation of culture activated pancreatic stellate cells and prevent ethanol induced pancreatic stellate cell activation

It is with great pleasure that I present to the readers of Gut this commentary accompanying the paper by McCarroll and colleagues1 published in this issue of the journal (see page 79). The field of pancreatic stellate cell (PSC) research has grown exponentially in the past five years and major advancements have been made since their first identification as a pathophysiological entity at the end of the 1990s.2,3 In those years, research on hepatic stellate cells (HSCs) and on their role in liver fibrogenesis had reached an elevated degree of sophistication. Therefore, the possibility of isolating stellate cells from rodent or human pancreas led to an almost automatic introduction of PSCs into a new research area: the cellular and molecular mechanisms of pancreatic fibrogenesis.

Fibrosis in the pancreas is consequent to necrosis/apoptosis, inflammation, or duct obstruction. The initial event that induces fibrogenesis in the pancreas is an injury that may involve the interstitial mesenchymal cells, duct cells, and/or acinar cells. Damage occurring in any of these tissue compartments is associated with cytokine triggered transformation of resident fibroblasts/pancreatic stellate cells into myofibroblasts and the subsequent production and deposition of extracellular matrix. The fibrogenic development depends on the site of injury and the involved tissue compartment. Deposition of excessive extracellular matrix is predominantly inter(peri)lobular (as in alcoholic chronic pancreatitis), periductal (as in hereditary pancreatitis), periductal and interlobular (as in autoimmune pancreatitis), or diffuse inter- and intralobular (as in obstructive chronic pancreatitis). In many ways, the development of pancreatic fibrosis recalls the different models of progressive scarring observed in liver tissue following chronic parenchymal damage or bile duct obstruction. Accordingly, it is likely that the two basic profibrogenic mechanisms known to be involved in hepatic scarring are also involved in pancreatic fibrogenesis: (1) chronic activation of the wound healing process with persistent chronic inflammation and progressive substitution of the parenchyma with fibrillar extracellular matrix; and (2) direct profibrogenic and proinflammatory effects of reactive oxygen species and oxidative stress end products (see Pinzani and Rombouts4 for review).

However, there are two main differences due to the different structure and reactivity of the hepatic and pancreatic tissue. Firstly, hepatocytes are able to regenerate and enter a cycle of cell divisions until the original functional mass of the organ is restored. This process is activated through similar basic mechanisms in the presence of both acute and chronic damage. As a consequence, the hepatic fibrogenic process is characterised by an abundant regenerative component that leads to the final cirrhotic outcome (regenerative parenchymal nodules surrounded by fibrous rings). In contrast, pancreatic tissue is characterised by limited regenerative potential and, as a result of its prevalent enzymatic content, is prone to significant fluid extravasation and tissue oedema. In addition, pancreatic tissue is more sensible than liver tissue to abnormal pressure developing within the ductal system.

The bulk of evidence produced in the past five years indicates that there are no major differences between the profibrogenic potential of HSCs and PSCs. Accordingly, PSCs undergo a process of activation and phenotypic modulation towards a “myofibroblast” phenotype following pathways previously described for HSCs. These include, for example, stimulation by proinflammatory cytokines,5 involvement of the peroxisome proliferator activated receptor (PPAR)-γ6 and Rho kinase,7 and the key role of oxidative stress and related products.8 The only different stimulus leading to activation of PSC is the increase in pressure exerted on primary cell culture, an experimental condition aimed at simulating an increase in pressure within the pancreatic tissue as in the case of ductal obstruction.9 Although it is likely that such a stimulus would induce the same effect in HSC cultures, the information appears relevant due to the established closer clinical association between ductal abnormalities and the presence of pancreatic damage. Sustained activation of PSC and their full profibrogenic role are then sustained by the same factors described for HSCs, and in particular platelet derived growth factor, transforming growth factor β1, and angiotensin II.10–13

In addition, as expected from previous research in HSC, the same intracellular signalling pathways mediating the biological effects of these factors are involved in PSC.14–16 Because of the possible major role of oxidative stress in pancreatic fibrogenesis, some studies have started delineating this aspect.17,18 Once again, the results of these studies lead to conclusions identical to those obtained by studies performed in liver tissue or in HSC cultures, and further studies in this direction are highly awaited. Finally, transcriptome analysis aimed at demonstrating whether or not HSCs and PSCs are part of the same lineage has shown that the two cell types are highly similar with minor organ specific variation, whose meaning should be further evaluated.19

All these new acquisitions on the biology of PSC are indeed of high technical and methodological value, particularly considering their rapid development. However, the scheme of development of this area of research had largely followed a track that lacks major originality (that is, most of the available knowledge on the pathogenic role of PSCs has been obtained using HSCs as a template rather than a term of comparison). In any case, it is true that knowledge of the biology of PSCs has reached a sound level of maturity, and research in this area is starting to move into regions more relevant for the understanding of the mechanisms that links chronic pancreatitis to pancreatic inflammation, fibrogenesis, and cancer. In this direction, it has recently been reported that activated PSCs express the protease activated receptor 2 which interacts with trypsin and tryptase, two key pancreatic enzymes involved in the pathogenesis of chronic pancreatitis.20 Trypsin and tryptase were able to induce PSC proliferation and collagen synthesis through activation of c-Jun N-terminal kinase and p38 mitogen activated protein kinase.

