Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Jul;29(7):1037-54.
doi: 10.1210/me.2014-1358. Epub 2015 Jun 8.

miR-22 and miR-29a Are Members of the Androgen Receptor Cistrome Modulating LAMC1 and Mcl-1 in Prostate Cancer

Affiliations
Free PMC article

miR-22 and miR-29a Are Members of the Androgen Receptor Cistrome Modulating LAMC1 and Mcl-1 in Prostate Cancer

Lorenza Pasqualini et al. Mol Endocrinol. .
Free PMC article

Abstract

The normal prostate as well as early stages and advanced prostate cancer (PCa) require a functional androgen receptor (AR) for growth and survival. The recent discovery of microRNAs (miRNAs) as novel effector molecules of AR disclosed the existence of an intricate network between AR, miRNAs and downstream target genes. In this study DUCaP cells, characterized by high content of wild-type AR and robust AR transcriptional activity, were chosen as the main experimental model. By integrative analysis of chromatin immunoprecipitation-sequencing (ChIP-seq) and microarray expression profiling data, miRNAs putatively bound and significantly regulated by AR were identified. A direct AR regulation of miR-22, miR-29a, and miR-17-92 cluster along with their host genes was confirmed. Interestingly, endogenous levels of miR-22 and miR-29a were found to be reduced in PCa cells expressing AR. In primary tumor samples, miR-22 and miR-29a were less abundant in the cancerous tissue compared with the benign counterpart. This specific expression pattern was associated with a differential DNA methylation of the genomic AR binding sites. The identification of laminin gamma 1 (LAMC1) and myeloid cell leukemia 1 (MCL1) as direct targets of miR-22 and miR-29a, respectively, suggested a tumor-suppressive role of these miRNAs. Indeed, transfection of miRNA mimics in PCa cells induced apoptosis and diminished cell migration and viability. Collectively, these data provide additional information regarding the complex regulatory machinery that guides miRNAs activity in PCa, highlighting an important contribution of miRNAs in the AR signaling.

