SEARCH

SEARCH BY CITATION

Keywords:

  • HCC;
  • hepatic stellate cell;
  • fatty liver;
  • Rho-kinase

Abstract

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

Fatty liver (FL) is associated with development of hepatocellular carcinoma (HCC). However, whether FL itself promotes the progression of HCC is unclear. We recently found that hepatic stellate cells (HSCs) were prominently activated in the steatotic liver. Here, we investigated whether steatotic livers promote HCC progression and whether HSCs of steatotic liver are associated with HCC progression. We implanted rat HCC cells into diet-induced steatotic livers in rats via portal vein injection. Thereafter, HSCs and HCC cells were co-implanted subcutaneously into nude rats. Migration and proliferation of HCC cells were measured, and activation of ERK and Akt in these cells was determined by western blotting. Chemokines secreted from HSCs and HCC cells were also evaluated by ELISA. Steatotic livers significantly promoted HCC metastasis compared with non-steatotic livers. Additionally, co-implantation of HCC cells with HSCs from steatotic livers produced significantly larger tumors in recipient rats as compared to those induced by HCC cells co-implanted with HSCs from normal livers (NLs). HSCs isolated from steatotic livers, compared with HSCs isolated from NLs, secreted greater amounts of interleukin-1α, vascular endothelial growth factor, and transforming growth factor-β. These cytokines may enhance the proliferation and migration of HCC cells by increasing the phosphorylation of ERK and Akt in HCC cells. Moreover, we noted that the Rho-kinase inhibitor deactivated activated HSCs and attenuated HCC progression. In conclusion, the rat steatotic liver microenvironment favors HCC metastasis, and this effect appears to be promoted by activated HSCs in the steatotic liver.

Abbreviations
CDD

choline-deficient diet

CM

conditioned media

DMEM

Dulbecco's modified Eagle's medium

ELISA

enzyme-linked immunosorbent assay

FL

fatty liver

FBS

fetal bovine serum

HCC

hepatocellular carcinoma

HGF

hepatocyte growth factor

HSCs

hepatic stellate cells

HSCFL

hepatic stellate cells isolated from fatty liver

HSCNL

hepatic stellate cells isolated from normal liver

HFD

high-fat diet

HIF

hypoxia inducible factor

IL-1α

interleukin-1α

IQGAP1

IQ motif containing GTPase activating protein 1

MMPs

matrix metalloproteinases

MTT

methyl thiazolyl tetrazolium

MAPK

mitogen-activated protein kinase

NAFLD

non-alcoholic fatty liver disease

NASH

non-alcoholic steatotic hepatitis

NL

normal liver

ROCK

Rho-associated kinase

SASP

senescence-associated secretory phenotype

SDF-1

stromal derived factor-1

TIMP-1

tissue inhibitor metalloproteinase-1

TGF-βRII

TGF-β receptor II

TGF-β

transforming growth factor-β

VEGF

vascular endothelial growth factor

Non-alcoholic fatty liver disease (NAFLD) is one of the most common hepatic disorders in developed countries. The epidemic of obesity in developed countries has increased along with its attendant complications, including metabolic syndrome and NAFLD. Recently, there is increasing evidence that NAFLD, including the more aggressive non-alcoholic steatohepatitis (NASH), is associated with hepatocellular carcinoma (HCC).[1-3] Diabetes and obesity are established independent risk factors for the development of HCC,[3] and obesity is reported to be an independent risk factor for HCC recurrence after curative treatment, such as hepatectomy in patients with NASH.[4] It has been reported that obesity and fatty liver (FL) promote a chemical carcinogen-induced hepatocarcinogenesis.[5, 6] However, the functional impact of FL on the progression and metastasis of HCC remains largely unexplored.

