Changing the Pathogenetic Roadmap of Liver Fibrosis? Where Did It Start; Where Will It Go?

Olav A. Gressner; Mohamed S. Rizk; Evgeniya Kovalenko; Ralf Weiskirchen; Axel M. Gressner


J Gastroenterol Hepatol. 2008;23(7):1024-1035. 

In This Article

Where Will It Go?

The 'canonical principle' of fibrogenesis based on generation of liver-resident MFB by cytokine-driven transdifferentiation of HSC is still thought to be the key pathogenetic mechanism of liver fibrosis, but studies of fibrosis in the kidney, lung, skin, and liver (see above) reveal a considerable heterogeneity of activated fibroblasts.[113,114] These observations suggest a number of supplementary origins of MFB in fibrogenesis.[115] The understanding of these mechanisms is essential for future development of innovative biomarkers and therapeutic approaches of fibrogenesis.

Contribution of Bone Marrow-derived Cells

In progressive renal fibrosis more than 30% of α-smooth muscle actin-positive interstitial myofibroblasts are derived from the bone marrow.[116] But also in the liver immature, multipotent bone marrow cells have the capacity to differentiate to hepatocytes, cholangiocytes, sinusoidal endothelial cells, and Kupffer cells, if the adequate microenvironment of the liver is present.[117,118] This was recently extended to HSC and (myo-)fibroblasts under experimental and clinical conditions. By transplantation of genetically tagged bone marrow or of male bone marrow (Y-chromosome) to female mice, it was estimated that up to 30% of HSC in the liver originate from the bone marrow and acquire the MFB phenotype under injurious conditions.[119] Another study indicates that up to 68% of HSC and 70% of MFB in CCl4-cirrhotic mice liver are derived from the bone marrow.[120] Even in human liver fibrosis, a significant contribution of bone marrow cells to the population of MFB has been proven, even though it is presently unclear which type of specific bone marrow cells or mesenchymal stem cells is relevant for the generation of hepatic (myo-) fibroblasts.[121] Another experimental study has shown that myelogenic fibrocytes are present in the liver, which can be differentiated by TGF-ß to collagen-producing MFB.[122] They are a subpopulation of circulating leukocytes displaying a unique surface phenotype with positivity for CD45 (hematopoietic origin), CD34 (progenitor cell), and type I collagen (capability of matrix synthesis).[123,124] They have potent immuno-stimulatory activities.[125] Fibrocytes represent a systemic source of contractile MFB in various fibrotic lesions, such as lung, keloids, scleroderma, and fibrotic changes of the kidney.[126,127] The mobilization of bone marrow cells and their recruitment into the damaged tissue is a central mechanism of tissue fibrosis and wound healing,[128] which is most likely regulated by colony-stimulating factors (CSF), such as granulocyte-CSF (G-CSF)[129] and chemokine receptors 4 (CXCR4), 2 (CCR2), and 7 (CCR7).[125,130] Thus, activated HSC probably play an important role since they secrete a broad spectrum of inflammatory mediators (chemokines, M-CSF, SCF, PAF) and leukocyte adhesion molecules (ICAM-1, VCAM-1, NCAM) required for recruitment, activation, and maturation of blood-born cells at the site of injury.[131] The homing of myelogenic cells in the damaged liver was claimed to also have a positive effect on the resolution of liver fibrosis, since these cells express MMPs, which augment the degradation of fibrotic extracellular matrix.[129]

Furthermore, a striking relationship between increasing hepatic fibrosis and periportal ductular reaction has been demonstrated.[132] The role of steatosis in chronic hepatitis C infection-related fibrosis is associated with an increase in both the number of cytokeratin-7-positive hepatic progenitor cells and the extent of the ductular reaction, providing a potential mechanism whereby steatosis contributes to the progression of portal fibrosis.[132]

