SEARCH

SEARCH BY CITATION

Keywords:

  • connective tissue growth factor;
  • CTGF;
  • liver fibrosis;
  • pathogenesis;
  • diagnosis;
  • therapy;
  • translational medicine

Abstract

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

Connective tissue growth factor (CTGF=CCN2), one of six members of cysteine-rich, secreted, heparin-binding proteins with a modular structure, is recognized as an important player in fibrogenic pathways as deduced from findings in non-hepatic tissues and emerging results from liver fibrosis. Collectively, the data show strongly increased expression in fibrosing tissues and transforming growth factor (TGF-β)-stimulated expression in hepatocytes, biliary epithelial cells and stellate cells. Functional activity as a mediator of fibre–fibre, fibre–matrix and matrix–matrix interactions, as an enhancer of profibrogenic TGF-β and several secondary effects owing to TGF-β enhancement, and as a down-modulator of the bioactivity of bone morphogenetic protein-7 has been proposed. By changing the activity ratio of TGF-β to its antagonist bone-morphogenetic protein-7, CTGF is proposed as a fibrogenic master switch for epithelial–mesenchymal transition. Consequently, knockdown of CTGF considerably attenuates experimental liver fibrosis. The spill-over of CTGF from the liver into the blood stream proposes this protein as a non-invasive reporter of TGF-β bioactivity in this organ. Indeed, CTGF-levels in sera correlate significantly with fibrogenic activity. The data suggest CTGF as a multifaceted regulatory protein in fibrosis, which offers important translational aspects for diagnosis and follow-up of hepatic fibrogenesis and as a target for therapeutic interventions. In addition, CTGF-promoter polymorphism might be of importance as a prognostic genetic marker to predict the progression of fibrosis.


Principles of liver fibrosis

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

Most environmental and genetic liver diseases cause a complex distortion of parenchymal [hepatocytes (PC)] and non-parenchymal [Kupffer cells, sinusoidal endothelial cells, hepatic stellate cells (HSC)] cell populations owing to an overlay of necrosis, apoptosis, proliferation and altered cellular functions. The loss of cellular homoeostasis is frequently accompanied by and causally related to scarring, i.e. increased production (fibrogenesis), deposition (fibrosis) and remodelling (fibrolysis) of the extracellular matrix (ECM).

In normal liver, HSC, which are pericytes found in the perisinusoidal space of the liver that are also known as Ito cells or fat-storing cells, are described as being in a quiescent state, in which about 85% of all hepatic vitamin A metabolites (retinoids) are dissolved and stored (1). Quiescent stellate cells represent 5–8% of the total number of liver cells.

The capability for transdifferentiation of resting HSC to activated, ECM-producing myofibroblasts (MFB) driven by necro-inflammatory conditions led to an unprecedented interest in these cells (and their activated form) as the main collagen- and ECM-producing cells in the process of liver fibrosis (2–4).

Very recently, the view on fibrogenesis and fibrolysis in the liver became highly complicated (5–8). Besides the ‘canonical pathway’ of HSC activation, several supportive pathogenetic mechanisms are suggested. These are epithelial–mesenchymal transition (EMT) of PC (9–11) and biliary epithelial cells (12–14) to (myo-)fibroblasts, influx of bone marrow-derived fibrocytes into injured liver tissue (15–18) and recruitment and transdifferentiation of peripheral blood cells, in particular, monocytes (19, 20). It is likely that these mechanisms follow a time-dependent sequence during fibrogenesis after the early-onset mechanisms of HSC activation and transdifferentiation. Furthermore, the relative contribution of these pathways might be dependent on the cause of fibrosis.

The fibrogenic mechanisms are dependent on the interplay of many pro- and anti-fibrotic/-inflammatory cytokines (21, 22). The hierarchy of pro-fibrogenic growth factors includes platelet-derived growth factor (PDGF) and transforming growth factor (TGF-β); the latter one is designated as ‘fibrogenic master cytokine’ with multiple effects on ECM turnover (23, 24), hepatocellular apoptosis (25–27), proliferation and liver regeneration (23, 28, 29), inflammation and immunosuppression (30) and cancerogenesis (31). The natural antagonist of many actions of TGF-β is bone-morphogenetic protein-7 (BMP-7), a member of the TGF-β superfamily (32). It counteracts TGF-β-induced EMT (10, 33, 34), ECM synthesis (35) and inhibition of PC-proliferation (36). Thus, the balance of both growth factors, i.e. TGF-β and BMP-7, will be crucial for the development of fibrosis and outcome of (chronic) liver disease.

Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

Connective tissue growth factor (CTGF) was originally discovered in 1991 as a secreted ‘PDGF-related mitogen’ in the conditioned medium of human umbilical vein endothelial cells and designated as CTGF (37). It is one of six unique proteins with a similar secondary structure and sequence homologies, of which cysteine-rich protein 61 (cyr61=CCN1), connective tissue growth factor (CTGF=CCN2) and nephroblastoma overexpressed protein (NOV=CCN3) are grouped to designate the CCN family of secreted, cysteine-rich (glyco-)proteins with a modular architecture (38, 39) (Table 1). The additional members of the CCN gene family are Wnt-1-induced proteins (WISP) as follows: WISP-1 (CCN4) and WISP-2 (CCN5), WISP-3 (CCN6) (40). The CTGF protein contains four structural modules, which follow a 37 amino acid leader peptide sequence: an insulin-like growth factor binding protein module (I), a von Willebrand factor type C module (II), a thrombospondin (TSP) type I homology module (III) and a carboxy-terminal cysteine knot motif, heparin-binding module (IV) (38, 39) (Figs 1 and 2). Modules I and II are predicted to comprise the N-terminal domain joined by the hinge region with the C-terminal domain of modules III and IV. The hinge region is a proteinase-sensitive cleavage site, which gives rise to N- and C-terminal fragments respectively (41). Chymotrypsin and plasmin are relevant proteinases, which not only cleave the hinge region but also modules I and IV (plasmin). Together with the C-terminal heparin-binding module IV of 10–12 kD CTGF immunoreactive fragments of 10, 16 and 20 kD in addition to the full-length CTGF molecule of 38 kD can be found in body fluids and conditioned media of some cell cultures (41). The structural modules fulfil specialized functions of CTGF with binding of insulin-like growth factor (IGF) (module I), TGF-β (II), certain integrins (α4β1, α5β1, α6β1 and αvβ3) and cell surface receptors (III) and proteoheparan sulphate proteoglycans like syndecan 4 and perlecan (IV).

Table 1.   Compilation of data on structure and function of connective tissue growth factor
  1. cAMP, cyclic adenosine monophosphate; CTGF, connective tissue growth factor; EGF, epidermal growth factor; HGF, hepatocyte growth factor; IGF, insulin-like growth factor; PDGF, platelet-derived growth factor; PPAR, peroxisome proliferator activated receptor; TGF, transforming growth factor; TNF, tumour necrosis factor; VEGF, vascular endothelial growth factor.

Discovery/terminologyDiscovered in 1991 in conditioned media of human umbilical vein endothelial cells (HUVECs) as a cysteine-rich, secreted protein having chemotactic and mitogenic activity and termed connective tissue growth factor One of the 6 members of the CCN family of cysteine-rich proteins: Cyr61 (CCN1), CTGF (CCN2), NOV (CCN3), WISP-1 (CCN4), WISP-2 (CCN5), WISP-3 (CCN6) share 40–60% sequence similarity
Gene structureChromosome 6q 23.1; 3 kb size; promoter polymorphism, 5 exons, 5 introns 5′-region regulatory elements: AP1, TATA box binding sites, HIF-binding site, BCE-1, CArG-like box, Sp1, Ets-1, TGF-β response and Smad-binding element Single 2.4 kb transcript
Protein structureMr 36–38 kD (microheterogeneity), 343–349 amino acid residues 37 amino acids signal peptide (secreted protein) N-linked glycosylation of residues 28–30 and 225–227 4 module structure (functional domains), heparin-binding site Substantial interspecies homology Proteolytic fragments of 10–20 kD: N-, C-terminal domains, isolated module IV
BiosynthesisMultiple mesenchymal and epithelial cell types: vascular smooth muscle cells, (myo-)fibroblasts, (sinusoidal) endothelial cells, mesangial cells, hepatic stellate cells, tubular epithelial cells, hepatocytes, bile duct epithelial cells, neuronal cells, astrocytes
RegulatorsPositive: TGF-β, VEGF, PDGF, endothelin-1, shear stress, cell stretch, hypoxia, H2O2, O2, high glucose, serotonin, lysophosphatidic acid, IGF-1, TGFα, EGF, HGF, angiotensin II Negative: TNFα, cAMP, NO, PPAR-γ
Clearanceα2-macroglobulin/low-density lipoprotein receptor-related protein (LRP), e.g. of hepatocytes, fibroblasts, hepatic stellate cells, endosomal degradation
Presence in body fluidsSerum, cerebrospinal, amniotic, peritoneal, follicular, uterine, eye, wound fluid
Biological functionsExtracellular matrix synthesis (TGF-β downstream modulator) DNA synthesis/proliferation Chemotaxis Cell migration Apoptosis/cell survival/cell differentiation Transdifferentiation Fibroblast adhesion
PathobiologyOverexpression in fibrotic lesions, fibrogenesis Wound healing Angiogenesis (neovascularization) Epithelial–mesenchymal transition (EMT)
image

Figure 1.  Structure of the human CTGF gene, its mRNA and protein as well as its module-related functions. Module II binds to TGF-β, module III to cell surface low-density lipoprotein receptor-related protein-1 (LRP-1)/α2-macroglobulin receptor and integrins, module IV binds to specific integrins, fibronectin and cell surface proteoheparan sulphate. Proteolytic cleavage sites at the hinge region and between modules III and IV are indicated. S, I, V, T and C denote the coding regions for the signal peptide, IGF-BP module, VWC module, TSP1 module and CT module respectively; IGF-BP, insulin-like growth factor binding protein module (I); ORF, open reading frame; CTGF, connective tissue growth factor; TGF, transforming growth factor.

Download figure to PowerPoint

image

Figure 2.  Modular structure, cellular binding sites and intracellular signal generation of CTGF. CTGF is internalized via LRP-1 and degraded in the endosomal/lysosomal compartment; a fraction might escape into the nucleus. Binding of CTGF to LRP-1 changes the Wnt/β-catenin signalling pathway by competing with specific Wnt-ligands. IGF-BP, insulin-like growth factor binding protein module (I); VWC, von Willebrand factor type C module (II); TSP1, thrombospondin type I homology module (III); CT, carboxy-terminal cysteine knot motif, heparin-binding module (IV); FAK, focal adhesion kinase; CTGF, connective tissue growth factor; TGF, transforming growth factor; LRP, low-density lipoprotein receptor-related protein-1.

Download figure to PowerPoint

The CTGF gene maps to chromosomal location 6q23.1 but the genomic region has not been fully characterized. Single nucleotide polymorphisms (SNPs) were identified; four of them are located in the promoter (42, 43). A specific SNP was recently shown to be significantly associated with susceptibility to systemic sclerosis (44). Several general transcription factor-binding sites have been predicted in the promoter region of the CTGF gene, such as AP-1, Sp1, hypoxia-inducible factor (HIF)-binding site, basal control element (BCE)-1, TATA box binding sites, Ets-1, nuclear factor (NF)-κB-binding site, an actin-sensitive CArG-like box, a unique TGF-β response and Smad-binding element (45–49). Consequently, CTGF expression is regulated by several signalling mechanisms including pathways of TGF-β/Smad (50), Ras/MEK/ERK, protein kinase C (PKC) (51, 52), NF-κB, JAK/STAT, cyclic adenosine monophosphate (AMP) (53), the small GTPase RhoA, serum response factor (49), phosphatidylinositol 3-kinase (PI3K) and reactive oxygen species (54). The final effects, i.e. up or downregulation of CTGF will depend on the cell type (mesenchymal vs. epithelial), the local environment of the cell and interacting regulatory factors. Several of these pathways act together in the TGF-β-driven induction of CTGF and an intensive crosstalk must also be supposed for other extracellular stimuli. External factors, which cell-type dependently upregulate CTGF expression, include, besides TGF-β as the major inducer, serum response factor (49), vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), fibroblast growth factor (FGF), thrombin, clotting factors VIIa, Xa, endothelin-1, lysophosphatidic acid, high glucose, hypoxia, oxidative and mechanical stress (55, 56). Negative regulators of CTGF expression comprise tumour necrosis factor (TNF) (57), prostaglandin E2 and cyclic AMP (53, 58, 59) and the peroxisome proliferator activated receptor (PPAR)-γ (60). Interestingly, overexpression of CTGF in scleroderma fibroblasts appears to be independent of the profibrogenic TGF-β/Smad signalling (47), but dependent on constitutively increased Sp1-binding activity (48). Intracellular TGF-β signalling requires Smad3 and is inhibited by Smad7 overexpression (47, 50).

Cellular biology of connective tissue growth factor

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

Although CTGF is localized in the cytosol and even in the nucleus of mesangial cells (61), the vast majority is secreted and fulfils functions in the extracellular space in the immediate micro-environment of the cells. The protein is also present in plasma, serum (62, 63) and urine (64, 65). A significant fraction is retained in platelets, but released upon activation during the coagulation process (66, 67). A content of 2.5 fg per single platelet was calculated (66). Thus, disaggregation of platelets at the site of injury is an important mechanism for local increase of CTGF, e.g. in fibrotic tissue lesions and atherosclerosis (66). The extracellular half-life time of CTGF in the circulation or interstitial space is not known. The uptake and clearance is mediated by the high-molecular-weight multi-ligand receptor, i.e. the low-density lipoprotein receptor-related protein/α2-macroglobulin receptor (LRP) (68). The LRPs are known to bind many diverse ligands including α2-macroglobulin (α2-M) and CTGF (69). LRP antibodies and cells that are genetically deficient for LRP are unable to bind CTGF (68). Bound CTGF is rapidly internalized and degraded in an LRP-dependent manner. However, in human mesangial cells it was shown that internalized CTGF is captured in endosomes and accumulates in juxtanuclear organelles, from which the growth factor is translocated into the cytosol. Here, it is phosphorylated by PKC and can be translocated into the nucleus possibly acting on transcription (61). Another effect of the binding of CTGF to LRP is related to the Wnt signalling pathway. It was demonstrated that CTGF regulates a signalling through the Wnt/β-catenin pathway in accordance with its ability to bind to the Wnt co-receptor LRP(6) (70). There CTGF might compete with Wnt family members for LRP-binding leading to inhibition of Wnt signalling. The findings suggest a further role for CTGF as a coordinator of multiple intracellular signalling pathways. With regard to CTGF clearance, it is of interest that the removal of external CTGF follows a similar pathway as circulating TGF-β, for which an α2-M-receptor-dependent mechanism was proven (71). It is not yet known whether CTGF in the circulation interacts with α2-M or other binding proteins potentially involved in the clearance mechanism. The LRP-binding site is located in module III of the C-terminal domain, which implies that the N-terminal fragment in the circulation might have a longer half-life time than the C-terminal part and the full-length CTGF molecule respectively. This assumption was proven by selective measurement of CTGF and its C- and N-terminal cleavage products in plasma of scleroderma patients. Only N-terminal CTGF was elevated, whereas no significant change of whole CTGF and of the C-terminal fragment was noticed (62).

Functional aspects of connective tissue growth factor

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

General aspects

It is well established that CTGF plays a crucial role in chondrogenesis and angiogenesis during skeletogenesis and, thus, in embryonic development (72). As expected, mice lacking the CTGF gene (CTGF−/−) die soon after birth because of multiple skeletal defects (72, 73). Fibroblasts derived from CTGF−/− knockout embryos have a significantly reduced expression of pro-adhesive, pro-inflammatory and pro-angiogenic genes, which is related to diminished expression of the proteoheparansulphate subtype syndecan 4 (74). In conclusion, CTGF proved to be a central regulator of ECM in cartilage and bone but also of vascularization during embryonic development with a widespread distribution in adult tissues and organs (for general reviews, see references (40, 75)).

Connective tissue growth factor affects important cellular functions, e.g. proliferative activity (58, 76), differentiation of cells, ECM synthesis (77), cell adhesion to ECM (78, 79), cell migration (78) and EMT (80). Most of these effects are not executed by CTGF alone, but in combination with specific growth factors or through direct interaction with ECM or cell-surface molecules.

The structural and functional subspecialization of the CTGF molecule provides multiple ways of interaction both with components of the ECM such as fibronectin, collagens or TSP through modules III and IV and membrane-bound molecules [e.g. integrins, low-density lipoprotein (LDL) receptor-related protein (LRP)] through module III. Direct interactions with growth factors (e.g. TGF-β, IGF and VEGF) (81) have been described for modules I–IV (Fig. 3).

image

Figure 3.  Interaction of each module in CCN2/CTGF with other growth factors. IGF-BP, insulin-like growth factor binding protein module (I); VWC, von Willebrand factor type C module (II); TSP1, thrombospondin type I homology module (III); CT, carboxy-terminal cysteine knot motif, heparin-binding module (IV); IGFs, insulin-like growth factors; VEGF, vascular endothelial growth factor; CTGF, connective tissue growth factor.

Download figure to PowerPoint

Interaction with extracellular matrix components

Covalent or non-covalent binding to matrix molecules such as fibronectin, proteoglycans and collagens has been described previously for a variety of pro-fibrogenic growth factors such as TGF-β, PDGF-B and PDGF-D, endothelin-1, several fibroblast growth factors (FGFs), insulin-like growth factor I, tumour necrosis factor (TNF)-α or adipocytokines (leptin, adiponectin) (reviewed in (6)). Thus, a similar mechanism of protein–protein interaction in fibrotic tissue may also be proposed for CTGF. For example, it was shown in chondrocytes that the interaction between the fibronectin type I repeat modules (the N-terminal fibrin, the gelatin–collagen and the C-terminal fibrin-binding domains) of fibronectin and the CT domain of CTGF (which also binds to integrin α5β1 in fibroblasts) is impaired in the presence of a monoclonal antibody against the CT domain and by an antibody against α5β1, thus suggesting a direct interaction of CTGF with the matrix molecule fibronectin, in which integrin α5β1 is involved. In the same study, an enhancing effect of CTGF on the binding of fibronectin to fibrin was demonstrated (82). In conclusion, it may thus be that CTGF contributes to the ECM accumulation in wound healing and fibrotic disorders and tissue fibrosis by promoting fibre–fibre, fibre–matrix and matrix–matrix interactions through direct molecular interactions with matrix components.

Interaction with the transforming growth factor-β cytokine superfamily

Connective tissue growth factor was described to modulate signalling responses of target cells to cytokines, such as TGF-β, FGF, EGF, VEGF, PDGF and others (39). Most importantly, CTGF synergizes the action of the fibrogenic master cytokine TGF-β and, thus, is recognized as a TGF-β downstream modulator protein (39, 83). This function is not dependent on the integral structure of CTGF, because the N-terminal domain with the TGF-β-binding module II is obviously sufficient whereas the C-terminal domain with module IV mediates fibroblast proliferation (84). Thus, depending on proteolytic cleavage at the hinge region CTGF fragments and the full-length molecule might induce quite different cellular responses (84). From a pathophysiological point of view the relation between TGF-β, the major stimulus for CTGF expression, and CTGF is of great significance. Overexpression of CTGF (and TGF-β) is observed in numerous fibrotic tissue and organ reactions, such as synovial fibrosis (85), gingival fibrosis (86), keloids (87), scleroderma (88) and fibrosis of the kidney (89), pancreas (90), lung (91), heart (92), liver (93) and bowel (94). CTGF is overexpressed in cells isolated from these lesions and some molecular defects, such as constitutive hyperactive Sp1-binding site in scleroderma fibroblasts, have been identified (48).

Connective tissue growth factor in epithelial–mesenchymal transition

Epithelial–mesenchymal transition describes a series of events in which epithelial cells are released from the surrounding tissue following loosening of cell–ECM and cell–cell interactions, in which the cytoskeleton is reorganized to confer the ability to move through a three-dimensional ECM, and in which a new transcriptional programme is induced to achieve a mesenchymal phenotype (95).

Current data provide compelling evidence that the accumulation of fibroblasts, leading to excessive collagen, and other matrix components at sites of chronic inflammation, which lead to scar tissue formation and progressive tissue injury, arise from an EMT of cells at the site of injury, and that EMT is likely involved in the progressive fibrotic diseases of the heart, lung, liver and kidney (96, 97).

The prototype of the currently most powerful inducer of EMT is TGF-β (98). The inducing function of TGF-β for the above-described mesenchymal transition of mouse hepatocytes was suggested to take place via an activation of Smad2/3 phosphorylation, inhibition by Smad4 silencing using siRNA and induction of the snail transcription factor (98).

The synergism between TGF-β and CTGF is at least partially because of an enhancement of TGF-β binding to its receptor, whereas that of BMP, the TGF-β antagonist, is inhibited (99). Both proteins bind directly to the von Willebrand module of CTGF. Thus, CTGF enhances Smad2 phosphorylation and TGF-β-activated target gene expression (99). Because the CTGF promoter contains a TGF-β response element, an autocrine upregulation of the TGF-β response of the target gene is feasible (under certain conditions, see below). This mechanism shifts the epithelial–mesenchymal balance greatly to the direction of EMT, which is governed by the bioactivity ratio of TGF-β/CTGF to BMP-7/ hepatocyte growth factor (HGF) (34, 80). A further supportive mechanism necessary for an autocrine upregulation of TGF-β response by CTGF as discussed below was recently shown in human mesangial cells (100). CTGF enhanced TGF-β/Smad signalling by transcriptional suppression of Smad7, a well-known inhibitory Smad, following a rapid induction of the transcription factor TIEG (100). The latter is a known suppressor of Smad7 transcription. Thus, CTGF blocks the negative feedback loop provided by TGF-β upregulation of Smad7. This allows continued and enhanced TGF-β/Smad signalling in the cell. Taken together, current knowledge provides evidence for at least three mechanisms of the synergistic action of CTGF to TGF-β and thus, for its pro-EMT/pro-fibrotic effect: (i) enhancement of TGF-β binding to its cognate receptor, (ii) downregulation of the negative feedback loop via Smad7 and (iii) inhibition of receptor binding and signalling of the physiological TGF-β antagonist BMP-7.

Cell-specific expression profile of connective tissue growth factor in liver diseases

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

Up to now this protein has received only little attention in liver (patho-)physiology and most of its functional roles are deduced from non-hepatic tissue. However, as anticipated from the reaction of CTGF in non-liver fibrotic tissues, it is not surprising that this protein is increasingly produced during human and experimental liver fibrogenesis (93). The transcript levels assessed by Northern blotting, reverse transcriptase-polymerase chain reaction, RNase protection assay and in situ hybridization as well as the tissue protein concentration of CTGF are significantly upregulated in human liver fibrosis/cirrhosis of various aetiologies (101–108). However, cellular distribution and transcripts, respectively, of CTGF in liver fibrosis are still a matter of debate and possibly dependent on the aetiology and time course of the disease. Whereas in situ hybridization in progressive fibrosis owing to biliary atresia reveals increased CTGF mRNA in HSC and PC (105), no appreciable staining was found in PC, HSC and sinusoidal endothelial cells in human liver with idiopathic portal hypertension (104). The most intense staining was noticed in proliferating bile ducts (104). Similarly, in experimental biliary fibrosis, proliferating bile duct epithelial cells were identified as the predominant source of CTGF mRNA (108) but CTGF transcripts were also noticed in desmin-positive HSC cells. Another report emphasizes strong CTGF mRNA expression in fibroblasts/MFBs of fibrous septa, in HSC, endothelial cells and ductular epithelial cells (102). HSC and fibroblasts were suggested in this report as the main source of CTGF, a conclusion that is supported by findings of a combined experimental and human in situ hybridization study (103). Surprisingly, the role of PC as a cell type contributing to CTGF expression in diseased liver remains controversial. Conflicting data were reported showing either no mRNA signal (106), no appreciable transcript staining (104) or marked expression (105). Other in situ hybridization studies did not mention PC at all (102, 103, 108). However, a recent immunocytochemical analysis of CTGF in CCl4-induced fibrotic rat liver clearly demonstrates patch-like positive immunostaining of CTGF in PC, which coincides with that of cytokeratin 18 but not with that of desmin (109) (Fig. 4). The competence of PC for the synthesis of CTGF was recently shown by detailed cell culture studies, which clearly demonstrate CTGF expression in parenchymal liver cells, which is sensitively upregulated by exogenous TGF-β (109, 110) but also occurs spontaneously under TGF-β-free culture conditions owing to intracellular activation of latent TGF-β (111) (Fig. 4). Thus, PC are now recognized as a quantitatively important source of CTGF, that responds to TGF-β. Cultured human and rat HSC are an established cell type for CTGF production (101, 109, 112–115). In contrast to PC, which express only the full-length CTGF, cultured HSC were shown to produce the C-terminal fragment (109) additionally, which increases during culture time (Fig. 5). It is suspected that the proteolytic fragmentation occurs intracellularly but details require further analysis (109). The expression level can be upregulated by TGF-β (60, 109, 115), endothelin-1 (109), glucocorticoids (112), VEGF, acetaldehyde, lipid peroxidation products (hydroxynonenal, malondialdehyde) and PDGF-BB (114) (Fig. 6). Additionally, under diseased liver hypoxic conditions, oxidative stress during inflammation, shear stress and cell stretch during scarring might contribute to CTGF upregulation in PC and HSC. Reduction of CTGF expression in HSC is accomplished by PPAR-γ ligands (60) and the potent antioxidant curcumin through increasing the cellular content of reduced glutathione (116). The HCV-core protein was recently reported to be pathophysiological potentially important stimulator of CTGF expression in HSC (117). Co-culture of HSC and a stable transfected HepG2-HCV core cell line significantly increased not only CTGF but also TGF-β1, suggesting a prominent mechanism in HCV-related fibrogenesis (117). In addition to CTGF, HSC were shown to express NOV (CCN3) increasingly during activation of HSC in culture (118). Taken together, in addition to mesenchymal liver cells [HSC, (myo-)fibroblasts, endothelial cells], epithelial cell types (PC, bile duct) are newly recognized sources of CTGF in diseased liver. This is potentially an important finding with respect to the ability of epithelial cells to change the phenotype during EMT to (myo-)fibroblasts. The expression of CTGF might serve as an important sensitizer, which promotes the TGF-β-driven phenotypic switch (see above). Whether CTGF also exerts mitogenic activity not only for HSC (76) but also for PC and bile duct epithelial cells remains an interesting aspect of future studies (119).

image

Figure 4.  (A) Western blot of CTGF in cultured rat hepatocytes. The intensity of the 38 kD fragment increases within the course of culture. (B) Immunocytochemical demonstration of CTGF in normal and bile-duct-ligated fibrotic rat liver (109). Localization of CTGF in cytokeratin 18-positive hepatocytes and in a few desmin-positive (myo-)fibroblasts is shown. CTGF, connective tissue growth factor.

Download figure to PowerPoint

image

Figure 5.  Western blot of CTGF in cultured rat hepatic stellate cells. Note the presence of the C-terminal fragment (molecular size about 19 kD), which is absent in hepatocytes. The intensity of the 19 kD fragment increases with an increase in the culture time, whereas the intensity of the 38 kD fragment remains largely stable. For details see reference (109). CTGF, connective tissue growth factor.

Download figure to PowerPoint

image

Figure 6.  Schematic presentation of liver cell types that produce CTGF, secrete it into the sinusoidal blood stream and remove it from the circulation via LRP-1. The most important exogenous stimulators (+) and inhibitors (−) are summarized. Several effector molecules are depicted from non-hepatic tissue; their relevance for the liver is not yet established. CTGF follows paracrine, autocrine and potentially endocrine routes. Possible excretion of CTGF into the bile is indicated. PC, hepatocytes; SEC, sinusoidal endothelial cells; HSC, hepatic stellate cells; BEC, biliary epithelial cells; LPOx, lipid oxidation products; LRP, low-density lipoprotein receptor-related protein; ROS, reactive oxygen species; PG, prostaglandin; CTGF, connective tissue growth factor.

Download figure to PowerPoint

Up to now, no data are available on SNPs and promoter polymorphisms of CTGF in fibrosing liver diseases. Such an analysis, however, would be valuable, because the information might have prognostic value for the assessment of the progression rate of fibrosis as done previously for TGF-β1 (120–122). In addition to CTGF production, the liver is likely to play a major role as a clearance organ for circulating CTGF. LRP, the unique multifunctional endocytosis receptor involved in uptake of CTGF (see above), is abundantly expressed by PC (69) (Fig. 6). HSC in addition express LRP and bind full-length CTGF and the C-terminal fragment (115). The isolated module IV is not bound to LRP. Cell surface proteoheparan sulphate of PC and HSC, respectively, are binding sites for CTGF without endocytotic activity (Fig. 2). As mentioned above binding of CTGF to LRP (6) interferes with Wnt/β-catenin signalling (70), which might have profound effects on the cross-talk of intracellular signalling pathways. Recently, anti-adipogenic Wnt signalling during HSC-activation was shown and, hence, the therapeutic potential of Wnt antagonism for liver fibrosis was suggested (123). Following reported data, suppression of Wnt signalling by CTGF (70) would inhibit HSC-activation.

Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

One of the ultimate goals of pathobiochemical research is focused on the development of novel therapeutic approaches to prevent, suppress and reverse hepatic fibrosis (124, 125). It is obvious that strategies against TGF-β and its intracellular signalling pathway are of primary interest (24, 126) applying ligand scavengers (e.g. soluble receptors), antagonist cytokines (HGF, BMP-7) and a variety of large and small molecule synthetic inhibitors of Ser/Thr-kinase of Alk-5 TGF-β receptors (127, 128). However, systemic application of these and other inhibitors of TGF-β is potentially hazardous and as such not suitable for human therapy (129). Based on the above-described data, CTGF is a more interesting target for future anti-fibrotic therapies (45). According to present knowledge, a down-modulation of CTGF activity would shift the TGF-β/BMP-7 balance in the direction of anti-fibrosis, e.g. inhibit ECM synthesis, EMT and HSC-activation, but increase ECM-degradation (fibrolysis). The beneficial effect of CTGF knockdown by gene silencing through siRNA has been shown independently in two toxic models of rat liver fibrosis (130–132). Activation of HSC and accumulation of ECM were considerably suppressed. Type I and type III collagens, laminin, TIMP-1 and TGF-β and the serum level of type III procollagen were markedly attenuated (130). Another study using CTGF-antisense oligonucleotides in CCl4-induced mouse liver fibrosis confirmed significant reduction of type I collagen mRNA but did not find an effect on transcript levels of TIMP-1 (132). While these results prove the great pathogenetic role of CTGF in liver fibrogenesis, the experiments are presently far away from clinical application. Other drugs that inhibit CTGF expression are natural and synthetic PPAR-γ ligands such as prostaglandin J2 (60, 133) and those related to cyclic AMP elevation and protein kinase A activity (53, 134). Among the latter ones prostacyclin derivatives (59, 135), prostaglanding E2 (136) and pentoxifyline (137) are drugs, that have been applied successfully in the treatment of scleroderma patients (59). Approaches such as those for liver fibrosis and possible adverse effects of sustained suppression of CTGF are not yet reported. Nonetheless, the limited set of data and the pivotal fibrogenic role of CTGF point to this molecule as a novel therapeutic target with great potential.

Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

The diagnosis, monitoring of progression and evaluation of effectiveness of therapeutic drugs of liver fibrogenesis depend strictly on serial assessment of grade (activity) and/or stage (extent) of liver fibrosis. As summarized in previous reviews, these requirements cannot be fulfilled with invasive, i.e. potentially hazardous liver biopsy and its histological evaluation (138–142). Owing to the high degree of sampling error depending on the quality of the specimen between 45 and 55% (143), there is an urgent clinical need for non-invasive, systemic, i.e. serum- or plasma-based parameters (140, 141). They should be reliable, cost-effective, relatively simple to measure and standardized to allow comparability between various laboratories. Unfortunately, these criteria are not met by both single analytes and multiparametric algorithms (panel tests) recommended recently (139). Furthermore, such an ideal non-invasive marker, recently categorized as a class I biomarker (140), should have a solid pathophysiological basis, which would facilitate its clinical interpretation.

The strong expression of CTGF in fibrotic tissue, its sensitive dependence on TGF-β, its metabolic behaviour as a secretory protein and its production in the liver not only in HSC and biliary epithelial cells but also in PC representing 65% of total liver cell population recommend CTGF in body fluids as an ideal read-out parameter of TGF-β bioactivity in fibrogenic tissues. Some reports from non-hepatic diseases confirm elevated serum levels in systemic sclerosis (63), and increased N-terminal CTGF levels in plasma and dermal interstitial fluid in scleroderma were found to be correlated with the severity of skin disease (144). Interestingly, C-terminal CTGF and full-length CTGF were reported to be not elevated in this condition (62), presumably because of accelerated clearance in comparison with the N-terminal fragment, which lacks the LRP-binding site (see above). Another study found the serum level of CTGF to be correlated with the extent of skin sclerosis and severity of pulmonary fibrosis (63). N-terminal CTGF plasma levels are elevated in patients with type I diabetic nephropathy; the extent is correlated positively with proteinuria and negatively with creatinine clearance (145). Similar data were obtained with plasma N-terminal CTGF levels and urinary CTGF excretion in diabetic mice (64, 65). The potential of serum CTGF as a biomarker of progressive kidney fibrosis in chronic allograft nephropathy has been emphasized in a clinical and experimental study (146). However, the dependence on the glomerular filtration rate (145) and the contribution of platelet-derived CTGF to the serum levels (66, 67) have to be kept in mind.

Surprisingly, only a few studies are available on systemic CTGF levels in fibrotic liver diseases. An early investigation reports on significantly elevated serum levels of CTGF in patients with biliary atresia, which correlated with the progression of hepatic fibrosis (147). Interestingly, patients with severe end-stage cirrhosis had lower levels than those with progressive cirrhosis. Both groups of patients, however, had significantly enhanced CTGF levels in comparison with healthy volunteers (147). A recent clinical study confirmed strong elevation of serum CTGF levels in patients with fibrosing liver diseases (148). Also in this study active fibrosis (ongoing fibrogenesis) was associated with significantly higher levels than fully developed cirrhosis, indicating CTGF to be a fibrogenic activity biomarker. Receiver-operating-characteristic (ROC) curves reveal for fibrosis and cirrhosis vs. healthy volunteers an area under the curve (AUC) of 0.955 and 0.887 respectively (Fig. 7). In the latter situation, AUC can be increased up to 0.946 if the CTGF level is related to the reduced number of platelets in cirrhotic patients (148). Sensitivities (84–100%) and specificities (85–89%) suggest CTGF as a potentially valuable biomarker of active fibrogenesis rather than the stage of fibrosis, which needs further extensive validation. Because presently a commercial ELISA is not available and standardization is lacking, the absolute concentrations of CTGF in body fluids measured with in-house immunoassays are not comparable. In addition, the analytic specificity of the antibodies used for determination of CTGF fragments of the mosaic protein or of the full-length CTGF should be clearly defined to allow detailed clinical interpretation of the results. Studies are needed to evaluate diagnostic criteria of the N-terminal CTGF-fragment in comparison with the full-length protein or its C-terminal domain. As described above, the N-terminal part of CTGF lacks the LRP-binding module III, which suggests a longer half-life time in the circulation than that of the C-terminal region and of intact CTGF. However, elevated serum CTGF levels in fibrosing liver diseases might be partially because of impaired clearance by the liver caused by liver cell insufficiency and/or disturbed blood flow. Taken together, CTGF is proposed as a pathophysiological well-suited innovative class I biomarker of liver fibrogenesis (140), that is measurable with routine immunological laboratory methods.

image

Figure 7.  Receiver-operating-characteristic (ROC) curves of the diagnostic power of serum CTGF for fibrosis (A) and cirrhosis (B), respectively, vs. healthy control population. AUC, area under the curve; CTGF, connective tissue growth factor. Data compiled from reference (148).

Download figure to PowerPoint

Conclusion and perspective

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References

Since the discovery of CTGF more than 15 years ago, the protein has proved to be a versatile regulator of fundamental functions during development, in health and in disease. It is evident that the originally chosen terminology ‘CTGF’ (37) covers only a small fraction of the functional pluripotency of this effector molecule of TGF-β and other growth factors as well. The mosaic nature of its structure, combined with distinct bioactivities of the modules, the occurrence of separate functional domains, their interaction with signaling receptors (integrins, LRP) and cell surface-binding sites (heparan sulphate), its binding to ECM and the sub-cellular localization, highlights multifaceted activities. Knowledge of the functions and metabolism of CTGF in liver pathophysiology, in particular, during fibrogenesis, is scanty, but emerging data suggest pivotal roles in this process. Indeed, CTGF might prove to be a master switch in tissue repair governing cell regeneration and matrix production. CTGF by itself or in concert with TGF-β and BMP-7 is likely to be a dominator of EMT of biliary epithelial cells and PC, recently recognized as an important (supplementary?) mechanism of fibrosis (10). Therefore, it is not surprising that CTGF was discovered as an important target for experimental anti-fibrotic trials. Analysis of promoter polymorphisms of the CTGF gene in liver fibrotic patients might give prognostic hints on the progression rate of fibrosis. In addition, this protein either as a full-length molecule or a N-terminal fragment is recommended as a diagnostically valuable read-out parameter in blood for TGF-β bioactivity in the liver (and other organs?). Taken together, the complexity of the intrinsic molecular structure, combined with multiple target (cell) interactions, offers a plenitude of translational options for this protein in the management of fibrosing liver diseases.

References

  1. Top of page
  2. Abstract
  3. Principles of liver fibrosis
  4. Overview on structure, expression regulation and protein–protein interactions of connective tissue growth factor in non-hepatic tissues
  5. Cellular biology of connective tissue growth factor
  6. Functional aspects of connective tissue growth factor
  7. Cell-specific expression profile of connective tissue growth factor in liver diseases
  8. Connective tissue growth factor as a target of therapeutic trials in fibrotic liver diseases
  9. Connective tissue growth factor as a non-invasive candidate biomarker of fibrotic liver diseases
  10. Conclusion and perspective
  11. References