Introduction

Hepatocellular carcinoma (HCC) is one of the major causes of cancer-related death throughout the world. Several studies have shown that the stroma of HCC is markedly infiltrated with myofibroblasts, which are identified through their positive staining with an antibody directed against α-smooth muscle actin (α-SMA) (Schmitt-Gräff et al. 1991; Enzan et al. 1994; Terada et al. 1996; Aishima et al. 2008). Myofibroblasts are a key player in the control of tumor cell behavior and are very rare in normal liver tissues (Desmoulière et al. 2004). They are, however, abundant in diseases such as liver fibrosis and cirrhosis (Svegliati-Baroni et al. 2008). In these cases, although some data show that they could be derived in part from portal fibroblasts (Tuchweber et al. 1996), most studies suggest that they originate through the phenotypic modulation of hepatic stellate cells (HSCs) (Svegliati-Baroni et al. 2008; Musso et al. 1997). HSCs constitute approximately 8–14% of cells in the normal liver and are the primary site for retinoid storage in the body (Bataller and Brenner 2005). Liver fibrosis is characterized by progressive accumulation of extracellular matrix (ECM) proteins, synthesized by activated HSCs. When HSCs are isolated and grown in dishes in DMEM/10% fetal bovine serum (FBS), they spontaneously undergo similar activation (Rockey et al. 1992). Moreover, rat and human tumoral hepatocyte-conditioned medium (CM) were able to directly activate cultured HSCs, which could lead to specific features that would distinguish it from fibrosis-type activation (Nhieu et al. 1998; Faouzi et al. 1999a, b). To discover genes involved in activation of HSCs, several studies used DNA microarrays to compare gene expression between normal and cirrhotic livers (Honda et al. 2005; Lau et al. 2005; De Minicis et al. 2007). The differentially expressed genes identified include ECM proteins, genes involved in liver regeneration, proinflammatory genes and genes encoding serum proteins. However, it is not known whether culture activation of HSCs can reproduce the underlying changes in gene expression of HSC activation in HCC, and whether it represents a suitable model to study in vivo HSC activation. Moreover, it is not known whether HSC activation in HCC results in a stimulus-specific activation of HSCs with certain gene expression patterns. To answer these questions, we prepared rat C5F cell line CM and induced HSC activation in vitro. Subsequently, we performed DNA microarray studies and compared gene expression profiles between culture-activated HSCs (cHSCs) and induction-activated HSCs (iHSCs), aiming to provide theoretical foundations for investigating the biology of HSCs in HCC.

Materials and methods

Animal and cell line

F344 rats (200–250 g) were obtained from Charles River Laboratory (USA). The rat C5F cell line was a gift from Dr. Kumiko Ogawa (First Department of Pathology, Nagoya City University, Japan). All animals received humane care according to the criteria outlined by Zhongshan Hospital, Fudan University, China.

Preparation of CM

Conditioned medium was prepared as described previously (Faouzi et al. 1999a, b), with some modifications. Briefly, C5F cells were cultured in Dulbecco’s modified Eagle medium (DMEM) containing 10% FBS. When subconfluent, cells were washed and incubated with serum-free DMEM for 2 h. This medium was then discarded and the cells were incubated with serum-free DMEM again. After 24 h, this CM was harvested and centrifuged at 400g for 8 min to remove debris, filtered through a 0.22 μm filter, and stored at −20°C until use.

Isolation and culture of rat HSCs

Rat HSCs were isolated from rat livers by perfusion of collagenase and pronase, followed by centrifugation over a Nycodenz gradient, as described previously (Siegmund et al. 2006). After isolation, the cells were cultured in DMEM supplemented with 10% FBS for 2 days, when they were harvested as quiescent HSCs (qHSCs) (Jiang et al. 2006). To prepare cHSCs, HSCs were cultured in uncoated plastic dishes in DMEM supplemented with 10% FBS for a total of 7 days (Jiang et al. 2006). To prepare in vitro iHSCs, HSCs were cultured in DMEM supplemented with 10% FBS for 24 h. Cells were then washed three times with serum-free DMEM and incubated with rat HCC-CM (CM diluted 1/2 with serum-free DMEM) for 6 additional days (Faouzi et al. 1999a, b). Cell purity was determined by desmin immunofluorescence staining and the typical light microscopic appearance of the lipid droplets as described previously (Liu et al. 2003). HSC activation was assessed by α-SMA immunofluorescence staining. Immunofluorescence staining for desmin and α-SMA were performed using anti-desmin and anti-α-SMA antibody (Boster, Wuhan, China). Cells were counterstained with fluorescein isothiocyanate-conjugated goat anti-mouse IgG (Molecular Probes, Inc., Eugene, OR). DAPI (Sigma, Germany) stain was used to stain the nuclei. Negative controls were performed without primary antibody.

DNA microarray analysis

Total RNA extracted from quiescent, culture-activated and induction-activated rat HSCs was used for DNA microarray analysis. Two independent isolations and microarray analysis (experiment 1 and experiment 2) were performed using 27K Rat Genome Array chips (SmartArray™, CapitalBio Corp., Beijing, China), according to the manufacturer’s instructions. Data analysis was performed using Tomodel and CB-MAS V4.0.

Real time RT-PCR analysis of gene expression

To quantify selected gene expression, HSCs were lysed with TRIzol reagent (Invitrogen Life Technologies) at a concentration of 0.1 ml/1 × 106 cells, and total RNA was extracted. mRNA was reverse transcribed using oligo(dT) 18–25 primer and Omniscript Reverse Transcriptase (TaKaRa, Danian, China). Selected gene transcripts were quantified by real-time RT-PCR using gene-specific primers and TaqMan Master Mix (TaKaRa). Expression of β-actin was used as an internal control. PCR amplification conditions were as follows: 95°C for 10 s, then 40 cycles of 95°C for 5 s and 60°C for 30 s. Samples were run in triplicate using a real-time PCR thermocycler (ABI PRISM 7000 Sequence Detection System; Applied Biosystems), and the results were analyzed by matched software. Relative expression of genes was determined by normalizing to β-actin expression in each set of samples according to the manufacturer’s instructions. The gene-specific primers and sizes of the expected PCR products are listed in Table 1.

Table 1 RT-PCR primers
Full size table

Western blot analysis

Total proteins were prepared by standard procedures. The total protein concentration was estimated by the Bradford assay, with BSA as the standard. The proteins were separated by SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane. Membranes were blocked for 1 h at room temperature in 5% nonfat milk in 0.1% Tween, washed, and incubated overnight at 4°C with one of the following anti-rat antibodies: VCAM-1, MMP-9, Ccr1 and TGF-β2 (all from BD Pharmingen™, Biosciences, New Jersey, USA). Immunodetection was performed using the ECL blotting detection system (Pierce, Rockford, IL).

Statistical analysis

All groups were compared with qHSCs. Data are given as mean ± SE. Statistical significance was determined using the Student’s t test. A value of P < 0.05 was considered significant. SPSS15.0 software for Windows was used.

Results

Identification of HSC

A pure population of viable HSCs with the quiescent phenotype was obtained using our protocol 2 days after isolation. More than 95% of the qHSCs showed positive desmin staining, indicating pure HSCs (Fig. 1a). To obtain cHSCs, the cells were cultured for 7 days and stained for α-SMA (Fig. 1c). To obtain iHSCs, the cells were cultured using C5F-CM for 7 days and stained for α-SMA (Fig. 1e). More than 95% of HSCs were α-SMA positive, indicating a pure population. qHSCs were determined by staining α-SMA negative (Fig. 1i).

Fig. 1
figure 1

Characterization of HSCs isolated from rat liver. After isolation, the cells were cultured for 2 days (quiescent HSCs, qHSCs), or for 7 days (culture-activated HSCs, cHSCs), or activated using C5F-CM for 7 days (induction-activated HSCs, iHSCs). a Immunofluorescence analysis of desmin expression in qHSCs. c α-SMA expression in cHSCs. e α-SMA expression-positive iHSCs. g Negative controls without primary antibody were performed in activated HSCs. i α-SMA expression-negative qHSCs. DAPI nuclear staining for all conditions is shown in b, d, f, h and j (×100)

Full size image

Gene expression in iHSCs was different from that in cHSCs

The isolation and activation were repeated twice and expression of 27,100 genes was compared between qHSCs, cHSCs and iHSCs using cDNA microarrays in two independent experiments. One thousand nine hundred sixty-seven genes were more than 1.5-fold up-regulated or 0.5-fold down-regulated in cHSCs and iHSCs (Fig. 2). According to go mapping analysis, these genes were classified into 14 or 15 groups and the fraction of genes in each group is shown in Fig. 3. Some genes encoding physiological process, cellular process and binding protein synthesis were preferentially up-regulated in cHSCs and iHSCs. 17.96% of up-regulated genes encoded physiological process proteins in cHSCs and 18.8% in iHSCs. The same was for cellular process proteins, where up-regulated genes represented 15.53% in cHSCs and 16.24% in iHSCs, and for genes involved in binding protein, with 14.08% of up-regulated genes in cHSCs and 11.11% of up-regulated genes in iHSCs. A less pronounced bias was observed for the genes related to enzyme regulator activity and the extracellular region. 2.43% in cHSCs and 1.71% in iHSCs of up-regulated genes were molecules involved in cell response to stimulus; 0.97% of up-regulated genes in cHSCs and 1.71% of up-regulated genes in iHSCs were extracellular region. Some genes encoding cellular process, physiological process and catalytic activity protein synthesis were preferentially down-regulated in cHSCs and iHSCs. 19.12% of down-regulated genes in cHSCs and 22.92% of down-regulated genes in iHSCs encoded cellular process proteins. The same was observed for physiological process proteins where down-regulated genes represented 15.44% in cHSCs and 15.63% in iHSCs and for genes involved in catalytic activity proteins, with 11.76% of down-regulated genes in cHSCs and 11.46% of down-regulated genes in iHSCs. A less pronounced bias was observed for the genes related to response to stimulus and signal transducer activity. 2.21% of down-regulated genes in cHSCs and 2.08% of down-regulated genes in iHSCs were cell response to stimulus molecules; 1.47% of down-regulated genes in cHSCs and 1.04% of down-regulated genes in iHSCs were signal transducer activity. Specifically up-regulated genes were related to organelles part, development and transcription regulator activity in cHSCs, but not in iHSCs. The up-regulated genes were involved in signal transducer activity and ECM in iHSCs but not in cHSCs. Specifically down-regulated genes were related to organelle part in cHSCs but not in iHSCs, and to synapses in iHSCs but not in cHSCs. Genes specifically up- and down-regulated in iHSCs were involved in chemokines, chemokine receptors, signal transduction, inflammation, proliferation, development and metabolism.

Fig. 2
figure 2

The gene expression pattern of cHSCs is different from that of iHSCs. HSCs were isolated from normal rat. HSCs were considered culture activated after 7 days of culture in 10% fetal bovine serum and induced after 7 days of culture in rat HCC conditioned medium. After isolation of mRNA and generation of cDNA, microarray was performed. Shown is a graph containing all genes that were more than 1.5-fold up-regulated (shown in red) or 0.5-fold down-regulated (shown in green) in comparison with qHSCs (t test, P < 0.05)

Full size image
Fig. 3
figure 3

One thousand nine hundred sixty-seven genes were differentially expressed in cHSCs and iHSCs. The up-regulated genes in cHSCs, but not in iHSCs, comprised 755 genes, the expression of which was increased more than 1.5-fold compared to qHSCs. The down-regulated genes in cHSCs, but not in iHSCs, included 501 genes, the expression of which was decreased more than 0.5-fold compared to qHSCs. The up-regulated genes in iHSCs but not in cHSCs comprised 163 genes, the expression of which was increased more than 1.5-fold compared to qHSCs. The down-regulated genes in iHSCs, but not in cHSCs, included 132 genes, the expression of which was decreased more than 0.5-fold compared to qHSCs. Relative fractions of up-regulated and down-regulated genes in HSCs, as classified into 14 or 15 functional groups, are shown. a Up-regulated genes in cHSCs but not in iHSCs, b down-regulated genes in cHSCs but not in iHSCs, c up-regulated genes in iHSCs but not in cHSCs, d down-regulated genes in iHSCs but not in cHSCs

Full size image

iHSC gene expression was partially similar to that in cHSCs and also partially different. We compared up- and down-regulated genes between cHSCs and iHSCs. Gene expression in cHSCs overlapped only partially with iHSCs. Accordingly, the Venn diagram of all 1967 differentially regulated genes (supplementary Tables 1–6) showed that 257 up-regulated and 159 down-regulated genes from cHSCs were concordantly up- and down-regulated in iHSCs (Fig. 4). Some genes, such as α-SMA, Wnt4, Ccr1, Tagln, Vcam1, Tgfb2, STAP-1, Enpp2, Il13ra2, CRHSP-1 and Lrrk2, were concordantly up-regulated in cHSCs and iHSCs. Many genes including Mnda, Il1a, Scin, Adam17, osteopontin, HIF-1α, LPTN, RT1-B, Crem and Ccl24 were concordantly down-regulated in cHSCs and iHSCs. Many genes, however, were up- or down-regulated in iHSCs but not in cHSCs, and vice versa (Fig. 4). Some genes, such as MT-2, ET-1, Adora2b, Prrg4, Col5α1, Ccnd1, Foxp1, Mki67 and Pdlim1, were up-regulated in cHSCs, but were down-regulated in iHSCs. The other genes, such as cystatin F, MMP-9, Jun, IgG-2a, Il7r, IAP1 and Igf1, were up-regulated in iHSCs, but were down-regulated in cHSCs (Tables 2, 3, 4, 5).