Figures

Figure 1.
Figure 1.
Identification of miR-22, miR-29a, and miR-17-92 cluster as members of the human androgenome. A, Venn diagram of potential androgen-regulated and AR-target miRNA host genes according to gene expression microarray and ChIP-seq analyses in androgen-stimulated DUCaP cells. B, ChIP-seq tracks of miR-22, miR-29a, and miR-17-92 cluster ARBSs genomic regions in DUCaP cells after androgen treatment. DNA reads of anti-AR and control IgG ChIP samples were aligned to the human consensus sequence (GRCh37/hg19) to identify ARBSs. Three ARBSs for miR-22, a single one for miR-29a, and 2 for miR-17-92 cluster were detected. Peaks height reflects the relative AR affinity for each binding site. A schematic depiction of the 3 host genes and the identified ARBSs is presented in Supplemental Figure 1. C, Validation of the ARBSs in the selected miRNAs via PCR amplification of independent anti-AR (α-AR) ChIP-samples obtained from DUCaP cells stimulated for 1 hour with 1nM R1881. Ethanol-treated (EtOH) and control IgG-immunoprecipitated (α-IgG) cells together with the inputs were used as controls. The binding sites present in miR-17-92 cluster host gene, which showed the weakest enrichment, were further confirmed by Sanger DNA sequencing.
Figure 2.
Figure 2.
Androgen regulation of the selected miRNAs and their host genes. A, Androgen regulation of miR-22, miR-29a, and miR-17-92 cluster components along with MIR17HG in DUCaP and LNCaP cell lines was confirmed by qPCR. LNCaP and DUCaP cells were treated either with 1nM R1881 or equivalent vehicle (Ctrl) before RNA isolation. For miR-22 and miR-29a, the mature isoform -3p was analyzed. B, Attenuation of androgen regulation of miR-22 and miR-29a by antiandrogen treatment. DUCaP cells were treated with 10μM MDV3100 either alone or in combination with 1nM R1881 for 24 hours, and, afterwards, miRNAs levels were analyzed by real-time PCR. C, Robust increase of MIR22HG expression but not of other genes adjacent to the ARBSs upon androgen stimulation, as shown by qPCR in DUCaP and LNCaP cell lines. TBP or HPRT1 were used to normalize gene expression. The direct AR target gene FKBP5 was included as positive control. A–C, miRNAs expression was normalized to the small nucleolar RNAs SNORD44 or SNORD38b. Each bar represents mean values ± SEM of at least 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001).
Figure 3.
Figure 3.
Basal levels of miR-22 and miR-29a are decreased in AR-positive PCa cell lines and in primary prostate tumors. A, miR-22 and miR-29a basal levels were measured by qPCR in various prostate cell lines. miRNA expression was normalized to the small nucleolar RNA, SNORD44. Each bar represents mean value ± SEM of 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001). Cell lines were grouped according to their AR and malignant/benign status, as specified below the graph. B, miR-22 and miR-29a expression determined by qPCR in a cohort of 41 matched nonmalignant and malignant tissue sections isolated from radical prostatectomy specimens. miRNA expression was normalized using a panel of normalizers: SNORD44, SNORD48, miR-151a–3p, and miR-320b, (paired Student's t test; *, P < .05; **, P < .01; ***, P < .001). C, Differential DNA methylation of miR-22 and miR-29a ARBSs genomic regions. DNA methylation data were obtained from benign (n = 53) and cancer (n = 51) DNA samples using MeDIP-seq and further analyzed for the presence of differences in the genomic region encompassing miR-22 and miR-29a ARBSs. Depicted are the MeDIP-seq values. Significance was tested with the Mann-Whitney test and further corrected using the Benjamini-Hochberg approach (Mann-Whitney test; *, Benjamini-Horchberg corrected P value (BHp) < .05; **, BHp < .01; ***, BHp < .001).
Figure 4.
Figure 4.
miR-22 and miR-29a target LAMC1 and Mcl-1 in PCa cells. A, Increased LAMC1 or Mcl-1 protein expression in the PC3 and DU145 cells 72 hours after transfection with 50nM antimiR-22 or antimiR-29a, respectively, but no change in their mRNA levels. B, Reduced LAMC1 or Mcl-1 protein content in LNCaP and DUCaP cells 48 hours after transfection with 50nM miR-22 mimic or miR-29a mimic, respectively, without any significant effects on their transcripts. A and B, Protein values were quantified using the Odyssey IR Imaging System and normalized to GAPDH, qPCR data to TBP, and levels of both were further normalized to the appropriate controls (neg. Ctrl, Ctrl mimic). Each bar represents mean value ± SEM of at least 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001). C, Validation of miR-22 and miR-29a target sites in the LAMC1 and MCL1 3′-UTRs by reporter gene assay. A luciferase reporter vector containing partial sequences of the LAMC1 or the MCL1 3′-UTRs, harboring the predicted miRNA target sites, in the 3′-UTR of a NanoLuc luciferase gene was used to confirm the regulation by miR-22 and miR-29a, respectively. PC3 cells were transfected for 36 hours with a combination of reporter and control vectors along with miR-22 mimic/antimiR-22 or miR-29a mimic/antimiR-29a (55nM). Afterwards, NanoLuc luciferase reporter gene and Firefly luciferase control reporter activities were measured using a Nano-Glo Dual-Luciferase assay. Reporter gene activity was normalized to the control reporter activity, and the values were expressed relative to the appropriate miRNA mimic or antimiR control treatment. Each bar represents mean value ± SEM of at least 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001).
Figure 5.
Figure 5.
Regulation of cell viability, migration, and apoptosis by miR-22 and miR-29a. A, Impairment of cell viability and migration. LNCaP or PC3 cells were transfected with 50nM miR-22 or miR-29a mimic. Cell viability was determined after 96 hours using the WST-1 colorimetric assay whereas the number of migrated cells was assessed 72 hours after transfection via cell counting of DAPI-stained migrated cells. Values were normalized to the mimic control. Each bar represents the mean value ± SEM of 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001). B, Apoptosis induction. The effect of miR-29a targeting the prosurvival factor Mcl-1 was evaluated in LNCaP and DUCaP cells. Transient transfection with miR-29a mimic enhanced PARP cleavage. Cleaved PARP (c-PARP) protein levels, determined by Western blot analysis, were quantified using the Odyssey IR Imaging System and normalized to the housekeeping protein GAPDH. Each bar represents the mean value ± SEM of at least 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001).
Figure 6.
Figure 6.
LAMC1 and Mcl-1 expression in PCa cell lines and in benign/cancer prostate tissues. A, Representative TMA cores of paraffin-embedded PCa cell lines stained with anti-AR, anti-LAMC1, and anti-Mcl-1 antibodies along with the immunostaining score values calculated using the HistoQuest software version 3.5. Each bar represents mean value ± SEM of 3 independent experiments (unpaired Student's t test; *, P < .05; **, P < .01; ***, P < .001). B, Anti-AR, anti-LAMC1, and anti-Mcl-1 immunostaining of representative cores of a primary tumor. Original magnification 20/0.5 digital camera. Scale bar, 100 μm. The immunostaining score values summarized in the graphs for tumor and benign tissue cores were obtained by visual inspection in a blinded fashion for anti-LAMC1 and anti-Mcl-1 (Fisher exact test; *, P < .05; **, P < .01; ***, P < .001) while for anti-AR were calculated with the HistoQuest software version 3.5 (Mann-Whitney test; *, P < .05; **, P < .01; ***, P < .001). C, LAMC1 and MCL1 mRNA expression according to gene expression analysis of an independent cohort of unmatched benign (n = 48) and cancer (n = 47) RNA samples isolated from primary tumors. Log 2 values of the normalized array signal are indicated. (Mann-Whitney test; *, P < .05; **, P < .01; ***, P < .001).

Similar articles

See all similar articles

Cited by 20 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback