The normal cell line RWPE-1 and the cancerous cell lines 22Rv1, LNCaP, MDA-PCa-2b, DU145, PC-3 and HEK-293 were obtained from the American Type Culture Collection. The following media were used for the cells: RWPE-1, Keratinocyte serum free medium (K-SFM) + EGF + bovine pituitary extract (BPE) (Gibco cat. no. 17005-042); 22Rv1, RPMI (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 11875-093) + 10% FBS (R&D systems; cat. no. S12450); LNCaP, RPMI (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 11875-093) + 10% FBS (R&D Systems; cat. no. S12450); MDA-PCa-2b, F12K (ATCC; cat. no. 30-2004) + 20% FBS (R&D systems; cat. no. S12450) + 25 ng/ml cholera toxin (Sigma-Aldrich; Merck KGaA; cat. no. C8052) + 10 ng/ml mouse epidermal growth factor (Corning, Inc.; cat. no. 354010) + 0.005 mM phosphoethanolamine (Sigma-Aldrich; Merck KGaA; cat. no. P0503) + 100 pg/ml hydrocortisone (Sigma-Aldrich; Merck KGaA; cat. no. H0135) + 45 nM sodium selenite (Sigma-Aldrich; Merck KGaA; cat. no. 9133) + 0.005 mg/ml human recombinant insulin (Thermo Fisher Scientific, Inc.; cat. no. 12585-014); DU145, MEM (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 11095-080) + 10% FBS (R&D systems, Inc.; cat. no. S12450); PC-3, F12 (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 11765-054) + 10% FBS (R&D systems, Inc.; cat. no. S12450); HEK-293, MEM (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 11095-080) + 10% FBS (R&D Systems, Inc.; cat. no. S12450). All cells were cultured at 37˚C in a humidified atmosphere containing 5% CO₂. A short tandem repeat analysis with GenePrint10 was performed for regular authentication and the cells were regularly tested for mycoplasma contamination.

Total RNA was isolated using TRIzol (Thermo Fisher Scientific, Inc.; cat. no. 15596026) according to the manufacturer's protocol. For RT-qPCR of miRNA, 10 ng total RNA was reverse transcribed using the TaqMan MicroRNA Reverse Transcription kit (Thermo Fisher Scientific, Inc.; cat. no. 4366596) according to the manufacturer's protocol. The expression levels of miR-769-5p were assessed in triplicate using TaqMan probes for miR-769-5p (Thermo Fisher Scientific, Inc.; cat. no. 4427975; assay ID 001998) and the internal standard reference U6 small nuclear RNA (Thermo Fisher Scientific, Inc.; cat. no. 4427975; assay ID 001973) with TaqMan Universal Master Mix II (not containing uracil-DNA glycosylase; Thermo Fisher Scientific, Inc.; cat. no. 4440041).

For RT-qPCR analysis of mRNAs, 1 µg total RNA was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4374967) according to the manufacturer's instructions. Expression levels of genes were assessed in triplicate using TaqMan probes for 1-acyl-sn-glycerol-3-phosphate acyltransferase α (Thermo Fisher Scientific, Inc.; cat. no. 4331182; assay ID Hs00965850_g1), Kremen protein 1 (Thermo Fisher Scientific, Inc.; cat. no. 4331182; assay ID Hs00230750_m1), cyclin-D1-binding protein 1 (Thermo Fisher Scientific, Inc.; cat. no. 4351372; assay ID Hs01050839_m1), ARHGAP10 (Rho GTPase-activating protein 10; Thermo Fisher Scientific, Inc.; cat. no. 4331182; assay ID Hs00226305_m1), FHL3 (four and a half LIM domains protein 3; Thermo Fisher Scientific, Inc.; cat. no. 4331182; assay ID Hs00916408_g1), KCTD11 (Potassium Channel Tetramerisation Domain Containing 11; Thermo Fisher Scientific, Inc.; cat. no. 4331182; assay ID Hs00922550_s1) and the internal standard reference 18s ribosomal (r)RNA (Thermo Fisher Scientific, Inc.; cat. no. 4331182; assay ID Hs99999901_s1) with TaqMan Gene Expression Master Mix (Thermo Fisher Scientific, Inc.; cat. no. 4369016).

Owing to the high abundance of 18s rRNA, total RNA was diluted 1:100 for the 18s rRNA assay yielding average C_qs of 13-20. Assays were run on a MicroAmp Optical 384-Well Reaction Plate with Barcode (cat. no. 4343814). The following thermocycling conditions were used: 1-2 min at 50˚C, 2-10 min at 95˚C, 3-15 sec at 95˚C and 4-60 sec at 60˚C, repeated x40. Results were then analyzed using 7900HT Fast Real-Time PCR System with a 384-well block module (Thermo Fisher Scientific, Inc.; cat. no. 4329001). To calculate normalized expression levels, the comparative C_q method was used and fold changes for miR-769-5p or each gene were obtained from the 2^ΔΔCq values (12).

Human prostate cells were seeded at 2x10⁵ cells/well in 6-well plates and transfected using one of two different protocols: i) At 24 h after seeding, the cells were transfected with 30 nM mirVana miRNA inhibitor, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464076) or mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084, assay ID MH11974) using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Inc.; cat. no. 13778150). By using RT-qPCR, a 40-50% reduction of miR-769-5p was verified, which was induced by the inhibitor relative to the negative control. ii) At 24 h after seeding, the cells were transfected with one of the following: a) 30 nM each of mirVana miRNA inhibitor, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464076) and Silencer Select negative control #1 siRNA (Thermo Fisher Scientific, Inc.; cat. no. 4390843); b) 30 nM each of mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084, assay ID MH11974) and Silencer Select negative control #1 siRNA (Thermo Fisher Scientific, Inc.; cat. no. 4390843); or c) 30 nM each of mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084, assay ID MH11974) and ARHGAP10 Silencer Select small interfering (si)RNAs (Thermo Fisher Scientific, Inc.; cat. no. 4392420, IDs s36028 and s36029) using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Inc.; cat. no. 13778150). Downregulation of miR-769-5p and upregulation of ARHGAP10 in response to mirVana miRNA inhibitor were verified by RT-qPCR. Furthermore, downregulation of ARHGAP10 in response to ARHGAP10 siRNAs was verified.

After 48 h of transfection, 4 columns of wells were seeded in a 96-well plate with each of the following for the 2 transfection groups: i) 100 µl of media/well with no cells, 1x10³ negative control-transfected cells/well or 1x10³ miR-769-5p inhibitor-transfected cells/well; ii) 100 µl of media/well with no cells, 1x10³ negative control miR- and siRNA control-transfected cells/well, 1x10³ miR-769-5p inhibitor- and siRNA control-transfected cells/well or 1x10³ miR-769-5p inhibitor- and ARHGAP10 siRNA-transfected cells/well. Each column of wells represented 8 replicates for a specific time-point (days 0, 1, 3 or 5).

Cell viability was measured on days 0, 1, 3 and5 using alamarBlue Cell Viability Reagent (Thermo Fisher Scientific, Inc.; cat. no. DAL1025) with a FLUOstar Omega microplate reader from BMG Labtech, using fluorescence (excitation, 544 nm/emission, 590 nm). Background fluorescence values of media-only wells were subtracted from the values of wells with treated cells. The subtracted fluorescence values of the 8 replicate wells from on day 5 were compared as follows for the 2 transfections: i) Negative control vs. miR-769-5p inhibitor-transfected cells. ii) miR control and siRNA control vs. miR-769-5p inhibitor and siRNA control. miR control and siRNA control vs. miR-769-5p inhibitor and ARHGAP10 siRNAs. The experiment was performed as three independent replicates for both transfection conditions.

The seeding and transfection conditions were as follows: i) Cells were seeded at 1.6x10⁶ cells/well in T25 flasks (n=9). At 24 h after seeding, the cells were transfected with 30 nM mirVana miRNA inhibitor, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464076) or mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084, assay ID MH11974) using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Inc.; cat. no. 13778150). A 40-50% reduction of miR-769-5p induced by the inhibitor relative to the negative control was verified by RT-qPCR. ii) PC-3 cells were seeded at 1.6x10⁶ cells/well in T25 flasks (n=7). At 24 h after seeding, the cells were transfected with one of the following: i) 30 nM each of mirVana miRNA inhibitor, negative control #1 (cat. no. 4464076) and Silencer Select negative control #1 siRNA (cat. no. 4390843); ii) 30 nM each of mirVana miR-769-5p inhibitor (cat. no. 4464084, assay ID MH11974) and Silencer Select negative control #1 siRNA (cat. no. 4390843); or iii) 30 nM each of mirVana miR-769-5p inhibitor (cat. no. 4464084, assay ID MH11974) and ARHGAP10 Silencer Select siRNAs (cat. no. 4392420, IDs s36028 and s36029) using Lipofectamine RNAiMAX (cat. no. 13778150), all from Thermo Fisher Scientific, Inc. Downregulation of miR-769-5p and upregulation of ARHGAP10 in response to mirVana miRNA inhibitor were verified by RT-qPCR. Furthermore, downregulation of ARHGAP10 in response to ARHGAP10 siRNAs was verified.

At 48 h after transfections, 1.2x10⁶ cell pellets from each treatment group were harvested and frozen at -80˚C. Caspase-3 levels, an indicator of apoptosis, were measured in cell pellets using the EnzChek Caspase-3 Assay Kit #2 (Thermo Fisher Scientific, Inc.; cat. no. E13184). and a FLUOstar Omega microplate reader from BMG Labtech, using fluorescence (excitation, 485 nm/emission, 520 nm). The caspase-3 levels were compared as follows for the 2 transfections: i) Negative control vs. miR-769-5p inhibitor-transfected cells. ii) miR control and siRNA control vs. miR-769-5p inhibitor and siRNA control. miR control and siRNA control vs. miR-769-5p inhibitor and ARHGAP10 siRNAs.

DU145 or PC-3 cells were seeded at 2x10⁵ cells/well in 6-well plates. At 24 h after seeding, the cells were transfected with 30 nM mirVana miRNA inhibitor, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464076) or mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084, assay ID MH11974) using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Inc.; cat. no. 13778150). A 40-50% reduction of miR-769-5p induced by the inhibitor relative to the negative control was verified (Fig SI).

After 48 h of transfection, 3x10⁴ DU145 cells were seeded for each treatment condition in replicate wells in the upper chamber of a cell invasion and migration (CIM) plate 16 (ACEA Biosciences; cat. no. 05665817001), whose bottom chamber contained media with 10% FBS (R&D systems, Inc.; cat. no. S12450) as an attractant. The chambers of a CIM plate 16 were assembled and loaded onto xCELLigence RTCA Systems (ACEA Biosciences; cat. no. 00380601050). The xCELLigence machinery quantified cell movements from the upper to the bottom chambers in 15-min intervals for 48 h. The experiment was performed as three independent replicates.

The steps described for the migration assay were reperformed with PC-3 cells with two additional steps: Prior to seeding of the cells onto the plate, the wells in the upper chamber of the CIM plate 16 (ACEA Biosciences; cat. no. 05665817001) were coated with 5% Matrigel Matrix (BD Biosciences; cat. no. 356234), and the upper chamber was placed in a 37˚C incubator for 4 h for the Matrigel to solidify.

DU145 and PC-3 cells were transfected with 100 nM of mirVana miRNA mimics, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464058) or miR-769-5p mimics (Thermo Fisher Scientific, Inc.; cat. no. 4464066, assay ID MC11974) in quadruplicates and submitted to RNAseq, whose workflow was as follows: The Sequencing Facility was at Leidos Biomedical Research, Inc., Frederick National Laboratory for Cancer Research performed RNA sequencing. For library preparation, the TruSeq V3 chemistry kit from Illumina, Inc. was used with 500 ng total RNA. The Illumina HiSeq 2000 system was utilized for sequencing. For each sample, ~50 million paired-end reads with a length of 101 bp were generated. Trimmomatic software version 0.36 (usadellab.org) was used for trimming reads for both adapters and low-quality bases and the Tophat software version 2.1.1 (http://ccb.jhu.edu/software/tophat/index.shtml) was used for aligning the trimmed reads with the reference human hg19 genome and gene annotation from the Ensembl database (http://grch37.ensembl.org/index.html). Picard software version 2.0.1 (https://broadinstitute.github.io/picard/) was used for calculating RNA mapping statistics and the average uniquely aligned reads were ~90% for all samples. The RNA-seq workflow module in Partek Genomics Suite 6.6 (Partek, Inc.) and the R/Bioconductor package DESeq2 version 3.12 (https://bioconductor.org/packages/release/bioc/html/DESeq2.html) were used to identify differentially expressed genes with a false discovery rate of <5% and a fold change (≤-1.5) as cut-offs. Partek was utilized to perform an analysis based on reads per kilobase per million mapped reads and it allowed the gene counts to be fitted to a negative binomial generalized linear model with DESeq2.

The TargetScan database (http://www.targetscan.org) was utilized to generate a list of predicted gene targets of miR-769-5p. In order to generate an overlap between TargetScan-predicted targets and the RNAseq-downregulated transcripts in miR-769-5p mimics-transfected DU145 or PC-3 cells [cut-off: P<0.01 and ≤-1.5 fold expression change (5)], the Oliveros, J.C. (2007-2015) Venny software (http://bioinfogp.cnb.csic.es/tools/venny/index.html), an interactive tool for comparing lists using Venn diagrams, was applied.

In order to reduce the list of genes from the overlap (322 candidate genes for DU145 and 280 for PC-3), the expression of genes in prostate tumor vs. non-cancerous tissues from 8 cohorts [Grasso et al (13), Lapointe et al (14), Singh et al (15), Taylor et al (5), The Cancer Genome Atlas (TCGA) (16), Tomlins et al (17), Wallace et al (18) and Yu et al (19)] was obtained from Oncomine (http://www.oncomine.org, December 2016, Thermo Fisher Scientific, Inc.). The gene lists for DU145 and PC-3 were compared with the gene lists from these 8 cohorts. Entries with an overlap with genes that had a higher expression in normal tissue vs. tumors in ≥2 cohorts were selected and they were searched in PubMed. Subsequently, the list of potential tumor suppressors targeted by miR-769-5p was further reduced.

The TargetScan database (http://www.targetscan.org) was utilized to locate the putative binding sites of miR-769-5p in the 3'-untranslated region (3'-UTR) of ARHGAP10 and to generate 4 constructs, 1 wild-type (WT) and 3 mutants (MUT). The reporter construct, pLenti-UTR-Luc, contains a cytomegalovirus (CMV) promoter followed by luciferase from Photinus pyralis, and it was digested with EcoRI and XhoI. The 3'-UTR sequences of ARHGAP10 WT, ARHGAP10 MUT1, ARHGAP10 MUT2 and ARHGAP10 MUT1+2 were amplified by PCR from human genomic DNA (isolated from HEK-293 using DNAzol, Thermo Fisher Scientific, Inc. cat. no. 10503027, according to the manufacturer's instructions) plus complementary DNA as follows: PCR was used to amplify the WT sequence with primers at both ends. The MUT1, MUT2 and MUT1+2 were amplified using outer primers together with internal primers that carried the mutation. A Ligation-Free Cloning kit (cat. no. E001) from ABM was used to assemble the 2 PCR amplicons into 1 mutant 3'-UTR, feasible due to the overlapping sequences of the 2 internal primers. Each 3'-UTR was inserted into the reporter construct using the Ligation-Free Cloning kit. The sequences of the 4 WT and MUT constructs were verified using Macrogen (https://dna.macrogen.com/main.do#) (Table SI).

A total of 7x10⁴ DU145 cells per well were seeded in 24-well plates and co-transfected with 100 nM mirVana miRNA mimics, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464058) or miR-769-5p mimics (Thermo Fisher Scientific, Inc.; cat. no. 4464066, assay ID MC11974), 100 ng of luciferase reporter pLenti-UTR-Luc ARHGAP10 WT, pLenti-UTR-Luc ARHGAP10 MUT1, pLenti-UTR-Luc ARHGAP10 MUT2 or pLenti-UTR-Luc ARHGAP10 MUT1+2 and 4 ng of pRL-CMV Renilla luciferase reporter (Promega; cat. no. E2261) using TransIT-X2 (Mirus; cat. no. 6004). Subsequently, the cells were cultured for 48 h and washed with PBS. Using a Luc-Pair Duo-Luciferase HS Assay kit (GeneCopoeia; cat. no. LF004), the cells were lysed, the lysates were loaded onto white 96 wells in quadruplicates and their luciferase/Renilla ratios were measured using a FLUOstar Omega microplate reader (BMG Labtech). The experiment was performed, n=5.

DU145 cells were seeded at 2x10⁵ cells/well in 6-well plates (n=6 experiments). At 24 h after seeding, the cells were transfected with one of the following: i) 30 nM each of mirVana miRNA inhibitor, negative control #1 (Thermo Fisher Scientific, Inc.; cat. no. 4464076) and Silencer Select negative control #1 siRNA (Thermo Fisher Scientific, Inc.; cat. no. 4390843); ii) 30 nM each of mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084; assay ID MH11974) and Silencer Select negative control #1 siRNA (Thermo Fisher Scientific, Inc.; cat. no. 4390843); or iii) 30 nM each of mirVana miR-769-5p inhibitor (Thermo Fisher Scientific, Inc.; cat. no. 4464084; assay ID MH11974) and ARHGAP10 Silencer Select siRNAs (Thermo Fisher Scientific, Inc.; cat. no. 4392420; assay IDs s36028 and s36029) using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Inc.; cat. no. 13778150). Downregulation of miR-769-5p and upregulation of ARHGAP10 in response to mirVana miRNA inhibitor transfection were verified. Furthermore, downregulation of ARHGAP10 in response to ARHGAP10 siRNAs was verified.

At 48 h after transfection, cell lysates were harvested on ice as per the instructions for the CDC42 G-LISA Activation Assay (Cytoskeleton, Inc.; cat. no. BK127). Protein concentrations were measured according to the manufacturer's instructions using Precision Red™ Advanced Protein Assay Reagent, absorbance at 600 nm. The equation used is as follows: C=A/εl; where C = protein concentration (mg/ml), A = absorbance, l = pathlength (cm) and ε = extinction coefficient [(mg/ml)^-1 cm^-1]. The samples were normalized accordingly. CDC42 activity was measured according to the manufacturer's protocol. CDC42 activity was compared as follows: miR control and siRNA control vs. miR-769-5p inhibitor and siRNA control; and miR control and siRNA control vs. miR-769-5p inhibitor and ARHGAP10 siRNAs.

A two-sided, unpaired t-test with Welch's correction was used to assess the endpoints, including differences in miRNA expression, proliferation, apoptosis and results of the luciferase assays. Welch's ANOVA with Dunnett's test was used for comparing >2 groups for miRNA and mRNA expression, CDC42 activity and apoptosis. Kaplan-Meier analysis was utilized to compare the differences in disease-free survival between groups of patients divided by miR-769-5p and/or ARHGAP10 expression.

TCGA data on prostate cancer were downloaded from the cBio Cancer Genomics Portal (http://cbio.mskcc.org/cancergenomics/prostate/data/) to obtain the expression of miR-769-5p and ARHGAP10 in patients. The data from Taylor et al (5) were downloaded from the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/) to obtain miR-769-5p expression, ARHGAP10 expression and disease-free survival data of patients according to the expression of miR-769-5p and ARHGAP10. Recurrence-associated transcripts from Taylor et al (5) were derived utilizing Partek Genomics Suite 6.6 from Partek using Cox proportional hazards regression. The Spearman correlation between ARHGAP10 and miR-769-5p expression in tumor samples from Taylor et al (5) was determined.

For t-tests, Welch's ANOVA with Dunnett's test, Kaplan-Meier and Spearman analyses, GraphPad Prism 7.0 (GraphPad Software, Inc.) was utilized. P<0.05 was considered to indicate statistical significance.

Combination analysis of the publicly available data set (5) and utilization of the strategy (3) to dichotomize miRs into low or high expression groups based on their median values indicated that patients with high expression of miR-769-5p had significantly decreased disease-free survival compared with that of patients with low expression (Fig. 1A). Furthermore, the expression of miR-769-5p in tumor tissues was significantly higher compared with that in non-cancerous tissues in the complete dataset (Fig. 1B). While this difference did not reach statistical significance when focusing on only the matching pairs (n=28), a stratified analysis into high- and low-expressing tumors using the median miR-769-5p value as the cutoff revealed a prevalent upregulation of this miR in the subset of prostate tumors with high expression of miR-769-5p. In addition, miR-769-5p was significantly upregulated in metastases when compared to primary tumors and the non-cancerous tissues in the dataset (Fig. 1C). This phenomenon of higher expression of miR-769-5p in tumor compared to non-cancerous tissues was also observed in the TCGA dataset matching pairs (Fig. 1D). Taken together, the survival and expression data pointed towards miR-769-5p being a candidate oncogene.

miR-769-5p expression was assessed in six human prostate cell lines by using RT-qPCR (Fig. 2A) and a trend toward a higher expression in the cancer cell lines was noted, specifically those derived from metastases compared to the non-tumorigenic RWPE1 and tumorigenic 22Rv1 cells (Fig. 2B). Therefore, it was reasoned that miR-769-5p may be most relevant in the metastatic cell lines and the two metastatic cell lines DU145 and PC-3 were used for the subsequent experiments.

miR-769-5p mimics were transfected into DU145 and PC-3 cells, followed by RNAseq and TargetScan analyses. TargetScan is a bioinformatics tool for miR research that may be used to select candidate target genes of miRNAs relevant in carcinogenesis. Protein-coding genes that were predicted to be targets of miR-769-5p according to TargetScan were focused on, and they were required to exhibit significant downregulation 24 h after transfection of the miR mimics.

Subsequently, the downregulated genes in the DU145 (n=322) and PC-3 (n=280) cell lines from the overlap with TargetScan predictions (Fig. 2C and Table I) were compared against a list of candidate tumor suppressors (decreased expression in tumors compared with normal tissue in two Oncomine datasets) and a literature search was performed, focusing on potential tumor suppressors yet to be extensively reported. The aim was to combine the powers of RNAseq, TargetScan, Oncomine and literature searches to select tumor suppressors targeted by miR-769-5p. This approach yielded several candidates (Table SII) and the top six are presented in Fig. 3.

Among the six candidates, only ARHGAP10, FHL3 and KCTD11 had the expected increase in expression upon treatment with a miR-769-5p inhibitor and decrease in expression following transfection of miR-769-5p mimics. ARHGAP10 expression had the largest fold change and the strongest association with the experimental miR-769-5p status. ARHGAP10 has been reported to be downregulated in ovarian cancer (10), having a tumor suppressor role. In addition, it is associated with a favorable prognosis for patients with prostate cancer (11). Therefore, subsequent experiments aimed to clarify the relationship between miR-769-5p and the candidate tumor suppressor ARHGAP10 in prostate cancer.

Moving forward, ARHGAP10 was indicated to be both a predicted target of miR-769-5p in TargetScan and a prostate cancer recurrence-associated transcript (Fig. 4A). When all recurrence-associated transcripts in the dataset by Taylor et al (5) were plotted according to their corresponding hazard ratios, ARHGAP10 congregated with the gene group with hazard ratios ≤1, indicating its likely tumor suppressor role. Additional survival and correlation analyses were performed to test the relationship between miR-769-5p and ARHGAP10, expecting that if ARHGAP10 is truly a target of miR-769-5p, the two entities would be inversely correlated. The patients who had low expression of ARHGAP10 in their tumor had reduced survival (Fig. 4B), which further confirmed its role as a tumor suppressor in prostate cancer. Furthermore, patients who expressed miR-769-5p at a high level and ARHGAP10 at a low level in their tumor exhibited poor survival (Fig. 4C), consistent with a functional relationship that affects patient survival, and the two entities were also inversely correlated (Fig. 4D), providing additional evidence for this relationship.

Based on the observations linking miR-769-5p and ARHGAP10, it was assessed whether miR-769-5p directly targets ARHGAP10. DU145 cells were transfected with luciferase reporter plasmids containing variations of the binding sequence from the 3'-UTR of ARHGAP10. The plasmids with the 3'-UTR sequences contained the WT or MUT binding sites for miR-769-5p and there were two predicted sites, which were mutated into MUT1, MUT2 or MUT1+2. miR-769-5p mimics significantly downregulated the luciferase activity of the plasmid containing the WT 3'-UTR, albeit moderately. By contrast, in the luciferase assays with the MUT1, MUT2 and MUT1+2 plasmids, the luciferase activity was not significantly affected by the miR-769-5p mimics (Fig. 5A), suggesting that both sites were binding targets of miR-769-5p.

When the cellular functions of miR-769-5p were assessed, it was noted that its inhibition consistently led to reduced proliferation of the 22Rv1, DU145 and PC-3 cell lines (Fig. 5B). Furthermore, it was assessed whether the reduced proliferation was a result of apoptosis by using caspase-3 as a readout. In 22Rv1 and PC-3 cells, caspase-3 increased significantly as a result of miR-769-5p inhibition (Fig. 5C). Therefore, miR-769-5p may have an oncogenic function to enhance the proliferation and inhibit apoptosis in at least a subset of human prostate cancer cell lines.

Inhibition of miR-769-5p increased ARHGAP10 (Fig. 3) and ARHGAP10 was reported to bind CDC42 to inhibit its activity (10). Therefore, it was hypothesized that inhibition of miR-769-5p, leading to high ARHGAP10 levels, may reduce CDC42 activity in prostate cancer cells, which was experimentally confirmed (Fig. 6A). Furthermore, the combined inhibition of miR-769-5p and knockdown of ARHGAP10 negated the decreased CDC42 activity brought about by miR-769-5p inhibition alone. Accordingly, it was hypothesized that, similar to the relationship between ARHGAP10 and CDC42 in ovarian cancer (10), miR-769-5p, ARHGAP10 and CDC42 have a regulatory interaction in prostate cancer cells, i.e., miR-769-5p inhibits ARHGAP10, which suppresses CDC42. As a result of a reduction in ARHGAP10, CDC42 would then be upregulated. Next, it was assessed whether the combined inhibition of miR-769-5p and knockdown of ARHGAP10 reduces the increase in caspase-3 caused by miR-769-5p inhibition alone. The combined miR-769-5p inhibition and ARHGAP10 knockdown modestly reduced the increased caspase-3 activity caused by miR-769-5p inhibition alone (Fig. 6B). In other words, ARHGAP10 knockdown partially rescued the increased caspase-3 activity induced by inhibition of miR-769-5p.