The immortalized benign prostatic epithelial cell line RWPE-1 and human malignant prostate cancer cell lines PC-3 and VCaP, were obtained from the American Type Culture Collection (ATCC), and maintained in regular medium supplemented with 10% fetal bovine serum (FBS) at 37˚C and under 5% CO₂ in a humidified incubator until 80% - 90% confluence.

MiR-141-3p mimic and inhibitor oligonucleotides (Biomics, Jiangsu China) were used at a 100 nM final concentration in the experiments. RWPE-1 cells were plated in 12-well plates at 1.5 - 2.5 × 10⁵ cells per well, grown for 24 h at 70% confluence, and then switched to antibiotic free media prior to transfection. Next, 5 µL oligos and 4.8 µL X-tremeGENE 9 DNA Transfection Reagent (Roche, Basel Switzerland) were added individually to 100 µL Opti-MEM (Invitrogen, Carlsbad USA), and then incubated for at least 20 min before being added to the cultured cells. The overexpression and inhibition of miR-141-3p were achieved by transfection of commercial mature miRNA and antisense miRNA with an appropriate miRNA control. The cells were incubated with the aforementioned oligonucleotides after 48 - 72 h. After transient transfection, total RNA extraction, genetic and functional characteristics were analyzed by qRT-PCR and WB.

For quantitative expression analyses of mRNA and miRNA, total RNAs were isolated from Formalin-Fixed and Parrffin-Embedded (FFPE) prostatic tissues (Table S1 and Table S2) or cultured cells using Trizol reagent (Life Technologies, Carlsbad USA) according to the manufacturer’s instructions. The RNA concentration and purity were measured with a Nanodrop 2000 (Thermo Scientific, Waltham USA). Total RNAs were reverse-transcribed by the M-MLV reverse transcriptase kit (Invitrogen), and residual DNA was removed by treatment with the DNA-free™ Kit (Ambion, Carlsbad USA). qRT-PCR was performed to detect AR mRNA and miR-141-3p levels using SYBR Green Premix DimerEraser (Takara, Japan) on a Roche 480 system. The primers used are listed in Table S3. Melting curves were determined following reactions using the following program: 30 s at 95˚C, followed by 40 cycles of 5 s at 95˚C, 30 s at 56˚C, and 30 s at 72˚C, and the melting curve was determined. GAPDH (for mRNA) and U6 (for microRNAs) levels were used as internal controls, and fold changes were calculated by relative quantification (2^−ΔΔCt). To minimize experimental variation, tumor and control samples were analyzed on the same reaction plate, and all reactions were measured in triplicate.

Transfected cells were lysed in radio-immunoprecipitation assay (RIPA) buffer (Thermo Scientific) supplemented with protease inhibitors (100 mM Tris-HCl at pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1% deoxycholate acid, 0.1% SDS, 2 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 2 mM DTT, 2 mM leupeptin, and 2 mM pepstatin). Lysates were centrifuged at 12,000 rpm for 10 min under 4˚C, and supernatants containing total proteins were collected. Protein concentrations were determined by the BCA method (Beyotime, Jiangsu China), and aliquots of 20 µg protein lysates were separated by SDS-polyacrylamide gel electrophoresis and then transferred to PVDF membranes (GE Healthcare Life Sciences, USA). Membranes were blocked with 5% non-fat milk solution for 2 h. Immunoblotting was performed using an anti- rabbit AR monoclonal antibody (diluted 1:1,000; Abcam, ER179(2)). GAPDH immunoblotting was performed using an anti-rabbit GAPDH antibody (Abcam, EPR6256). The secondary antibody was goat anti-rabbit HRP-linked (Abcam, ab6721), and blots were developed using enhanced chemiluminescence Detection System (Thermo Scientific). Protein levels were quantified using ImageJ2× software and normalized to GAPDH levels.

To construct reporter plasmids, a DNA fragment of the AR ORF containing the putative miR-141-3p binding site (5’-GCCATTGAGCCAGGTGTAGTGTG-3’) was amplified by PCR from human cDNA. The AR ORF fragment lacking the miR-141-3p binding site (5’-GCCA-------AG-3’) was used as negative control. DNA fragments were inserted downstream of the reporter gene of pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega) after SalI and XbaI digestion. The sequences and cloning direction of these plasmids were validated by DNA sequencing. For luciferase reporter assay, 293T cells (4 × 10⁴ per well) were seeded into 24-well plates and cultured for 24 h. The cells were then co-transfected with reporter plasmids and 100 nM chemically synthesized miR- 141-3p mimic or miRNA negative control (miR-NC). After 48 h, cells were harvested and lysed with passive lysis buffer (Promega, Madison USA). Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) on a Fluroskan Ascent FL Microplate Fluorometer (Thermo Electron) following the manufacturers’ instructions. Luciferase activities were expressed as the ratio of firefly to Renilla luciferase activity and normalized to the levels detected in control transfections.

To assess the contribution of miR-141-3p to PCa cell proliferation, transiently transfected RWPE-1 cells were seeded into 96-well plates (3000 cells/well). Cells were incubated for 12 h to allow them to attach to the bottom of the well before the addition of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, St. Louis USA) to measure cell growth at different time points (1 d, 2 d, 3 d, 4 d, 5 d and 6 d). According to the manufacturer’s instructions, 10 µL of MTT solution was added to the cultured cells, and incubated for 4 h at 37˚C. The supernatant was removed, and 200 µL of DMSO was added to each well to solubilize the water-insoluble purple formazan crystals before the absorbance at 490 nm was measured. Experiments were performed in triplicate.

Cells were plated in 12-well plates (1.5 - 2.5 × 10⁵ cells/well) and transfected with the miR-141-3p mimic or inhibitor. After 48 - 72 h, cells were harvested by trypsinization, and 1 × 10⁶ cells were used for cell cycle analysis. The cells were washed with PBS, fixed in 70% ice-cold ethanol overnight at 4˚C, then washed with PBS again and incubated with 1 mL staining solution (20 µg/mL propidium iodide; 10 U/mL RNaseA) for 30 min at room temperature. The DNA content was measured by flow cytometry on a FACS Calibur system (Becton Dickinson, New Jersey USA), and cell cycle distributions of the different populations were determined using FlowJo software (Verity Software House).

Transfected cells were cultured to 80% - 90% confluence in 6-well plates. Cell layers were scratched using a 10 µL pipette tip to form wound gaps and then washed twice with PBS. The wound-healing was photographed at different time points (1 d, 2 d and 3 d). Each wound was analyzed by measuring the distance migrated by cells in three different areas. Data are presented as the mean ± standard deviation (SD) for experiments and compared to control miRNA transfected cells.

Potential miR-141-3p targets were predicted and analyzed using the following four publicly available algorithms: RNAhybrid (http://bibiserv.techfak.uni-bielefeld.de/rnahybrid), PicTar (http://pictar.mdc-berlin.de/), TargetScan (http://www.Targetscan.org), and miRanda (http://www.microrna.org/). Conservative analysis was performed by NCBI-blast (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Statistical analyses of data were performed with SPSS 20.0. All data are expressed as the mean ± SD of at least three separate experiments. All statistical tests were two-sided, and P-values < 0.05 were considered to indicate statistical significance. Unless otherwise noted, the differences between groups were analyzed using either the unpaired Student’s t-test when only two groups were compared or a one-way analysis of variance (ANOVA) when more than two groups were compared.

We previously investigated small RNA transcriptomes by deep-sequencing technology to evaluate three pooled libraries [9] . A total of 13,896,705, 16,768,094 and 17,184,337 raw sequence reads were produced for the malignant PCa (G > 7), early PCa (G ≤ 7), and BPH groups, respectively. After filtering out low quality reads and trimming off adaptors, 10,494,173, 14,577,221 and 14,614,835 sequence reads were obtained [9] . The normalized counts of sequencing reads (specific miRNA/total sequencing tags in the library) were used to quantify miRNA expression levels among the PCa (G > 7), PCa (G ≤ 7) and BPH groups. The statistical significance (P-value) was inferred based on the Bayesian method, which was developed for analysis of digital gene expression profiles. We found a 1.26-fold reduction in miR-141-3p expression in early PCa compared to BPH (P < 0.05), while compared to the early PCa samples, the average miR-141-3p level increased by 3.90-fold in advanced tumors (P < 0.05) (Figure 1(a)). A number of previous reports describe the effects of miR-141-3p overexpression in human PCa [4] [5] [6] [10] , although reports of the effects of inhibited miR-141-3p expression in early PCa are rare.

