SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The identification of molecular mechanisms involved in the maintenance of the transformed phenotype of hepatocellular carcinoma (HCC) cells is essential for the elucidation of therapeutic strategies. Here, we show that human HCC cells display an autocrine loop mediated by connective tissue growth factor (CTGF) that promotes DNA synthesis and cell survival. Expression of CTGF was stimulated by epidermal growth factor receptor (EGFR) ligands and was dependent on the expression of the transcriptional coactivator, Yes-associated protein (YAP). We identified elements in the CTGF gene proximal promoter that bound YAP-enclosing complexes and were responsible for basal and EGFR-stimulated CTGF expression. We also demonstrate that YAP expression can be up-regulated through EGFR activation not only in HCC cells, but also in primary human hepatocytes. CTGF contributed to HCC cell dedifferentiation, expression of inflammation-related genes involved in carcinogenesis, resistance toward doxorubicin, and in vivo HCC cell growth. Importantly, CTGF down-regulated tumor necrosis factor–related apoptosis-inducing ligand (TRAIL) receptor 2 expression and was involved in the reduced sensitivity of these cells toward TRAIL-mediated apoptosis. Conclusion: We have identified autocrine CTGF as a novel determinant of HCC cells' neoplastic behavior. Expression of CTGF can be stimulated through the EGFR-signaling system in HCC cells in a novel cross-talk with the oncoprotein YAP. Moreover, to our knowledge, this is the first study that identifies a signaling mechanism triggering YAP gene expression in healthy and transformed liver parenchymal cells. (HEPATOLOGY 2011)

In spite of significant technical improvements in surgical and percutaneous interventions, the prognosis of patients with hepatocellular carcinoma (HCC) remains very poor. HCCs are molecularly heterogeneous tumors; nevertheless, active research has exposed common genomic alterations that allowed the subclassification of these neoplasms and the identification of driving oncogenic mechanisms.1-3 Knowledge of the oncogenic processes and the signaling pathways regulating HCC cell proliferation, survival, invasion, and metastasis has led to the development of targeted therapies with positive results, as exemplified by the multitarget inhibitor, sorafenib.2 However, notwithstanding the survival benefits of sorafenib, there is still room for improvement, and current research is aiming to combine molecular therapies. To this end, increasing knowledge on the signaling pathways critical for HCC progression is of crucial importance.

Connective tissue growth factor (CTGF) is a cysteine-rich secreted protein that interacts with a variety of extracellular matrix components and cell-surface proteins, such as fibronectin, proteoglycans, and integrins, strongly influencing cellular behavior.4 CTGF participates in many biological processes, including cell proliferation, survival, migration, angiogenesis, wound healing, and cancer development.4-6 In the liver, CTGF has been recognized as a key profibrogenic factor, its expression is increased in fibrotic human and rat liver, and the manipulation of CTGF levels in experimental fibrosis modulates the course of the disease.6 More recently, CTGF expression was reported to be elevated in HCC tissues, and HCC patients with high serum CTGF levels show reduced survival, attesting to the potential relevance of CTGF in HCC progression.7, 8 Transforming growth factor beta (TGF-β) is regarded as the major inducer of CTGF expression, and evidence suggests that CTGF, indeed, mediates many of the pathological effects of TGF-β in liver disease, including fibrosis development and HCC progression.4, 9 In view of all this, the understanding of CTGF regulation and biological effects in liver cancer cells would be important for the characterization of this factor as a molecular target in HCC.

Previously, we demonstrated that CTGF expression in experimental liver fibrosis was affected by the epidermal growth factor receptor (EGFR) ligand, amphiregulin (AR), and that AR directly promoted CTGF expression in human and mouse fibrogenic cells.10 Signaling through the EGFR is regarded as an important mechanism in hepatocarcinogenesis and as a target for molecular therapies.2, 3, 11 Here, we demonstrate that CTGF expression in HCC cells participates in cell proliferation, survival, and inflammatory gene expression and is regulated by EGFR activation in a novel cross-talk with the oncogenic transcriptional coactivator, Yes-associated protein (YAP).12, 13

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Materials and methods are described in Supporting Materials and Methods.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

CTGF Expression in Human Liver Tissues.

