RNF7 knockdown inhibits prostate cancer tumorigenesis by inactivation of ERK1/2 pathway

Development of castration resistance is a key contributor to mortality in patients with prostate cancer. High expression of RING finger protein 7 (RNF7) in cancer cells is known to play a key role in tumor progression. However, the role of RNF7 in prostate cancer progression is not well elucidated. In this study, we silenced RNF7 by shRNA interference in two castration resistant prostate cancer (CRPC) cell lines, DU145 and PC3. RNF7 knockdown attenuated proliferation and enhanced sensitivity of prostate cancer cells to cisplatin treatment. Invasive property of DU145 and PC3 cells was also attenuated by RNF7 silencing. The underlying mechanisms appear to be associated with accumulation of tumor suppressive proteins p21, p27 and NOXA, while inactivation of ERK1/2 by RNF7 knockdown. We demonstrated that RNF7 knockdown induced growth suppression of prostate cancer cells and inactivated ERK1/2 pathway, which suggested RNF7 might be a potential novel therapeutic target for CRPC.

The role of RNF7 in prostate cancer, especially in castration resistant prostate cancer (CRPC), is not yet clear. In this study, we found that knockdown of RNF7 in two CRPC cell lines, DU145 and PC3, enhanced the sensitivity of these cells to cisplatin treatment. The underlying mechanisms were likely associated with increased cell apoptosis and inhibited ERK1/2 activity.

Results
RNF7 was efficiently silenced by RNA interference. DU145 and PC3 cell lines were transfected with shRNF7 retroviruse (shRNF7-1 or shRNF7-2) or with negative control retrovirus (shCON). Protein and mRNA levels of RNF7 were determined by Western blot and qRT-PCR, respectively. RNF7 expression was significantly decreased after interference with shRNF7-1 or shRNF7-2 in both DU145 and PC3 cells (Fig. 1a).

RNF7 Knockdown inhibited prostate cancer cell growth.
Rapid cell division is one of the properties of cancer cells. To determine the effect of RNF7 knockdown on prostate cancer cell proliferation, we seeded 1 × 10 5 cells in 12 well plates and cultured them for 7 days. Cell numbers were recorded every day. Cell proliferation in the shRNF7-2 interfered group was significantly attenuated than that in the control (shCON) group from the third day onwards (P < 0.01) ( Fig. 2a and b). On the 7 th day, the number of DU145 cells decreased to 78.4% by RNF7 interference (Fig. 2a), and 69.6% of PC3 cells (Fig. 2b). The results indicated significant attenuation of DU145 and PC3 cell proliferation after RNF7 knockdown.

RNF7 Knockdown induced cell cycle arrest in prostate cancer cells.
To investigate the underling mechanism of growth suppression caused by RNF7 knockdown, cell cycle distribution of shRNF7 and shCON cells were measured by PI staining on flow cytometer. The mean percentage of negative control cells in G2 phase, in which RNF7 were normally expressed, were about 25.9% and 21.3% ( Fig. 3a and b), respectively. However, mean percentage of G2 phase in RNF7 silenced DU145 and PC3 cells were about 41.2% and 38.7% ( Fig. 3a and b), respectively. These results showed that RNF7 knockdown induced significantly G2 phase arrest in both DU145 and PC3 cells, P < 0.001. RNF7 knockdown accelerated prostate cancer cell apoptosis and cell death upon administration of chemotherapy. To investigate whether RNF7 knockdown had synergetic effect with chemotherapy on prostate cancer treatment, DU145 and PC3 cells transfected with shRNF7-2 or shCON were treated with cisplatin. Apoptosis cells were positively stained with Annexin V or double positively stained with Annexin V and PI, while dead cells were only PI positive (Fig. 4a). In DU145 cells, cisplatin combined with shRNF7-2 interference induced significantly higher apoptosis or necrosis indicated by Annexin V and/or PI staining positive (25.7%) as compared to that induced by cisplatin alone (15.3%) (P < 0.0001) (Fig. 4b). Also, cisplatin induced significantly higher cell apoptosis or necrosis (26.3%) in shRNF7-2 silenced PC3 cells compared to shCON control group (19.4%), P < 0.0001 (Fig. 4b). So, RNF7 silencing enhanced the sensitivity of prostate cancer cell lines to cisplatin.

RNF7 Knockdown inhibited prostate cancer cell invasion. Cell invasion plays a critical role in cancer
relapse and causes death. The role of RNF7 in DU145 and PC3 cell invasion were analyzed by invasion assay. Our results showed that the invasive ability reduced significantly after RNF7 silencing in DU145 (Fig. 5a) and PC3 ( Fig. 5b) cells, P < 0.0001.

RNF7 Knockdown inhibited prostate cancer cell colonization and tumorigenesis.
Colony formation indicates the ability of tumorigenesis. Clonogenic survival assay was performed to detect the role of RNF7 in prostate cancer cell colony. As inspected, DU145 ( Fig. 6a left panel) and PC3 (Fig. 6b left panel) cells exhibited higher colony forming ability. However, this ability was significantly compromised by RNF7 interference, P < 0.0001 ( Fig. 6a and b, middle and right panels). The size and weight of tumors developed from DU145 ( Fig. 6c) or PC3 (Fig. 6d) cells in nude mice were significantly decreased by RNF7 silencing, P < 0.0001. These data demonstrated that RNF7 knockdown inhibited tumorigenesis in prostate cancer cells.

RNF7 Knockdown induced accumulation of p21, p27 and NOXA, and inhibition of ERK1/2 activity.
In both DU145 and PC3 cells, the expression of pro-apoptosis protein NOXA and tumor suppressor proteins p21 and p27 was highly up-regulated by RNF7 silencing (Fig. 7a). ERK plays a critical role in tumor cell survival and proliferation. And phosphorylation indicates activation of ERK. To investigate the effects of RNF7 knockdown on ERK activation, prostate cancer cells were activated by epidermal growth factor (EGF) for 5 to 60 minutes and ERK phosphorylation was detected by Western blot. Our study showed that it was the activity not the expression of ERK that was inhibited by RNF7 knockdown in DU145 and PC3 cells (Fig. 7b). These data demonstrated that RNF7 knockdown up-regulated expression of pro-apoptosis and/or tumor suppressor proteins, while inhibited the activity of ERK pathway.
To verify that proliferation suppression was associated with accumulation of p21, p27, and NOXA in RNF7 silenced prostate cancer cell, p21, p27, or NOXA was silenced by siRNA in RNF7 knockdown DU145 and PC3 cells. Four days after siRNA interference, cell proliferation was measured by MTT assay. Our results showed that cell proliferation increased significantly by silencing of p21, p27, or NOXA in both RNF7 knockdown DU145 Figure 2. RNF7 knockdown inhibited cell proliferation. Cell numbers of DU145 (a) and PC3 (b) transfected with shCON or shRNF7-2 were monitored for 7 days. Data were mean ± SD. **P < 0.01; ***P < 0.001.
( Fig. 8a) and PC3 cells (Fig. 8b). These results showed that the suppressive effect of RNF7 knockdown on cell growth was likely attributable to accumulation of P21, P27 and NOXA.  In conclusion, our study demonstrated that RNF7 knockdown could suppress proliferation and inhibit tumor formation of prostate cancer cells. The underlying mechanisms were associated with accumulation of tumor suppressive proteins p21, p27 and NOXA, and inactivation of ERK1/2 pathway.