To further validate the sequencing data, we independently chose 32 early PCa, 14 malignant PCa and 16 BPH tissues. Total RNAs were prepared and analyzed by qRT-PCR. We determined whether miR-141-3p downregulation was common in clinical early PCa tissues. The boxplot of qRT-PCR analyses is shown in Figure 1(b). The average miR-141-3p level was 1.17-fold less in early PCas than that in BPHs (P = 0.2053) (Figure 1(b)). This may be due to the limited number of cases analyzed. Taken together, these data provide evidence that miR-141-3p expression is not evidently changed in early PCa.

AR overexpression is well known to play an important role in the pathogenesis of PCa. We investigated whether certain miRNA(s) contribute to AR upregulation. From RNAhybrid 2.0 [11] , we predicted that the region spanning 2563 -

2585 bp of the AR mRNA (GenBank: M23263.1) was a potential miR-141-3p binding site (Figure 2(a)). Sequence alignment of this putative site showed evolutionary conservation within mammalian species and AR mutants (Table 1).

To verify that the putative miR-141-3p binding site in the ORF of AR mRNA is responsible for regulation by miR-141-3p, the region spanning 2550 - 3000 bp of the AR ORF (5’-GCCATTGAGCCAGGTGTAGTGTG-3’) and its mutant (5’-GCCA-------AG-3’) were cloned separately into the pmirGLO luciferase reporter vector, and then cotransfected with the miR-141-3p mimic or control miR-NC into 293T cells. Luciferase activity was measured after two days. As shown in Figure 2(b), the miR-141-3p mimic significantly reduced luciferase levels when cotransfected with the wild-type AR ORF (8.9% inhibition, P <

0.05), but not with the mutant AR ORF, indicating a direct interaction between miR-141-3p and AR mRNA.

To further confirm that miR-141-3p downregulates the levels of AR mRNA and protein, RWPE-1, PC-3 and VCaP cells were transfected with the miR-141- 3p mimic or control miR-NC, and AR mRNA and protein levels were evaluated by qRT-PCR and WB analyses, respectively (Figure 2(c) and Figure 2(d)). Our data showed that miR-141-3p downregulated AR mRNA and protein levels. Similarly, miR-141-3p knockdown led to recovered AR expression in these cells (Figure 2(c) and Figure 2(d)). Taken together, these data indicate that miR- 141-3p targets the AR and affects AR activity.

To analyze the role of miR-141-3p in early progression of human PCa, we upregulated or downregulated miR-141-3p to observe its influence on cell proliferation, cell cycle and mobility ability. We chose immortalized normal human prostatic epithelial cell line RWPE-1 cells to study the functions of miR-141-3p, because the genome and phenotype of these cells are similar to those of normal prostatic epithelial cells. In this cellular system, we were able to reproduce the development process of early PCa induced by miR-141-3p and AR. RWPE-1 cells were transfected with the miR-141-3p mimic or miR-NC, and analyzed for cell growth, cell cycle and mobility. MiR-141-3p-transfected RWPE-1 cells expressed the miRNA at a level 106-fold higher than that of the control (Figure 2(c)). Proliferation assays showed that cell growth was reduced in miR-141-3p- transfected cells compared with miR-NC-transfected control cells (Figure 3(a)). In FCM analysis, the percentage of cells in the G0/G1 phase increased from 73.29% in controls to 75.75% in miR-141-3p-transfected cells, whereas the percentage of cells in S phase decreased from 18.70% to 16.20% (Figure 3(b)), indicating that miR-141-3p overexpression effectively inhibited the transition from G₀/G₁ to G2-S-M phase in the progression of human early PCa. We analyzed the effect of miR-141-3p overexpression on the mobility behavior of RWPE-1 cells. MiR-141-3p overexpression decreased the mobility of cells (Figure 4). In loss experiments of miR-141-3p, miR-141-3p knockdown led to enhanced cell growth and mobility as shown in Figure 3 and Figure 4.

To further evaluate the association of miR-141-3p with PCa progression, we chose two human malignant prostate cancer cell lines PC-3 and VCaP to assay their cell growth, cell cycle and mobility. As shown in Figure 3 and Figure 4, the same alteration was observed, but miR-141-3p’s effection to cell functions in PC-3 and VCaP cells was slightly compared to that in RWPE-1 cells.