RhoE is frequently down-regulated in hepatocellular carcinoma (HCC) and suppresses HCC invasion through antagonizing the Rho/Rho-Kinase/Myosin phosphatase target pathway


  • Wei Ma,

    1. State Key Laboratory for Liver Research, University of Hong Kong
    2. Department of Pathology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong
    Search for more papers by this author
  • Carmen Chak-Lui Wong,

    1. State Key Laboratory for Liver Research, University of Hong Kong
    2. Department of Pathology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong
    Search for more papers by this author
  • Edmund Kwok-Kwan Tung,

    1. State Key Laboratory for Liver Research, University of Hong Kong
    2. Department of Pathology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong
    Search for more papers by this author
  • Chun Ming Wong,

    1. State Key Laboratory for Liver Research, University of Hong Kong
    2. Department of Pathology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong
    Search for more papers by this author
  • Irene Oi-Lin Ng

    Corresponding author
    1. State Key Laboratory for Liver Research, University of Hong Kong
    2. Department of Pathology, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong
    • Room 127B, University Pathology Building, Department of Pathology, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
    Search for more papers by this author
    • fax: 852-2872-5197

  • Potential conflict of interest: Nothing to report.

  • Funded in part by Hong Kong Research Grants Council Collaborative Research Fund (HKU 7/CRF/09) and General Research Fund (HKU775811M). I.O.L. Ng is Loke Yew Professor in Pathology.


Deregulation of Rho guanosine triphosphatase (GTPase) pathways plays an important role in tumorigenesis and metastasis of hepatocellular carcinoma (HCC). RhoE/Rnd3 belongs to an atypical subfamily of the RhoGTPase, the Rnd family, as it lacks the intrinsic GTPase activity and remains always in its active GTP-bound form. In this study we investigated the role of RhoE in HCC. We examined the expression of RhoE in primary HCC samples from patients predominantly infected with the hepatitis B virus (HBV) and found that the RhoE messenger RNA (mRNA) level was frequently down-regulated (83.1%, 59/71) in HCCs. Low expression of RhoE in the tumors was significantly associated with shorter disease-free survival (P = 0.020) of the patients. Knockdown of RhoE by short-hairpin RNA using a lentiviral approach led to increased cell motility and invasiveness in SMMC7721 and BEL7402 HCC cells. Moreover, in vivo an orthotopic liver injection model in nude mice further demonstrated that knockdown of RhoE enhanced local invasion of HCC cells in the livers, with more invasive tumor front and increased incidence of venous invasion. Mechanistically, stable knockdown of RhoE in HCC cells significantly enhanced the phosphorylation of myosin phosphatase, promoted assembly of stress fibers, and increased the formation of plasma membrane blebbings, all these changes and activities being associated with activation of the Rho/Rho-kinase (ROCK) pathway. Conclusion: RhoE was frequently down-regulated in predominantly HBV-associated HCCs and this down-regulation was associated with a more aggressive HCC phenotype. RhoE regulated the cytoskeleton remodeling and suppressed HCC motility and invasiveness by way of inhibiting the Rho/ROCK axis. (HEPATOLOGY 2013)

Hepatocellular carcinoma (HCC) is a common malignancy worldwide and the third leading cause of cancer-related mortality.1 Frequent intrahepatic and extrahepatic metastases are major factors contributing to the high mortality of HCC patients. Gain of invasive phenotype of tumor cells and modification of the tumor microenvironment are shown to be influential factors leading to cancer metastasis.

Rho guanosine triphosphatases (GTPases) are key regulators of the actomyosin cytoskeleton and therefore crucial in modulating tumor cell motility. Deregulation of Rho GTPase pathways has been reported to play an important role both in tumorigenesis and cancer metastasis of HCC.2, 3 RhoE, also known as Rnd3, belongs to an atypical Rnd subfamily of the Rho GTPases and antagonizes RhoA activity. The well-characterized functions of RhoE are to enhance the activity of p190RhoGAP, bind directly to Rho-kinase 1 (ROCK1), and inhibit its kinase activity.4, 5 RhoE has been shown to be involved in regulating many different aspects in tumor development, including cell transformation,6 cell proliferation and apoptosis,7-9 and cell motility.10-12 Nevertheless, the exact role of RhoE in cancers is controversial and seems to be cell origin-dependent. On the one hand, RhoE is overexpressed in desmoplastic medulloblastomas, nonsmall-cell lung cancer, and pancreatic cancer.13-15 On the other hand, it is found to be down-regulated in prostate and gastric cancers and sarcomas.9, 10, 16 Although RhoE has been reported to be down-regulated in HCC,17-19 the exact roles of how RhoE regulates HCC cell invasion have not been fully elucidated.

In this study we investigated the functional implications and molecular mechanisms of RhoE in HCC invasion. We found that RhoE was frequently down-regulated in human HCCs and its down-regulation was associated with shorter disease-free survival of HCC patients. Using both in vitro cell and in vivo nude mouse models, we provide evidence that RhoE suppressed HCC invasion and metastasis through antagonizing the Rho/ROCK pathway.


GTPase, guanosine triphosphatase; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; MYPT1, myosin phosphatase target subunit; PCR, polymerase chain reaction; ROCK1, Rho-kinase 1; ROCK2, Rho-kinase 2.

Materials and Methods

Patient Samples.

Seventy-one Chinese patients (57 men and 14 women; age ranging from 28 to 82 years) who had surgical resection at Queen Mary Hospital, the University of Hong Kong from 1991 to 2000, were randomly selected for study. Normal liver samples were collected from patients who had resection of liver metastasis from colorectal carcinoma. Fifty-five (77.5%) of the 71 HCC patients were serum hepatitis B surface antigen (HBsAg)-positive, whereas four (5.63%) of the 71 HCC patients were serum antihepatitis C virus (anti-HCV)-positive. One (0.01%) of the patients was positive for both HBsAg and anti-HCV. Their resected specimens were collected at the time of surgical resection, snap-frozen in liquid nitrogen, and kept at −80°C.

Cell Lines.

Human HCC cell lines, SMMC-7721 and BEL-7402, were obtained from Shanghai Institute of Cell Biology, Chinese Academy of Sciences. Human HCC cell line MHCC-97L was a gift from Fudan University (Dr. Z.Y. Tang) of Shanghai. The HCC cell lines SMMC-7721, BEL-7402, and MHCC-97L were maintained in Dulbecco's modified Eagle's medium (DMEM) with high glucose (Gibco-BRL, Grand Island, NY) supplemented with 10% fetal bovine serum.

Establishment of RhoE Stable Knockdown Cells and Inducible RhoE-Expressing cells.

Small-hairpin RNA (shRNA) targeting RhoE was ordered from Sigma (St. Louis, MO) with the sequence 5′-CCGGGATCCTAATCAGAACGTGAAACTCGAGTTTCACGTTCTGATTAGGATCTTTTTG-3′ in lentiviral plasmid. To package the virus, 2 × 105 293FT cells were seeded onto 35-mm 6-well plates 1 day before transfection. shRhoE plasmid was transfected into cells together with lentiviral packaging mix and Lipofectamine 2000 (Invitrogen, Grand Island, NY) according to the manufacturer's protocol. One day after transfection the virus was collected in the culture medium; 2 × 105 cells were seeded onto 35-mm 6-well plates 1 day before viral infection. Then 300 μL of virus-containing culture medium was added to the cells. After 72 hours of infection, infected cells were spread onto a 100-mm culture dish at 1:100 dilution. For selection of stable clones, cells were cultured in DMEM with high glucose with Puromycin (Sigma). Stable knockdown of RhoE was confirmed by way of western blot analysis.

Real-Time Quantitative Polymerase Chain Reaction (q-PCR).

q-PCR was performed in triplicate, using the Taqman Universal PCR Master Mix kit and the Applied Biosystems 7900HT Fast Real-Time PCR system (Applied Biosystems, Carlsbad, CA) according to manufacturer's instructions. Taqman probes for RhoE (4331182, Applied Biosystems) and hypoxanthine phosphoribosyltransferase (HPRT; 4453320, Applied Biosystems) were used. RhoE mRNA levels were normalized with HPRT. RhoE mRNA levels above the median among the 71 patients were considered as having high RhoE expression and those below were considered as having low RhoE expression.

Protein Preparation and Western Blot Analysis.

For western blotting, cells were lysed with sodium dodecyl sulfate (SDS) buffer (for detection of phosphor-MYPT and MYPT) or NET buffer containing 1% NP40 (for detection of RhoE), both with addition of 1× complete protease inhibitors (Roche Molecular Biochemicals, Indianapolis, IN). Proteins from human samples were extracted with RIPA buffer. Proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) for western blotting analysis. Commercially available primary antibodies against RhoE (for human samples) (Cell Signaling Biotechnology, Danvers, MA), phosphor-MYPT, MYPT, RhoE (for cell lines) (Millipore, Billerica, MA), α-tubulin and β-actin (Sigma) were used for protein analysis.

In Vitro Cell Migration and Invasion Assays.

Transwell assay was performed as described.20 Cell invasion assay was performed by self-coating Matrigel (BD Biosciences, Sparks, MD) on the upper surface of the Transwell chamber. The invaded cells at the lower surface of the chamber were fixed with 75% methanol and stained with crystal violet. Three randomly selected fields of the fixed cells were captured and the cells were counted. Three independent experiments were performed.

Immunofluorescence Microscopy and Scanning Electron Microscopy.

Cells were seeded onto coverslips and incubated overnight at 37°C in a CO2 incubator. For immunofluorescence study, cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) and then permeabilized with 0.05% Triton X-100 in PBS. Fixed cells were incubated with 1:5,000 fluorescein isothiocyanate-conjugated phalloidin (Sigma) or antibodies as indicated. Cell nuclei were stained with 4,6-diamidino-2-phenylindole (Calbiochem, San Diego, CA) and mounted with Vectashield antifade mountant (Vector Laboratories, Burlingame, CA). Images were captured using a fluorescence microscope connected to a charge-coupled device camera (Leica, Wetzlar, Germany) under 1,000× magnification. For scanning-electron microscopy, cells were first fixed with 2.5% glutaraldehyde and then were ethanol-dehydrated in a stepwise manner. The slides were critical-point-dried and mounted on silver paste. Images were scanned and captured under 2,700× magnification using a Stereoscan 440 scanning electron microscope (Oxford Instruments, Cambridge, UK).

Orthotopic Liver Injection in Nude Mice and Histopathological Analysis.

One × 106 MHCC-97L cells were mixed with 15 μL of 1× Matrigel and the mixture was then injected orthotopically into the left lobes of the livers of the nude mice. The animals were sacrificed 4 weeks after surgery and examined. Livers and lungs from the sacrificed mice were harvested and fixed in 10% formalin followed by 75% ethanol. The tissues were then paraffin-embedded. Five-μm-thick paraffin sections were cut and stained with hematoxylin and eosin for histological analysis. For ex vivo organ imaging, mice were injected with 100 mg/kg D-luciferin intraperitoneally 5 minutes before necropsy. The excised organs were imaged with the IVIS 100 Imaging System (Xenogen, Hopkinton, MA). All experiments were performed according to the Animals (Control of Experiments) Ordinance (Hong Kong) and the Institute's guidance on animal experimentation.


RhoE Was Frequently Down-regulated in Human HCCs and Underexpression Was Associated with Shorter Disease-Free Survival.

We examined the RhoE mRNA expression levels in 71 pairs of human HCCs and their corresponding nontumorous livers using real-time qPCR. RhoE was found to be frequently (83.1%, 59/71) and significantly down-regulated in the HCCs by 2- or more fold (P < 0.001, Mann-Whitney U test) (Fig. 1A). In representative cases, the RhoE protein levels by western blot analysis correlated with the transcript levels (Fig. 1B).

Figure 1.

RhoE was down-regulated in human HCC. (A) q-PCR analysis of 71 pairs of HCCs. HPRT was used as an endogenous control. T, tumor, NT, nontumorous liver, NL, normal liver. (P < 0.001, Mann-Whitney U test). (B) Western blot analysis of representative HCC cases with RhoE down-regulation. β-Actin was used as a loading control. (C) HCC patients with RhoE expression level below median exhibited significantly lower disease-free survival rate (P = 0.020, Mantel-Cox test). (D) There was no significant difference between the overall survival rates of HCC patients with higher and lower RhoE expression (P = 0.383, Mantel-Cox test).

The 1-, 3-, and 5-year disease-free survival rates of the patients with lower RhoE expression levels were 55.2%, 31.0%, and 17.2%, respectively, with a median of 16.5 months, whereas those of the patients with higher RhoE expression levels were 79.2%, 54.2%, and 48.8%, respectively, with a median of 53.4 months. A lower RhoE expression level was found to be associated with shorter disease-free survival (P = 0.020, Mantel-Cox test) (Fig. 1C). The 1-, 3-, and 5-year overall survival rates of the patients with lower RhoE expression levels were 85.7%, 65.7%, and 57.1%, respectively, with a median of 63.1 months, whereas those of the patients with higher RhoE expression levels were 75.0%, 65.2%, and 51.6%, respectively, with a median of 63.8 months. There was no significant difference between the groups of patients with lower and higher RhoE expression in terms of overall survival rates (P = 0.383, Mantel-Cox test) (Fig. 1D). As for the clinicopathological correlation, there was no significant association between the RhoE mRNA expression levels and the clinical and pathological parameters (Table 1).

Table 1. Clinicopathological Correlation of High RhoE Expression in Human HCC
 No. of cases with high RhoE expressionNo. of cases with low RhoE expressionP-value
Tumor size   
No. of tumor nodules   
Hepatitis B surface antigen   
Cellular differentiation by Edmondson grading   
Venous invasion   
Tumor microsatellite formation   
Direct liver invasion   
Tumor stage   

Down-regulation of RhoE Enhanced Cell Migration and Invasion of HCC Cells In Vitro.

To investigate the effects of RhoE on HCC cell motility and invasiveness, we stably knocked down the RhoE expression in both SMMC-7721 and BEL-7402 HCC cells with shRNA using the lentiviral approach. The knockdown was efficient and the protein expression level of RhoE was reduced by 82.9% to 85.5%, as shown by western blot analysis (Fig. 2A; Supporting Fig. S2A). RhoE knockdown cells showed no difference in their proliferation rates as compared to the nontarget control cells (P = 0.267 and 0.344 for SMMC-7721 and BEL-7402, respectively, Student's t test) (Supporting Fig. S1). At the same time, we observed that RhoE stable knockdown cells migrated through the transwell chamber significantly faster than the nontarget control cells (P = 0.002 and 0.029 for SMMC-7721 and BEL-7402, respectively, Student's t test) (Fig. 2B; Fig. S2B). Moreover, using a Matrigel-coated transwell chamber, we found that RhoE knockdown cells invaded through the matrix faster than their nontarget control counterparts (P = 0.001 and 0.026 for SMMC-7721 and BEL-7402, respectively, Student's t test) (Fig. 2C; Fig. S2C). Consistently, overexpressing RhoE in a hepatitis B virus (HBV)-positive cell line PLC/PRF/5 suppressed both cell migration and cell invasion (P = 0.022 and 0.001 for migration and invasion, respectively, Student's t test) (Fig. S3). These findings indicate that down-regulation of RhoE enhances HCC cell motility and invasiveness.

Figure 2.

Down-regulation of RhoE enhanced HCC cell migration and invasion. (A) RhoE expression in RhoE stable knockdown cells. Stable clones were generated from SMMC-7721 and BEL-7402 HCC cells. A high efficiency of RhoE knockdown was achieved in shRhoE-16 clones as compared to the nontarget control.(B) Transwell cell migration assay; 5 × 104 cells were seeded onto the upper chamber of the transwell and allowed to migrate for 12 hours. The number of migrated cells was significantly increased in shRhoE-16 stable knockdown cells compared with the nontarget control (P = 0.002 and 0.029 for SMMC-7721 and BEL-7402, respectively, Student's t test). (C) Cell invasion assay; 1 × 105 cells were seeded onto Matrigel-coated transwell chamber and allowed to invade for 24 hours. The number of invaded cells was significantly increased in shRhoE-16 stable knockdown cells compared with the nontarget control (P = 0.001 and 0.026 for SMMC-7721 and BEL-7402, respectively, Student's t test). Results were from three independent experiments.

Stable Knockdown of RhoE in HCC Cells Suppressed Tumor Invasion In Vivo in Nude Mice.

To assess the effects of RhoE on HCC cell invasion in vivo, RhoE stable knockdown MHCC-97L cells were injected orthotopically into the livers of nude mice. The HCC cell line MHCC-97L has been shown to be invasive in orthotopic liver implantation model.21 It was sequentially transfected with shRNA targeting RhoE and firefly luciferase reporter and the nontarget control cells were transfected with scrambled shRNA. The stable knockdown of RhoE was confirmed with western blotting (Fig. 3A). MHCC-97L shRhoE or nontarget control cells were mixed with Matrigel and injected into the left lobes of the livers of the nude mice. Four weeks after injection, the mice were sacrificed and their livers and lungs were examined with bioluminescence imaging and then histologically. Histological analysis revealed a significant effect of RhoE in local invasion of the HCC tumors in the livers. The tumors of the MHCC-97L shRhoE group more frequently had invasive growth fronts with irregular tumor borders, whereas those of the nontarget control group more often had expansive tumor growth fronts with more regular and less invasive borders (P < 0.001, Fisher's exact test) (Fig. 3B). In addition, venous invasion was observed mostly in the shRhoE group (P = 0.037, Student's t test) (Fig. 3C). The presence of tumor microsatellites was observed in both the shRhoE knockdown and nontarget control groups without statistical significant difference (P = 0.163, Student's t test) (Fig. S4A). Pulmonary metastasis was present in all mice of both nontarget control and shRhoE groups, as detected by bioluminescence and histological analysis (Fig. S4B,C). Taking these data together, stable knockdown of RhoE suppressed HCC invasion in vivo.

Figure 3.

Knockdown of RhoE in MHCC-97L enhanced HCC invasion in vivo. (A) RhoE expression was significantly reduced in shRhoE-16 MHCC-97L clones. (B) Tumors formed in livers from MHCC-97L shRhoE-16 clone had more invasive tumor growth fronts with irregular tumor margins, whereas tumors formed in the livers from the MHCC-97L nontarget control clone displayed more expansive growth fronts (hematoxylin and eosin sections) (P < 0.001, Fisher's exact test). Arrows mark the tumor fronts. (C) Multiple foci of venous invasion (arrows) were present more frequently in the livers of the MHCC-97L shRhoE-16 knockdown group than in the nontarget control group with statistical significance (P = 0.037, Student's t test).

Knockdown of RhoE Enhanced Phosphorylation of Myosin Phosphatase Target Subunit.

To investigate the underlying molecular mechanism of RhoE in regulating HCC cell motility and invasiveness, we examined the Rho/ROCK pathway in the RhoE stable knockdown HCC cells. Cytoskeleton remodeling or formation of stress fiber bundles is required to provide mechanical force for the contractile motion of cancer cells. It has previously been shown that ROCK can phosphorylate myosin phosphatase target subunit (MYPT1) at Thr 853 and this phosphorylation is essential for myosin II phosphorylation and therefore actomyosin contractility. With western blotting, we demonstrated that the phosphorylation level of MYPT1 was significantly increased in RhoE stable knockdown HCC cells as compared with their nontarget control (Fig. 4A; Fig. S5A). Consistently, this increase in MYPT1 phosphorylation was reversed by ROCK-specific inhibitor Y27632, suggesting that RhoE up-regulates MYPT1 phosphorylation through regulating ROCK.

Figure 4.

Knockdown of RhoE impaired actomyosin contractility through ROCK. (A) Knockdown of RhoE enhanced the phosphorylation level of MYPT1 at Thr 853. Western blot analysis showed that shRhoE-16 knockdown cells had enhanced MYPT1 phosphorylation (Thr 853) (arrow) as compared with the nontarget control, indicating deregulated actomyosin contractility. The increased phosphorylation was reversed by addition of ROCK inhibitor Y27632. (B) Addition of the ROCK inhibitor Y27632 suppressed the accelerated cell migration induced by knockdown of RhoE (P <0.001, Student's t test). The results were from three independent experiments.

ROCK-Specific Inhibitor Reversed the Enhanced Cell Migration with RhoE Knockdown.

To further confirm that RhoE regulated HCC cell motility through ROCK, we performed transwell cell migration assays using RhoE stable knockdown cells with or without the addition of Y27632. Our result showed that RhoE knockdown enhanced cell migration in SMMC-7721 cells. Furthermore, this enhancement was suppressed by Y27632 (Fig. 4B; Fig. S5B), supporting that RhoE regulates HCC cell motility through regulating ROCK activity.

RhoE Suppressed Stress Fiber Formation in HCC Cells in a Transient Manner.

Next, we investigated the impact of RhoE-mediated cytoskeleton remodeling by examining stress fiber formation in RhoE-overexpressing cells. Rho/ROCK pathway is a key regulatory player regulating cytoskeleton remodeling. As previously reported by other groups, the effect of RhoE on stress fiber formation may be transient and disappear within hours. Therefore, we used a doxycycline-inducible system to overexpress RhoE (Tet-On RhoE). We transfected SMMC-7721 and BEL-7402 HCC cells with this system using a lentiviral approach. With western blot analysis, we confirmed that RhoE protein was expressed only upon the addition of doxycycline in Tet-On-RhoE cells (Fig. 5A; Fig. S6A). Six hours after the addition of doxycycline, the Tet-On-RhoE cells were fixed and stained with phalloidin. We observed that RhoE-expressing HCC cells showed loss of stress fiber formation, whereas the vector control cells showed no disruption in the stress fiber networks (Fig. 5B,C; Fig. S6B). Sixteen hours after the addition of doxycycline, stress fibers reappeared in RhoE-expressing cells, indicating that the effects of RhoE on stress fiber formation were transient (data not shown).

Figure 5.

Overexpression of RhoE suppressed stress fiber formation in HCC cells. (A) Ectopic expression of Myc-tagged RhoE fusion protein in SMMC-7721 stable clones. One μg/mL doxycycline was added to the culture medium for the indicated lengths of time. Myc-RhoE fusion expression was confirmed with western blot analysis using anti-Myc antibody. (B) Expression of RhoE suppressed stress fiber formation. Myc-tagged RhoE protein was stained by anti-Myc antibody (green) and stress fibers (arrows) was stained with phalloidin (red). Tet-On RhoE cells with RhoE staining exhibited a loss of the stress fibers, whereas the vector control cells showed nondisrupted stress fiber networks. (C) Intensity plots of phalloidin staining of SMMC-7721 cells in (B) processed by ImageJ. The vector control showed distinct patterns (arrows) which represent stress fiber staining, whereas the Myc-tagged RhoE-expressing cells showed a diffused staining pattern, indicating loss of stress fibers.

Knockdown of RhoE Increased Plasma Membrane Blebbing.

To further investigate the effect of RhoE-mediated cytoskeleton reorganization on cellular changes in relation to migratory ability and invasiveness of HCC cells, we compared the cell morphology of stable nontarget control and RhoE knockdown BEL-7402 cells with scanning electron microscopy. We observed a significant augmentation of plasma membrane blebbing on the cell surfaces of RhoE stable knockdown cells, as compared with the nontarget control cells (Fig. 6A). Plasma membrane blebs are dynamic cytoskeleton-regulated protrusions and have been shown to be involved in cell movement and invasion.22 Several studies have also suggested that the presence of plasma membrane blebs is closely associated with the ameboid mode of cancer cell motility that requires activation of the Rho/ROCK pathway.23, 24 Our result further indicates that knockdown of RhoE may regulate cellular events related to cancer invasion through the Rho/ROCK pathway.

Figure 6.

Knockdown of RhoE augmented formation of plasma membrane blebbing (arrows) in both SMMC-7721 and BEL7402 cells (scanning electron microscopy, magnification 2,700×).


In this study we found that RhoE protein was frequently and significantly down-regulated in human HCCs. With real-time q-PCR, the RhoE mRNA was found to be frequently (83.1%) and significantly down-regulated in the 71 pairs of primary HCC cases. Very recently, a group from France has also reported that RhoE was significantly down-regulated in HCCs in their cohort of patients who were predominantly non-HBV related.19 Our finding added another layer of evidence showing that RhoE is also down-regulated in HCCs from patients with predominant HBV infection (77.5%). Taken together, the data suggest that down-regulation of RhoE is a common event in primary HCCs. Moreover, our results suggest that RhoE expression is deregulated in HCC transcriptionally, consistent with the notion that it may be under the regulation of microRNAs, p53, or HIF-1α,8, 12, 25 as reported by other groups. In addition, RhoE is also reported to be epigenetically regulated by histone deactylation in gastric cancer.16 Further study is required to elucidate the exact mechanism of RhoE down-regulation in HCC.

In addition to the down-regulation of RhoE in human HCCs, significantly, we found that down-regulation of RhoE correlated with shorter disease-free survival in HCC patients. In HCC patients, disease-free survival is closely related with tumor recurrence due to intrahepatic metastasis. Consistently, we found that knockdown of RhoE promoted cell motility and invasiveness in HCC cells in vitro. This is also in concordance with the finding by Grise et al.19 showing increased invasiveness of RhoE-knockdown HCC cells in vitro. Significantly, our animal model using orthotopic injection of MHCC-97L cells into the livers of nude mice confirmed that knockdown of RhoE was associated with a more locally invasive tumor phenotype in vivo. Interestingly, down-regulation of RhoE seemed to affect mainly the local invasion of HCC cells in the livers, as we only observed a more aggressive tumor front and increased incidence of venous invasion in our mouse model. On the other hand, there was no significant difference in the incidence of pulmonary metastasis with RhoE knockdown. This may be explained by the fact that after tumor cells have escaped from their primary site and invaded into the veins, they still need to be seeded in the lungs and be able to colonize. These steps require modification of the microenvironment in the lungs and/or formation of the premetastatic niche.26 Therefore, it is possible that, because RhoE may not be involved in regulating these steps, even if the shRhoE HCC cells could enter the bloodstream, they might not be able to form distant colonies in the lungs. With these results taken together, RhoE is closely involved in regulating HCC cell movement and invasion both in vitro and in vivo.

The Rho/ROCK pathway has been shown to be a key regulatory player that regulates cytoskeleton remodeling and its deregulation is related to enhanced motility and invasiveness of cancer cells.11, 27 The specific ROCK inhibitor Y27632 has been demonstrated to suppress HCC cell migration, invasiveness, and intrahepatic metastasis of HCC.28, 29 We previously reported that ROCK2 was overexpressed in HCC and ROCK2 overexpression correlated with more aggressive tumor behavior.30 However, this does not exclude the possibility that ROCK1 and its modulators participate in regulating HCC progression and metastasis. RhoE is an antagonist for ROCK1 but not ROCK2, and is shown to directly bind ROCK1 and inhibit its activity.4 Down-regulation of RhoE in HCC suggests that the activity of ROCK1 may be augmented and ROCK1 may play significant roles as ROCK2 in regulating HCC motility. Our data demonstrated that RhoE knockdown up-regulated the phosphorylation level of MYPT1, implying that ROCK activity was increased in these cells. Furthermore, the intense plasma membrane blebbing observed in our RhoE knockdown cells is supportive that ROCK was activated. Because RhoE only interacts with ROCK1 but not ROCK2,4 it is reasonable to argue that it was ROCK1 that was activated. However, as RhoE can also suppress RhoA by way of activating p190RhoGAP,31 it is also possible that down-regulation of RhoE can indirectly activate ROCK2 through the p190RhoGAP/RhoA pathway. Further experiments are needed to find out the relative importance of ROCK1 and ROCK2 in HCC invasion in order to shed light on the development of antimetastasis drugs that target ROCK1/2 for HCC patients.