Consistent with previous reports,7 we observed increased CTGF messenger RNA (mRNA) levels in HCC tissues, compared with samples from individuals with healthy or only small changes in liver function tests (Supporting Fig. 1A). Interestingly, CTGF expression was also elevated in samples of peritumoral noncirrhotic and cirrhotic liver tissues (Supporting Fig. 1A,B).

CTGF Expression Is Induced by EGFR Activation.

Treatment with EGFR ligands activated EGFR phosphorylation and elicited a time- and dose-dependent increase in CTGF mRNA in Hep3B cells, which was abolished by the EGFR inhibitor, PD153035 (Fig. 1A). Similar observations were made in Huh7 cells (not shown). CTGF expression by EGFR triggering likely involved transcriptional activation, because it was prevented by actinomycin-D (Act-D) (Fig. 1A). CTGF protein was increased also in cells and conditioned media (Fig. 1B). CTGF gene transcription upon EGFR triggering was further demonstrated in HCC cells transfected with a CTGF promoter-luciferase reporter construct (Fig. 1C). Previously, we described the existence of an autocrine loop in HCC involving AR release and EGFR stimulation.14, 15 Currently, we observe that AR knockdown significantly reduces CTGF expression, suggesting an important role for AR in the basal expression of CTGF (Fig. 1D). Next, we examined whether EGFR activation could lead to CTGF expression in nontransformed human liver parenchymal cells. According to previous findings in rat hepatocytes,16 we observed that TGF-β stimulated CTGF expression in human hepatocytes, and that activation of EGFR by AR and heparin-binding epidermal growth factor (HB-EGF) treatment shared this effect (Fig. 2A,B).

thumbnail image

Figure 1. EGFR ligands stimulate CTGF gene expression in HCC cells. (A) Time- and dose-dependent induction of CTGF mRNA in Hep3B cells by AR and HB-EGF in the absence or presence of Act-D (5 μg/mL) or the EGFR inhibitor, PD153035 (10 μmol/L). Inset shows phospho-EGFR (Tyr1068) levels upon AR treatment (50 nmol/L, 15 minutes). (B) CTGF protein in Hep3B cells (upper panel) and conditioned media (lower panel) upon AR (50 nmol/L) or HB-EGF (50 nmol/L) treatment (24 hours). (C) Luciferase reporter assay in Hep3B cells transfected with a CTGF gene promoter (−485 to +15)-luciferase construct (pCTGF) or control vector (pGL3-enhancer) treated with AR (50 nmol/L) or HB-EGF (50 nmol/L) (10 hours). (D) Effect of AR knockdown on basal CTGF expression in Hep3B and PLC/PRF/5 cells. mRNA levels were measured 48 hours after transfections. AU, arbitrary units; siGL, control siRNA; siAR, AR-specific siRNA.

Download figure to PowerPoint

thumbnail image

Figure 2. TGF-β and EGFR ligands induce CTGF expression in primary human hepatocytes. Cells were treated with TGF-β (5 ng/mL) (A) or AR (50 nmol/L) and HB-EGF (50 nmol/L) (B), and CTGF mRNA (upper panels) and protein levels in conditioned media were determined after 3 and 24 hours of treatment, respectively. Inset shows phospho-EGFR (Tyr1068) levels upon AR or HB-EGF treatment (15 minutes).

Download figure to PowerPoint

Regulation of CTGF Gene Expression in HCC Cells: Implication of the Transcriptional Coactivator, YAP.

Different transcription factors and coactivators participate in the complex regulation of the CTGF promoter, among them antimothers against decapentaplegic homolog 2/3, activator protein 1 (AP-1), and YAP/TEAD (TEA domain), play important roles in basal and/or growth-factor–triggered CTGF expression.4, 16-19 YAP was recently identified as an oncogene overexpressed in liver cancer,12, 13, 20 and three putative YAP/TEAD-binding sites (TB1, TB2, and TB3) exist in the human CTGF promoter (Fig. 3A).18, 19 This led us to test first whether these motifs were involved in basal CTGF expression in HCC cells. Though mutation of the most 5′ TB element (TB1) did not change reporter gene activity, mutation of the two adjacent TB sites, TB2 and TB3, caused a significant reduction in basal CTGF promoter activity (Fig. 3B). Conversely, mutation of the AP-1 site, present in our CTGF promoter construct, did not affect basal activity in Hep3B cells (Fig. 3B). Similar findings were obtained in HepG2 cells (Fig. 3B); however, in this cell line, mutation of TB2, or the combined mutation of TB2 and TB3, but not mutation of TB3 alone, was found to be important for basal CTGF expression. This difference among cell lines may be attributed, in part, to the significantly higher basal levels of YAP expression found in HepG2 cells (see below), which could compensate, in part, for the reduced binding of YAP/TEAD complexes to a mutant TB3 site. Moreover, in addition to YAP, the expression of an active mutant β-catenin in HepG2 cells may also be an important determinant for CTGF expression in this cell line, because CTGF is a Wnt/β-catenin target gene.5 Next, we examined whether these elements were required for EGFR-mediated CTGF gene expression. Though AR-mediated transactivation of the CTGF promoter was significantly attenuated when the TB2 and TB3 sites were mutated, it was not affected by AP-1 site mutation (Fig. 3C). In agreement with this, we found that AR treatment increased the recruitment of YAP to the CTGF promoter region encompassing the YAP/TEAD-binding sites (Fig. 4A). Furthermore, YAP knockdown in Hep3B cells significantly reduced basal and AR or HB-EGF-stimulated CTGF gene expression (Fig. 4B). In line with these observations, we found a close correlation between YAP and CTGF mRNA levels in five human HCC cell lines and primary hepatocytes (Fig. 4C,D). In accord with previous reports,20 we observed that YAP expression was increased in HCC tissues and, also, in peritumoral noncirrhotic and cirrhotic liver tissues (Supporting Fig. 1A). Moreover, YAP mRNA levels significantly correlated with those of CTGF in our collection of healthy and diseased liver samples (r = 0.53, P = 0.0019). Accordingly, the expression of both proteins was detected in cirrhotic and HCC tissues (Supporting Fig. 1B).

thumbnail image

Figure 3. TEAD-YAP-binding sites in the CTGF promoter are involved in basal and AR-stimulated CTGF promoter activity in HCC cells. (A) Position of AP1 and TEAD-YAP-binding sites (TB1, TB2, and TB3) in the human CTGF proximal promoter. (B) Luciferase activity in Hep3B and HepG2 cells transfected with control vector (pGL3-enhancer), wild-type CTGF promoter-luciferase reporter (pCTGF), or the indicated mutants. (C) Transactivation by AR (50 nmol/L, 10 hours) of wild-type pCTGF or the indicated mutant constructs transfected in Hep3B cells.

Download figure to PowerPoint

thumbnail image

Figure 4. YAP expression influences basal and EGFR-activated CTGF expression. (A) Chromatin immunoprecipitation assay measuring the binding of YAP to the −286 to −52 CTGF promoter region encompassing the three TEAD-YAP-binding sites, in Hep3B cells treated with AR (50 nmol/L, 12 hours). Immunoprecipitated DNA fragments were quantified by real-time polymerase chain reaction. A representative ethidium-bromide–stained gel of amplified DNA fragments is shown. (B) Basal and AR (50 nmol/L, 3 hours) or HB-EGF (50 nmol/L, 3 hours)-induced CTGF expression in Hep3B cells transfected with control or YAP-specific small interfering RNAs (siRNAs). Inset shows YAP protein levels in Hep3B cells 48 hours after transfections. (C and D) Basal levels of YAP and CTGF mRNAs in primary human hepatocytes and HCC cell lines determined by real-time polymerase chain reaction.

Download figure to PowerPoint

YAP gene expression is increased in over 60% of HCCs and is frequently detected in nontransformed tissues surrounding tumor nodules.20, 21 However, the mechanisms underlying the regulation of YAP gene transcription are not known. Given the important role played by YAP in basal and EGFR-triggered CTGF expression, we examined whether EGFR activation could influence YAP expression. We found that AR or HB-EGF treatment increased YAP mRNA and protein levels (Fig. 5A,B), an effect abolished in the presence of Act-D (Fig. 5A). Stimulation of YAP expression by AR was mediated through the EGFR receptor and the downstream activation of extracellular regulated kinase kinase-1 (MEK1), because it was prevented by the PD153035 and UO126 inhibitors, but not by PI3K inactivation (Fig. 5C). The effect of AR and these inhibitors on EGFR downstream signaling is shown in Fig. 5C. Interestingly, EGFR and MEK1 inhibitors also reduced basal YAP mRNA levels (Fig. 5C). In primary human hepatocytes, EGFR activation also elevated YAP mRNA and protein levels (Fig. 6A) and consistently up-regulated integrin β2 (ITGB2) expression, a direct transcriptional target of YAP18 (Fig. 6B). YAP overexpression in HCC tissues is accompanied by increased accumulation of the protein in transformed cell nuclei.20, 21 Immunofluorescence staining of hepatocytes treated with AR showed a clear nuclear accumulation of YAP protein (Fig. 6C). Reduced YAP-Ser127 phosphorylation has been associated with its nuclear translocation.21 However, we did not appreciate changes in pYAP-Ser127 upon AR treatment (not shown), suggesting that the observed YAP nuclear localization could be related to its overexpression, as found in YAP-transgenic mice.22 These observations were reproduced in the nontransformed breast epithelial cells, MCF-10A (Supporting Fig. 2). It has been recently shown that microRNA (miRNA)-375, which is down-regulated in HCC, is able to reduce YAP expression.23 In view of this, we explored the effect of AR on miR-375 levels in HCC cells and cultured human hepatocytes. We found that AR treatment did not change miR-375 expression (Supporting Fig. 3).

thumbnail image

Figure 5. EGFR activation stimulates YAP gene expression in HCC cells. (A) YAP mRNA levels in Hep3B cells treated with AR (50 nmol/L) or HB-EGF (50 nmol/L) for 3 hours with or without Act-D (5 μg/mL). (B) YAP protein levels in Hep3B cells treated with AR or HB-EGF, as indicated above. (C) YAP mRNA levels in Hep3B cells treated with AR (50 nmol/L, 3 hours) in the presence or absence of the EGFR inhibitor, PD153035 (10 μmol/L), the MEK1 inhibitor, UO126 (10 μmol/L), or the PI3K inhibitor, LY294002 (10 μmol/L). Inset shows p-Erk1/2 (Tyr204) and p-Akt (Ser473) levels under the indicated conditions (AR treatment, 50 nmol/L; 15 minutes).

Download figure to PowerPoint

thumbnail image

Figure 6. EGFR activation stimulates YAP expression in primary human hepatocytes. (A) YAP mRNA levels in cells treated with AR (50 nmol/L) or HB-EGF (50 nmol/L) for 3 hours. Lower panel shows YAP protein levels in hepatocytes treated with AR (50 nmol/L) analyzed by western blotting. (B) Effect of AR and HB-EGF treatment (50 nmol/L, 6 hours) on ITGB2 mRNA levels. (C) Immunofluorescent detection of YAP protein in control and AR (50 nmol/L, 12 hours)-treated hepatocytes.

Download figure to PowerPoint

Autocrine CTGF Contributes to HCC Cells Neoplastic Phenotype.

CTGF production by HCC cells enhances tumor growth by promoting cross-talk between HCC and stromal cells.9 In the present study, we evaluated whether CTGF could also have an autocrine effect on HCC cells. We observed that CTGF knockdown significantly reduced DNA synthesis under serum-free conditions (Fig. 7A,B), decreased anchorage-independent cell growth, and significantly reduced the tumorigenicity of PLC/PRF/5 cells in vivo (Supporting Fig. 4A,B). Moreover, the stimulatory effect of AR on DNA synthesis was also influenced by the concomitant expression of CTGF (Fig. 7C). In line with these effects, we observed that treatment with recombinant CTGF activated extracellular signal-regulated kinases 1/2 (Erk1/2) signaling and stimulated DNA synthesis (Fig. 7D).

thumbnail image

Figure 7. CTGF stimulates DNA synthesis in HCC cells. (A) CTGF protein levels 72 hours after transfection with control (siGL) or CTGF-specific (siCTGF) siRNAs determined by western blotting in conditioned media. (B) Hep3B and PLC/PRF/5 cells were transfected with siGL or siCTGF siRNAs; after 48 hours, [3H]thymidine was added to cultures, and radioactivity incorporation into DNA was measured 24 hours later. (C) DNA synthesis in Hep3B cells transfected with control or CTGF-specific siRNAs in response to AR (50 nmol/L, 24 hours). Values are relative to [3H]thymidine incorporation in control siRNA-transfected cells without AR treatment. (D) Upper panel: analysis of Erk1/2 phosphorylation in Hep3B cells upon CTGF (250 ng/mL) treatment. Lower panel: DNA synthesis in Hep3B cells treated with CTGF or 10% fetal bovine serum for 24 hours. [3H]thymidine was added for the last 10 hours of treatments.

Download figure to PowerPoint

To further explore the relevance of CTGF on HCC cell biology, we performed a microarray gene-expression analysis in Hep3B cells upon CTGF knockdown. The expression of 189 genes was up-regulated, whereas 419 genes were inhibited upon CTGF knockdown to 40% of basal levels. Analysis with the Ingenuity Pathway Analysis Network identified genes mostly associated with lipid and bile acid metabolism, amino acid and small-molecule biochemistry, including membrane transporters, as well as cell cycle, DNA replication, and cell-to-cell signaling and interaction (Supporting Table 1). The differential expression of genes selected by their potential physiopathologic significance was validated in independent transfections. Up-regulated genes included genes normally expressed in the healthy differentiated human liver, such as bile acid coenzyme A, amino acid N-acyltransferase, UDP-glucuronosyltransferase-2B15, and tryptophan dioxygenase-2 (Supporting Fig. 5). Other up-regulated genes perhaps more directly implicated in hepatocarcinogenesis included the latent TGF-β binding protein-1, which influences the bioavailability of TGF-β and is down-regulated in HCC,24, 25 the tumor necrosis factor–related apoptosis-inducing ligand (TRAIL) receptor 2 (TRAIL-R2), which is reduced in HCC and may determine resistance to TRAIL-induced apoptosis,26, 27 the transcriptional coactivator, K(lysine) acetyltransferase 2B/P300/CBP-associated factor, also down-regulated in HCC cells,28 the cell-cycle regulator, SET domain containing lysine methyltransferase 8 histone-methyltransferase,29 and p21-activated protein kinase-2 involved in apoptotic signaling.30 Down-regulated transcripts comprised genes involved in inflammation and acute-phase response, including the chemokines, lipopolysaccharide-induced CXC chemokine (CXCL)5 and CXCL12, orosomucoid 1 and 2, and serpina3, supporting the emerging role of CTGF as an inflammation modulator.31 Interestingly, membrane transporters known to confer growth and survival advantages to tumor cells were also down-regulated upon CTGF knockdown, including the amino acid transporters, SLC6A14 and SLC6A6,32, 33 and the ABC-type transporters, ABCC1 and ABCC2, involved in drug resistance.34, 35 Other repressed genes implicated in cancer progression were lysyl-oxidase,36 carbonic anhydrase XII,37 connexin 43/GJA1,38 vimentin,39 and fibroblast growth factor receptor 240 (Supporting Fig. 5).

In view of this, we examined whether CTGF expression could modulate HCC survival and sensitivity to doxorubicin. CTGF knockdown resulted in increased basal apoptosis and sensitized Hep3B cells to the drug (Fig. 8A), whereas recombinant CTGF protected from doxorubicin-induced apoptosis (Fig. 8B). In agreement with our microarray data showing enhanced TRAIL-R2 mRNA levels upon CTGF knockdown, we observed increased TRAIL-R2 protein levels in Hep3B cells upon CTGF down-regulation, whereas TRAIL-R2 expression was reduced by recombinant CTGF treatment (Fig. 8C). Consistent with these changes, cells underwent apoptosis when challenged with TRAIL after CTGF knockdown (Fig. 8D).

thumbnail image

Figure 8. CTGF expression contributes to HCC cell resistance to doxorubicin and TRAIL-induced apoptosis. (A) Hep3B cells were transfected with siGL or siCTGF siRNAs; 48 hours later, cells were treated with doxorubicin (Dox; 24 hours), and apoptosis was measured. (B) Hep3B cells were pretreated with CTGF in serum-free medium for 2 hours, then doxorubicin was added to cultures. Apoptosis was measured 24 hours later. (C) Western blotting analysis of TRAIL-R2 expression in Hep3B cells transfected with siGL or siCTGF siRNAs for 72 hours (upper panel) or in Hep3B cells treated with CTGF (250 ng/mL, 24 hours) (lower panel). (D) Hep3B cells were transfected with control or siCTGF siRNAs; 48 hours later, cells were treated with TRAIL (24 hours), and apoptosis was measured.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This work provides new insights into the regulation and biological function of CTGF in human HCC cells. We confirm previous observations on increased CTGF expression in human HCC and chronically injured liver.7, 9 Interestingly, CTGF expression was also high in normal liver tissues from patients that had developed HCC and were free of any known risk factor for liver cancer. It is tempting to speculate that CTGF overexpression might contribute to tumor development in this small, but sizeable, minority of HCCs.41 We demonstrate that EGFR signaling regulates CTGF expression in HCC cells, and that autocrine AR significantly contributes to elevated CTGF levels. Different EGFR ligands, including AR and HB-EGF, are up-regulated in HCC.11 The reasons for the apparently nonredundant role of autocrine AR in promoting the expression of CTGF in HCC cells are not currently known, but our findings are in agreement with the previously shown predominant role for AR in HCC cell biology.14, 15 One possible explanation could be the distinct interaction of AR with the EGFR, leading to unique signaling patterns and biological responses.42 EGFR ligands also triggered CTGF expression in primary hepatocytes, suggesting that the activation of EGFR during liver injury and inflammation11 can contribute to the expression of CTGF not only in fibrogenic cells, but also in the parenchyma.10 We found that basal and EGFR-regulated CTGF gene expression depended, in part, on TEAD-YAP-binding elements present in the CTGF promoter and on the expression of the YAP transcription coactivator. The correlation between CTGF and YAP expression across our collection of healthy and diseased liver tissues further supports this notion. Though these findings are novel for HCC cells, and for EGFR-mediated gene regulation, similar control of CTGF expression by TEAD-YAP has been previously reported for other cell types.18, 19 However, perhaps one of our most compelling findings was the observation that YAP gene expression was induced by EGFR activation not only in tumor cells, but also in hepatocytes and other nontransformed epithelial cells (e.g., MCF-10A). Indeed, EGFR stimulation promoted YAP mRNA up-regulation and the accumulation of YAP protein in hepatocytes' nuclei. The finding that YAP gene expression can be triggered by EGFR/MEK1 signaling contributes to understanding the complex regulation of YAP and highlights the importance of growth-factor–activated pathways in the regulation of this gene. This observation adds to recent findings showing the modulatory effects of MEK1/Erk and phosphatidylinositol 3-kinase (PI3K) pathways on Mst1/2 kinase activity, the upstream regulator of YAP protein.21, 43

YAP is currently considered as an oncogene up-regulated in liver cancer that is able to promote cell proliferation, survival, and anchorage-independent growth.12, 13, 21 In the healthy liver, YAP mRNA levels are low, but are significantly increased in HCCs.20, 21 This is not entirely the result of YAP genomic amplification, because focal amplification on chromosome 11q22, encompassing the YAP gene, is found in <10% of HCCs.12 Together with the recently reported negative effects of miRNA-375 on YAP levels,23 our study contributes toward explaining the elevation of YAP gene expression in liver cancer cells. Moreover, our observations in primary human hepatocytes suggest that YAP up-regulation by EGFR signaling might occur already at preneoplastic stages, when expression of EGFR ligands is elevated and there is enhanced hepatocellular proliferation.3, 10, 11 On the other hand, AR has been recently characterized as a transcriptional target for YAP in MCF10A cells.44 However, AR expression was not affected by YAP knockdown in HCC cells (not shown), indicating that YAP is not a major determinant for the constitutive expression of AR in transformed liver cells.

Our data also support the existence of a CTGF-mediated autocrine loop contributing to HCC cells' malignant phenotype, including basal HCC cell proliferation and survival. CTGF knockdown reduced the aggressiveness of HCC cells, as shown by impaired growth in soft agar and reduced in vivo tumorigenesis. Interestingly, CTGF expression was also necessary for AR-induced DNA synthesis in HCC cells. Therefore, CTGF could be acting as a modulator of AR effects, enhancing the responses to the EGFR ligand, which, in turn, would increase CTGF expression in a feedforward loop. A similar type of interaction has been firmly established between TGF-β and CTGF in the activation of liver stellate cells.45

In line with the attenuation of neoplastic traits upon CTGF knockdown, we also observed a significant modification of HCC cell gene-expression profile. CTGF down-regulation partially reversed HCC cell dedifferentiation and also modified the expression of genes putatively involved in hepatocarcinogenesis. Among them were latent TGF-β binding protein-1, lysyl-oxidase, connexin43/GJA1, and vimentin,24, 25, 36, 38, 39 and the chemokine, CXCL12, recently demonstrated to promote HCC autocrine growth and survival.46 We also obtained evidence on the functional implications of CTGF-modulated genes in HCC cell behavior. These included an increased sensitivity to doxorubicin-induced apoptosis upon CTGF knockdown, which can be related to the CTGF-promoted expression of the drug efflux pumps, ABCC1 and ABCC2.34, 35 Moreover, of special significance was the finding that CTGF regulated the expression of TRAIL-R2 in HCC cells. TRAIL is a major mediator of acquired immune tumor surveillance that is able to induce apoptosis in tumor cells and is a promising candidate for cancer therapy.27 However, TRAIL resistance is a common feature among HCC cells and, therefore, is a major limitation in the therapeutic application of TRAIL receptor agonists.27 Importantly, loss of TRAIL receptor expression is a mechanism for acquired resistance to TRAIL.26, 27 In this context, our current observations showing that CTGF knockdown increases TRAIL-R2 expression and sensitizes to TRAIL-mediated apoptosis may be of special relevance, identifying CTGF as a target for the successful application of TRAIL in the treatment in HCC.

