miR-93-3p inhibition suppresses clear cell renal cell carcinoma proliferation, metastasis and invasion

miRNA dysregulation is associated with many human diseases, including cancer. This study explored the effects of miR-93-3p on clear cell renal cell carcinoma (ccRCC). We found that miR-93-3p is upregulated an average of 38-fold in 138 ccRCC specimens compared to matched normal kidney tissues, which correlated with poor patient outcome. miR-93-3p inhibition reduced ccRCC cell growth, invasion, and migration in vitro and in a mouse xenograft model. A search of the TargetScan, miRanda, and PicTar databases revealed that miR-93-3p is predicted to regulate pigment epithelium-derived factor (PEDF). A direct PEDF-miR-93-3p interaction was confirmed via dual-luciferase reporter assays. Like miR-93-3p inhibition, PEDF overexpression induced cell apoptosis and inhibited migration and invasion. Additionally, co-transfection with PEDF siRNA reversed the effects of miR-93-3p inhibition in ccRCC cells. Thus, miR-93-3p is a likely ccRCC oncogene that acts by regulating PEDF. These results suggest that miR-93-3p may predict ccRCC patient clinical outcome and serve as a novel anti-ccRCC therapeutic target.


INTRODUCTION
Renal cell carcinoma is one of the 10 most common cancers in both men and women, and clear cell renal cell carcinoma (ccRCC) is the most frequent histological subtype. ccRCC represents approximately 70% of renal cell carcinomas. Even with surgical resection, 20-50% of patients diagnosed with localized tumors develop local recurrence or metastases within three years, and five-year survival in patients with stage IV disease is approximately 10% [1][2][3][4]. Current first line treatments for metastatic ccRCC are limited to the angiogenesis and tyrosine kinase inhibitors, pazopanib and sunitinib. Anti-angiogenic therapy is efficacious against metastatic ccRCC. However, the ability of anti-angiogenic drugs to delay tumor progression and extend patient survival is limited, due to either innate or acquired drug resistance [5][6][7]. Despite the application of multimodal therapies, including surgical excision and chemotherapy, ccRCC patient prognoses remain unfavorable. Thus, novel predictive markers and effective therapeutic strategies against this malignancy are urgently needed.
MicroRNAs (miRNAs) are a class of endogenous, small noncoding RNAs that regulate gene expression by directly degrading or inhibiting translation of target mRNAs through base pairing to partially complementary sites. Dysregulation of miRNA expression and activity is associated with a variety of human diseases, and many www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 47), pp: 82824-82834 Research Paper www.impactjournals.com/oncotarget miRNAs are potential ccRCC prognostic markers and therapeutic targets [3].
Pigment epithelium derived factor (PEDF) is a multifunctional serpin with antiangiogenic activity in tumor cells [8,9]. This study showed for the first time that miR-93-3p inhibitor induced ccRCC cell apoptosis, inhibited metastasis, and upregulated PEDF expression. We also found that miR-93-3p overexpression was associated with poor prognosis in a large ccRCC patient cohort from a single institution.

Anti-miR-93-3p suppressed ccRCC cell tumorigenesis and metastasis in vivo
786-O cells were implanted subcutaneously in the flanks of nude mice, which were then injected with either control or anti-miR-93-3p lentivirus. anti-miR-93-3p injection inhibited tumor growth and reduced tumor weights and sizes compared to the control ( Figure 6A-6C).
786-O cells transduced to express luciferase were injected into mouse tail veins, followed by control or anti-miR-93-3p lentivirus. Four weeks following tumor cell injection, mice were injected intraperitoneally with luciferin and imaged using a Xenogen IVIS imaging system to detect tumor cell metastases. Multiple tumors were present in all mice. anti-miR-93-3p suppressed 786-O cell metastasis and proliferation ( Figure 6D). These data indicated that miR-93-3p knockdown may serve as a potential anti-ccRCC therapeutic strategy.

DISCUSSION
This study presents the first evidence that miR-93-3p is overexpressed in ccRCC. miR-93-3p upregulation was correlated with poor prognosis and shorter overall survival time in a large ccRCC patient cohort. We found that miR-93-3p inhibition induced ccRCC cell apoptosis and decreased invasion and metastasis in vivo and in vitro. We also showed that miR-93-3p inhibits PEDF expression by binding the PEDF 3'-UTR, and miR-93-3p and PEDF levels are negatively correlated in ccRCC patients. The effects of miR-93-3p knockdown in ccRCC cells were abrogated by PEDF siRNA. Thus, the oncogenic activities of miR-93-3p in ccRCC are likely mediated by PEDF. Additionally, ccRCC tumor imaging in live mice demonstrated that miR-93-3p inhibition suppressed tumor growth and metastasis in vivo.
Many ectopically expressed miRNAs have been shown to promote carcinogenesis by regulating tumor cell proliferation, apoptosis, and metastasis. Inhibition of these miRNAs represents a new avenue for anti-cancer treatments [12][13][14]. Thus far, only two published studies directly reported on miR-93 in ccRCC, and both focused on miR-93-5p. Shi, et al. found that TGF-β induces RBL2 expression and arrests renal cancer cell growth by downregulating miR-93-5p, and indirectly showed that miR-93-5p suppression induces G1 phase arrest. Wulfken, et al. demonstrated that miR-93-5p is upregulated in renal cancer patient serum and tissues, but the miR-93-5p function was not well understood [15,16]. miRNAs also modulate renal carcinoma tumorigenesis and their expression patterns can distinguish between renal carcinoma subtypes [10, 17,18]. While miR-93 is upregulated in several solid tumors and is positively correlated with malignancy [19][20][21], its oncogenic mechanisms in ccRCC have not been addressed. The present study associated miR-93-3p with ccRCC patient survival via univariate and multivariate analyses, and assessed the biological functions of miR-93-3p in vitro. ccRCC cell transfection with an miR-93-3p expression vector did not affect cell viability or apoptosis, possibly due to the already-high viability and low apoptosis levels of these cells, but did decrease damage induced by H 2 O 2 treatment. Additionally, miR-93-3p inhibition suppressed ccRCC cell motility and invasion, and increased apoptosis. Our results demonstrate that miR-93-3p is an oncogene in ccRCC, and may act by regulating PEDF.
PEDF is a serpin with multiple biological functions, including anti-angiogenic, tumor growth inhibitory, cell apoptosis promoting, and anti-metastatic activities [9]. PEDF inhibits melanoma cell growth by inducing apoptotic cell death, and PEDF loss in melanoma cells is associated with metastasis and poor prognosis [22]. PEDF inhibits breast cancer growth and metastasis by downregulating fibronectin via laminin receptor/AKT/ ERK signaling [23]. PEDF also triggers lung cancer cell apoptosis by regulating p53 protein-driven Fas ligand and Fas protein cell surface translocation [24]. Mirochnik, et al. found that a short PEDF peptide effectively blocked renal carcinoma cell growth in vivo [25]. Jiang, et al. reported that most ccRCC patients are positive for PEDF, which is an independent favorable prognostic factor inversely correlated with grade and pT stage [26].
In conclusion, we presented the first evidence that miR-93-3p inhibition suppresses ccRCC cell proliferation, metastasis, and invasion, and promotes apoptosis by regulating PEDF. We confirmed that miR-93-3p overexpression is common in ccRCC. Our results suggest that miR-93-3p may predict ccRCC patient clinical outcome and serve as a novel anti-ccRCC therapeutic target.

Patients with ccRCC and their surgically excised specimens
Surgically excised normal and tumor specimens were collected from 138 clear cell renal cell carcinoma patients (matched normal kidney tissues) who underwent surgery in the Department of urinary surgery at the Forth Affiliated Hospital of Harbin Medical University in China between 2008 and 2014. Tissue samples were routinely processed for histological diagnosis in strict accordance with World Health Organization criteria. Normal kidney tissues were obtained from normal adjacent tissues away from tumor tissues or non-neoplastic kidney diseases and were histologically confirmed to be free of any pathological lesions. The follow-up data were available for 66 cases. The study was approved by the Institutional Review Board of Harbin Medical University, and the participants gave informed consent.

Cell culture
The

Cell viability assays
The cells were cultured in 96-well plates, and each well was seeded with 4×10 3 cells. After transfection with 50 nM miR-93-3p, NC, anti-miR-93-3p, the viability of the cancer cells was detected with 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT; Roche; Switzerland). 20 μl of MTT solution (5 mg/ml, Sigma, USA) was added to each well, and the mixtures were incubated for 4 h at 37°C. Then, 200 μl of dimethyl sulfoxide (DMSO) was added to the wells. The absorbance was measured using an ELISA plate reader at 490 nm. A high absorbance indicated a greater viability.

Apoptosis assay
The apoptosis ratio was analyzed using flow cytometry and TUNEL.
The cells were transfected for 48 h then collected and washed twice with cold PBS (phosphate buffered saline). The cells were resuspended in 500 μl of binding buffer at a concentration of 10 6 cells/ml and then mixed with 10 μl of Annexin V (BestBio, China) for 15 min in the dark at room temperature (RT). Then 5 μl of PI was added to the cells. After incubation for 5 min in the dark at RT, the samples were analyzed using a FACSAria flow cytometry (BD Biosciences, USA).
The TUNEL assay was performed according to the protocol of death detection kit (Roche, Switzerland). Briefly, the cells were fixed in 4% paraformaldehyde in PBS, washed three times with PBS, and washed with 3% H 2 O 2 in methanol for 10 min at RT. After incubating with 0.1% Triton X-100 in 0.1% sodium citrate for 2 min on ice, the cells were incubated with the TUNEL reaction mixture for 1 h and then with DAPI for 5 min in the dark at RT. Fluorescence microscope (Nikon, Japan) were used for data analysis.

Wound healing assay
Cells were seeded in six-well plates in culture medium. Cells were grown to 70% confluence, rinsed with phosphate-buffered saline (PBS), and then starved for 12 hours in serum-free medium. A sterile 200 μL pipette tip was used to create wounds. Furthermore, cells were transfected with NC, miR-93-3p and anti-miR-93-3p respectively. The migration of the cells across the wound line was assessed after 48 hours.

Migration and invasion assay
Cellular invasiveness was quantified using a modified Matrigel Boyden chamber assay, as previous described [27]. 1×10 5 cells seeded into upper chambers. The chambers were then inserted into transwell apparatus (Costar, USA). The upper chambers were coated with Matrigel (BD Biosciences, USA) when cell invasion assay was done. Medium with 10% FBS was added to the lower chamber. After 48 h, cells on the bottom of the inserts were fixed in 4% paraformaldehyde and stained with 0.05% crystal violet. Then cells that invaded into the lower surface were counted. Each experiment was repeated at least three times.

RNA extraction and quantitative real-time polymerase chain reaction
Total RNA was isolated from cultured cells or tissue sections (5 to 10 of 10 μm-thick) using a Trizol standard protocol (Invitrogen, USA). For formalin-fixed, paraffinembedded (FFPE) ccRCC samples, total RNA was extracted from 5 to 10 of 10 μm-thick tissue sections using the Ambion RecoverAll kit (Ambion, USA) according to the manufacturer's instructions. Quantitative real-time polymerase chain reaction (qRT-PCR) was performed in triplicate in the ABI 7500 fast real-time PCR System (Applied Biosystems, USA) and normalized with U6 and Actin endogenous control. Total RNA from Normal kidney tissues was used as a control. miR-93-3p, U6 levels and endogenous mRNA levels of PEDF and Actin were detected using SYBR Green PCR Master Mix kit in accordance with the manufacturer's instructions (Applied Biosystems, USA). U6 was used as an internal control for miR-93-3p calculation and Actin were used as an internal control for detection of PEDF mRNA. The expression of miR-93-3p and PEDF mRNA were calculated using the comparative Ct method. Relative expression intensity values were calculated as 2 -∆∆ Ct.

Luciferase assays
HEK293 cells were co-transfected on 24-well plates by Lipofectamine 2000 reagent (Invitrogen, USA) with 0.5 μg of constructed reporter plasmid and miR-93-3p or control miRNA at a final concentration of 50 nM. Luciferase assays were performed using the Dual-Luciferase Reporter Assay System (Promega, USA) according to the manufacturer's instructions.

Protein extraction and western blot
The proteins were extracted from human ccRCC cells and specimens. Lysate was separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis, and the gel was blotted onto polyvinylidene fluoride (PVDF) membrane (Millipore, USA). The membrane was blocked in 5% nonfatmilk, and then incubated with either rabbit anti-human PEDF (ab180711, Abcam, USA), Bcl-2 (sc-56015, Santa, USA), Bax (sc-20067, Santa, USA), MMP9 (sc-53630, Santa, USA), MMP2 (sc-13595, Santa, USA), or Actin (3700, Cell Signaling Technology, USA). After washing, the membrane was incubated with the fluorescence-conjugated anti-mouse or anti-rabbit IgG (Invitrogen, USA). The bound secondary antibody was quantified using the Odyssey v1.2 software (LI-COR, USA) by measuring the band intensity (area×optical density) for each group and then normalized with Actin. The final results are expressed as fold changes by normalizing the data to control values.

Immunohistochemistry
Paraffin-embedded sections of excised ccRCC specimens were immunostained for PEDF protein.
Staining was performed with the streptavidinbiotin peroxidase complex method according to the manufacturer's recommendation (Dako, Denmark). Rabbit anti-human PDEF primary antibody (Abcam, USA) was administered, followed by secondary goat anti-mouse IgG (Dako, Denmark). Negative controlswere performed throughout the entire immunohistochemistry procedure.

Tumorigenicity assays in nude mice
5×10 6 786-O cells suspension was subcutaneously injected into the flank of 5-week-old female athymic BALB/c nude mice (SLAC Laboratory Animal Company, China). Then, control lentivirus or anti-miR-93-3p lentivirus delivering approximately 2×10 7 transforming units of recombinant lentivirus were injected into mice once through the tail vein. Growth rates were determined by measuring tumor size over time. Tumor size was measured with calipers after the tumor cell injection every 7 days for a period of 4 weeks. Tumor volume was determined using the formula: volume = length × width 2 /2.
For experimental metastasis models, control and stable knockdown miR-93-3p cells (stable luciferasetransfected 786-O cells) were injected into the tail vein of nude mice. To monitor metastasis, tumors derived from stable luciferase-transfected 786-O cells were imaged to observe luciferase expression on day 28 after tumor cell injection. Briefly, the animals were anesthetized and then injected IP with luciferin at 150 mg/kg in a volume of 100 μL. Images were captured at a peak time of 15-20 min after injection using an IVIS-200 Imaging System (Xenogen Corporation, USA) and then processed using Living Image software. All animal procedures were approved by the Harbin Medical University Animal Committee.

Statistical analysis
Statistical analysis was performed with SPSS13.0 software. Student t test, ANOVA, or chi-square analysis was applied, where appropriate. Survival rates were estimated using the Kaplan-Meier method, and survival curves were compared using the log-rank test. Survival data were evaluated by using univariate and multivariate Cox regression analyses. A probability of <0.05 (*) or <0.001 (**) was considered significant.

Author contributions
W.X. and L.W. designed and supported the study; L.W., Z.A. and P.G. prepared figures and wrote the manuscript; X.Z., G.C., G.L., collected and analyzed clinical data; H.W., P.S., L.P. Y.G. and W.W. conducted most experiments; Z.W. and Y.B. performed experiments in vivo. All authors reviewed and revised the manuscript.