Discussion
Prostate cancer is one of the leading causes of death in males worldwide. While most prostate cancer patients respond to ADT initially, a vast proportion of them will eventually relapse due to castration resistance [2][3][4] . Hence, development of treatment regimens that reverse castration resistance is a key priority in CRPC research.
Chemotherapy is one of the main therapeutic modalities for CRPC. It was reported that docetaxel could prolong survival time 2 to 3 months 18 . Other chemotherapeutic agents used for improving treatment efficacy either singly or in combination with docetaxel include, cisplatin 19,20 , estramustine 21 , 153Sm-lexidronam 22 , prednisolone 23 , mitoxantrone 24,25 , epirubicin 26 , bortezomib 27 , zoledronic acid 28 , capecitabine 29 , and vinorelbine 30 . Other regimens were used or studied in clinical trials include, combination of estramustine phosphate, ifosfamide and cisplatin 20 ; prednisone plus cabazitaxel and prednisone plus mitoxantrone 25 ; epirubicin and cisplatin 31 ; and cisplatin plus prednisone 32 . Although these regimens enhanced the 50% survival time up to 15-20 months, more efforts are needed for improving the treatment efficacies. It was reported that cisplatin induced reactive oxygen species (ROS) in prostate cancer cells. And the production of ROS in hormone-sensitive LNCap cells was significantly higher than that of CRPC DU145 and PC3 cells 33 . Increasing the ROS production in CRPC cells may improve efficiency of chemotherapy especially cisplatin treatment.
RNF7 is highly expressed in various human cancers, such as lung, liver and stomach cancers 15 . Also we found high expression level of RNF7 in prostate DU145 and PC3 cells. RNF7 was proved to be an antioxidant by Yi Sun and his colleagues 5 . Then we hypothesized that silencing of RNF7 might have a synergistic effect with cisplatin chemotherapy. So we sought to investigate the effect of RNF7 knockdown on sensitivity of DU145 and PC3 cells to cisplatin treatment. Retrovirus based shRNA were used to silence RNF7 expression. The knockdown efficiency was comparable to that reported in previous studies 7,8 . RNF7 knockdown significantly inhibited DU145 and PC3 cell proliferation starting from the third day, which may be associated with RNF7 knockdown induced cell cycle arrest in G2 phase. Consistent with our study, Tan et al. recently showed that RNF7 knockdown by lentivirus-based siRNA significantly decreased prostate cancer cell proliferation 34 . We also investigated the effect of RNF7 knockdown on CRPC cell sensitivity to cisplatin. RNF7 knockdown combined with cisplatin chemotherapy induced significantly higher percentage of Annexin V positively stained apoptosis cells as compared to that observed with cisplatin treatment only. The underlying mechanism might be associated with increased expression of pro-apoptosis protein NOXA in RNF7 silenced cells. We further investigated the effect of RNF7 silencing on cancer cell invasion. RNF7 knockdown significantly inhibited invasive properties of DU145 and PC3 cells. Consistent with our study, Tan et al. also showed that RNF7 knockdown significantly inhibited prostate cancer cell migration and clone formation in vitro 34 . Based on the above results, we hypothesized that RNF7 silencing might counteract CRPC tumorigenesis. Our study showed that clonogenic forming ability of DU145 and PC3 in vitro, and its tumor formation ability in nude mice were significantly decreased by RNF7 interference. These results were consistent with increased expression of tumor suppressor proteins p21 and p27. Similar findings have been reported from a previous study on lung cancer cells 8 . We also sought to figure out the potential signal  shRNA interfered cells (shRNF7-2 group) were significantly lower than those formed by negative controls (shCON group). DU145 (c) or PC3 cells (d) (both shNRF7 interfered and shCON negative controls) were implanted subcutaneously into the back of nude mice. Each group contained 6 nude mice. Mice were sacrificed six weeks later. Xenografts were collected and their weights were measured. Representative photos and statistical data were showed. Data were mean ± SD. ***P < 0.001. transduction pathway involved in RNF7 knockdown mediated cell growth inhibition. While Tan et al. showed that RNF7 knockdown decreased AKT activation 34 , we observed inactivation of ERK1/2 activity by RNF7 silencing. Multiple signaling pathways might be involved in the inhibition of cell growth and tumorigenesis mediated by RNF7 knockdown. To address further whether ERK1/2 inactivation is associated with inhibition of RAS activity, we detected RAS activity of shRNF7 and shCON cells, and we did not observe RAS inactivation caused by RNF7 interference (data not shown), indicating ERK1/2 inactivation by RNF7 knockdown might not due to RAS inactivation by potential NF1 accumulation. More investigations are needed to uncover the underlying mechanisms. To our knowledge, this is the first study to demonstrate that ERK1/2 activation is regulated by RNF7 in cancer cells.
In conclusion, our study demonstrated that RNF7 knockdown inhibited prostate cancer cell proliferation and tumorigenesis, suggesting that RNF7 might be a promising target for CRPC treatment. The underlying mechanisms might be associated with accumulation of tumor suppressive proteins p21, p27 and NOXA, and inactivation of ERK1/2 activity.

Packaging of Retroviruses in HEK293T cells and transfection of prostate cancer cell lines.
Recombinant LMP vectors containing shRNF7-1 or shRNF7-2 combined with pCL10A1 retrovirus packaging vector were transfected into HEK293T cells using lipofectamine 2000 (Life Technologies, Carlsbad, CA, USA), according to manufacturer's instructions. After 48 h incubation, retroviruses containing shRNF7-1 or shRNF7-2 were collected from culture supernatant. pCL10A1 retrovirus and LMP vector without shRNF7 sequences served as the negative control (shCON).
Prostate cancer cells were transfected with retroviruses in the presence of 10 μ g/ml polybrene and selected by 1 μ g/ml puromycin. Total RNA of prostate cancer cells was extracted using Trizol reagent (Invitrogen, Madison, WI, USA). To determine the efficiency of RNA interference, RNF7 mRNA levels were assessed by quantitative RT-PCR (qRT-PCR), and RNF7 protein levels were assessed by Western blot analysis.
Cell proliferation assay. DU145 and PC3cell lines transfected with or without RNF7 shRNA were seeded into 12 well plates at a concentration of 2 × 10 5 cells/ml with a total volume of 0.5 ml/well. Cell numbers were recorded every day using Bio-Rad cell counter (Bio-Rad Laboratories, Hercules, CA, USA). Growth curves of cells (number of cells versus time) were plotted for up to seven days. RNF7 knockdown DU145 and PC3cells interfered by siRNA targeting p21 (SASI_Hs01_00025255, sigma), p27 (SASI_Hs01_00113637) or NOXA (SASI_ Hs01_00136187) were seeded to 96 wells with 2 × 10 4 /ml in a total volumn of 200 μ l. Cell viability was detected by Methylthiazolyldiphenyl-tetrazolium bromide (MTT, M5655, sigma) after 4 days culturing.
Cell cycle assay. Both RNF7 knockdown and the negative control DU145 and PC3 cells were harvested and fixed by 70% ethanol at 4 °C for 24 h. Cells were incubated with 50 μ g/ml propidium iodide (PI) containing 400 U/ml RNase A for 30 min at room temperature in the dark. Cell cycle distribution was measured by FACSCalibur flow cytometer.
Apoptosis assay. DU145 and PC3 cells transfected with or without RNF7 shRNA were seeded into 24-well plates in 0.5 ml culture medium (2 × 10 5 cells/ml). Cells were treated with or without 10 μ g/ml cisplatin (P4394, Sigma) for 24 h. Cells were harvested by centrifugation and stained with Annexin V and 1 μ l propidium iodide (PI) (V13242, ThemoFisher Scientific) according to the manufacturer's instruction. Cells were detected by FACSCalibur C6 (BD Biosciences, San Jose, CA, USA) and analyzed using FlowJo software (FlowJo LLC, Ashland, OR, USA).

Cell invasion assay.
Cell invasion was assessed using 24-well Boyden chamber plates which contained polycarbonate membrane filter inserts with 8-μ m pore size (Costar Group, Washington, DC, USA). Diluted Matrigel (BD Biosciences) was added to the interior of the transwell inserts for mimicking of basement membrane. The upper chambers were seeded with 1 × 10 5 cells and the lower chambers were filled with 1 ml DMEM containing 10% FBS, and cells were cultured for 24 h. Subsequent to removal of non-migrating cells present on the upper chamber surface, invaded cells at the bottom of the membrane were fixed with methanol and stained with crystal violet. Invaded cell numbers were recorded under microscope.
Clonogenic survival assay. Ten thousand prostate cancer cells transfected with shRNF7 retrovirus or negative controls were resuspended in DMEM medium containing 10% FBS and 0.3% agarose. Cells were plated onto a solidified bottom layer in DMEM medium containing 10% FBS and 0.5% agarose in 60-mm culture dishes. The colonies that formed at day14 were recorded following fixation with methanol, and stained with 0.05% methylene blue.
Statistical analysis. Data were presented as mean ± Standard Deviation (SD). All experiments were repeated at least three times independently with three or more replicates. Statistical differences were determined by GraphPad Prism 5.0 software (GraphPad Software Inc., CA, USA). Gaussian distribution data were analyzed by two-tailed Student's t test, while non-Gaussian distribution data by Mann-Whitney of nonparametric test.