In summary, here, we demonstrate that a CTGF-mediated autocrine loop exists in HCC cells contributing to the malignant phenotype. EGFR signaling promotes CTGF expression in HCC cells through a novel cross-talk with the YAP oncogene.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors thank the Bioinformatics Unit of the CIMA-Universidad de Navarra for microarray data analysis. The help of Dr. Laura Guembe, from the Morphology Service CIMA, is highly appreciated.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_24587_sm_SuppFig1.tif9040KSupporting Information Figure 1. Expression of CTGF and YAP in healthy and diseased human liver. (A) CTGF and YAP mRNA levels were measured in healthy livers (HL, n-9), healthy liver tissue from organs harbouring a hepatocellular carcinoma (HL+HCC, n-12), cirrhotic liver tissue from organs harbouring a hepatocellular carcinoma (Cirr+HCC, n-17) and hepatocellular carcinoma (HCC, n-20). “+” indicates the mean value of CTGF or YAP expression for each set of samples. For CTGF expression statistical significance was: HL vs HL+HCC p-0.002; HL vs Cirr+HCC p- 0.0002; HL vs HCC p-0.026. For YAP expression statistical significance was: HL vs HL+HCC p-0.0003; HL vs Cirr+HCC p-0.0001; HL vs HCC p-0.0001. (B) Immunohistochemical detection of CTGF and YAP proteins in sections from HL, Cirr+HCC and HCC liver tissues. Representative photomicrographs are shown (magnification × 200).
HEP_24587_sm_SuppFig2.tif4972KSupporting Information Figure 2. EGFR activation stimulates CTGF and YAP expression in MCF-10A cells. (A) CTGF mRNA levels in cells treated with AR (50 nmol/L) or HB-EGF (50 nmol/L) for 3h. (B) YAP mRNA levels in cells treated with AR (50 nmol/L) or HB-EGF (50 nmol/L) for 3h. (C) YAP protein levels in cells treated with AR (50 nmol/L) or HB-EGF (50 nmol/L) for 12h analyzed by Western blotting. (D) Immunofluorescent detection of YAP protein in control and AR (50 nmol/L, 12h) treated MCF-10A cells.
HEP_24587_sm_SuppFig3.tif85KSupporting Information Figure 3. Effect of AR on miRNA-375 expression. Levels of miRNA-375 in control and AR (50 nmol/L, 3h) stimulated Hep3B cells (A) and primary human hepatocytes (B).
HEP_24587_sm_SuppFig4.tif788KSupporting Information Figure 4. Effect of CTGF knockdown on anchorage-independent cell growth in soft agar and on in vivo growth in nude mice. (A) PLC/PRF/5 cells were transfected with control (siGL) or CTGF specific (siCTGF) siRNAs. After 48h cells were harvested, counted, resuspended in 0.2% soft agar, and seeded onto 0.4% soft agar in DMEM supplemented with 10% fetal calf serum (104 cells per plate). After 3 weeks, colonies were stained with crystal violet and counted. (B) PLC/PRF/5 cells were transfected with control (siGL) or CTGF specific (siCTGF) siRNAs. After 48h cells were harvested and subcutaneously injected into nude mice (10 × 106 cells per mouse). Tumor volumes (mm3) were monitored over time. Data are means ± SEM (n-10 mice per group). Asterisks indicate p<0.05 vs. siCTGF.
HEP_24587_sm_SuppFig5.tif185KSupporting Information Figure 5. Validation of selected genes identified by microarray analysis as differentially expressed in Hep3B cells upon CTGF knockdown. Quantitative real-time PCR analysis of gene expression in total RNA samples from Hep3B cells 48h after transfection with siCTGF or control siGL siRNAs. Values are expressed in relative transcript levels as compared with control siGL siRNA. (A) Genes downregulated upon CTGF knockdown included G protein-coupled receptor 160 (GPR160), connexin 43/GJA1 (GJA), lysyl oxidase (LOX), orosomucoid 1 and 2 (ORM1, ORM2) and fibroblast growth factor receptor 2 (FGFR2). (B) Genes upregulated upon CTGF knockdown included bile acid CoA: amino acid N-acyltransferase (BAAT), the PR-Set7 histone methyltransferase (SETD8), latent TGFβ binding protein-1 (LTBP1), p21-activated protein kinase-2 (PAK2), lysine acetyltransferase 2B (KAT2B/PCAF), UDP-glucuronosyltransferase-2B15 (UGT2B15), tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) receptor 2 (TNFRSF10B or TRAIL-R2) and tryptophan dioxygenase (TDO). Differences were statistically significant (p values at least <0.05) with respect control transfections for all genes shown.
HEP_24587_sm_SuppInfo.doc47KSupporting Information
HEP_24587_sm_SuppTabS1.doc57KSupporting Information Table 1. Selection of genes with differential expression in Hep3B cells upon CTGF knockdown identified by microarray analysis.
HEP_24587_sm_SuppTabS2.doc71KSupporting Information Table 2. Primers used in this study to measure gene expression by quantitative RT-PCR.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.