Fig. 4
figure 4

One thousand nine hundred sixty-seven genes were more than 1.5-fold up-regulated or 0.5-fold down-regulated in cHSCs and iHSCs. The Venn diagram shows overlapping genes that were significantly (P < 0.01) and at least 1.5-fold up-regulated or 0.5-fold down-regulated in cHSC and iHSC. a 257 genes were concordantly up-regulated in cHSCs and iHSCs. 755 genes were up-regulated in cHSCs but not in iHSCs. 163 genes were up-regulated in iHSCs but not in cHSCs. b 159 genes were concordantly down-regulated in cHSCs and iHSCs. 501 genes were down-regulated in cHSCs but not in iHSCs. 132 genes were down-regulated in iHSCs but not in cHSCs

Full size image
Table 2 Concordantly up-regulated genes in cHSCs and iHSCs
Full size table
Table 3 Concordantly down-regulated genes in cHSCs and iHSCs
Full size table
Table 4 Genes up-regulated in iHSCs but not in cHSCs
Full size table
Table 5 Genes up-regulated in cHSCs but not in iHSCs
Full size table

Verification of novel genes associated with HSC activation

To assess the validity of the obtained microarray data, we analyzed whether typical HSC activation markers were up-regulated both by microarray and real-time RT-PCR, and confirmed the strong up-regulation of α-SMA mRNA levels. In addition, we assessed the expression of the other selected genes whose expression has been associated with HSC activation (Fig. 4). We identified genes involved in proliferation (cystatin F, c-Myc, ET-1), inflammation (Ccl24, Il-1a), ECM organization and degradation (Colla1, osteopontin, MMP9), cell surface receptors (Ccr1, Il-1Ra, Robo-1, PIP4), cell adhesion molecules (Vcam1), signal transduction factors (Wnt4, Jun, Raf1, Rac2), and immunity factors (CD44, CD72, Tgfb2, Ifngr1, Tlr4, Tlr7). MMP9 was previously reported to be down-regulated in activated HSCs in vivo and in vitro (Jiang et al. 2006; Tsuneyama et al. 2002). We confirmed that the expression of MMP9 was down-regulated in cHSCs, and up-regulated in iHSCs. Ccr1, Colla1, Tgfb2, Vcam1 and Rac2 were concordantly increased in cHSCs and iHSCs. Ccl24, Il-1a, Il-1Ra and Tlr4 were down-regulated in both cHSCs and iHSCs. Robo-1, Raf1 and CD72 were up-regulated in iHSCs, but down-regulated in cHSCs. ET-1, c-Myc and PIP4 were up-regulated in cHSCs, but down-regulated in iHSCs (Fig. 5). These results suggest that the data obtained by cDNA microarrays are reliable.

Fig. 5
figure 5

Identification of previously known and novel differentially regulated genes during HSC activation in CM and in culture. Expression of differentially regulated genes was confirmed in qHSCs, cHSC, iHSCs by real-time RT-PCR. Results were normalized to β-actin expression and are expressed as fold induction ± standard error of the mean in comparison with qHSCs. Each group of HSC samples consisted of at least three different isolations (paired t test, P < .05). a Concordantly up-regulated genes, b concordantly down-regulated genes, c genes up-regulated in iHSCs but not in cHSCs, d genes up-regulated in cHSCs but not in iHSCs

Full size image

Expression patterns of several genes by Western blot analysis

Expression patterns of four genes were confirmed by Western blot analysis. Expression patterns of all investigated proteins were similar to our PCR and microarray analysis data, with MMP-9 being up-regulated in iHSC but not in cHSCs, Ccr1 being significantly up-regulated in cHSCs but not in iHSCs, and TGF-β2 and Vcam1 being up-regulated in both cHSC and iHSCs (Fig. 6).

Fig. 6
figure 6

Expression of selected genes was confirmed by Western blot analysis. Each Western blot is based on HSCs isolated from three normal rats, and is representative of at least two independent experiments

Full size image

Discussion

Myofibroblasts are major components of the stroma of many tumors, such as those of colon, pancreas, oral cavity and breast (Adegboyega et al. 2004; Masamune et al. 2008; Kellermann et al. 2007; Surowiak et al. 2007). Although their precise functions remain to be determined, several arguments point to their role in favoring tumor progression (Dimanche-Boitrel et al. 1994; De Wever et al. 2008). Studies have already shown that human hepatic myofibroblasts are able to greatly increase the invasiveness of HCC cell lines through the secretion of hepatocyte growth factor and that myofibroblasts are also involved in the synthesis of many ECM components of the HCC stroma (Neaud et al. 1997; Guirouilh et al. 2000; Faouzi et al. 1999a, b). By activating qHSC into myofibroblasts, HCC cells themselves contribute to tumor progression (Faouzi et al. 1999a, b; Guirouilh et al. 2000). However, the precise mechanism of HSC activation in HCC is not well understood at present. Accordingly, we induced HSC activation by the rat HCC cell line C5F, and used cDNA microarray analysis to investigate gene expression levels of HSCs, to understand the role of iHSCs in this process. Our results showed that HCC cells were able to induce the in vitro activation of HSCs. HSC activation was characterized by typical morphologic modifications, proliferation, and increased expression of α-SMA.

Our microarray study detected a large number of previously known and unknown genes that were up- or down-regulated during HSC activation, thus confirming the ECM-producing and ECM-degrading, proliferative, and inflammatory phenotype of activated HSCs. Although cHSCs up-regulated typical HSC activation markers such as α-SMA, they did have the same changes of gene expression patterns seen in iHSCs. The finding that induced activation of HSCs up-regulates a different set of genes than cultured activation of HSCs is not completely surprising, in view of the fact that the microenvironment of HSCs in the CM is complex and exposes HSCs to a number of cytokines secreted by HCC cells, whereas culture activation occurs in an artificial environment and does not include interactions with cytokines secreted by HCC cells. Culture activation with FBS is by far the most common model of HSC activation. Our data suggest that the common practice of evaluating genetic manipulations during HSC culture activation may produce results that do not reflect effects on HSCs and HCC in vivo. Our study compared gene expression patterns of cHSCs and iHSCs to that of qHSCs, and thus had a higher capacity to detect differentially regulated HSC-specific genes. Specific gene expression patterns of iHSCs indicated their distinct role in HCC. In addition to detecting classical HSC activation genes, such as α-SMA, collagen α1(I), cyclin D2 and fibronectin1, our study identified a large number of genes that have not been previously described to be associated with HSC-induced activation. Among novel differentially regulated genes identified in iHSCs are genes involved in (i) ECM formation and degradation, e.g. Col5α1 (down-regulated) and MMP-9 (up-regulated); (ii) apoptosis including IAP1 (up-regulated); (iii) proliferation, such as cystatin F (up-regulated); (iv) inflammation, e.g. Il-1a (up-regulated) and osteopontin (down-regulated); (v) adhesion, such as Vcam1 (up-regulated); (vi) signal transduction, including Wnt6, Jun, Raf1 (all up-regulated), and (vii) immune factors, such as CD44, Tgfb2 and Ifngr1 (all up-regulated). In addition, we identified differential expression of several receptors such as Ccl24, PIP4 (both down-regulated), Robo1, Il7r, CD72, Tlr7 and Ccr1 (all up-regulated). Increased expression of VCAM1 can facilitate the adherence of HCC cells to the vessel wall (Futakuchi et al. 2002). MMP-9 expression can degrade ECM and promote invasion and metastasis of HCC (Hong et al. 2005). Up-regulated TGF-β2 expression can induce HCC immune inhibition (Maggard et al. 2001), suggesting that iHSCs may indirectly facilitate recurrence and metastasis of HCC by inhibiting organism immune function. The IAP family member IAP1 was up-regulated in iHSCs, suggesting a potential role for IAP1 in protecting HSCs from apoptosis during HCC progression.

The fact that cHSCs and iHSCs display a significantly different gene expression pattern suggests that iHSCs might play an important role in HCC different from that of cHSCs. Moreover, cHSCs dramatically up-regulated genes such as Ccr1, Sdc2, CD44 and TGF-β2, which were slightly up-regulated or down-regulated in iHSCs, suggesting that in vitro culture artificially drives the expression of some genes, and thus contributes to aberrant gene expression during culture activation. Our data indicate that, to some extent, iHSCs can reflect interactions with HCC in vivo, but cHSCs cannot truly reflect interactions with HCC. Therefore, this study has provided theoretical foundations for investigating the biology of HSCs in HCC. Nevertheless, the microenvironment of HSC in CM is not completely identical to the microenvironment in the liver. The in vivo activation of HSCs should be considered as the standard for the study of HSC biology in HCC.