Hepatic stellate cells (HSCs) are key contributors to liver fibrosis and portal hypertension.[7-9] Recently, these cells were postulated to form a component of the pro-metastatic liver microenvironment because they can transdifferentiate into highly proliferative and motile myofibroblasts, which have been implicated in desmoplastic reactions and metastatic growth.[10, 11] Moreover, HSC activation has been shown to correlate with the severity of hepatic steatosis.[12, 13] Therefore, activated HSCs in FL may enhance the progression of HCC, but this possibility has not been fully explored. Therefore, we investigated whether FL in rats has a microenvironment that can promote the progression of HCC, and whether the HSCs in FL enhance the progression of HCC.

Material and Methods

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

Animals

Four-week-old male Buffalo and F344 nude rats were purchased from Clea Japan, (Tokyo, Japan), and F344 rats were purchased from Charles River Breeding Laboratories (Osaka, Japan). Four-week-old rats were fed a choline-deficient diet (CDD) (Oriental Yeast Co., Tokyo, Japan) for 6 weeks or a high-fat diet (HFD) for 16 weeks (F2HFD2, 82% kcal fat; Oriental Yeast Co.) to promote the development of FL. All animal experiments were performed according to the guidelines set by the United States National Institutes of Health (1996).

Cell lines

The rat HCC cell line McA-RH7777 was obtained from the American Type Culture Collection (Rockville, MD). The rat HCC cell lines C1 and L2 were kindly provided by Dr. K. Ogawa, National Institute of Health Sciences (Tokyo, Japan). The McA-RH7777 cell line originated in Buffalo rats, whereas the C1 and L2 cell lines originated in F344 rats. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) in a humidified atmosphere of 5% CO2 at 37°C.

Isolation of HSCs

HSCs were isolated from rat livers according to previously described procedures.[14] The purity of the cells was estimated through ordinal light and fluorescence microscopic examinations and by indirect enzyme immunoreactivity with an anti-desmin antibody (Dako, Versailles, France).

Conditioned medium

Conditioned medium (CM) was harvested from cultured HSCs after incubation in serum-free DMEM for 48 hr. At the end of the incubation period, the medium was stored at −80°C until use.

Proliferation assay

HCC cells were seeded at 5,000 cells per well in 96-well plates and cultured overnight in DMEM supplemented with 10% FBS. The medium was then changed to serum-free DMEM and CM. Incubations continued for 24 hr before the addition of 3-(4, 5-methylthiazol-2-yl)-2, 5-diphenyl-tetrazolium bromide [methyl thiazolyl tetrazolium (MTT), nonradioactive proliferation assay; Promega Corp, Madison, WI] for 4 hr. Cellular MTT was solubilized with acidic isopropanol, and the optical density was measured at 570 nm by using a 96-well plate reader. The survival fraction was then quantified.

Migration assay

For studies of HCC cell migration, 8-μm-pore size Transwell chambers (Corning, NY) were used. In total, 5 × 105 HSCs were seeded into the lower chamber that was coated with collagen type I in 1 mL of medium containing 10% FBS, and cultured for 48 hr. No HSCs were added to the control wells. The medium was changed to 750 µL of RPMI supplemented with 0.1% bovine serum albumin, and 2 × 104 HCC cells in 200 µL of RPMI with 0.1% bovine serum albumin were added to the upper chamber. After incubation at 37°C in 5% CO2 for 24 hr, the non-migrating cells on the upper surface of the membrane were removed with a cotton swab. Cells were fixed in 4% paraformaldehyde and stained with propidium iodide solution (Dojindo, Kumamoto, Japan). Migrating cells were counted at 200× magnification in nine adjacent microscope fields for each membrane.

Enzyme-linked immunosorbent assay

The amount of vascular endothelial growth factor (VEGF), tissue inhibitor of metalloproteinases 1 (TIMP-1), matrix metalloproteinase-9 (MMP-9), transforming growth factor-β1 (TGF-β1), and interleukin 1α (IL-1α) were quantified using ELISA kits, according to the manufacturer's instructions (R&D Systems, Minneapolis, MN).