Studies have shown contradictory results regarding the role of RhoE in cancers. In some cell types, RhoE was shown to be linked to cell proliferation and prosurvival effects.6, 7, 32 In melanoma, p53-mediated induction of Rnd3 inhibits cancer cell invasion and loss of p53 promotes cell invasion by way of increased Rho/ROCK signaling.12 Moreover, reduced expression of RhoE in human sarcoma cells increases their invasiveness and metastatic potential.10 On the other hand, RhoE has been shown to be required for collective cell migration and invasion.33 RhoE-mediated suppression of Rho/ROCK signaling is required for reducing actomyosin assembly at cell-cell contacts, which favors collective cell migration. Recently, the report by Grise et al.19 also echoes our results by showing that down-regulation of RhoE in HCC cells enhanced HCC cell invasiveness, but by down-regulating E-cadherin and inducing epithelial-mesenchymal-transition. Their patient cohort was predominantly non-HBV-related, whereas ours had predominant chronic HBV infection. As both of our studies showed that RhoE was down-regulated in HCC as compared with the corresponding nontumorous liver tissues, this suggests that down-regulation of RhoE may be a global event in HCC irrespective of etiology. They showed that down-regulation of RhoE in Hep3B cells decreased miR200b/c expression and increased ZEB2 expression, which subsequently led to E-cadherin suppression and epithelial-mesenchymal transition (EMT). Interestingly, although they suggested that RhoE knockdown cells should be using the amoeboid mode of migration, as they did not require the metalloproteinase, they also showed that the cell movement depended on Rac1 and CDC42, but not RhoA. This differs from the findings from other reports that the amoeboid mode of migration involves the Rho but not Rac pathway.24, 34 On the other hand, the findings from our study suggest that RhoE knockdown in SMMC-7721 and BEL-7402 cells regulates a different pathway that is Rho/ROCK-dependent. This is supported by both the in vitro and in vivo mouse model. Given that Hep3B is an HCC cell line established from American patients, whereas SMMC-7721, BEL-7402, and MHCC97L were established from Chinese patients, this suggests that down-regulation of RhoE may regulate cell migration and invasion through different pathways in HCC cells with different genetic backgrounds. In our study, when RhoE protein was depleted the Rho/ROCK signaling pathway was activated and promoted the ameboid movement, as in the case of squamous cell carcinoma and HCC.10, 19 With amoeboid movement, cancer cells migrate with a rounded morphology and in an integrin-independent manner.35 These cells exhibit weak cell-substrate adhesions and migrate through the extracellular matrix (ECM) by changing shape and squeeze into spaces between the ECM. Therefore, amoeboid migration does not require degradation of the ECM by metalloproteases.36 The force required for squeezing is mainly provided by actomyosin contractility. Therefore, active Rho/ROCK signaling is essential for amoeboid migration.37 Rounded amoeboid-migrating cells are associated with bleb-like protrusions at plasma membranes and these blebs are similar to those observed during apoptosis, which also requires constitutively active ROCK.22, 24 Our data have demonstrated that RhoE regulated actomyosin contractility by regulating the phosphorylation of MYPT1 and inhibition of ROCK by the ROCK-specific inhibitor Y27632 reversed the accelerated cell migration induced by knockdown of RhoE. On the other hand, overexpression of RhoE in HCC cells significantly suppressed stress fiber formation and knockdown of RhoE enhanced formation of bleb-like protrusions at the plasma membranes. Altogether, our data indicate that down-regulation of RhoE enhances HCC cell motility and invasiveness through regulating the Rho/ROCK pathway and adopting an amoeboid mode of migration.

In conclusion, we have provided evidence that RhoE is a potential metastatic suppressor in HCC and it suppresses HCC invasiveness by way of inhibiting the Rho/ROCK axis. Because RhoE has a diverse expression pattern in different cancer types, it implies that cancers with different cell origins require different levels of activation of Rho pathway. Our results have added knowledge of the complex RhoGTPase pathways in regulating HCC progression.