Contribution of Peripheral Blood Cells

Recent studies indicate a highly developed multidifferentiation potential of a subgroup of circulating blood monocytes, which can be recruited quickly for tissue repair processes.[133] In addition, the content of circulating myelogenic stem cells in the blood is suggested to be important for regenerative mechanisms as found in ischemic and degenerative diseases (i.e. myocardial infarction). In vitro investigations over recent years have shown that peripheral blood monocytes can differentiate in hepatocyte-like cells if they are exposed to macrophage-colony stimulating factor (M-CSF) and specific interleukins (monocyte-derived neo-hepatocytes).[134,135] Although not yet proven for liver fibrogenesis, subgroups of monocytes may also differentiate into fibroblast-like cells (fibrocytes) after entering the damaged tissue. Here, they participate in fibrotic processes, e.g. of the lung and kidney. The differentiation is positively influenced by G-CSF, M-CSF, monocyte chemotactic peptide 1 (MCP-1), and other chemokines and hematopoietic growth and differentiation factors, which are also expressed and secreted by activated HSC[50,136,137,138] and other liver cell types.[139] It is of interest that recently an inhibitory effect of the acute-phase protein serum amyloid P (SAP) on the process of differentiation of monocytes to fibrocytes could be established[140] and, consequently, a preventive effect of SAP-injections on the development of bleomycin-induced lung fibrosis was reported.[141] Since SAP is synthesized in hepatocytes, severe liver injury might facilitate the monocyte-fibrocyte differentiation process due to reduction of the inhibitory SAP. It should be emphasized that this mechanism is presently speculative for the liver, but circulating monocytes might be a pool for immediate repair processes of liver damage. However, beside special monocytes as source of fibroblasts in the fibrotic liver, also circulating stem cells should be considered, which are positive for CD34 and CXCR4 (a chemokine receptor).[133] G-CSF and the stromal derived factor (SDF)-1 are probably the most important regulators of stem-cell mobilization from the bone marrow and their integration into the damaged tissue followed by differentiation to fibroblasts and other cells (see above).

Epithelial-mesenchymal Transition (EMT)

Beside activation and transdifferentiation of HSC, which developmentally derive from the septum transversum mesenchyme, from endoderm, from the mesothelial liver capsule[142] or even from other embryonic origins,[143] recent studies point to a potentially important mechanism for the enlargement of the resident pool of fibroblasts during the fibrotic reaction of the damaged organ, e.g. in kidney and lung.[144] This process, designated as epithelial-mesenchymal transition (EMT), is well known in the context of embryonic development, but is now recognized as an important mechanism in the generation of fibroblasts during fibrogenesis in adult tissues[145] (Figure 7). It has been proven that in fibrotic kidney disease tubulus epithelial cells can transdifferentiate to fibroblasts expressing collagens and the fibroblast-specific protein 1 (FSP-1), also known as S100A4 calcium-binding protein.[145] It is estimated that in kidney fibrosis about 40% of all fibroblasts are the result of EMT.[146] Similarly, alveolar epithelial cells of the lung are subject to EMT and even cardial endothelial cells can switch to fibroblasts under conditions of damage (mesenchymal-mesenchymal transition). In vitro and in vivo observations made in blood vessels following sustained inflammation support the hypothesis that vascular endothelial cell transformation to myofibroblast-like cells may increase matrix proteins in fibrotic diseases.[147] Recent studies further provide evidence for the importance of EMT in liver fibrogenesis, as evidenced by transition of albumin-positive hepatocytes to FSP-1 positive and albumin-negative fibroblasts.[148] It is claimed that up to 45% of hepatic fibroblasts are derived from hepatocytes, and up to 60% of FSP-1-positive hepatocytes are colabeled with albumin indicating an intermediate transitional stage of EMT of hepatocytes.[148] Another recent report has shown EMT of mature mouse hepatocytes in vitro and of the mouse hepatocyte cell line AML12.[149] Here, the EMT-state was indicated by strong up-regulation of α1(I) collagen mRNA expression and type I collagen deposition. Thus, hepatocytes are capable of EMT changes and type I collagen synthesis and might be a source of a substantial population of (myo-)fibroblasts in fibrogenesis. A further target for EMT are cholangiocytes (bile duct epithelial cells). In primary biliary cirrhosis (PBC) it has been proven that bile duct epithelial cells express FSP-1 (S100A4) and vimentin as early markers of fibroblasts.[150] The bidirectional consequence of EMT for cholangiocytes are ductopenia (reduction of bile ducts) and enlargement of the pool of portal fibroblasts, which significantly contribute to portal fibrosis. In vitro studies with cultured human cholangiocytes have confirmed the clinical observations described. Thus, EMT as a result of hepatocellular pluripotency proves to be a general pathogenetic principle of chronic cholestatic liver diseases.[151] In addition, activation and proliferation of portal/periportal mesenchymal cells to peribiliary MFB, which are stimulated in a paracrine manner by bile duct epithelial cells via TGF-ß, PDGF-BB, and endothelin-1[152] are important pathogenetic mechanisms of portal fibrosis and septa formation in cholestatic liver diseases.[114] Indeed, only a minority of ECM-producing MFB in obstructive cholestatic injuries are derived from HSC.[114,153] This also underlines the heterogeneous origin of MFB in fibrogenesis and emphasizes the importance of the underlying fibrogenic liver disease.[112]

Figure 7.

Currently known mechanisms of fibrogenesis, i.e. HSC activation by liver cell necrosis, EMT, influx of fibrocytes from the blood, and activation of periportal fibroblasts at sites of injury. Expression of typical cellular markers is indicated. FSP, fibroblast-specific protein. Explanation in the text.

The molecular inducers of EMT are TGF-ß,[145] epidermal growth factor (EGF), insulin-like growth factor (IGF)-II, and fibroblast growth factor (FGF)-2, which promote the genetic and phenotypic programming of epithelial cells to mesenchymal cells (fibroblasts). The prototype of the most powerful inducer of EMT is TGF-ß.[148] The inducing function of TGF-ß for the above-described mesenchymal transition of mouse hepatocytes was shown by activation of Smad2/3 phosphorylation, inhibition by Smad4 silencing using siRNA, and induction of the snail transcription factor.[149] Interestingly, TGF-ß induces EMT only of those hepatocytes, which escape from the pro-apoptotic effects of this cytokine.[154,155] The subpopulation of surviving hepatocytes exhibits a strong overexpression of snail by TGF-ß conferring resistance to programmed cell death.[156] Several additional pathways are involved in the generation of apoptosis resistance, e.g. proteinkinase A[154] and epidermal growth factor (EGF)/transforming growth factor-alpha (TGF-α).[155] Thus, EMT of hepatocytes is dependent on the balance between apoptotic and survival mechanisms. The process of EMT requires the action of metalloproteinases and a TGF-ß dependent snail-mediated down-regulation of E-cadherin both contributing to the release of epithelial cells from cell-cell and cell-basement membrane binding (Figure 7). The most important molecular counterpart is the bone morphogenetic protein (BMP)-7, which belongs to the TGF-ß superfamily. BMP-7 does not only inhibit EMT, but can even induce a mesenchymal-epithelial (retro-)transition (reverse EMT = MET).[157] BMP-7 has been shown to inhibit TGF-ß dependent EMT of hepatocytes and the progression of experimental fibrosis in mice.[148] It has also antiapoptotic properties, anti-inflammatory, and proliferation-stimulating effects.[158] BMP-7 inhibits TGF-ß signaling via Smads,[159] which transduce the effect of the latter cytokine from its receptor, a serine/threonine kinase, to the Smad-binding element (SBE) of respective target genes in the nucleus.[160] In addition, several trapping proteins such as the small proteoglycans decorin and biglycan, latency associated peptide (LAP), BAMBI (BMP- and activin-membrane-bound inhibitor), KCP (kielin-chordin-like protein), gremlin, and α2-macroglobulin change the balance between TGF-ß and BMP-7 in favor of an anti-EMT effect by binding and neutralization of TGF-ß.[161] Similarly, the important downstream-modulator protein connective tissue growth factor (CTGF/CCN2),[162] which is expressed in hepatocytes, HSC, portal fibroblasts, and cholangiocytes[163,164] changes the functional TGF-ß/BMP-7 ratio.[165] CTGF is over-expressed in experimental and human liver cirrhosis,[164,166,167] which is mediated mainly by TGF-ß, but also by endothelin-1,TNF-α, vascular endothelial growth factor (VEGF), nitrogen oxide (NO), prostaglandin E2, thrombin, high glucose, and hypoxia.[168] CTGF inhibits BMP, but activates TGF-ß signaling by modulation of the receptor-binding of these ligands.[165] The prominent functional role of CTGF is supported by very recent data, which show sustained antifibrotic effects if CTGF expression is reduced by siRNA.[169,170] Taken together, EMT, but also MET (mesenchymal-epithelial transition), and in special conditions, even MMT (mesenchymal-mesenchymal transition, e.g. of vascular endothelial cells to fibroblasts), and the fine tuning of the bioactive TGF-ß/BMP-7 ratio and of their adaptor- and trapping proteins, offer multiple regulatory possibilities of influencing fibrogenesis. These mechanisms are known in some detail for the kidney,[171] but still need more experimental proof for the liver, in particular with regard to their quantitative contribution to fibrogenesis.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: