Identification of Mir-133B and RB1CC1 As Independent Predictors for Biochemical Recurrence and Potential Therapeutic Targets for Prostate Cancer

Published OnlineFirst March 7, 2014; DOI: 10.1158/1078-0432.CCR-13-1588

Clinical Cancer Human Cancer Biology Research

Identiﬁcation of miR-133b and RB1CC1 as Independent Predictors for Biochemical Recurrence and Potential Therapeutic Targets for Prostate Cancer

Xia Li1, Xuechao Wan1, Hongbing Chen3, Shu Yang1, Yiyang Liu1, Wenjuan Mo1, Delong Meng1, Wenting Du1, Yan Huang1, Hai Wu1, Jingqiang Wang1, Tao Li2, and Yao Li1

Abstract Objective: We aimed to investigate the contribution of microRNA-133b (miR-133b) in prostate cancer cell proliferation, cell cycle, and apoptosis. We also examined expression of miR-133b in prostate cancer tissues, and evaluated the prognostic significance of miR-133b, as well as its target gene RB1CC1 in patients with prostate cancer after radical prostatectomy. Experimental Design: miR-133b mimics (miR-133bm) and anti–miR-133b were transfected into LNCaP and PC-3 cells. CCK-8 was used to look at cell proliferation, flow cytometric analysis was carried out to study cell cycle, and apoptosis was determined by caspase-3 activity. miR-133b expression was assessed by real-time reverse transcription PCR and in situ hybridization in prostatic cell lines and 178 prostate tissue samples, respectively. The protein level of RB1CC1 was examined by Western blot and immunohistochemistry in prostatic cell lines and prostate tissue samples, respectively. Results: Overexpression of miR-133b in LNCaP cells boosted cell proliferation and cell-cycle progression, but inhibited apoptosis; in contrast, miR-133bm promoted cell apoptosis, but suppressed cell proliferation and cell-cycle progression in PC-3 cells. In LNCaP cells, silencing of RB1CC1, a target of miR-133b, inhibited cell apoptosis, and promoted cell-cycle progression. Moreover, miR-133b expression was significantly inversely correlated with RB1CC1 expression in prostate cancer tissues. Multivariate Cox analysis indicated that miR-133b and RB1CC1 might be two independent prognostic factors of biochemical recurrence. Conclusions: miR-133b might enhance tumor-promoting properties in less aggressive LNCaP cells, whereas this miR may act as a tumor suppressor in more aggressive PC-3 cells. miR-133b and RB1CC1 were independent prognostic indicators for prostate cancer. Clin Cancer Res; 20(9); 2312–25. 2014 AACR.

Introduction 92% are present as local and low-stage cases, as indicated by Prostate cancer is the most frequently diagnosed cancer prostate-specific antigen (PSA) testing (2). To date, radical and the second leading cause of cancer-related deaths in the prostatectomy (RP) is the standard treatment for localized male population of the United States (1). It has been prostate cancer (3). However, about 25% to 50% post-RP reported that nearly 217,730 new cases of prostate cancer patients on long-term follow-up have biochemical recur- were diagnosed in the United States in 2010, among which rence (BCR; ref. 4). This is generally the earliest indicator of the recurrence. Therefore, the early risk detection of BCR with sensitive methods and specific tumor markers may Authors' Affiliations: 1State Key Laboratory of Genetic Engineering, effectively improve the prognosis of patients with prostate Institute of Genetics, School of Life Sciences, Fudan University; 2Depart- cancer. ment of Urology, Tongji Hospital, Tongji University School of Medicine, Androgen has been reported to play a pivotal role in Shanghai; and 3Department of Urology, Hefei First People's Hospital, Hefei, Anhui, People's Republic of China androgen pathways and prostate carcinogenesis (5). Andro- gen receptor (AR) regulates androgen-dependent gene tran- Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). scription (6). Many AR-regulated genes are crucial regula- Corresponding Authors: Yao Li, Room 604 Science Building, State Key tors of prostate development and maintenance (7). Recent- Laboratory of Genetic Engineering, School of Life Sciences, Fudan Uni- ly, emerging evidence suggests the potential involvement of versity, 220 Handan Road, Shanghai 200433, People's Republic of China. altered regulation of microRNAs (miRNA) in AR signaling Phone: 86-21-65642047; Fax: 86-21-65642502; E-mail: [email protected]; and Tao Li, Department of Urology, Tongji Hospital, and the pathogenesis of prostate cancer (8–11). Tongji University School of Medicine, Shanghai 200433, People's Republic miRNAs are short, endogenous RNAs in cells, promoting of China. Phone: 86-21-66111533; E-mail: [email protected] translational repression and/or destabilizing target mRNAs, doi: 10.1158/1078-0432.CCR-13-1588 to optimize protein levels in numerous biologic processes 2014 American Association for Cancer Research. (12). MicroRNA-133b (miR-133b), located at chromosome

2312 Clin Cancer Res; 20(9) May 1, 2014

The Role of miR-133b in Prostate Cancer

study. The normal prostate tissue samples were derived Translational Relevance from patients with bladder cancer after total cystectomy; The identification of novel molecular prognostic bio- PIN and prostate cancer samples were derived from markers and therapeutic targets for prostate cancer is patients with prostate cancer after RP and regional lymph fundamental toward improvement of patient outcome. node dissection in Tongji Hospital, subsidiary of Tongji MicroRNAs represent a new class of biomarkers and drug University (Shanghai, China), between January 2001 and targets. Here, we identify and validate the dual role of November 2012. None of the patients received any pre- microRNA-133b (miR-133b) as a mediator of aggressive operation treatment. Cases were classified according to phenotype in LNCaP (androgen-dependent) cells and as World Health Organization criteria and staged according a tumor suppressor in PC-3 (androgen-independent) to tumor, node, metastasis classification and the Gleason cells. We have also identified miR-133b target, RB1CC1 grading system. The Research Ethics Committee of Tongji gene, which functions as a tumor suppressor in LNCaP Hospital approved this protocol, and verbal consents cells. We performed in situ hybridization and immuno- were obtained from all patients. BCR was defined as a histochemistry on tissue microarrays from 178 prostate postoperative serum PSA concentration 0. 2 ng/mL tissue specimens, showing that miR-133b was associated (22). with poor prognosis, whereas RB1CC1 was associated with favorable prognosis. The expression of miR-133b Cell culture and androgen treatment was negatively correlated with RB1CC1 protein levels. The prostate cancer cell lines LNCaP, 22Rv1, PC-3, and Taken together, these results suggested that miR-133b DU145, and the immortalized noncancerous prostatic might play crucial roles in cellular processes of prostate stromal cell line WPMY-1 were obtained from Cell Bank cancer progression, and that miR-133b and RB1CC1 are of Chinese Academy of Sciences (Shanghai, China) where independent biomarkers of prognosis in prostate cancer. they were authenticated by mycoplasma detection, DNA fingerprinting, isozyme detection, and cell vitality detection. The four prostate cancer cell lines were maintained in RPMI 1640 medium, and WPMY-1 cells were cultured in Dulbecco’s Modified Eagle’s medium. The androgen 6p12.2, is downregulated in several solid cancers (13–15). treatment assay was performed as described previously miR-133b has been reported to function as an oncomiR or (23). According to previous reports (24–26), 10 nmol/L tumor suppressor in different types of cancers (16–18). dihydrotestosterone (DHT) was found to be the optimal However, few studies have focused on the role of miR- concentration to stimulate LNCaP cell proliferation. 133b in prostate cancer, especially the significance of miR- 133b expression in clinical prognostic. miRNA mimics and inhibitor molecule treatment and RB1-inducible coiled-coil 1 RB1CC1 ( ), identified as a RNA interference RB1 potential regulator of gene (19), is found in chromo- Synthetic miR-133b mimic (miR-133bm) and its scram- some 8q11, a region containing several loci of putative bled control, control miRNA (miR-NC), siRNA against RB1CC1 tumor suppressors. hasbeenreportedasacritical RB1CC1 gene, and its negative control (an siRNA lacking signaling node that coordinately regulates cell growth, homology to the genome) were from GenePharma (Shang- cell proliferation, apoptosis, and cell spreading/migration hai, China), and used at 50 nmol/L concentration. Anti– (20). It is also associated with various types of human miR-133b (si133b; catalog number IH-300709-06-0005) cancer, especially primary breast cancer (21). However, and its scrambled control, control anti-miRNA (siNC; cat- RB1CC1 theroleof in prostate cancer and the significance alog number IN-001005-01-05) were from Thermo Fisher of RB1CC1 expression in clinical diagnosis were rarely Scientific, and at 100 nmol/L concentration. All sequences reported. of synthetic oligonucleotides are listed in Supplementary In this study, we investigated miR-133b expression in Table S3. cell lines and clinical samples, and analyzed the potential association between miR-133b and various clinicopath- Transient transfection ologic factors of patients with prostate cancer. We iden- Transient transfection was performed as described previ- tified that miR-133b displayed tumor-promoting proper- ously (27). The Opti-MEM medium and Lipofectamine ties in LNCaP cells, whereas acted as a tumor suppressor 2000 were both from Life Technologies. in PC-3 cells. miR-133b and RB1CC1, a target of miR- 133b, were independent prognostic indicators for pros- Real-time quantitative reverse transcription PCR tate cancer. Real-time quantitative reverse transcription PCR (qRT- PCR) for mRNAs was performed as described in (23). For Materials and Methods miRNA quantification, cDNA was synthesized from total Patients RNA with the MiScript Reverse Transcription Kit (Qia- A total of 135 specimens of patients with prostate cancer, gen). Specific primers for mature miR-133b were from 18 patients with prostatic intraepithelial neoplasia (PIN), Qiagen (catalog number MS00031430). The primers used and 25 normal prostate tissue samples were included in this for qPCR were listed in Supplementary Table S4.

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2313

Li et al.

Chromatin immunoprecipitation assay Annexin V Apoptosis Detection Kit (Becton-Dickinson) on Chromatin immunoprecipitation (ChIP) was performed a FACScalibur flow cytometer following the manufacturer’s as described in (23). With minor modification, LNCaP instructions. cells were treated with vehicle, DHT (100 nmol/L), antiandrogen Casodex (10 mmol/L), or a combination of DHT Intracellular caspase and caspase-3 activities and antiandrogen for 4 hours before harvesting. The primers Intracellular caspase and caspase-3 assays were per- of androgen response element (ARE) for qPCR were listed formed as described previously (32). Activities of the total in Supplementary Tables S5 and S6. KLK3 (PSA) enhancer caspase and caspase-3 were assayed with the CaspGLOW was used as the positive control (28), whereas XBP-1 pro- Fluorescein Caspase Staining Kit and CaspGLOW Fluores- moter and a DNA region adjacent to miR-133b gene without cein Caspase-3 Staining kit (BioVision), respectively, and a putative ARE served as the negative controls (29). detected by FACScalibur flow cytometer.

Expression constructs Tissue microarray construction Expression construct was performed as described previ- Tissue microarray (TMA) construction was performed as ously (30). Briefly, the wild-type 30UTR (untranslated described previously (33). Two replicate tumor samples region) of RB1CC1was amplified by PCR from LNCaP cell, (1 mm in diameter) were taken from the donor tissue blocks and was inserted downstream of the stop codon of lucifer- in a highly representative fashion and arrayed into a recip- ase. Two pairs of primers, Mut-1 and Mut-2, were used to ient paraffin block (35 mm 622 mm 65 mm) with a generate a 5 nucleotides substitution mutation of the puta- tissue microarrayer (Beecher Instrument Inc.), as described tive miR-133b binding sites with the KOD-Plus-Mutagen- by Kononen (34). esis Kit (Toyobo). The primers are listed in Supplementary Table S7. Locked nucleic acid in situ hybridization and immunohistochemistry of TMA Dual luciferase assay Locked nucleic acid in situ hybridization (LNA-ISH) LNCaP cells were cotransfected in 24-well plates with 20 was performed as described previously (33) on a set of pmol miR-133bm or miR-NC, together with 0.8-mg firefly prostate cancer TMA slides using LNA probes for miR-133b luciferase report construct containing the wild-type or (Exiqon; sequence: /5DigN/TAGCTGGTTGAAGGGGAC- mutant RB1CC1-30-UTR and 8-ng control vector pRL-TK CAAA/3Dig_N/). Expression of RB1CC1 was detected in (Promega). Forty eight hours after transfection, firefly and immunohistochemistry assays performed as described pre- renilla luciferase activities were measured with GloMax 96 viously (35), with polyclonal antibody for RB1CC1 (1:100; Microplate Luminometer (Promega). Proteintech).

Western blot Image acquisition and management Western blot was performed as described previously (23) Digital images were captured using Nikon DS-Ri1 with antibodies against RB1CC1 (1:1,000; Sigma-Aldrich), ECLIPSE Series (Nikon) with a 10 objective. TMA b-actin (1:4,000; Sigma-Aldrich), and EGFR (1:1,000; images were managed using NIS Element D4.00 software Proteintech). (Nikon).

Cell proliferation assay Statistical analysis Cell proliferation was performed as described in (27). The BCR-free survival time was defined as the time from Briefly, cells were transfected with miR-133bm or anti–miR- the date of surgery to that of BCR, and the follow-up data 133b, and their negative control, respectively, and exam- were updated until November 2012. The differential expres- ined at 0, 24, 48, and 72 hours. At each time point, CCK-8 sions of RB1CC1 between groups were analyzed using c2 reagent was added into cells and incubated for 2 hours. The analysis. The probability of survival was estimated accord- absorbance was measured at 450 nm using a multimode ing to the Kaplan–Meier method. With the log-rank test, we microplate reader (BioTek). analyzed the differences between the survivals of different patients groups. To analyze the correlation between miR- Cell-cycle assay 133b and RB1CC1 expression and clinicopathologic char- Cell cycle was performed as described in (31). Briefly, acteristics, the t test, c2 analysis, Mann–Whitney U test, and cells were incubated with propidium iodide (10 mg/mL) for Fisher exact test were used according to the test condition. A 15 minutes in the dark. The fractions of viable cells in G1,S, Cox proportional hazards model was performed to estab- and G2 phases of cell cycle were measured with a FACStar lish independent factor(s) for patients’ survival. Statistical flow cytometer (Becton-Dickinson). significance was defined as P < 0.05, and all tests were two- tailed. The statistical analysis was performed based on the Apoptosis assays SPSS software, version 15.0 (SPSS Inc.). miR-133bm, miR-NC, anti–miR-133b, or anti–miR-NC For experiments in cell lines, mean and SDs of individual were transfected as described above. Forty eight hours after groups (n 3) were calculated. P values were assessed by transfection, cells were collected and assayed with the FITC performing the two-tailed Student t test.

2314 Clin Cancer Res; 20(9) May 1, 2014 Clinical Cancer Research

The Role of miR-133b in Prostate Cancer

Results upstream and downstream 10 kb from pre–miR-133b’s Androgen induces the expression of miR-133b transcription start site. Seven AREs, formed 5 clusters We have previously analyzed miRNA and mRNA expres- (named A–E), were detected in the proximal DNA regions sion in LNCaP cells treated with 10 nmol/L DHT at different of miR-133b (Fig. 1D). By ChIP–qPCR, we found that AR time points (23). cDNAs reverse transcribed from total recruitment onto six of seven AREs was induced significant- RNAs at each time point were hybridized to Illumina Sentrix ly after 100 nmol/L DHT stimulation for 4 hours compared Human WG-6_V2 expression BeadChip arrays (for mRNA) with the vehicle (Fig. 1E). Treatment with solo-Casodex or a or MicroRNA Expression Profiling Panels (for miRNA). The combination of DHT and Casodex significantly reduced AR PSA raw data have been uploaded to the Gene Expression recruitment to levels comparable with vehicle. The XBP-1 Omnibus public repository (http://www.ncbi.nlm.nih. enhancer and promoter served as positive control gov/geo/; Gene Expression Omnibus series number and negative control for AR binding, respectively (Fig. 1F). GSE21245). Data preprocessing, filtering criteria, and sta- Taken together, our results suggested that androgen–AR tistical analysis were as described in (23). Of the 137 signaling mediated the regulation of miR-133b in LNCaP differentially expressed miRNAs, miR-133b was found to cells. be androgen-responsive in early stage after DHT stimulation (20 minutes, 8 hours), and had higher ARE enrichment miR-133b influenced cell proliferation and cell-cycle according to genomatix database prediction (36). Further, status of prostate cancer miRDB database was applied to predict the sequence-based Because miR-133b was directly regulated by AR, we miRNA targets (37), which were also androgen-responsive analyzed the role of miR-133b on various cellular processes genes, and GenMAPP for the pathway enrichment of miR- of prostate cancer. First, we evaluated the effect of miR-133b 133b. As a result, miR-133b was valued with significant on cell proliferation of LNCaP, 22Rv1, and PC-3 cells. These pathway enrichment in key cellular processes, such as cells were transfected with miR-133bm or anti–miR-133b regulation of cell size, cell growth, and cell cycle (P < (si133b). We observed that when cells were treated with 0.05, data not shown). miR-133bm, cell proliferation was significantly increased Our microarray data analysis identified miR-133b to be compared with miR-NC in LNCaP and 22Rv1 cells (Fig. 2A upregulated before 4 hours, with a peak at 1 hour after DHT and C), whereas anti–miR-133b partially decreased cell stimulation (Fig. 1A). Further validation of miR-133b proliferation compared with anti–miR-NC in these two cell expression was performed by qRT-PCR in LNCaP cells lines (Fig. 2B and D). In contrast, in PC-3 cells, miR-133bm under the same experimental conditions. The expression inhibited cell proliferation, whereas si133b boosted cell of miR-133b was increased compared with the vehicle proliferation (Fig. 2E and F). The transfection efficiency of control (Fig. 1B). The expressions of two androgen-respon- these assays was analyzed by qRT-PCR (Supplementary Fig. sive genes PSA and KLK2 were used as positive controls S1C–S1E). (Supplementary Fig. S1A and S1B). As cell-cycle distribution is a parameter reflecting the growth of cells, we assessed the function of miR-133b on cell-cycle profile of LNCaP and PC-3 cells. Overexpression Differential expression of miR-133b in prostatic cell of miR-133b in LNCaP cells induced an increase in S phase lines and a decrease in G1 phase compared with miR-NC, where- Next, we evaluated the basal expression of miR-133b in as knockdown of miR-133b resulted in a reduction in S noncancerous prostatic cells WPMY-1 and four prostate phase and an increase in G1 phase (Fig. 2G and Supple- cancer cell lines LNCaP, 22Rv1, PC-3, and DU145. LNCaP mentary Fig. S2A). However, in PC-3 cells, miR-133bm is androgen-dependent prostate cancer (ADPC) cell line; caused an increase in G1 phase and a decrease in S phase, 22Rv1 is weakly androgen-dependent; PC-3 is androgen- whereas si133b decreased the proportion of cells in G1 independent prostate cancer (AIPC) cell line without AR phase and increased the proportion in S phase (Fig. 2H expression; and DU145 is AIPC cell line with AR expression. and Supplementary Fig. S2C). The transfection efficiency of We observed that the less aggressive LNCaP cells had the these assays is shown in Supplementary Fig. S2B and S2D. least expression of miR-133b, whereas the more aggressive Collectively, miR-133b boosted cell proliferation and PC-3 cells showed the highest expression, which was 395.5- growth of ADPC cells LNCaP by promoting cell-cycle pro- fold over that of LNCaP cells (Fig. 1C). The finding raised gression, but suppressed cell proliferation and growth of the possibility that differentially expressed miR-133b might AIPC cells PC-3. contribute to the progression of prostate cancer. In this study, we focused on miR-133b and investigated its involve- miR-133b affected cell apoptosis of prostate cancer ment in prostate cancer. We further explored the role of miR-133b on cell apoptosis of prostate cancer. LNCaP and PC-3 cells were treated Verification of the binding of AR to miR-133b with miR-133bm, si133b, or their negative control, respec- As we have confirmed the expression of miR-133b is tively, and followed by flow cytometry analyses. In LNCaP upregulated by androgen, we wondered whether AR could cells, miR-133bm boosted the fraction of living cells and bind to the DNA region of the miR-133b locus. We used decreased that of late apoptotic cells compared with the Genomatix database (36) to predict AREs within the miR-NC, whereas si133b increased both the fraction of

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2315

Li et al.

Figure 1. miR-133b expression to response androgen and the basal expressionindifferentprostaticcelllines.A,timecourseprofile of miR-133b expression in response to 10 nmol/L DHT stimulation in LNCaP cells. The Y-axis referred to normalized signal intensity of the miRNA microarray. B, LNCaP cells were treated with 10 nmol/L DHT or ethanol as vehicle at different time points as indicated. Total RNA was prepared as described in Materials and Methods and analyzed for miR-133b expression by qRT-PCR. All expression levels were normalized to vehicle control at each time point, each experiment was performed in triplicate, and values are mean SD. C, qRT-PCR analysis of miR-133b expression level in different prostatic cell lines as indicated. The data were expressed as fold change of miR-133b expression in each cell line versus LNCaP cells. D, schematic diagram of miR-133b's ARE sites at the 50 and 30 regions. ARE site is indicated by arrows. E and F, occupancy of AR on candidate AREs of miR-133b, PSA enhancer, and XBP-1 promoter with antiandrogen treatment. LNCaP cells were treated with vehicle, 10 mmol/L antiandrogen Casodex (CDX), 100 nmol/L DHT, or a combination of 100 nmol/L DHT and 10 mmol/L Casodex (indicated as DHTþCasodex) for 4 hours before harvesting for ChIP assays. Chromatin samples were immunoprecipitated with anti-AR antibody, and bound DNA was quantified as described in Materials and Methods. E, AR occupancy on the candidate AREs of miR-133b. A DNA region adjacent to miR-133b gene (9982–9895-bp upstream from pre–miR-133b transcription start site) that does not contain a putative ARE served as negative control (indicated as NC). F, AR occupancy on the PSA enhancer (positive control) and XBP-1 promoter (negative control). Values, expressed as percentages of input DNA, are the mean SD of at least three independent experiments. Significance is defined as P < 0.05 (, P < 0.05; , P < 0.01; , P < 0.001).

early apoptotic and late apoptotic cells (Fig. 3A and Sup- To clarify the mechanism by which miR-133b affected plementary Fig. S3A). On the other hand, in PC-3 cells, miR- cell apoptosis of prostate cancer, we detected the activity of 133bm reduced the fraction of living cells and boosted that total caspase and caspase-3 in LNCaP and PC-3 cells. In of late apoptotic cells, but si133b increased the proportion LNCaP cells, miR-133bm signiﬁcantly reduced the fractions of living cells and decreased that of late apoptotic cells of activated caspase (Fig. 3C and Supplementary Fig. S4A) (Fig. 3B and Supplementary Fig. S3C). The transfection and caspase-3 (Fig. 3E and Supplementary Fig. S4E), as efﬁciency of these assays is shown in Supplementary Fig. compared with miR-NC; however, si133b boosted the S3B and S3D. proportion of activated caspase. In PC-3 cells, miR-133bm

2316 Clin Cancer Res; 20(9) May 1, 2014 Clinical Cancer Research

The Role of miR-133b in Prostate Cancer

Figure 2. miR-133b played critical roles in cell proliferation and cell cycle in prostate cancer cells. A to F, cell proliferation analysis. Cells were seeded into a 96-well plate at 5,000 cells/well and examined at 0, 24, 48, and 72 hours after transfection. Each experiment was performed in sextuplicate (n ¼ 6). miR-133b overexpression boosted cell proliferation in LNCaP (A) and 22Rv1 (C), and inhibited proliferation in PC-3 cells (E); knockdown of miR-133b suppressed cell proliferation in LNCaP (B) and 22Rv1 (D), and promoted proliferation in PC-3 cells (F). G and H, cell-cycle analysis. Cells were transfected with miR-133bm, miR-NC, si133b, or siNC for 48 hours, and stained with propidium iodide (PI), and then subjected to cell-cycle analysis. G, miR-133b overexpression caused a decrease in G1 phase and an increase in S phase in LNCaP cells; knockdown of miR-133b caused an increase in G1 phase and a decrease in S phase. H, miR-133b overexpression caused an increase in G1 phase and a decrease in S phase in PC-3 cells; anti–miR-133b caused a decrease in G1 phase and an increase in S phase. signiﬁcantly increased the fractions of activated caspase decreased the proportion of activated caspase. The trans- (Fig. 3D and Supplementary Fig. S4C) and caspase-3 fection efﬁciency is shown in Supplementary Fig. S4B, S4D, (Fig. 3F and Supplementary Fig. S4G), whereas si133b S4F, and S4H, respectively. Taken together, we found that

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2317

Li et al.

Figure 3. miR-133b involvedin prostate cancer cell apoptosis. Cells were transfectedwith miR-133bm or miR-NC, si133b or siNC for 48 hours, and then subjected to cell apoptosis (stained with PI and FITC-Annexin V), caspase (stained with FITC-VAD-FMK), and caspase-3 (stained with FITC-DEVD-FMK) analyses. A, in LNCaP cells, miR-133b overexpression caused a decrease in late apoptotic cells and an increase in living cells; knockdown of miR-133b increased the proportions of both early and late apoptotic cells. B, in PC-3 cells, miR-133b overexpression caused a decrease in living cells and an increase in late apoptotic cells; anti–miR-133b caused an increase in living cells, and a decrease in late apoptotic cells. C, in LNCaP cells, miR-133b overexpression suppressed the activities of intracellular caspase; knockdown of miR-133b promoted the activity of intracellular caspase. D, in PC-3 cells, miR-133b overexpression promoted the activities of intracellular caspase; anti–miR-133b suppressed the activity of intracellular caspase. E, miR-133b overexpression suppressed the activities of caspase-3 in LNCaP cells. F, miR-133b overexpression boosted the activities of caspase-3 in PC-3 cells. Each experiment was performed in triplicate.

2318 Clin Cancer Res; 20(9) May 1, 2014 Clinical Cancer Research

The Role of miR-133b in Prostate Cancer

miR-133b functioned as an oncomiR, which promoted sion was found, with a Pearson correlation coefficient of cell proliferation and suppressed apoptosis of ADPC cells 0.831 (P < 0.05; Fig. 4G). LNCaP; however, this miR acted as a tumor suppressor by Because androgen induced expression of miR-133b, and inhibiting cell proliferation and promoting apoptosis of miR-133b downregulated RB1CC1, we went further to AIPC cells PC-3. evaluate changes of RB1CC1 mRNA and protein levels after 10 nmol/L DHT stimulation for 24 hours in LNCaP cells. As RB1CC1 is a target of miR-133b a result, androgen significantly decreased mRNA and pro- To identify potential target of miR-133b, we conducted tein level of RB1CC1, relative to vehicle control (Supple- a computational screen for genes with complementary mentary Fig. S5B). Collectively, these results suggested that sites of miR-133b in their 30UTR using open-access soft- expression of RB1CC1 and miR-133b was closely linked in ware [TargetScan (www.targetscan.org), PicTar (http:// prostate cancer cells, supporting RB1CC1 as an endogenous pictar.mdc-berlin.de/), and miRDB (www.mirdb.org)]. target of miR-133b. We have performed expression profile in LNCaP cells and found several androgen-responsive miRNAs and Knockdown of RB1CC1 expression boosted cell-cycle mRNAs. To precisely evaluate the actual regulation of progression and inhibited cell apoptosis each predicted miRNA–mRNA pair, the modulation score To evaluate how RB1CC1 functioned in cell cycle and (MS) was proposed (23). For each predicted miRNA– cell apoptosis processes of prostate cancer, we used mRNA pair whose both members were androgen respon- RB1CC1 siRNA to knock down the expression of RB1CC1. sive, MS was calculated and the P value for each MS was As shown in LNCaP cells, knockdown of RB1CC1 signif- evaluated. Subsequently, four mRNAs were identified as icantly decreased the proportion of cells in G1 phase and candidate targets of miR-133b: RB1CC1, PTPRK, SESN1, increased that of cells in S phase (Supplementary Fig. S6A and CPNE3. qRT-PCR analysis in LNCaP cells indicated and S6B), similar to that observed in the overexpression that the expression of RB1CC1 was the most significantly of miR-133b (Fig. 2G). affected by overexpression or knockdown of miR-133b Cell apoptosis analysis revealed that knockdown of among these putative targets (Supplementary Fig. S5A). RB1CC1 boosted the fraction of living cells and reduced Further, we focused on RB1CC1, which was reported to that of late apoptotic cells compared with negative control, be a tumor suppressor and activate the expression of RB1 similar to that observed in the overexpression of miR-133b in breast cancer (38), but its role in prostate cancer is (Supplementary Fig. S6C and S6D). The transfection effi- unclear. ciency of RB1CC1 siRNA is shown in Supplementary Fig. To confirm RB1CC1 asthedirecttargetofmiR-133b, S6E. we performed luciferase assays in LNCaP cells. There is Taken together, we found that RB1CC1, a target of miR- an miR-133b eight-nucleotide seed match region at posi- 133b, functioned as a tumor suppressor by inducing G1–S- tion 252–259 of the RB1CC1 30-UTR(Fig.4A).Cotrans- phase arrest and promoting cell apoptosis of LNCaP cells. fection of LNCaP cells with the wild-type 30-UTR-reporter and miR-133bm showed significantly decreased lucifer- Expression of miR-133b and RB1CC1 in prostate tissue ase activity as compared with miR-NC (Fig. 4B). Inter- samples action between the binding site in RB1CC1 30-UTR and To assess the clinical relevance of these observations in miR-133b was sequence-specific, because mutation of the cell lines, we evaluated the expression of miR-133b and seed sequence restored luciferase activity to values com- RB1CC1 on prostate cancer TMA. LNA-ISH was per- parable with the empty vector (Fig. 4B). These results formed to detect miR-133b expression for 178 prostate denoted that the 30 UTR of RB1CC1 transcript was a tissue specimens mounted on a set of prostate cancer TMA target of miR-133b. slides. ISH score of miR-133b was estimated by scanning each TMA spot. Figure 5A and B showed representative Ectopic expression of miR-133b affected endogenous miR-133b staining in TMAs. Among the 135 prostate RB1CC1 cancer samples, 92 cases (68.1%) showed positive expres- After we transiently transfected LNCaP, 22Rv1, and PC-3 sion of miR-133b, compared with 14 of 18 PIN (77.8%) cells with miR-133m, si133b, or their negative controls, samples and 7 of 25 normal prostate tissues (28.0%), respectively, miR-133bm was shown to decrease the mRNA respectively, which was statistically significant (c2 ¼ and protein level of RB1CC1 relative to miR-NC, whereas 16.432; P < 0.0001). si133b increased the mRNA and protein level of RB1CC1 Immunohistochemistry assay was conducted to analyze compared with siNC (Fig. 4C and D and Supplementary RB1CC1 protein level in 178 prostate tissue specimens. Fig. S5C). Figure 5C and D showed the representative staining of To further address the physiologic relationship between RB1CC1 protein in TMAs. Among the 135 prostate cancer endogenous miR-133b and RB1CC1 in prostate cancer cells, samples, 54 cases (40.0%) showed positive expression of RB1CC1 protein level and miR-133b expression were eval- RB1CC1 protein, compared with 6 (33.3%) of 18 PIN uated by Western blot (Fig. 4E) and qRT-PCR (Fig. 4F) in samples and 20 (80.0%) of 25 normal prostate tissues, four prostatic cell lines, respectively. A significantly inverse respectively, which was statistically significant (c2 ¼ correlation between RB1CC1 protein and miR-133b expres- 14.730; P < 0.001).

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2319

Li et al.

Figure 4. miR-133b directly targeted RB1CC1. A, the binding site of miR-133b to the 30-UTR of RB1CC1. Vertical arrows represent the mutated bases in the RB1CC1-30UTR mutant reporter constructs. B, LNCaP cells were seeded into a 24-well plate, RB1CC1 reporterconstruct (wild-typeor mutant) ortheemptyreporter vector was cotransfected with miR-133bm and pRL-TK, or cotransfected with miR-NC and pRL-TK. miR-133b overexpression decreased the luciferase activity of RB1CC1-30UTR wild-type (RB1CC1), but not that of RB1CC1-30UTR-mutant (MUT) or the empty reporter vector (empty). (Continued on the following page.)

2320 Clin Cancer Res; 20(9) May 1, 2014 Clinical Cancer Research

The Role of miR-133b in Prostate Cancer

Figure 5. miR-133b and RB1CC1 as potential prognostic markers for prostate cancer. A to D, miR-133b and RB1CC1 expression in prostate cancer samples. A, the representative positive expression of miR-133b was mainly detected in prostatic epithelial cells. B, the representative negative staining of miR-133b in prostatic epithelial cells. C, the representative positive staining of RB1CC1 on tissue arrays. D, RB1CC1 staining was negative in prostate cancer samples. The magniﬁcation is 100. E and F, the Kaplan–Meier analysis showed BCR-free survival of patients with prostate cancer patients. E, miR-133b. F, RB1CC1.

Correlation between expression of miR-133b and and S9). We also examined the relationship between the RB1CC1 protein in prostate cancer samples and their expression of miR-133b, RB1CC1, and various clinico- relationship to clinical variables pathologic features in patients with prostate cancer after To explore the correlation between miR-133b and RP. As shown in Supplementary Table S1, positive miR- RB1CC1 expression, we performed both the Pearson and 133b expression was significantly associated with patho- Spearman rank correlation coefficient analyses. A signif- logic stage (P ¼ 0.012) and BCR (P ¼ 0.038), whereas icantly inverse correlation between RB1CC1 protein and positive RB1CC1 protein expression was significantly miR-133b expression was found, with a correlation coef- associated with preoperation PSA (P ¼ 0.021), Gleason ficient of 0.180 (P ¼ 0.036; Supplementary Tables S8 score (P ¼ 0.012), and BCR (P ¼ 0.024).

(Continued.) C and D, qRT-PCR and Western blotting showed theendogenous expression of RB1CC1 in LNCaP cells (C) and PC-3 cells (D) followed by transfection with miR-133bm,miR-NC, si133b,orsiNC. The results revealed that overexpression of miR-133brepressed the endogenous expression ofRB1CC1 compared with that of miR-NC control, and anti–miR-133b could partially restore the expression of RB1CC1. E, Western blot analysis of RB1CC1 protein levels in different prostatic cell lines as indicated. Image quantiﬁcation was performed using Quantity One software. Data were normalized using b-actin protein level. Fold change was presented as relative RB1CC1 protein level of each cell line versus LNCaP cells. F, qRT-PCR analysis of miR-133b basal expression in different prostatic cell lines as indicated. Data were normalized using RNU6B expression. Fold change was presented as relative miR-133b expression (log10) of each cell line versus LNCaP cells. G, correlation between RB1CC1 protein levels and miR-133b expression showed a high degree of reverse correlation (R ¼0.831; P ¼ 0.023). Each experiment was performed in triplicate. Values are mean SD.

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2321

Li et al.

Table 1. Multivariate Cox regression analysis of BCR–free survival for all patients with prostate cancer after RP

Variables P HR for recurrence 95% CI miR-133b expression 0.045 1.775 1.013–3.108 RB1CC1 expression 0.019 0.534 0.316–0.902 Preoperation PSA 0.012 1.009 1.002–1.016 Tumor margins 0.045 0.470 0.224–0.984 Gleason score 0.010 2.242 1.215–4.137

NOTE: Signiﬁcant contributing factors in multivariate Cox regression analysis were miR-133b expression, RB1CC1 expression, tumor margins, preoperation PSA, and Gleason score.

miR-133b and RB1CC1 were prognostic factors for BCR miR-133b, in cervical carcinoma cells, it enhanced cell after RP proliferation and colony formation, suggesting that it could To further examine whether there was significantly dif- act as an oncomiR (16). On the other side, miR-133b was ferent outcome between patients with negative and positive reported to be a tumor suppressor in tongue squamous cell miR-133b expression, the risk of BCR was evaluated using carcinoma (SCC; ref. 13), esophageal SCC (14), and colo- the Kaplan–Meier method, and a statistically significant rectal cancer cells (15). separation of survival curves was observed (Fig. 5E). Fur- Our previous study had shown that miR-133b signifi- thermore, the 5-year BCR-free survival probability rate cantly boosted AR-induced cell viability of LNCaP cells, and and mean survival time were estimated (Supplementary miR-133b was required for mediating AR signaling to Table S2). We also estimated the probability of survival in prostate cancer cell viability and survival (23). In this study, patients with negative and positive RB1CC1 expression, and we found that miR-133b stimulated proliferation of LNCaP the mean BCR-free time and 5-year BCR-free survival rates. and 22Rv1 cells, whereas inhibited apoptosis of LNCaP A statistically significant separation of survival curves was cells. Therefore, miR-133b played an oncogenic role in observed (Fig. 5F). ADPC cells LNCaP and 22Rv1. On the other hand, two Meanwhile, the Cox proportional hazards model was previous articles demonstrated tumor suppressive functions used to assess independent predictors of BCR-free survival, of miR-133b in AIPC cells PC-3 and DU145 (41, 42). Both the factors including preoperation PSA, Gleason score, of these studies showed that miR-133b contributed greatly surgical margin, lymph node metastasis, capsular invasion, to the suppression of proliferation in PC-3 cells. Consistent pathologic stage, and the expression of miR-133b and with their reports, we observed that miR-133b inhibited cell RB1CC1 protein. As a result, miR-133b was identified as proliferation and cell-cycle progression, whereas promoted an independent prognostic factor of outcome for patients apoptosis of PC-3 cells. It has been shown that the same with prostate cancer who underwent RP [HR (95% CI), miRNA could have opposing roles in different cellular 1.775 (1.013–3.108), P ¼ 0.045], as well RB1CC1 [HR settings (43). Collectively, these data indicated that miR- (95% CI), 0.534 (0.316–0.902), P ¼ 0.019], preoperation 133b might display tumor-suppressing properties in AIPC PSA [HR (95% CI), 1.009 (1.002–1.016), P ¼ 0.012], tumor cells PC-3, but act as an oncomiR in ADPC cells LNCaP and margins [HR (95% CI), 0.470 (0.224–0.984), P ¼ 0.045], 22Rv1. and Gleason score [HR (95% CI), 2.242 (1.215–4.137), P ¼ At present, we asked which factor affects or regulates 0.010; Table 1]. Thus, our findings suggested that miR-133b the cellular abundance of miR-133b in prostate cancer expression and RB1CC1 protein level could independently cells. We have found that androgen stimulation induced predict the risk of BCR for patients with prostate cancer after an enhanced abundance of miR-133b in LNCaP cells. RP. Furthermore, we observed that the AR is recruited to the ARE region of miR-133b in the presence of DHT. Discussion These findings suggest that androgen–AR signaling The miRNA expression profiles of many solid malignan- may mediate the regulation of miR-133b in AR-positive cies have been reported (33). Compared with traditional prostate cancer cells. AR may associate with additional mRNA and protein markers, miRNA expression patterns are regions or may require other factor(s) to associate effi- more reliable and sensitive to changes in cell biology. ciently with the miR promoter. Further characterization mRNA and its translated protein level are not often pro- of androgen-dependent and androgen-independent reg- portional, an important reason for which is the regulatory ulations of miRNA expression is currently in progress in influences of epigenetic mechanisms, including those medi- our group. ated by miRNAs (39). A previous study demonstrated that EGFR was a target of Specifically, miRNAs can act as oncomiRs or tumor miR-133b in prostate cancer (42). In line with their results, suppressors in a multitude of cancers (40). As referred to we found that miR-133b downregulated the expression of

2322 Clin Cancer Res; 20(9) May 1, 2014 Clinical Cancer Research

The Role of miR-133b in Prostate Cancer

EGFR in LNCaP and PC-3 cells (Supplementary Fig.S5D and of this study, patients with prostate cancer with low expres- S5E). sion of miR-133b or high protein level of RB1CC1 should In 2012, Patron and colleagues performed whole-genome undergo regular monitoring of serum PSA and clinical microarray expression analysis after transfection of miR- symptoms. In contrast, patients with high expression of 133bm in HeLa cells and matched the list of downregulated miR-133b or low protein level of RB1CC1 could benefit genes with miRecords (41). However, RB1CC1 was not on from obtaining additional clinical data including ultra- the list, because HeLa cells are cervical cancer cells, and our sound-guided biopsy, computed tomography, magnetic data were obtained in prostate cancer cells. As we know, the resonance imaging, and bone scans. same miRNA can target different mRNAs in different cellular In conclusion, our results ascertained miR-133b as a context (43). Further experiments will be necessary to clarify direct target of AR in LNCaP cells. miR-133b acted as a whether RB1CC1 is the target of miR-133b in HeLa cells. potent antiapoptotic and progrowth molecule with onco- Notably, recent studies suggested that RB1CC1 could miRs properties in ADPC cells LNCaP, but functioned as a represent a potential diagnosis and drug target for a broad potential tumor suppressor in AIPC cells PC-3. miR-133b range of cancers. Nuclear expression of RB1CC1 was shown targeted and downregulated RB1CC1 in prostate cancer to predict a better clinical outcome in salivary gland cancers cells. The expression of miR-133b and RB1CC1 protein is (44). In breast cancer, RB1CC1 was reported to predict long- inversely correlated, and they can be used to predict the risk term survival together with RB1 and p53 (45). On the basis of BCR in patients with prostate cancer after RP. Although of our results, we speculated that RB1CC1 might function as evaluations of miR-133b expression and RB1CC1 protein in a tumor suppressor, which was opposite to the oncomiR larger populations are still needed, our results indicated that role of miR-133b in ADPC cells LNCaP, yet further study is miR-133b and RB1CC1 might be promising candidates as needed to fully illustrate the function of RB1CC1 in AIPC molecular prognostic biomarkers and therapeutic targets cells. for prostate cancer. To our knowledge, this is the first study showing the RB1CC1 significance of miR-133b and in relation to clinical Disclosure of Potential Conflicts of Interest data on patients with prostate cancer. We found that the No potential conflicts of interest were disclosed. expression of miR-133b and RB1CC1 protein was inversely correlated. Because none of the patients involved in our Authors' Contributions study received any preoperation treatment, the clinical Conception and design: X. Li, X. Wan, Y. Li samples can be regarded as androgen dependent. Hence, Development of methodology: X. Li, H. Chen, H. Wu our findings in patients’ specimens confirmed the results we Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X. Li, X. Wan, H. Chen, Y. Liu, W. Du, H. Wu obtained in ADPC cells LNCaP, which indicated an onco- Analysis and interpretation of data (e.g., statistical analysis, biostatist- miR role of miR-133b and a tumor suppressor function of ics, computational analysis): X. Li, X. Wan, S. Yang, Y. Liu, W. Mo, D. Meng, RB1CC1 W. Du, J. Wang, T. Li . Finally, multivariate Cox hazard regression anal- Writing, review, and/or revision of the manuscript: X. Li, Y. Huang, ysis revealed that miR-133b and RB1CC1 were identified as J. Wang, T. Li independent prognostic factors of outcome for patients Administrative, technical, or material support (i.e., reporting or orga- nizing data, constructing databases): X. Li, X. Wan, W. Mo, W. Du, Y. Li with prostate cancer after RP. Study supervision: Y. Li Despite the combination of increasingly refined surgical techniques and reduced incidence of surgical complica- Acknowledgments tions, the variable disease course leads to eventual recur- The authors thank Drs. Aditya Dutta, Shazia Irshad, Chee Wai Chua, and rence of about one third of patients following RP (46). Songyan Han for the comments, thoughtful suggestions, and revision of Pound and colleagues reported that distant or local recur- the English language in the article. They also thank Chuan Li for helpful discussions. rence of prostate cancer has not occurred in males without BCR (47). Therefore, to achieve long-term disease-free survival for patients with prostate cancer, BCR risk should Grant Support be assessed. Recent investigations have tried to apply the This work is supported by grants 31071142 and 31171246 from the National Natural Science Foundation of China. improved understanding of tumor cell biology to determine The costs of publication of this article were defrayed in part by the payment additional tumor characteristics as potential prognostic of page charges. This article must therefore be hereby marked advertisement in factors. Such prognostically important factors might help accordance with 18 U.S.C. Section 1734 solely to indicate this fact. determine the optimal treatment strategy based on the Received June 17, 2013; revised January 26, 2014; accepted February 25, biologic potential of individual tumor (48). On the basis 2014; published OnlineFirst March 7, 2014.

References 1. Sun D, Lee YS, Malhotra A, Kim HK, Matecic M, Evans C, et al. miR-99 2. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer family of MicroRNAs suppresses the expression of prostate-speciﬁc J Clin 2010;60:277–300. antigen and prostate cancer cell proliferation. Cancer Res 2011;71: 3. Hyun JS. Prostate cancer and sexual function. World J Mens Health 1313–24. 2012;30:99–107.

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2323

Li et al.

4. Molitierno J, Evans A, Mohler JL, Wallen E, Moore D, Pruthi RS. 24. Wang Q, Li W, Liu XS, Carroll JS, Janne OA, Keeton EK, et al. A Characterization of biochemical recurrence after radical prostatecto- hierarchical network of transcription factors governs androgen my. Urol Int 2006;77:130–4. receptor-dependent prostate cancer growth. Mol Cell 2007;27: 5. Li X, Huang Y, Fu X, Chen C, Zhang D, Yan L, et al. Meta-analysis of 380–92. three polymorphisms in the steroid-5-alpha-reductase, alpha poly- 25. Lin MF, Meng TC, Rao PS, Chang C, Schonthal AH, Lin FF. Expression peptide 2 gene (SRD5A2) and risk of prostate cancer. Mutagenesis of human prostatic acid phosphatase correlates with androgen-stim- 2011;26:371–83. ulated cell proliferation in prostate cancer cell lines. J Biol Chem 6. Culig Z, Bartsch G. Androgen axis in prostate cancer. J Cell Biochem 1998;273:5939–47. 2006;99:373–81. 26. Horoszewicz JS, Leong SS, Kawinski E, Karr JP, Rosenthal H, Chu TM, 7. Xiao J, Gong AY, Eischeid AN, Chen D, Deng C, Young CY, et al. et al. LNCaP model of human prostatic carcinoma. Cancer Res miR-141 modulates androgen receptor transcriptional activity in 1983;43:1809–18. human prostate cancer cells through targeting the small hetero- 27. Li Y, Zhang D, Chen C, Ruan Z, Huang Y. MicroRNA-212 displays dimer partner protein. Prostate 2012;72:1514–22. tumor-promoting properties in non-small cell lung cancer cells and 8. Lin SL, Chiang A, Chang D, Ying SY. Loss of mir-146a function in targets the hedgehog pathway receptor PTCH1. Mol Biol Cell hormone-refractory prostate cancer. RNA 2008;14:417–24. 2012;23:1423–34. 9. Mercatelli N, Coppola V, Bonci D, Miele F, Costantini A, Guadagnoli M, 28. Wang Q, Carroll JS, Brown M. Spatial and temporal recruitment of et al. The inhibition of the highly expressed miR-221 and miR-222 androgen receptor and its coactivators involves chromosomal looping impairs the growth of prostate carcinoma xenografts in mice. PLoS and polymerase tracking. Mol Cell 2005;19:631–42. ONE 2008;3:e4029. 29. Carroll JS, Liu XS, Brodsky AS, Li W, Meyer CA, Szary AJ, et al. 10. Musiyenko A, Bitko V, Barik S. Ectopic expression of miR-126,an Chromosome-wide mapping of estrogen receptor binding reveals intronic product of the vascular endothelial EGF-like 7 gene, regulates long-range regulation requiring the forkhead protein FoxA1. Cell prostein translation and invasiveness of prostate cancer LNCaP cells. 2005;122:33–43. J Mol Med 2008;86:313–22. 30. Li T, Li D, Sha J, Sun P, Huang Y. MicroRNA-21 directly targets 11. Shi XB, Xue L, Yang J, Ma AH, Zhao J, Xu M, et al. An androgen- MARCKS and promotes apoptosis resistance and invasion in pros- regulated miRNA suppresses Bak1 expression and induces androgen- tate cancer cells. Biochem Biophys Res Commun 2009;383: independent growth of prostate cancer cells. Proc Natl Acad Sci U S A 280–5. 2007;104:19983–8. 31. Zhang D, Chen C, Li Y, Fu X, Xie Y, Huang Y. Cx31.1 acts as a tumour 12. Valencia-Sanchez MA, Liu J, Hannon GJ, Parker R. Control of trans- suppressor in non-small cell lung cancer (NSCLC) cell lines through lation and mRNA degradation by miRNAs and siRNAs. Genes Dev inhibition of cell proliferation and metastasis. J Cell Mol Med 2006;20:515–24. 2012;16:1047–59. 13. Wong TS, Liu XB, Chung-Wai Ho A, Po-Wing Yuen A, Wai-Man Ng R, 32. WangX,GuoY,YangS,WangC,FuX,WangJ,etal.Cellularand Ignace Wei W. Identification of pyruvate kinase type M2 as potential molecular mechanisms of photodynamic hypericin therapy for oncoprotein in squamous cell carcinoma of tongue through microRNA nasopharyngeal carcinoma cells. J Pharmacol Exp Ther 2010; profiling. Int J Cancer 2008;123:251–7. 334:847–53. 14. Kano M, Seki N, Kikkawa N, Fujimura L, Hoshino I, Akutsu Y, et al. miR- 33. Li T, Li RS, Li YH, Zhong S, Chen YY, Zhang CM, et al. miR-21 as an 145, miR-133a and miR-133b: tumor-suppressive miRNAs target independent biochemical recurrence predictor and potential thera- FSCN1 in esophageal squamous cell carcinoma. Int J Cancer peutic target for prostate cancer. J Urol 2012;187:1466–72. 2010;127:2804–14. 34. Kononen J, Bubendorf L, Kallioniemi A, Barlund M, Schraml P, 15. Hu G, Chen D, Li X, Yang K, Wang H, Wu W. miR-133b regulates the Leighton S, et al. Tissue microarrays for high-throughput molecular MET proto-oncogene and inhibits the growth of colorectal cancer cells profiling of tumor specimens. Nat Med 1998;4:844–7. in vitro and in vivo. Cancer Biol Ther 2010;10:190–7. 35. Li T, Huang S, Dong M, Gui Y, Wu D. Prognostic impact of SUMO- 16. Qin W, Dong P, Ma C, Mitchelson K, Deng T, Zhang L, et al. MicroRNA- specific protease 1 (SENP1) in prostate cancer patients undergoing 133b is a key promoter of cervical carcinoma development through the radical prostatectomy. Urol Oncol 2013;31:1539–45. activation of the ERK and AKT1 pathways. Oncogene 2012;31: 36. Cartharius K, Frech K, Grote K, Klocke B, Haltmeier M, Klingenhoff 4067–75. A, et al. MatInspector and beyond: promoter analysis based 17. Townley-Tilson WH, Callis TE, Wang D. MicroRNAs 1, 133, and on transcription factor binding sites. Bioinformatics 2005;21: 206: critical factors of skeletal and cardiac muscle development, 2933–42. function, and disease. Int J Biochem Cell Biol 2010;42:1252–5. 37. Wang X. miRDB: a microRNA target prediction and functional anno- 18. Haas JD, Nistala K, Petermann F, Saran N, Chennupati V, Schmitz S, tation database with a Wiki interface. RNA 2008;14:1012–7. et al. Expression of miRNAs miR-133b and miR-206 in the Il17a/f locus 38. Ikebuchi K, Chano T, Ochi Y, Tameno H, Shimada T, Hisa Y, et al. is co-regulated with IL-17 production in alphabeta and gammadelta T RB1CC1 activates the promoter and expression of RB1 in human cells. PLoS ONE 2011;6:e20171. cancer. Int J Cancer 2009;125:861–7. 19. Chano T, Ikegawa S, Kontani K, Okabe H, Baldini N, Saeki Y. Iden- 39. Filipowicz W, Bhattacharyya SN, Sonenberg N. Mechanisms of post- tification of RB1CC1, a novel human gene that can induce RB1 in transcriptional regulation by microRNAs: are the answers in sight? Nat various human cells. Oncogene 2002;21:1295–8. Rev Genet 2008;9:102–14. 20. Gan B, Guan JL. FIP200, a key signaling node to coordinately regulate 40. Shenouda SK, Alahari SK. MicroRNA function in cancer: oncogene or a various cellular processes. Cell Signal 2008;20:787–94. tumor suppressor? Cancer Metastasis Rev 2009;28:369–78. 21. Dahiya R, Perinchery G, Deng G, Lee C. Multiple sites of loss of 41. Patron JP, Fendler A, Bild M, Jung U, Muller H, Arntzen MO, et al. MiR- heterozygosity on chromosome 8 in human breast cancer has 133b targets antiapoptotic genes and enhances death receptor- differential correlation with clinical parameters. Int J Oncol 1998; induced apoptosis. PLoS ONE 2012;7:e35345. 12:811–6. 42. TaoJ,WuD,XuB,QianW,LiP,LuQ,etal.microRNA-133inhibits 22. Freedland SJ, Humphreys EB, Mangold LA, Eisenberger M, Dorey FJ, cell proliferation, migration and invasion in prostate cancer cells by Walsh PC, et al. Risk of prostate cancer-specific mortality following targeting the epidermal growth factor receptor. Oncol Rep 2012; biochemical recurrence after radical prostatectomy. JAMA 2005; 27:1967–75. 294:433–9. 43. Garzon R, Calin GA, Croce CM. MicroRNAs in cancer. Annu Rev Med 23. Mo W, Zhang J, Li X, Meng D, Gao Y, Yang S, et al. Identification of 2009;60:167–79. novel AR-targeted microRNAs mediating androgen signalling through 44. Tameno H, Chano T, Ikebuchi K, Ochi Y, Arai A, Kishimoto M, et al. critical pathways to regulate cell viability in prostate cancer. PLoS ONE Prognostic significance of RB1-inducible coiled-coil 1 in salivary gland 2013;8:e56592. cancers. Head Neck 2012;34:674–80.

2324 Clin Cancer Res; 20(9) May 1, 2014 Clinical Cancer Research

The Role of miR-133b in Prostate Cancer

45. Chano T, Ikebuchi K, Tomita Y, Jin Y, Inaji H, Ishitobi M, et al. RB1CC1 47. Pound CR, Partin AW, Eisenberger MA, Chan DW, Pearson JD, Walsh together with RB1 and p53 predicts long-term survival in Japanese PC. Natural history of progression after PSA elevation following radical breast cancer patients. PLoS ONE 2010;5:e15737. prostatectomy. JAMA 1999;281:1591–7. 46. Shariat SF, Svatek RS, Kabbani W, Walz J, Lotan Y, Karakiewicz PI, 48. Li T, Gui Y, Yuan T, Liao G, Bian C, Jiang Q, et al. Overexpression of et al. Prognostic value of syndecan-1 expression in patients treated high mobility group box 1 with poor prognosis in patients after radical with radical prostatectomy. BJU Int 2008;101:232–7. prostatectomy. BJU Int 2012;110:E1125–30.

www.aacrjournals.org Clin Cancer Res; 20(9) May 1, 2014 2325

Identification of miR-133b and RB1CC1 as Independent Predictors for Biochemical Recurrence and Potential Therapeutic Targets for Prostate Cancer

Xia Li, Xuechao Wan, Hongbing Chen, et al.

Clin Cancer Res 2014;20:2312-2325. Published OnlineFirst March 7, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-13-1588

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2014/03/11/1078-0432.CCR-13-1588.DC1

Cited articles This article cites 48 articles, 9 of which you can access for free at: http://clincancerres.aacrjournals.org/content/20/9/2312.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/20/9/2312.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/20/9/2312. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.