Low Expression of CFTR in Hepatocellular Carcinoma and Hepatocellular Carcinoma Cell and Its Correlation with Metastasis
The expression level of CFTR in HCC were predicted by TGCA, GEPIA and GTEx databases. According to the prediction of database, CFTR expression in HCC is significantly lower than that in normal adjacent tissues (Fig. 1A-C). The correlation of CFTR expression with metastasis and prognosis were predicted by TGCA, GEPIA and Kaplan-Meier plotter databases. According to the prediction of database, it was further found that CFTR was negatively correlated with the histological stage of HCC and the clinical prognosis (Disease-free survival, DFS; Recurrence-free survival, RFS) (Fig. 1D-F).
Then, HL7702 normal hepatocytes cells and HepG2, MHCC97L and MHCC97H HCC cells with different metastatic potential were selected to detect the expression of CFTR protein by Western Blot. The results showed that the expression level of CFTR in HepG2, MHCC97L and MHCC97H cells were lower than that in HL7702 cells, and its expression level was negatively correlated with the metastatic potential of hepatoma cells (MHCC97H lower than MHCC97L lower than HepG2) (Fig. 1G).
CFTR negatively regulates the proliferation of hepatocellular carcinoma
In order to further verify the correlation between CFTR and HCC proliferation and metastasis, HepG2 sh-CFTR cell line was constructed by using HepG2 cells with low metastatic potential and high CFTR expression, and MHCC97H CFTR cell line was constructed by using MHCC97H cells with high metastatic and low CFTR expression. After preliminary verification of CFTR overexpression and interference, the effects of CFTR on proliferation of HCC cells were detected by MTT methods. The MTT results showed that the proliferation level of HepG2 sh-CFTR group was significantly higher than that of HepG2 sh-Ctrl group, while that of MHCC97H CFTR group was significantly lower than that of MHCC97H Ctrl group. The inhibitory effect of CFTR on HCC was time-dependent, and the cell proliferation effect of 72h was the most significant (Fig. 2A).
PI staining and flow cytometry were used to detect the HCC cell cycle. The results showed that over-expression of CFTR significantly increased the proportion of S phase cells and decreased the proportion of G2 phase cells in MHCC97H cells, which inhibited the transformation of MHCC97H cells from S phase to G2 phase. After interfering with CFTR expression, the proportion of S phase cells in HepG2 cells decreased significantly, while the proportion of G2 phase cells increased significantly, which promoted the transformation of MHCC97H cells from S phase to G2 phase, but had no significant effect on the proportion of G1 phase cells (Fig. 2B).
At last, the expression level of related cell cycle proteins was detected by Western Blot. The results showed that overexpression of CFTR inhibited the expression of Cyclin B1 in MHCC97H cells, and interference with CFTR promoted the expression of Cyclin B1 in HepG2 cells, while CFTR had no significant effect on the expression of Cyclin D1 in both MHCC97H and HepG2 cells (Fig. 2C).
CFTR negatively regulates metastasis of hepatocellular carcinoma
Transwell was used to detect the effects of CFTR on invasion and migration of hepatocellular carcinoma. The results showed that overexpression of CFTR significantly inhibited the invasion of MHCC97H cells, while interference with CFTR significantly promoted the invasion of HepG2 cells. However, it showed no significant effects on migration of both MHCC97H and HepG2 cells (Fig. 2D).
Therefore, expression level of invasion-related cell adhesion proteins and metastasis-related proteins were tested by Western Blot. The results showed that overexpression of CFTR inhibited the expression of MMP2, MMP9, and N-cadherin in MHCC97H cells, and interference with CFTR promoted the expression of MMP2, MMP9, and N-cadherin in HepG2 cells. Meanwhile, overexpression of CFTR promoted the expression of E-cadherin in MHCC97H cells, and interference with CFTR inhibited the expression of E-cadherin in HepG2 cells. (Fig. 2E)
Further exploration of potential pathways of CFTR regulating proliferation and metastasis of hepatocellular carcinoma
In order to further explore the related mechanism of CFTR affecting the proliferation and metastasis of HCC, informatics methods were used to predict the potential mechanism. The differentially expressed genes of HCC were obtained through TCGA database, and at the same time, 20 direct (primary) related genes and 229 indirect (secondary) related genes of CFTR were obtained through String database. After intersecting the differential genes of HCC and indirect CFTR related genes, 11 genes were obtained. Two potential pathway genes, TERT and DNAJC6, were obtained by further screening with Genecards database and related researches (Fig. 3A and B).
To validate the predictive results, the protein expression levels of TERT, DNAJC6 and HSP90 (corresponding CFTR-related genes) in HCC cells were detected by Western Blot. The results showed that overexpression of CFTR inhibited the expression of TERT, DNAJC6, and HSP90 in MHCC97H cells, and interference with CFTR promoted the expression of them in HepG2 cells (Fig. 3C).
In order to further verify the mechanism of CFTR regulating of HCC, geldanamycin (GA), a double inhibitor of HSP90 and TERT, was selected. The results showed that the proliferation and invasion of HepG2 cells interfering with CFTR expression were significantly inhibited after GA administration by 100 nmol/L, 1µmol/L, and 10µmol/L (Fig. 3D and E). Compared with HepG2 sh-Ctrl group, the inhibitory effect of HepG2 sh-CFTR group was more significant. It is suggested that this inhibitory effect is closely related to the decrease of CFTR expression level (Fig. S1).
Negative regulation of CFTR on proliferation and metastasis of hepatocellular carcinoma in vivo
In vivo model of orthotopic HCC transplanted in nude mice was used to further verify the negative regulation of CFTR on proliferation and migration of HCC and its related mechanism. Firstly, four HCC cell lines (MHCC97H Ctrl, MHCC97H CFTR, HepG2 sh-Ctrl, HepG2 sh-CFTR) were transplanted into BALB/c nude mice in situ. After 6 weeks of tumor bearing, the liver and other tissues of mice were taken out, and the growth rate and metastasis rate of mice were counted, meanwhile, the tumor volume was measured and the tumor weight was weighed.
The results showed that MHCC97H CFTR group significantly reduced tumor weight, tumor volume and tumor metastasis rate in liver tissue of mice compared with MHCC97H Ctrl group. Compared with HepG2 sh-Ctrl group, the tumor weight, tumor volume and tumor metastasis rate in liver tissue of HepG2 sh-CFTR group were significantly increased, and most of the metastatic tissues were subcutaneous and intestinal metastasis (Fig. 4A-D).
To further validate the potential mechanism results in vitro, the related protein expressions were detected by either Western Blot or Immunohistochemistry. The results showed that the expression level of MMP2, MMP9, N-cadherin, and Cyclin B1 in MHCC97H CFTR group were lower than that in MHCC97H Ctrl group, while the expression levels of these proteins in HepG2 sh-CFTR group were higher than that in HepG2 sh-Ctrl group. The expression levels of E-cadherin in MHCC97H CFTR group were higher than that in MHCC97H Ctrl group, while the expression levels of E-cadherin in HepG2 sh-CFTR group were lower than that in HepG2 sh-Ctrl group (Fig. 4E). The potential pathway proteins were also verified, and the results showed that the expression level of TERT, DNAJC6 and HSP90 in MHCC97H CFTR group were all lower than that in MHCC97H Ctrl group, while the expression levels of them in HepG2 sh-CFTR group were higher than that in HepG2 sh-Ctrl group (Fig. 4F).
Prediction of drugs targeting CFTR to inhibit the growth and metastasis of hepatocellular carcinoma
Since CFTR seems to be one potential therapeutic target for HCC due to its significant regulatory effect on proliferation and metastasis of HCC. Drugs targeting CFTR might become potential drugs for the treatment of HCC. Therefore, potential drugs targeting CFTR were predicted through CLUE Command database. The results showed that 10 drugs (DMP-543, mesalazine, valaciclovir, otenzepad, parecoxib, nicorandil, cytarabine, tosufloxacin, midazolam, secnidazole) with potential effects were obtained from CLUE Command database, and all these 10 compounds were positively correlated with CFTR, which seemed to be potential activators of CFTR (Fig. 5A). Then, autodock was used for further prediction. The results showed that, all these 10 compounds could bind to CFTR with different binding energy, among which, parecoxib showed the least binding energy (Fig. 5B and C) (Fig. S2). Considering the binding energy, binding site and the type of bond, parecoxib might be supposed to a potential therapeutic for HCC by targeting CFTR.