| BackgroundHepatocellular carcinoma(HCC) is one of the most common malignancies in liver cancer. Infection by hepatitis B and C viruses, exposure to aflatoxi B1 and liver fibrosis are the major risk factors during HCC development and progression. Surgical resection, liver transplantation, chemoradiotherapy are the common therapies for HCC currently. However, most HCC patients still have the symptoms of metastasis and recurrence after surgery, which leads to a poor prognosis.Accumulating evidence suggests that epithelial to mesenchymal transition(EMT), during which epithelial cells lose their polarity and become mobile mesenchymal cells, plays a crucial role in tumor metastasis. HCC cells undergo EMT triggered by various stimulating factors and acquire chemo-resistance, migration/invasion, and stemness. TGF?, as a multifunctional cytokines, could induce activation of Smad signaling and those non-Smad signaling pathways in EMT. Therefore, TGF? has been commonly accepted as the most potent inducer of EMT.Cancer stem cells(CSCs) or Tumor initiating cells(TICs) are a small population of cancer cells with stem cell features and the ability to seed heterogeneous tumors. CSCs could generate a large number of tumor cells through self-renewal and differentiation. In addition, CSCs are resistant to chemoradiotherapy and only a small number of CSCs could generate tumors in experimental animals. A number of researches have indicated that EMT is associated with acquisition of stemness. For example, ectopic expression of Twist or Snail transcription factors or TGF?1 stimulation in human mammary epithelial cells can induce EMT and acquire CD44high/CD24 low breast cancer cells. Similarly, breast cancer stem cells(CSCs) can also be generated from CD8+T cells-induced EMT in epithelial breast cancer cells.Abnormal expression of ribosomal proteins was found in most tumors. Ribosomal protein S5(RPS5), a eukaryotic 40 S small ribosomal subunit, is mainly responsible for protein synthesis. RPS5 is highly expressed in hepatocellular carcinoma, and downregulated in colorectal cancer progression. In recent years, RPS5 has been found to have extra-ribosomal functions. For example, RPS5 could inhibit murine erythroleukemia(MEL) cell differentiation. In our previous studies, RPS5 has been shown to prevent activation and EMT process of hepatic stellate cells(HSCs) and attenuate experimental liver fibrosis. These effects are associated with reducing phosphorylation of AKT at Ser473 and Thr308 sites as well as the subsequent dephosphorylation of downstream molecular GSK3? and P70S6 K. In addition, RPS5 is found to be an intercellular target for MASM. MASM exerts anti-fibrosis function through stabilizing RPS5 protein in the hepatic stellate cells and the liver of experimental animal models induced by DMN and BDL. However, there is still no report about the molecular mechanism of RPS5 in hepatoma cell mesenchymal transition and cancer stem cells generation. ObjectivesThe present study focused on the role of RPS5 in the EMT process of hepatoma cells and liver cancer stem cells generation. Considering RPS5 as a target protein of MASM, we will further evaluate effect of MASM on liver cancer stem cells. Results will clarify the relationship between RPS5 and hepatocellular carcinoma, and provide the possibility of RPS5 as a potential novel therapeutic target for liver cancer treatment. Methods 1. Role of RPS5 in hepatoma cell mesenchymal transition(1) RPS5 expression in TGF?1-induced EMT in hepatoma cellsWe used different concentrations of TGF?1 to induce EMT in hepatoma cells Huh7 and LM3. The morphological changes of cells were observed, and total proteins and RNA were extracted at indicated time points. Western blots and Real time PCR were used to detect the expression change of EMT markers(ZO-1, E-cadherin, N-cadherin and vimentin) and RPS5 at protein and mRNA levels.(2) The effect of RPS5 knockdown on EMT process in hepatoma cells(1) Huh7 cells were transfected withRPS5siRNA1, RPS5siRNA2, RPS5siRNA3 and ControlsiRNA, respectively. Total proteins and RNA were extracted after transfection 72 h. Western blots and Real time PCR were used to assess the efficacy of RPS5 knockdown and the expression change of EMT markers E-cadherin and vimentin.(2)To generate a lentivirus containing shRNA against the RPS5 gene, the shRNA design was based on the RPS5 siRNA screening. The RPS5 shRNA sequences or scramble normal control(NC) shRNA were cloned into the pLKO.1 lentiviral vector. The constructed plamid were further co-transfected into HEK-293 T cells with lentiviral packaging plasmids to generate a RPS5 shRNA expressing lentivirus(Lenti-shRPS5) or control vector shRNA expressing lentivirus(Lenti-shNC).Huh7 cells were treated with viral supernatant for 72 h and then treated with or without TGF?1 for another 48 h. Then total proteins and RNA were extracted. Real time PCR and Western blots were used to detect the expression of EMT markers and RPS5.In some experiments, Huh7 cells were seeded on coverslips and infected with Lenti-sh NC and Lenti-shRPS5. Immunofluoresence was used to detect the expression of EMT markers at protein levels.(3) The molecular mechanism of RPS5 downregulation induced EMTHuh7 were infected with Lenti-shRPS5 and Lenti-shNC for 72 h and then treated with or without TGF?1 for another 48 h. In parallel, Huh7 cells were treated with 2.5ng/ml TGF?1 for 48 h or 72 h. Cell proteins were extracted and the phosphorylation levels of PI3 K, AKT, GSK3?, ERK and Smad2 were assessed by western blots.(4) RPS5 overexpression reverses the RPS5 shRNA-induced EMTConstruct the control GFP expression vector and the RPS5 overexpression vector named RPS5-mutant with the sequence of RPS5 not be recognized by RPS5 shRNA, and package a control vector GFP expressing lentivirus(Lenti-GFP) and RPS5 expressing lentivirus(Lenti-RPS5-mutant). Huh7 cells were infected with Lenti-shNC and Lenti-shRPS5 for 24 h, and then infected with Lenti-GFP and Lenti-RPS5-mutant for another 96 h. Proteins was extracted. The expression levels of RPS5, EMT markers(E-cadherin and vimentin) and the EMT-related signaling proteins were detected by western blots.(5) The effect of RPS5 overexpression on the migration ability of hepatoma cellsConstruct the RPS5 overexpression vector with the same cDNA sequence of human RPS5 genes, and generate a RPS5 expressing lentivirus(Lenti-RPS5) and control vector GFP expressing lentivirus(Lenti-GFP). After Huh7 and LM3 cells were treated with TGF?1 for different time, or infected with Lenti-GFP and Lenti-RPS5 lentivirus for 72 h, cell migration ability was evaluated by transwell migration assays. 2. Role of RPS5 in the generation of HCC cancer stem cells(1)The effect of RPS5 overexpression on HCC cancer stem-like cells generation(1) Huh7 cells were infected with Lenti-GFP and Lenti-RPS5, then puromycin were added to select the stably expressing RPS5 cells(Huh7-RPS5) and control cells(Huh7- GFP). Western blots were used to confirm the high expression level of RPS5 protein. The percentage of EpCAM+ and CD133+ cells were detected by flow cytometry.(2) After Huh7-GFP and Huh7-RPS5 cells were cultured in serum-free stem cell conditioned culture medium for 7 days, the number of the primary sphere cells was counted and the sphere cells were photographed. A single cell suspension was obtained by trysinization of the primary sphere cells and cultured for another 7 days to yield second passages of cells and the number of the sphere cells was counted.(3) RNA was extracted from Huh7 cells transiently infected with Lenti-GFP and Lenti-RPS5 for 5 days or Huh7-GFP and Huh7-RPS5 stable cell lines. Real time PCR was used to detect the expression level of some genes relevant to stemness such as EpCAM, CD133, Sox2 and Oct3/4.(4)Huh7-GFP and Huh7-RPS5 stable cell lines were treated with different concentrations of doxorubicin(DOX) for 72 h. Cell proliferation was assessed by a Cell Counting Kit-8 to evaluate the sensitivity of cells to DOX.(2) RPS5 expression levels in 8 hepatoma cell linesProteins were extracted from non-tumorigenic, immortalized human hepatocytes LO2 and 8 hepatoma cell lines(MHCC-97 H, MHCC-97 L, Hep3 B, PLC/PRF/5, SMMC-7721, HepG2, HCCLM3 and Huh7). Western blots were used to detect the expression levels of RPS5 protein.(3) Comparisons of the stemness between two hepatoma cell lines with different RPS5 expression levelsLM3 and Huh7 with different expression levels of RPS5 were used in the following experiments.(1)The percentage of EpCAM+ and CD133+ cells in both cell lines were detected by flow cytometry.(2)After LM3 and Huh7 cells were cultured in serum-free stem cell conditioned culture medium for 7 days, the number of the primary sphere cells was counted and the spheres were photographed.(3)Real time PCR was used to detect the expression level of some genes relevant to stemness such as EpCAM, CD133, Oct3/4 and Sox2.(4)After Doxorubicin and 5-FU treatment at different concentrations for 72 h, cell proliferation was assessed by a Cell Counting Kit-8. 3. Effect of Matrine derivative MASM on HCC cancer stem cells(1) The effect of MASM on HCC CSCs in vitro(1)After Huh7 and Hep3 B sphere cells were treated with MASM(0, 2, 10 and 20 ?M) for 72 h, flow cytometry was used to detect the percentage of EpCAM+ and CD133+ cells, as well as the apoptic cells. Real time PCR was used to detect the expression of genes relevant to stemness(CD133, EpCAM, Sox2 and Oct3/4) and some genes related to hepatocytes such as ALB, CYP1A3 and G-6-P.(2)After Huh7 and Hep3 B sphere cells were treated with MASM(0, 2, 10 and 20 ??) for 7 days, the primary spheres were photographed and the number of the spheres was counted. The second and third passages were grown in the absence of drug. After 7 days of culture, the number of spheres was counted.(2) Effect of MASM on the growth of the xenograft tumorsHuman Huh7 cells were subcutaneously injected into the nude mice. Drugs were administrated after 24 h cells injection. The nude mice were administrated either the saline, MASM(10 mg/kg, orally by gavage daily), DOX(6 mg/kg, intraperitoneal injection, once a week), or MASM in combination with DOX for 21 days. Tumors length and width were monitored every three days and tumors volumes were calculated using the formula 1/2ŚLW2, where L represents the longest surface length(mm) and W indicates width(mm). The growth curve of xenograft tumors was made. Then tumor tissues were excised, photographed and weighted at the end of the experiment. A portion of the tumor tissues from MASM-treated nude mice and the control mice was dissociated enzymatically to obtain single-cell suspensions for spheroid formation assays. The remaining tumor tissues were used for Real time PCR analysis to detect the expression levels of CD133, EpCAM, Sox2, Oct3/4, ALB, CYP1A3 and G-6-P.(3) The studies of possible mechanisms of MASM on HCC CSCs(1) Hep3 B and Huh7 cells were treated with different concentrations of MASM(0, 2, 10 and 20 ?M) for 72 h. Western Blots were used to detect the expression levels of PI3K/AKT/mTOR/P70S6 K and GSK3?/?-catenin relevant signaling molecules.(2) After Hep3 B and Huh7 cells were treated with 10 ?M CHIR(GSK3? specific inhibitor), 10 ?M MASM and 10 ?M CHIR in combination with 10 ?M MASM for 72 h, the proteins were extracted. Western blots were used to detect the expression of GSK3?/?-catenin signaling molecules. Results 1. Role of RPS5 on hepatoma cell mesenchymal transition(1) RPS5 is downregulated in TGF?1-induced EMTAfter TGF?1 treatment for 48 h or 72 h, the morphology of Huh7 cells obviously changed from typical cuboidal cobblestone-like characteristic to an elongated and spindle-shaped mesenchymal cell phenotype. Western blots and Real time PCR results showed that the epithelial cells marker E-cadherin expression was decreased in TGF?1-treated Huh7 cells, but the expression level of the mesenchymal cells marker vimentin and N-cadherin was increased significantly(p<0.01). Meanwhile, RPS5 expression was obviously decreased at mRNA and protein levels(p<0.01).After TGF?1 treatment for 5 days, the morphology of LM3 cells obviously changed into an elongated and spindle-shaped mesenchymal cell phenotype. Meanwhile, the expression of vimentin was significantly increased at mRNA and protein levels accompanying with obviously decreased expression of ZO-1 and RPS5(p<0.05). Those suggested that RPS5 may play an important role in EMT.(2) RPS5 knockdown promotes EMT in hepatoma cells(1) RPS5 siRNA promotes EMT in Huh7 cellsAfter Huh7 cells were transfected with three RPS5 siRNA for 72 h, only RPS5 siRNA3 could significantly decrease expression of RPS5 at mRNA and protein levels(p<0.05). RPS5 knockdown decreased the expression of E-cadherin and increased N-cadherin and vimentin expression(p<0.01). These results indicate that decreased expression of RPS5 alone could induce the EMT process in HCC cells.(2)Lenti-shRPS5-mediated RPS5 knockdown induces EMT in Huh7 cellsSimilarly to the effect of RPS5 siRNA, Lenti-shRPS5-mediated RPS5 knockdown markedly decreased E-cadherin expression and increased vimentin expression at both mRNA and protein levels(p<0.05), which is independent of TGF?1. Immunofluoresence results further confirmed the effect of RPS5 knockdown on the expression of E-cadherin and vimentin.(3) The molecular mechanism of RPS5 downregulation-induced EMTIn accordance with the previous studies, TGF?1 treatment activated AKT, ERK and Smad2 signaling pathways in Huh7 cells as evidenced by the significantly increased phosphorylation level of AKT, ERK and Smad2. RPS5 knockdown markedly increased the phosphorylation levels of AKT, ERK and Smad2. These results suggest that RPS5 downregulation induced EMT associated with activation of AKT, ERK and Smad2 signaling pathways.(4) RPS5 overexpression reverses RPS5 shRNA-induced EMT and activates cell signaling pathways(1)RPS5 overexpression reverses RPS5 shRNA-induced EMTTo further confirm the cause-effect relationship of RPS5 loss and EMT, RPS5-silenced cells were infected with a lentivirus vector containing a mutated RPS5 nucleotide sequence that could not be targeted by the RPS5 shRNA or a lentiviral control vector. The results show that reduced E-cadherin expression and increased vimentin expression in RPS5-silenced cells were partially recovered by re-expression of mutant RPS5.(2)RPS5 overexpression inhibits the AKT and ERK signaling pathways activated by RPS5 shRNATo identify the specific pathways leading to EMT by RPS5 knockdown, we re-introduced the mutant-RPS5 into RPS5-silenced cells. RPS5 overexpression significantly reduced the phosphorylation of AKT(Ser473), mTOR(Ser2448) and ERK(Thr202/Tyr204), but not Smad2(Ser465/467). These results indicate that RPS5 knockdown-induced EMT was mediated through AKT and ERK pathways.(5) RPS5 overexpression inhibits the migration ability of hepatoma cellsTranswell migration experiments showed that the number of migrated cells in TGF?1-treated Huh7 and LM3 was significantly more than that of the control-treated cells, which was consistent with the previous reports. RPS5 overexpression in Huh7 and LM3 cells could significantly reduce the number of migrated cells. Thus, it is suggested that RPS5 overexpression could inhibit the migration ability of HCC cells, which may be related with the inhibitory effect of RPS5 on EMT. 2. Role of RPS5 on HCC CSC generation(1) RPS5 overexpression inhibits HCC cancer stem-like cells generation(1)RPS5 overexpression inhibits cancer stem-like cells generation in Huh7 cellsRPS5 overexpression in stably expressing RPS5 Huh7 cell line was confirmed by Western blots. Flow cytometry results showed that the percentage of CD133+ cells in Huh7-RPS5 cell line was less than that in Huh7-GFP cell line(70.90% versus 83.76%, p<0.05), while the percentage of EpCAM+ cells had no difference in both cell lines(96.51% versus 98.77%).These results suggest that RPS5 could inhibit HCC cancer stem-like cell generation.(2)RPS5 overexpression inhibits the sphere-forming ability and self-renewal ability of Huh7 cellsSphere formation assay showed that the number of the primary sphere forming cells in Huh7-RPS5 cell line was significantly less than that in Huh7-GFP cells(99±8 versus 70±4, p<0.05), which indicated that RPS5 overexpression inhibited the sphere-forming ability of Huh7 cells. Moreover, the number of the second passage of sphere cells in Huh7-RPS5 was still less than that in Huh7-GFP(50±2 versus 38±3, p<0.05), which indicate that RPS5 overexpression inhibited the self-renewal ability of Huh7 sphere cells.(3) RPS5 overexpression inhibits the expression of genes relavent to stemnessReal time PCR results showed that EpCAM and CD133 mRNA was obviously decreased in Huh7-RPS5 cell lines(p<0.05) and no alteration was observed in the expression of Sox2 and Oct3/4. When Huh7 cells were transiently infected with Lenti-RPS5 for 5 days, expression of EpCAM, Sox2 and Oct3/4 mRNA was significantly decreased(p<0.05), while expression of CD133 was not changed.(4)RPS5 overexpression reduces the chemoresistance of Huh7Drug sensitivity assays showed that RPS5 overexpression could slightly increase the sensitivity of Huh7 cells to Doxorubicin. The IC50 value of Doxorubicin to Huh7-GFP was 0.34 ?M, while that of Doxorubicin to Huh7-RPS5 was 0.23 ?M.(2) The expression levels of RPS5 in 8 hepatoma cell linesWestern blots results showed that RPS5 expression levels in LM3, PLC/PRF/5 and HepG2 cells were relatively high. However, the expression levels of RPS5 in Huh7, SMMC-7721, Hep3 B, MHCC-97 H and MHCC-97 L cells were relatively low, equal to that in LO2 cells.(3) Comparision of stemness features between Huh7 and LM3 hepatoma cellsWe chose Huh7 cells with relatively low level of RPS5 and LM3 cells with relatively high level of RPS5 to compare their stemness. Flow cytometry results showed that the percentage of CD133+ cells in LM3 was obviously smaller than that in Huh7(7.33% versus 81.11%, p<0.01), while the percentage of EpCAM+ cells in both cell lines has no significant difference(89.46% versus 96.26%). Moreover, the number of sphere-forming cells in LM3 cells was obviously less that in Huh7 cells(62±2 vs. 104±4, p<0.05). Real time PCR results showed that levels of EpCAM and CD133 mRNA was lower in LM3 than that in Huh7 cells(p<0.01), while there was no difference of Sox2 and Oct3/4 mRNA between Huh7 and LM3 cells. In addition, LM3 was more sensitive to DOX than Huh7. The IC50 values of DOX to LM3 and Huh7 were 0.12 ?M and 0.90 ?M(p<0.05), respectively. However, LM3 was also more sensitive to 5-FU than Huh7, with IC50 values of 5-FU to LM3 and Huh7 were 76.5 ?M and 454.5 ?M(p<0.05). 3. Matrine derivative MASM inhibits HCC cancer stem cells(1) MASM inhibits HCC CSCs in vitro(1) MASM reduces the percentage of EpCAM+/CD133+ cells in sphere cellsAfter Hep3 B and Huh7 sphere cells were treated with MASM at the concentrations of 10 ?M and 20 ?M for 72 h, the percentage of EpCAM+/CD133+ cells was markedly reduced. The percentage of EpCAM+/CD133+ cells in Hep3 B spheres was significantly reduced from 97.0% to 86.5% and 76.5%, while the percentage of EpCAM+/CD133+ cells in Huh7 spheres was decreased from 94.1% to 82.1% and 73.4%.(2) MASM induces apoptosis of sphere cellsMASM(10 ?M ? 20 ?M) treatment could induce more or less apoptosis of Hep3 B and Huh7 sphere cells. The percentage of apoptotic cells in Hep3 B spheres was increased from 12.0% to 21.1% and 27.4%, while the percentage of apoptotic cells in Huh7 spheres was increased from 10.4% to 13.5% and 17.8%.(3) MASM inhibits the formation and self-renewal of hepatoma sphere cellsThe number and the size of the primary sphere cells were significantly reduced after Hep3 B and Huh7 sphere cells were treated with MASM(10 ?M ? 20 ?M) for 7 days. These results indicate that MASM could inhibit the formation of hepatoma sphere cells. Furthermore, a significant decrease in the number of sphere-forming cells in subsequent two passages cultured without drug indicate that MASM reduced self-renewal capacity of these stem/progenitor cells.(4) MASM induces differentiation of HCC cancer stem cells into hepatocytesMASM(0-20 ?M) treatment for 72 h decreased the levels of EpCAM, CD133, Sox2 and Oct3/4 mRNA in a dose-dependent manner(p<0.05) and significantly increased the levels of ALB, CYP1A3 and G-6-P(p<0.05). These results indicate that MASM could induce differentiation of HCC CSCs into hepatocytes.(2) MASM inhibits the growth of xenograft tumors(1) MASM inhibits the growth of xenograft tumors in vivoIn vivo experiments showed that MASM(10 mg/kg) administration for 21 days could significantly inhibit the growth of Huh7 xenograft tumors, which was equal to the inhibitory effect of DOX(6 mg/kg). When administered in combination with DOX, MASM exerted more effectively inhibitory effect. The tumors were deprived, photographed and weighed at the end of the experiment. The results of the size and the weight of the tumors were in accordance with the abovementioned tumor growth curves results. Those results indicate that MASM could inhibit the growth of xenograft tumors in vivo.(2) MASM reduces the sphere-forming ability of cancer stem cells in vivoThe tumor tissues were trypsinized into single-cell suspension and cultured in serum-free stem cell conditioned culture medium for 7 days. The results showed that the number of sphere cells dissociated from xenograft tumors in MASM treatment group was significantly less than that in control group(26±5 versus 75±7, p<0.05), indicating that MASM could reduce HCC cancer stem cells in vitro.(3) MASM promotes the differentiation of HCC CSCs in tumor tissues into hepatocytesReal time PCR results showed that the expression level of EpCAM, CD133, Sox2 and Oct3/4 in the tumor tissues from MASM treatment group was significantly decreased(p<0.05), while the expression level of CYP1A3, G-6-P and ALB in tumor tissues from MASM treatment group was obviously higher than that from control group(p<0.05). These results are in accordance with those in vitro.(3) The mechanism of MASM inhibiting liver CSCs(1) MASM inhibits the PI3K/AKT signaling pathwayMASM treatment for 72 h significantly decreased the phosphorylation levels of PI3K(Tyr458), AKT(Ser473 and Thr308), mTOR(Ser2448) and P70S6K(Thr389) in Hep3 B and Huh7 cells dose-dependently.(2) MASM inhibits the ?-catenin signaling pathway?-catenin is phosphorylated at Ser33/Ser37/Thr41, but not at Ser675 site, by glycogen synthase kinase 3?(GSK3?), an AKT downstream molecule, for its degradation. Therefore, we detected whether MASM affected the phosphorylation of GSK3? and ?-catenin(Ser33/Ser37/Thr41), as well as the expression of ?-catenin and its target gene CyclinD1. The results showed that MASM treatment for 72 h resulted in the decreased phosphorylation levels of GSK3?(Ser9) in parallel with increased phosphorylation of ?-catenin(Ser33/Ser37/Thr41) and decreased expression of ?-catenin and CyclinD1. However, MASM rendered no evident influence on the phosphorylation of ?-catenin at Ser675 sites.We further used GSK3? inhibitor CHIR to confirm whether MASM could specifically inhibit ?-catenin signaling pathway. As expected, GSK3? inhibitor CHIR increased the phosphorylation level of GSK3?(Ser9), suppressed the phosphorylation of ?-catenin(Ser33/Ser37/Thr41) but not ?-catenin(Ser675), and increased the protein level of ?-catenin.When MASM cotreated with CHIR, the phosphorylation level of ?-catenin(Ser33/Ser37/Thr41) was lower than that of MASM treatment alone accompanied by higher level of ?-catenin protein expression. These results suggest that MASM-induced ?-catenin degradation was possibly mediated through activation of GSK3? because GSK3? phosphorylation at Ser9 results in its inhibition, thereby stabilizing ?-catenin. Conclusions1. TGF?1 induced EMT process in hepatoma cells accompanied by downregulation of RPS5.2. RPS5 downregulation induced EMT process through activation of AKT and ERK signaling pathways.3. RPS5 overexpression could inhibit the migration ability of hepatoma cells and HCC cancer stem-like cells generation.4. The anti-tumor effect of MASM was related with its inhibition of HCC CSCs self-renewal and the induction of differentiation into hepatocytes. Furthermore, we found the inactivation of PI3K/AKT/mTOR and GSK3?/?-catenin signaling pathways by MASM as the possible mechanisms for its efficacy. |
PDF Full Text Request