Prostate cancer tumor specimens and tumor normal adjacent tissues were obtained from Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (Wuhan, China) after informed consent and ethics committee approval. After surgical removal, fresh samples were fixed in 4% paraformaldehyde immediately and then paraffin-embedded for immunohistochemistry.

Human prostate cancer cell lines DU145, PC3 and LNcaP (ATCC) were maintained in RPMI-1640 medium (Hyclone, Logan, UT, USA) and normal prostate epithelial cells RWPE-1 (ATCC) were maintained in Keratinocyte-SFM (Gibco, Grand Island, NY, USA), respectively. The medium was supplemented with 10% fetal bovine serum (FBS) (Gibco) and cells were cultured in a humidified atmosphere of 5% CO₂ at 37°C.

A candidate miRNA complementary to CDKN1A promoter was predicted by the miRanda program on the Linux operating system (11). In addition, the selected miRNA possessed no putative target sites in 3′ or 5′ untranslated regions (UTR) of the p21 transcript.

RNAs and 5′- or 3′-biotin covalently linked RNAs were manufactured by Shanghai GenePharma Co., Ltd. (Shanghai, China). A dsControl lacking significant homology to all known human sequences was used as a non-specific control and dsP21-322 was used as the positive control (Table I). Cells were plated in growth medium at a density of 50–60%. The transfection was carried out by using Lipofectamine RNAiMax (Invitrogen, Carlsbad, CA, USA) 24 h later according to the manufacturer's protocol. The final concentration of RNAs was 50 nM for each well.

Total RNA from cells was extracted by using TRIzol reagent (Invitrogen). Total RNA (500 ng) was reverse transcribed and real-time quantitative PCR was performed on the Mx3000P system (Stratagene, Santa Clara, CA, USA) according to the manufacturer's instructions. Mature miRNA were reverse transcribed and then quantitated by using All-in-one™ miRNA qPCR kit (GeneCopoeia, Guangdong, China). GAPDH and U6 snRNA were used as internal controls. The primers were provided by Invitrogen or GeneCopoeia. Some primer sequences used are available in Table II.

Formalin-fixed, paraffin-embedded tissue sections (5 µm) were deparaffinized in xylene and rehydrated with gradient concentrations of ethanol. The tissue sections were stained with specific antibodies against p21 (Cell Signaling Technology, Danvers, MA, USA) (1:400). The sections incubated with secondary antibodies in the absence of primary antibodies were used as negative control. Hematoxylin was used for counterstaining. Slides were viewed and photographed under a light microscope.

FISH was determined as described previously (12). Briefly, a 5′ and 3′-digoxin-conjugated miR-3619-5p probe (Exiqon, Vedbaek, Denmark) was used for miRNA in situ hybridization of miR-3619-5p. Sections were first blocked with prehybridization buffer for 20 min at 22–25°C below the predicted probe melting temperature in a humidified chamber, and then the miR-3619-5p probe (10 ng/ml) in the hybridization buffer was hybridized with the tissue sections overnight at the same temperature. After washing the slides with washing buffer, the sections were stained with anti-digoxin rhodamine conjugate (1:100, Exon Biotech Inc., China) at 37°C for 1 h away from light. Then, the sections were stained with DAPI for nuclear staining. All fluorescence images were captured on a fluorescence microscope (Leica, Mannheim, Germany).

Cells were harvested at 72 h following transfection. Protein were separated by 10% SDS/PAGE and transferred on to PVDF membranes (Millipore, Billerica, MA, USA). After blocking the membranes were incubated overnight at 4°C with appropriate dilutions of specific primary antibodies as follows: p21 (1/2000) (Cell Signaling Technology), cyclin D1 (1/2000) (Affinity, USA), CDK4 (1/1000) (Affinity), CDK6 (1/2000) (Affinity), GAPDH (1/500) (Boster, Wuhan, China) and α-tubulin (1/500) (Boster). Then membranes were incubated with corresponding second antibody and detected by enhanced chemiluminescence (ECL) assay kit (Millipore).

Cells transfected with biotin-labelled RNAs were processed for ChIP assay (Millipore) at 72 h according to the manufacturer's instructions with further modifications. Approximate 3×10⁶ cells were used for one immunoprecipitation. Degradation of RNAs was avoided by RNase inhibitor (Thermo Fisher Scientific, Waltham, MA, USA) with a final concentration of 50 U/ml. An antibody against biotin (Santa Cruz Biotechnology, Santa Cruz, CA USA) and negative control normal rabbit IgG (Millipore) were used for immunoprecipitation. Primers are available in Table II.

At 72 h after transfection, cells were fixed by 70% cold ethanol, incubated with RNase A (Sigma, St. Louis, MO, USA) and stained by propidium iodide (PI) (Nanjing KeyGen Biotech Co., Ltd., Nanjing, China) staining solution. After staining, the cells were analyzed on a FACSort flow cytometer (BD Biosciences, San Diego, CA, USA). The data were processed by CELL Quest software (BD Biosciences).

After transfection for 24 h, cells were trypsinized, counted and reseeded in 6-well plates at the density of 1000 cells/well for 10 days. The colonies were fixed and stained with 0.5% crystal violet (Sigma). The colony formation rate = number of colonies/number of seeded cells × 100%.

Cell proliferation was assessed by using the CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay kit (Promega, Madison, WI, USA) as previously described (13). Briefly, RNA transfected cells were grown in 96-well plates at a density of 1000 cells/well. Cell growth was measured daily for 4 days. At each time point, 20 µl of CellTiter 96 AQ_ueous One Solution was added and incubated. Absorbance was detected by a microplate reader (Bio-Rad, Berkeley, CA, USA) at 490 nm.

The data are expressed as the mean ± standard deviation (SD) for three independent experiments. Differences between groups were analyzed by t-tests using SPSS version 13.0 software (SPSS Inc., Chicago, IL, USA). P-value <0.05 was considered to be statistically significant.

Firstly, we scanned 1 kb of CDKN1A promoter for sites complementary to known human miRNAs from miRBase database. A candidate miRNA (miR-3619-5p) was identified with a higher sequence complementarity score (171) and a lower minimum free energy (MFE) (−33.24 kcal/mol) through the miRanda algorithm. In addition, the putative target of CDKN1A promoter position −193/−176 relative to the transcription start site (Fig. 1A).

Then we performed quantitative real-time PCR to analyze the miR-3619-5p and CDKN1A expression patterns in prostate cancer cell lines. As shown in Fig. 1B, compared with normal prostate epithelial cells RWPE-1, prostate cancer cell lines DU145 and PC3 had significant lower levels of miR-3619-5p expression). However, the prostate cancer LNcaP cells exhibited higher expression of miR-3619-5p than RWPE-1. Furthermore, CDKN1A mRNA expression was downregulated in the 3 prostate cancer cell lines compared to RWPE-1.

Moreover, we evaluated the p21 protein levels in prostate tissues and we found that p21 was lower expressed in prostate cancer than the corresponding adjacent control (Fig. 1C). Additionally, the FISH assay was conducted to determine miR-3619-5p expression in tissues and indicated that miR-3619-5p expression level in tumor tissues was lower compared to adjacent non-tumor tissues (Fig. 1D). Based on these results, miR-3619-5p may act as a tumor suppressor in prostate cancer cells.

To identify candidate miR-3619-5p that may be involved in positively regulating CDKN1A expression, synthetic miR-3619-5p mimics or controls (dsP21-322 and dsControl) were transfected into DU145 and PC3 cells. Compared to mock and control treatment, dsP21-322 and miR-3619-5p led to a significant induction of p21 mRNA levels in both cell lines at 72 h after transfection (Fig. 2A and B). Moreover, this induction was further confirmed by immunoblot analysis (Fig. 2C).

We and other researchers have reported that the binding of miRNAs to the promoter is one of the possible mechanisms to promoter-targeted specific gene activation (11,14). To explore whether miR-3619-5p interact directly with CDKN1A promoter, we performed ChIP assay via biotinylated RNAs. Biotin was covalently linked to either 5′-end or 3′-end of miR-3619-5p and dsControl antisense strand. After 72 h with individual transfection, the target promoter DNA was pulled down by using a well-characterized biotin antibody and then amplified by real-time PCR. Two sets of primers were used to amplify promoter DNA sequences at different locations, in which one primer set amplifying sequences from −1309 bp to −1150 bp relative to the TSS served as a negative control (Fig. 3A).

Finally, miR-3619-5p pulled down promoter DNA more effectively than the dsControl RNA both in DU145 and PC3 cells (Fig. 3B-E). In contrast, there was no difference in the binding of miR-3619-5p and dsControl to DNA upstream of the CDKN1A promoter (−1309 to −1150). These results indicate that the miR-3619-5p interacts directly with CDKN1A promoter and enhances its expression.

Next, we investigated whether p21 was mainly responsible for tumor suppression after miR-3619-5p transfection. To address this, a small interfering RNA (siP21) was used to silence the p21 expression in DU145 and PC3 cells (Fig. 6A and B). Then we performed flow cytometry to measure cell cycle distribution. In both tested cell types, transfection with miR-3619-5p led to a significant increase in the G0/G1 population compared with dsControl group (Fig. 4A-C). Moreover, knockout of p21 expression markedly attenuated G0/G1 cycle arrest mediated by miR-3619-5p in DU145 (Fig. 4A and C) and PC3 cells (Fig. 4B and C).

In order to assess the cell growth mode after p21 upregulation activated by miR-3619-5p, the colony formation assay was applied and showed miR-3619-5p transfected cells formed significantly fewer colonies in number and smaller in size (Fig. 5A). Additionally, the colony formation rates of miR-3619-5p transfected cells were remarkably lower than dsControl treatment in both cell lines (Fig. 5B). Moreover, the colony formation ability of the prostate cancer cells was restored after co-treatment of siP21 (Fig. 5A and B).

CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay indicated that compared to dsControl group, both DU145 and PC3 cells transfected with miR-3619-5p exhibited progressive retarded growth from the next day following transfection (Fig. 5C). Then we verified whether depletion of p21 could affect the inhibitory function of miR-3619-5p in prostate cancer cells. As illustrated in Fig. 5C, silencing of p21 evidently attenuated the anti-proliferative effect mediated by miR-3619-5p in both cell types. Taken together, these results suggest that miR-3619-5p possesses the capacity to inhibit prostate cancer cell growth through activating CDKN1A expression.

Subsequently, we determined the effects of miR-3619-5p transfection on the expression of downstream genes associated with cell cycle in prostate cancer cells. As shown in Fig. 6C, transfection of miR-3619-5p caused a significant decrease of mRNA levels of cyclin D1, CDK4 and CDK6 in both DU145 and PC3 cells. The suppression effects on protein levels of these genes were further verified by western blot analysis (Fig. 6D). We also blocked p21 expression by cotransfecting the siP21 in both cell lines. Real-time PCR revealed that miR-3619-5p failed to downregulate cyclin D1, CDK4 and CDK6 mRNA levels after siP21 cotreatment (Fig. 6C). The protein analysis of immunoblotting further proved this (Fig. 6D). These findings manifested that miR-3619-5p manipulated prostate cancer cell cycle-associated genes largely by upregulating CDKN1A expression.