CAF-derived HGF promotes cell proliferation and drug resistance by up-regulating the c-Met/PI3K/Akt and GRP78 signalling in ovarian cancer cells

Abstract

The tumour microenvironment is a highly heterogeneous entity that plays crucial roles in cancer progression. As the most prominent stromal cell types, cancer-associated fibroblasts (CAFs) produce a variety of factors into the tumour microenvironment. In the present study, we firstly isolated CAFs from tumour tissues of the patients with ovarian cancer and demonstrated that the hepatocyte growth factor (HGF) was highly expressed in the supernatants of CAFs. CAF-derived HGF or human recombinant HGF promoted cell proliferation in human ovarian cell lines SKOV3 and HO-8910 cells. Western blotting analysis also showed that CAF-derived HGF or recombinant HGF activated c-Met/phosphoinositide 3-kinase (PI3K)/Akt and glucose-regulated protein 78 (GRP78) signalling pathways in ovarian cancer cells, and these effects could be abrogated by anti-HGF and c-Met inhibitor INCB28060. Moreover, HGF in CAF matrix attenuated paclitaxel (PAC)-caused inhibition of cell proliferation and increase in cell apoptosis through activating c-Met/PI3K/Akt and GRP78 pathways in SKOV3 and HO-8910 cells. The results in vitro were further validated in nude mice. These findings suggest that CAF-derived HGF plays crucial roles in cell proliferation and drug resistance in ovarian cancer cells.

Keywords: cancer-associated fibroblasts; cell proliferation; drug resistance; hepatocyte growth factor; ovarian cancer.

Figure 1. Isolation and characterization of primary CAFs and NFs

(a) mRNA levels of α-SMA and FAP were determined by Q-PCR by using specific primers. U6 served as an internal control. (b) Immunofluorescence staining of FAP and α-SMA in CAFs and NFs. Primary CAFs and NFs were isolated from ovarian tumour tissue and adjacent normal tissue, respectively and cultured on the coverslips. Cells were stained with anti-FAP or anti-α-SMA antibody. Cell nuclei were visualized by DAPI. (c) FAP and α-SMA were analysed by Western blotting. β-actin served as a loading control. Data were representative images of three independent experiments; *, P<0.05.

Figure 2. Effect of CAF matrix on cell proliferation

(a) The mRNA levels of HGF in CAFs, NFs, SKOV3 and HO-8910 cells were determined by Q-PCR by using specific primers. U6 served as an internal control for Q-PCR; **, P<0.01. (b) The concentration of HGF secreted by CAFs, NFs, SKOV3 and HO-8910 cells were determined by ELISA. The culture medium of CAFs, NFs, SKOV3 and HO-8910 cells were collected and analysed. (c) The total protein level of HGF was analysed by Western blotting. β-actin served as a loading control. Data were representative images of three independent experiments. (d) SKOV3 or HO-8910 cells were cultured with CM, NF matrix or CAF matrix. Cell viability was monitored by MTT assay. (e) SKOV3 or HO-8910 cells were treated with 10 μg/ml HGF. Cell viability was monitored by MTT assay. (f) SKOV3 or HO-8910 cells were treated CAF matrix with or without 30 μg/ml anti-HGF. Cell viability was monitored by MTT assay. Each bar is a mean ± S.D. of three independent experiments; *, P<0.05; **, P<0.01.

Figure 3. HGF in the CAF promoted the c-Met/PI3K/Akt activation and up-regulated GRP78 expression in SKOV3 and HO-8910 cells

(a) SKOV3 or HO-8910 cells were cultured in the condition as described above. Phosphorylated c-Met and total c-Met were analysed by Western blotting. β-actin served as a loading control. Data are representative images of three independent experiments. (b) SKOV3 or HO-8910 cells were cultured with CAF matrix with or without c-Met inhibitor INCB28060 (60 nM). CM served as a negative control. Cell viability was monitored by MTT assay. Each bar is a mean ± S.D. of three independent experiments; *, P<0.05. (c) Phosphorylated c-Met/PI3K/Akt, total Met/PI3K/Akt and GRP78 protein level were analysed by Western blotting. β-actin served as a loading control. Data are representative images of three independent experiments.

Figure 4. HGF in the CAF matrix decreases sensitivity to PAC in SKOV3 and HO-8910 cells

(a) SKOV3 or HO-8910 cells were treated with CM, combination of CM and 1.5 μM PAC (CM + PAC), the combination of PAC and CAF matrix (CAFs + PAC), the combination of PAC and HGF (HGF + PAC) or the combination of PAC, CAF matrix and c-Met inhibitor INCB28060 (CAFs + c-Met inhibitor + PAC) for 24h. Cell viability was monitored by MTT assay. (b) SKOV3 cells were treated with CM, CM + PAC, CAFs + PAC, HGF + PAC and CAFs + c-Met inhibitor + PAC for 24 h and then cells were stained with Annexin-V-FITC and propidium iodide. The apoptotic rate was determined by flow cytometry. (c) Quantificative analysis of SKOV3 apoptosis. The graph shows the summarized data. (d) HO-8910 cells were treated with CM, CM + PAC, CAFs + PAC, HGF + PAC and CAFs + c-Met inhibitor + PAC for 24 h and then cells were stained with Annexin-V-FITC and propidium iodide. The apoptotic rate was determined by flow cytometry. (e) Quantificative analysis of SKOV3 apoptosis. Each bar is a mean ± S.D. of three independent experiments; *, P<0.05; **, P<0.01; ***, P<0.001.

Figure 5. Western blotting analysis of c-Met/PI3K/Akt activation and GRP78 expression in SKOV3 and HO-8910 cells

(a) SKOV3 or HO-8910 cells were cultured in the condition as described above. Phosphorylated c-Met/PI3K/Akt, total c-Met/PI3K/Akt and GRP78 protein level were analysed by Western blotting. β-actin served as a loading control. Data are representative images of three independent experiments. (b) Quantificative analysis of Western blotting data. Each bar is mean ± S.D. of three independent experiments.

Figure 6. The validation in xenograft model

(a) Images of tumour size in different groups of mice on day 40 following tumour inoculation. (b) Growth curves of tumour in different groups of mice; *, P<0.05 compared with control group.