Western blot analysis

Cells were cultured in DMEM without 10% FBS overnight. Thereafter, the cells were incubated in the presence of CM for 10 min before being homogenized in lysis buffer (Cell Lysis Buffer; Cell Signaling Technology, Danvers, MA). Western blot analysis was performed as described previously.[14] Antibodies to β-actin were purchased from Abcam (Tokyo, Japan). Antibodies to Akt, p-Akt, anti-p44/42, mitogen-activated protein kinase (MAPK), and anti-phospho-p44/42 MAPK antibodies were purchased from Cell Signaling Technology (Beverly, MA). The phosphorylation levels were normalized to the levels of total Akt or MAPK protein expression.

Experimental model of intrahepatic HCC metastasis

HCC cells (5 × 106 cells or 5 × 105 cells/body) were implanted into the livers of rats via portal vein injection. In rats that were fed on a CDD, normal diets were given after the injection of the HCC cells, whereas in rats fed with a HFD, the HFD was continued until the study was completed. At the end of the experiment, rats were humanely sacrificed. The area of the liver occupied by tumor was calculated by averaging the percentage of the liver area occupied by tumor in microscopic sections continuously cut at 5-mm intervals.

Confocal immunofluorescence and histology

Phalloidin staining of isolated HSCs was performed as described previously.[14] Samples were observed under a conventional fluorescence microscope or a laser confocal microscope. For histological analysis, formalin-fixed liver tissue sections were cut, stained with hematoxylin–eosin, and examined microscopically. To assess the grade of the steatosis, sections were stained with oil red O.

HSC/tumor cell co-implantation model

C1 cells (5 × 106 cells/body) were implanted into the subcutis of F344 nude rats, either alone or in combination with HSCs isolated from F344 rats (5 × 106 cells/body). Successful implantation of HSCs was determined via frozen section analysis of the co-implantation of fluorescently labeled HSCs and HCC cells. The HSCs were fluorescently labeled with red fluorescent linker dye (PKH26 Red Fluorescent Cell Linker Kit; Sigma, Sigma-Aldrich, St Louis, MO) according to the manufacturer's instructions. At the end of the experiment, rats were humanely sacrificed. The largest tumor diameter and the tumor weight were measured 4 weeks after implantation. The survival of HSCs and cellular proliferation were assessed by immunohistochemical analysis of desmin and Ki-67 (BD Pharmingen, San Jose, CA), respectively.

Treatment of HSCs by Y-27632

The specific Rho-associated kinase (ROCK) inhibitor, Y-27632, was purchased from Wako (Osaka, Japan). Activated HSCs were deactivated by incubation with Y-27632 (10 μM) in serum-free DMEM for 3 hr.

Statistical analysis

The survival rates of rats were compared using the Kaplan–Meier method and were analyzed using the log-rank test. The tumor engraftment rates were compared by the chi-square test. One-way analysis of variance was used for multiple comparisons. All the data are expressed as the average (±SE). p Values less than 0.05 were considered statistically significant. Statistical analyses were performed with the SPSS software, version 16 (SPSS Japan, Tokyo, Japan).

Technical and material details of the cytokine assay are given in the Supporting Information Materials.

Results

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

CDD induced FL, activation of HSCs and increased secretion of cytokines

Feeding on a CDD for 6 weeks resulted in severe steatotic changes (60% macrosteatosis) (Supporting Information Fig. S1a). The purity of the isolated HSCs was estimated by fluorescence microscopic examination and by indirect enzyme immunoreactivity with an anti-desmin antibody, and was found to be >90% (Supporting Information Fig. S1b). HSCs isolated from FL (HSCFL) of rats fed on a CDD for 6 weeks had significantly increased stress fiber formation compared to HSCs isolated from normal liver (HSCNL) (Supporting Information Fig. S1c). We assessed the mediators secreted by cells in monoculture by performing cytokine arrays on CM samples. TIMP-1 and VEGF were detected in CM harvested from HSCFL, whereas they were not detected in CM harvested from HSCNL (Supporting Information Fig. S1d–S1f).

CDD-induced FL has a permissive microenvironment for HCC metastasis

To assess the effects of FL on HCC metastasis, we first implanted McA-RH7777 cells (5 × 105 cells/body) via the portal vein into the livers of syngeneic Buffalo rats. One of seven rats with normal livers (NLs) developed a small, single nodular HCC; all seven rats survived more than 8 weeks after inoculation of the HCC cells. In contrast, all ten rats with FL fed on a CDD for 6 weeks developed diffusely distributed tumors, and five of ten died of HCC within 8 weeks (Figs. 1a1c). Volumes of the HCC tumors were significantly greater in rats with FL than in rats with NL (Fig. 1d).

image

Figure 1. Choline deficient diet-induced FL has a permissive microenvironment for HCC metastasis. (a) Macrography of liver tumors in NL and FL. One of seven rats with NL developed a single nodular tumor (arrow) after inoculation of HCC cells (McA-RH7777 cells, 5 × 105 cells/body), whereas all ten rats with FL developed diffusely distributed liver tumors. (b) Hematoxylin-eosin-stained images of liver tumors in NL and FL. (c) Kaplan–Meier curve documenting the survival of rats inoculated with HCC cells. (d) The graph shows the percentage of tumor volume (tumor volume/liver volume). Results are presented as means (±SE).

Download figure to PowerPoint

In other experiments, L2 cells (5 × 106 cells/body) were injected into the portal veins of F344 rats that were fed on either a CDD or a normal diet. In all eight rats fed on a CDD for 6 weeks, multiple nodular liver tumors developed within 8 weeks, and four of these rats also developed pulmonary metastases. In contrast, none of the five rats fed on a normal diet developed tumors (Supporting Information Figs. S2a and S2b).

HSCFL stimulate HCC cell proliferation and migration in vitro

First, we investigated whether HSCFL could induce the proliferation of McA-RH7777 and C1 HCCs. When these cells were cultured with CM harvested from syngeneic HSCFL, their rate of proliferation increased, whereas when the cells were incubated with CM harvested from HSCNL, no increase in tumor cell proliferation occurred (Fig. 2a, Supporting Information Fig. S3a). We then examined whether HSCFL could induce migration of HCCs. HSCFL promoted HCC migration to a significantly greater extent than did HSCNL (Fig. 2b, Supporting Information Fig. S3b).

image

Figure 2. HSCFL promoted HCC progression in vitro. (a and b) HSC-FL promoted HCC proliferation and migration. (c) Protein lysates from McA-RH7777 cells treated with conditioned media from HSCNL or HSCFL were analyzed for Akt and ERK1/2 activation. The phosphorylation levels were normalized to total protein expression. (dh) Cytokine secretion from McA-RH7777 cells and HSCs in monoculture and co-culture, using ELISA kits. Data shown are presented from more than 5 independent experiments. *p < 0.05, **p < 0.01.

Download figure to PowerPoint

Factors secreted from HSCFL enhanced activation of MAPK and Akt pathways in HCC cells

To determine the signaling pathways that may be involved in the tumor-promoting effects of HSCFL, we examined McA-RH7777 cells treated with CM harvested from a monoculture of HSCFL for activation of Akt and MAPK by western blotting. The phosphorylation levels of Akt and ERK were significantly increased in cells treated with CM harvested from monoculture of HSCFL in comparison with CM harvested from HSCNL (Fig. 2c).

HCC-HSC cross-talk is bidirectional

We analyzed the secretion of VEGF, TGF-β1, IL-1α, MMP-9 and TIMP-1 by HCC cells and HSCs in monoculture and co-culture, using ELISA kits. Only an additive effect of VEGF and TGF-β1 was noted, and their levels were significantly increased in the CM harvested from co-culture of HCC cells and HSCFL compared with the levels in CM harvested from co-culture of HCC cells and HSCNL (Figs. 2d and 2e). Although HSCNL, HSCFL, and HCC cells barely secreted IL-1α and MMP-9, both factors were secreted in CM harvested from co-culture of HCC cells and HSCNL. Furthermore, the concentrations of IL-1α and MMP-9 were significantly increased in CM harvested from co-culture of HCC cells and HSCFL, due to a synergistic effect, compared to concentrations in CM harvested from co-culture of HCC cells and HSCNL (Figs. 2f and 2g). However, the concentration of TIMP-1 was significantly lower in CM harvested from co-culture of HCC cells and HSCFL compared to CM harvested from monoculture of HSCFL because of an inhibitory effect (Fig. 2h). These results suggested that cytokine secretions are altered through HCC-HSC interactions.

Co-implantation with HSCFL promotes HCC growth in vivo

We examined the effect of HSCFL on HCC growth in vivo. C1 cells, originating from F344 rat HCC cells, were implanted to the subcutis of syngeneic F344 nude rats, either alone or in combination with HSCFL or HSCNL from F344 rats. HSCNL and HSCFL were implanted to a similar extent, according to frozen section analysis, 2 days after the co-implantation of HSCs labeled by PKH26 and non-labeled HCC cells (Supporting Information Figs. S4a and S4b). When C1 cells (5 × 106 cells/body) were implanted into the subcutis of syngeneic F344 nude rats, three of the nine rats implanted with HCC cells alone and six of nine rats implanted with HCC cells and HSCNL developed tumors at the site of implantation, but all nine rats implanted with HCC cells and HSCFL developed tumors. Moreover, the rats implanted with HSCFL in addition to HCC cells developed significantly larger tumors than those implanted with HCC cells and HSCNL (Figs. 3a3d). The presence of desmin-positive cells was determined by immunohistochemical analysis. The number of desmin-positive cells in the tumors of rats co-transplanted with HSCFL was comparable with those co-transplanted with HSCNL, whereas only a few desmin-positive cells were noted in tumors of rats implanted with C1 cells alone (Supporting Information Figs. S4c and S4e). In addition, we used immunohistochemistry to examine the expression of nuclear Ki-67, a cellular proliferation marker. The number of Ki-67-positive cells was significantly greater in the tumors of rats co-transplanted with HSCNL than in the tumors of rats transplanted with C1 cells alone. Furthermore, the number of Ki-67-positive cells was significantly increased in the tumors of rats co-transplanted with HSCFL compared with those co-transplanted with HSCNL (Supporting Information Figs. S4d and S4f).

image

Figure 3. HSCFL promoted HCC progression in vivo. (a) Macrography of subcutaneous tumors in each group. (b) The table shows the engraftment rate in each group. (c) Tumor sizes were measured weekly. Data are presented as means (±SE). *p < 0.05, **p < 0.01 compared to the C1+HSCNL group (d) The graph shows the average tumor weight 4 weeks after co-implantation. Data are presented as means (±SE).

Download figure to PowerPoint

Rho-kinase inhibitor attenuates HCC progression through deactivation of HSC-FL

We have previously shown that HSCFL exhibited increased stress-fiber formation and F-actin expression as compared to HSCNL,[14] and increased stress-fiber formation and F-actin expression in HSCFL were suppressed by treatment with the ROCK inhibitor, Y-27632. However, HSCFL were re-activated within 48 hr of Y-27632 administration (Fig. 4a). In this study, we investigated whether the ROCK inhibitor can deactivate activated HSCFL and suppress tumor progression in co-culture. First, we investigated the effect of Y-27632-treated HSCFL on the proliferation of McA-RH7777 cells by performing the MTT assay. When McA-RH7777 cells were cultured with CM harvested from Y-27632-treated HSCFL, cell proliferation was suppressed compared with that in CM harvested from untreated HSCFL (Fig. 4b). Thereafter, a migration assay showed that the migration of McA-RH7777 cells was significantly suppressed in co-culture with Y-27632-treated HSCFL compared with co-culture with untreated HSCFL (Fig. 4c). Furthermore, the enhancement in the proliferation and migration by Y-treated HSCFL were significantly greater than those of HSCNL (Figs. 4b and 4c). In addition, the phosphorylation levels of Akt and ERK were significantly decreased in HCC cells treated with CM harvested from monoculture of Y-27632-treated HSCFL compared with phosphorylation levels in HCC cells treated with CM harvested from untreated HSCFL (Fig. 4d). Moreover, an ELISA study showed that the levels of TGF-β1 were significantly lower in CM harvested from Y-27632-treated HSCFL than levels in CM harvested from untreated HSCFL (Fig. 4e). Moreover, the concentration of IL-1α was significantly decreased in CM harvested from a co-culture of HCC and HSCFL treated with Y-27632, compared to that in CM harvested from a co-culture of HCC and untreated HSCFL (Fig. 4f). However, the levels of VEGF and TIMP-1 in CM harvested from Y-27632-treated HSCFL and the levels of MMP-9 and TIMP-1 in CM harvested from a co-culture of HCC and HSCFL treated with Y-27632 were similar to those in CM harvested from untreated HSCFL and in CM harvested from a co-culture of HCC and untreated HSCFL, respectively (Supporting Information Figs. S5a–S5d). These results indicated that Rho-kinase inhibitor did not inhibit the overall features of HSC, but partially inhibited the activation of HSC.

image

Figure 4. The Rho-kinase inhibitor, Y-27632, attenuated the HCC-promoting effect of HSCFL. (a) Stress fiber and F-actin expression in HSCFL and Y-27632-treated HSCFL (Y-treated HSCFL). (b) Conditioned media harvested from Y-treated HSCFL (Y-treated HSCFL-CM) significantly suppressed the proliferation of McA-RH7777 compared with that of untreated HSCFL-CM. (c) Migration of McA-RH7777 under the stimulation of Y-treated HSCFL was significantly suppressed compared with that of untreated HSCFL. (d) The phosphorylation levels of Akt and ERK were significantly suppressed in HCC cells treated with Y-treated HSCFL-CM, compared to those in HCC cells treated with untreated HSCFL-CM. (e) Y-27632 suppressed TGF-β1 secretion of HSCFL. (f) Y-27632 suppressed IL-1α secretion on co-culture of HCC and HSCFL.

Download figure to PowerPoint

High fat diet-induced FL also has a permissive microenvironment for HCC metastasis

We assessed the effect of HFD-induced FL (non-CDD-induced FL) on HCC metastasis. Rats fed on a HFD for 16 weeks developed microvesicular steatosis [approximately 20–30% steatosis, (Supporting Information Fig. S6a)], although the HFD-induced FL appeared less fibrotic compared to CDD-induced FL (Supporting Information Fig. S6b). Furthermore, the serum levels of AST were significantly lower in rats with HFD-induced FL than in those with CDD-induced FL (Supporting Information Fig. S6c). All the six rats fed on a HFD for 16 weeks developed several nodular tumors at 8 weeks after the inoculation of McA-RH7777 cells (5 × 105 cells/body). In contrast, one of the seven rats fed on a normal diet for 16 weeks developed several small tumors (Figs. 5a and 5b). Volumes of the HCC tumors were significantly greater in rats with HFD-induced FL than in rats with NL (Fig. 5c).

image

Figure 5. High fat diet-induced FL also provides a permissive microenvironment for HCC metastasis and progression. (a) Macrography of liver tumors in NL and high fat diet (HFD)-induced FL. Only one of seven rats with NL developed several nodular tumors after inoculation of HCC cells (McA-RH7777 cells, 5 × 105 cells/body), whereas all six rats with FL induced by feeding of a HFD for 16 weeks developed multiple nodular liver tumors. (b) Hematoxylin–eosin-stained images of liver tumors in NL and HFD-induced FL. (c) The graph shows the percentage of tumor volume in the liver. (d and e) HSCs derived from HFD-induced FL stimulated HCC migration and proliferation in vitro.

Download figure to PowerPoint

HSC derived from HFD-induced FL stimulate HCC cell proliferation and migration in vitro

We investigated whether HSCFL of rats fed on a HFD for 24 weeks (HSCHFD) could induce proliferation and migration of HCC cells. HSCHFD promoted HCC migration to a significantly greater extent than did HSC isolated from NL of rats fed on a normal diet for 24 weeks (Fig. 5d). In addition, HSCHFD promoted HCC proliferation to a significantly greater extent than did HSC isolated from NL of rats fed on a normal diet for 24 weeks (Fig. 5e).

Discussion

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

A rapidly growing literature indicates that NAFLD, including NASH, is associated with HCC.[1-3] However, whether NAFLD itself promotes the progression and metastasis of HCC is unclear. Therefore, we investigated whether FL either promotes or suppresses HCC progression. Our results have shown that both CDD- and HFD-induced FL have pro-metastatic microenvironments in the model of portal vein HCC cell inoculation. Furthermore, HSCs were activated in both CDD- and HFD-induced FL, and these activated HSCs enhanced the proliferation and migration of HCC cells. In addition, Y-27632, a Rho-kinase inhibitor, partially reduced the progression of HCC through deactivating activated HSCs. These results indicated that the steatotic liver microenvironment favors HCC progression and metastasis through the activation of HSCs.

In the current study, we have shown that CDD-induced FL activates HSCs to enhance the proliferation and migration of HCC in co-culture and co-implantation models through the secretion of paracrine signaling molecules such as VEGF. HSCs can transdifferentiate into highly proliferative and motile myofibroblasts during the activation process that follows liver injury.[8] In addition, we noted that the HSCs derived from HFD-induced FL, which appeared less fibrotic compared with CDD-induced FL, promoted the migration and proliferation of HCC cells, even though the HSCs may not be fully activated. Free fatty acids such as oleate and palmitate reportedly stimulate the activation of fibrosis-related genes (i.e., TGF-β, TIMP-1) in HSCs.[15] Activated HSCs produce growth factors and cytokines, such as TGF-β, hepatocyte growth factor (HGF), stromal-derived factor-1 (SDF-1), and IL-1, to stimulate the proliferation, adhesion, and migration of cancer cells.[16, 17] It has been postulated that HSCs are a component of the prometastatic liver microenvironment.[10, 11] Neaud et al.[18] showed that myofibroblasts increased the invasiveness of HCC cells by secreting HGF. Recently, Liu et al. demonstrated that the IQ motif containing GTPase-activating protein 1 (IQGAP1) binds to TGF-β receptor II (TGF-βRII) and suppresses TGF-βRII-mediated signaling in HSCs, thus preventing myofibroblastic differentiation. IQGAP1 deficiency in HSCs promoted myofibroblast activation, tumor implantation, and metastatic growth via upregulation of paracrine signaling molecules, including SDF-1/CXCL12 and HGF.[19] Our results are consistent with that report. Yoshimoto et al.[6] showed that senescence-associated secretory phenotype (SASP)[20] plays crucial roles in promoting obesity-associated HCC development in mice. In the current study, some of the SASP factors, such as IL-1α, CXCR2-binding chemokines, and TGF-β were increased in activated HSCs. These activated HSCs may be senescing, and this effect may be related to the promotion of metastatic growth of HCCs, but further study of this possibility is necessary. Sancho-Bru et al.[21] have examined the effect of hepatocarcinoma cells on HSCs in a co-culture system, and have reported the interaction of HSCs and HCC cells. In that study, co-culture of the cells reduced the expression of fibrogenic factors, such as procollagen-αI(I). Those results may be consistent with our results indicating the presence of decreased levels of TIMP-1 and increased levels of MMP-9 in CM harvested from co-culture of HCC cells and HSCs. Coulouarn et al. also showed that hepatocyte-HSC cross-talk generated a permissive proangiogenic microenvironment by inducing VEGF and MMP9 expression in HSCs.[22] Our results indicate that HCCs and HSCs have bidirectional cross-talk; that is, this interaction is proangiogenic and tumorigenic, but also antifibrogenic. These paradoxical results may have important implications for the progression of HCC. It is speculated that the remodeling of the extracellular matrix, along with the formation of new vessels, contributes to the invasiveness of HCC.

We have also shown that a ROCK inhibitor converted activated HSCs to inactivated HSCs, thereby suppressing the progression of HCC. The Rho signaling pathway and actomyosin system are reportedly involved in the motility and invasion of various cells, including cancer cells.[23-25] It is known that Rho signaling is involved in HSC activation, and a specific ROCK inhibitor, Y-27632, inhibited the activation of HSCs by regulating the formation of actin fibers and focal adhesion.[25-27] Our results are consistent with those reports. However, the production of TGF-β and IL-1α was suppressed by Y-27632, whereas HSCs reverted by treatment with Y-27632 still secreted several cytokines, including VEGF and MMP-9. These data indicated that HSCs did not fully revert to a quiescent state, but rather retained a preactivated intermediate state.[28] Several reports, including ours,[29] have also shown that treatment of tumor-bearing rats with a ROCK inhibitor suppressed peritoneal dissemination of cancer cells and intrahepatic metastasis of HCC cells.[30, 31] This suggested that the ROCK inhibitor is implicated in suppressing HCC progression through not only a direct action on HCC (by inhibiting actomyosin contractility of HCC cells), but also through another indirect action within the cancer stroma, which includes HSCs.

Several limitations to our study should be considered. In the present study, we used hepatoma cells derived from rats. The animals develop large tumors within a few weeks of HCC inoculation, which diffusely infiltrate the liver. This may lead to the development of a different tumor microenvironment and different interactions between HCC and HSCs compared to those present in tumors that arise endogenously. Furthermore, we have shown that the progression of HCC in FL is associated with activated HSCs. However, the activation of HSCs is probably one of several mechanisms associated with the tumorigenic environment in FL. It has been shown that fatty change in hepatocytes induces hypoxic environments in the liver.[32] Indeed, fat droplet accumulation in the cytoplasm of hepatocytes is associated with an increase in cell volume, which may result in partial or complete obstruction of the hepatic sinusoidal space and reduction in sinusoidal blood flow. A state of chronic cellular hypoxia persists in FL, which induces hypoxia-inducible factor (HIF).[33] HIF can induce a vast array of gene products controlling energy metabolism, neovascularization, survival, and cell migration, and is recognized as a strong promoter of tumor growth.[34] The sinusoidal endothelial cells are injured in cases of FL.[32] The number of adherent leukocytes in the injured sinusoid cells induced by a methionine-and CDD was found to be significantly increased, compared with that in normal sinusoid cells. The injured sinusoid cells in the FL may promote tumor cell arrest and extravasation into the hepatic parenchyma. Additional studies of the other mechanisms by which a FL promotes tumor progression are needed. We have also shown that FL with approximately 10–20% steatosis, which was induced by a long-term feeding (16 weeks) of a HFD, promoted the progression of HCC. However, it remains unexplored whether FL with less than 10–20% steatosis promotes the progression of HCC.

In conclusion, our results indicate that the rat steatotic liver microenvironment favors HCC metastasis. This effect appears to be promoted through the activation of HSCs in the steatotic liver.

Acknowledgements

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

The authors like to thank K. Ogawa, National Institute of Health Sciences (Tokyo, Japan), for the generous gift of the HCC cell lines, as well as M. Kiyokawa and Y. Ishida for their excellent technical assistance.

References

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

Supporting Information

  1. Top of page
  2. Abstract
  3. Material 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
ijc29096-sup-0001-suppinfo.docx18KSupporting Information
ijc29096-sup-0002-suppfig1.tif20957KSupporting Figure 1
ijc29096-sup-0003-suppfig2.tif16719KSupporting Figure 2
ijc29096-sup-0004-suppfig3.tif6410KSupporting Figure 3
ijc29096-sup-0005-suppfig4.tif22441KSupporting Figure 4
ijc29096-sup-0006-suppfig5.tif10994KSupporting Figure 5
ijc29096-sup-0007-suppfig6.tif7428KSupporting Figure 5

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.