The potential contribution of PSCs to the development and progression of pancreatic cancer appears indeed fundamental and sound advancements have been made in this area. The first important observation is that malignant cells can actively alter the microenvironment of the pancreatic tissue by modulating the composition of the extracellular matrix in a tumour favourable way through synthesis and release of soluble factors.21 Accordingly, recent evidence suggests that pancreatic cancer promotes the activation/proliferation of PSCs and the consequent increase in extracellular matrix synthesis.22,23 Marked accumulation of fibrillar extracellular matrix, and particularly collagen type I, in peritumoral areas leads to the so-called desmoplastic reaction, often observed in pancreatic cancer. PSCs have been shown to represent a key cellular component in this type of stromal reaction.24

Although the desmoplastic reaction is classically indicated as a phenomenon limiting the expansion of the cancer mass, there are data indicating that collagen type I is able to promote the malignant phenotype of pancreatic adenocarcinoma.25 It is therefore likely that PSCs can influence the organisation and progression of pancreatic cancer, providing key components of the tumour stroma. Along these lines, it is worth investigating the possible production of proteases and other factors involved in tumour invasion by PSCs. This area of investigation is now very active and major advances, potentially transferable to PSCs, have recently been made for HSCs.26 Finally, a recent important observation has been provided by a study demonstrating that pancreatic cancer cells are able to increase expression of cyclooxygenase 2 (COX-2) in PSCs, and COX-2 expression is associated with several human cancers, including pancreatic adenocarcinoma.27

The last topic worth addressing is the potential implications for therapy of chronic pancreatitis arising from the advances in PSC research. Firstly, in the context of the relevant role of PPAR-γ in PSC activation, two studies have shown that troglitazone, a PPAR-γ agonist, reduced the profibrogenic activity of PSCs and progression of chronic pancreatitis in mice.28,29 Interestingly the antifibrogenic effect of troglitazone seemed to be independent of PPAR-γ.28 Glitazones, pioglitazone in particular, are currently indicated as potential therapeutic agents for liver diseases such as chronic alcoholic and non-alcoholic steatohepatitis.30–32 It is therefore relevant that the same class of drugs could be used to reduce fibrogenic progression in both the liver and pancreas in those patients in which the two organs are affected by the same aetiological agent.

Other pharmacological agents that have been shown to produce a potential antifibrogenic effect in PSC cultures or animal models of chronic pancreatitis include plant derived polyphenolic antioxidants such as epigallocatechin-3-gallate33 and ellagic acid,34 and the trypsin inhibitor camostat mesilate.35,36

McCarroll and colleagues1 investigated the effect of retinol and its metabolites on the activation state of PSCs. They demonstrated that these compounds can reverse activation of culture activated PSCs and prevent ethanol induced PSC activation, both effects being mediated through the MAPK pathway. The study contains novel and original information and not just in the field of PSC biology. Indeed, several findings emerging from this work should be confirmed in HSCs which are clearly more involved in vitamin A metabolism and are similarly exposed to ethanol under conditions of chronic abuse. What is debatable about this otherwise excellent study is the title and the last paragraph in the discussion (that is, the possible use of retinoids for the treatment of pancreatic fibrosis). The possibility of employing retinoids for treating hepatic fibrosis in humans has been a key issue for some time since the beginning of the 1990s. However, this option was abandoned for two main reasons: (1) the need to chronically use very high doses in order to achieve less than 50% of the serum concentration effective in animal models and less than 1% of the concentration effective in cell cultures; and (2) the toxic effect of vitamin A accumulation, which is paradoxically able to induce extensive fibrosis and cirrhosis of the liver or non-cirrhotic portal hypertension. These concerns obviously apply to the proposed use for chronic pancreatitis.

In conclusion, I believe that the PSC area of research has become mature, and will develop in directions more relevant to the pathophysiology of the pancreas. A final warning for liver fibrosis researchers: be alerted, the kids have become adults!

Vitamin A and its metabolites can reverse activation of culture activated pancreatic stellate cells and prevent ethanol induced pancreatic stellate cell activation


View Abstract


  • * Pinzani M. New kids on the block: pancreatic stellate cells enter the fibrogenesis world. Gut 1999;44:451–2.

  • Conflict of interest: None declared.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles