Abstract
Serine Peptidase Inhibitor, Kazal type-1 (SPINK1) overexpression defines the second largest subtype of prostate cancer (PCa), however, molecular mechanisms underlying its upregulation remains poorly understood. Here, we identified a critical role of miRNA-338-5p and miRNA-421 in post-transcriptional regulation of SPINK1. We show that SPINK1-positive PCa patients also exhibit overexpression of Polycomb group member EZH2, which confers repressive trimethylation marks on lysine 27 of histone 3 (H3K27me3) on the regulatory regions of these miRNAs. Further, we demonstrate that oncogenic lncRNA MALAT1 interacts with EZH2, which in turn are targeted by miRNA-338-5p/miRNA-421, thus reinforcing a repressive molecular circuitry. Moreover, ectopic expression of miRNA-338-5p/-421 in SPINK1-positive PCa cells abrogate oncogenic properties including EMT, stemness and drug resistance, resulting in reduced tumor growth and distant metastases in mice. Collectively, we show that restoring miRNA-338-5p/miRNA-421 expression using epigenetic drugs or synthetic miRNA mimics could serve as a potential adjuvant therapy for treatment of SPINK1-positive malignancies.
Significance SPINK1 overexpression is associated with aggressive prostate cancer subtype. We demonstrate EZH2-mediated epigenetic silencing of miR-338-5p/miR-421 leads to oncogenic overexpression of SPINK1. Ectopic expression of miRNA-338-5p/miRNA-421 in SPINK1+ cancer cells attenuate oncogenicity by targeting multiple pathways. Further, restoring miR-338-5p/miR-421 expression using synthetic mimics or epigenetic drugs could abrogate SPINK1-mediated oncogenicity.
Introduction
Prostate Cancer (PCa) is characterized by extensive molecular heterogeneity and varied clinical outcomes (1). Multiple molecular subtypes involving recurrent genetic rearrangements, DNA copy number alterations, and somatic mutations have been associated with this disease (2–6). Majority of the PCa patients harbor gene rearrangements between members of the ETS transcription factor family and the androgen-regulated transmembrane protease serine 2 (TMPRSS2), most recurrent (~50%) being TMPRSS2-ERG, a gene fusion involving the v-ets erythroblastosis virus E26 oncogene homolog (ERG) (5,7). The ERG transcription factor encoded by TMPRSS2-ERG fusion is known to drive cell invasion and metastases, induce DNA damage in vitro, and focal pre-cancerous prostatic intraepithelial neoplasia (PIN) lesions in transgenic mice (8,9).
While TMPRSS2-ERG fusion forms the most frequent molecular subtype, a significant subset of ETS-negative(−) PCa show overexpression of Serine Peptidase Inhibitor, Kazal type-1 (SPINK1) in ~10-15% of the total PCa patients, a distinct subtype defined by overall higher Gleason score, shorter progression-free survival and biochemical recurrence (10,11). SPINK1 promotes cell proliferation and invasion through autocrine/paracrine signaling and mediate its oncogenic effects in part through EGFR interaction by activating downstream signaling. Monoclonal antibody against EGFR showed only a marginal decrease in the growth of SPINK1-positive (+) xenografts in mice, supporting involvement of EGFR-independent oncogenic pathways (12).
Although, genomic events such as genetic rearrangements and somatic mutations constitute most recurrent oncogenic aberrations, many could also be attributed to epigenetic alterations. Earlier studies have shown that aberrant expression of Enhancer of Zeste Homolog 2 (EZH2) owing to genomic loss of miRNA-101 (13) or hypermethylation of miR-26a (14) constitutes a common mechanism across several solid cancers including prostate. Moreover, EZH2, a key component of the Polycomb-Repressive Complex 2 (PRC2) mediates trimethylation on the histone 3 lysine 27 (H3K27me3), leading to gene silencing (15). However, phosphorylated form of EZH2 is known to switch its function from Polycomb repressor to transcriptional coactivator of androgen receptor in castration-resistant prostate cancers (CRPC) (16).
Nodal regulators such as the EZH2 are themselves under the control of other key players, such as lncRNAs and miRNAs, which regulate gene expression by mRNA degradation or translational inhibition (13,17). For instance, long non-coding RNA (lncRNA) such as Metastasis-associated Lung Adenocarcinoma Transcript 1 (MALAT1) specifically interacts with EZH2 and enhances EZH2-mediated gene repression in PRC2-dependent and - independent manner (18), imparting critical role in cancer progression and metastases. Moreover, PRC2 is known to epigenetically repress the expression of miR-181a/b, miR-200b/c, and miR-203, while these miRNAs in turn directly target PRC1 members, namely BMI1 and RING2, in breast and prostate cancer (19).
Although SPINK1+ subtype forms a well-defined and second most prevalent subset of PCa, but the underlying mechanism involved in its upregulation is poorly understood and remains a matter of conjecture. Further, overexpression of SPINK1 is not ascribed to chromosomal rearrangement, deletion, or amplification (10), and thus alludes to a possible transcriptional or post-transcriptional regulation. The present study uncovers the molecular mechanism involved in SPINK1 overexpression and shows how the SPINK1 expression is regulated by miRs-338-5p and -421, which in turn are regulated by EZH2. The mechanism provides compelling evidence that EZH2 acts as an epigenetic switch, which promotes transcriptional silencing of miR-338-5p/miR-421 by establishing H3K27me3 repressive marks, thus leading to SPINK1 overexpression. Taken together, our findings suggest potential benefits with epigenetic drugs such as EZH2 inhibitors or synthetic miR-338-5p/-421 mimics as an adjuvant therapy for the treatment of aggressive SPINK1+ malignancies.
Results
Identification of differentially expressed miRNAs in SPINK1+/ERG-fusion-negative prostate cancer
We employed four miRNA prediction algorithms, namely PITA (omicstools.com), miRmap (mirmap.ezlab.org), miRanda (microRNA.org) and RNAHybrid (BiBiserv2-RNAhybrid) to examine putative binding of miRNAs to the 3’ untranslated region (3’UTR) of SPINK1 transcript. Notably, three miRs -338-5p, -421 and -876-5p were predicted as strong candidates by all four algorithms (Fig. 1A and Supplementary Table S1), and hence were taken forward for further investigation. To determine, whether these three miRNAs show any differential expression between SPINK1+ and ERG+ PCa patients’ specimens, RNA-seq data available at public repository, The Cancer Genome Atlas Prostate adenocarcinoma (TCGA-PRAD) was analyzed. Interestingly, hierarchical clustering of TCGA-PRAD RNA-Seq dataset exhibit reduced expression of miR-338-5p and miR-421 (miR-338-5p/-421) in SPINK1+/ERG-negative patient specimens (Fig. 1B). To validate further, we examined the expression of miR-338-5p/-421 and miR-876-5p in our PCa patients’ specimens. A significant lower expression of miR-338-5p and miR-421 was observed specifically in SPINK1+ as compared to ERG+ specimens (Fig. 1C), while no difference in miR-876-5p expression was noticed (Fig. 1C and Supplementary Fig. S1A). To understand the clinical significance of miR-338-5p/-421, we stratified TCGA-PRAD patients’ data into high and low miRNAs expressing groups, intriguingly group with low miR-338-5p expression show significant (P=0.0024) association with decreased survival probability compared to high miRNAs group (Supplementary Fig. S1B), while no such association was found in case of miR-421. Moreover, lower expression of miR-338-5p also associate with higher Gleason score, advanced clinical T score and lymph node status (Supplementary Fig. S1C). An association of higher Gleason score with lower expression of both miRNAs was further confirmed in another independent cohort (GSE45604) (Supplementary Fig. S1D). In summary, SPINK1+ subtype show lower expression of miR-338-5p/-421, which strongly associate with over-all poor survival and aggressiveness of the disease.
MiR-338-5p and miR-421 directly target SPINK1 and modulate its expression
Having established an association between miR-338-5p/ -421 and SPINK1 expression in PCa specimens (Fig. 1, B-C), we next examined the ability of these miRNAs to bind to the 3’-untranslated region (3’UTR) of SPINK1. The wild-type (3’-UTR-WT) and mutant (3’-UTR-mut) SPINK1 3’-UTR cloned in Firefly/Renilla dual-luciferase reporter vectors were co-transfected with synthetic mimics for miR-338-5p or miR-421 in HEK293T cells, a significant reduction in the luciferase activity was noted with 3’-UTR-WT, while 3’-UTR-mut constructs failed to show any suppressive effect (Fig. 1D). We next evaluated the expression of these miRNAs in various PCa cell lines including 22RV1 (SPINK1+), ETS-fusion positive VCaP (TMPRSS2-ERG+) and LNCaP (ETV1+) cells. Supporting our observation in clinical specimens the cell line data also showed lower expression of miR-338-5p/-421 in the 22RV1 cells relative to fusion-positive cell lines (Supplementary Fig. S1E). To further ascertain that miR-338-5p/miR-421 specifically regulates SPINK1, we used antagomiRs to abrogate miR-338-5p and miR-421 expression (anti-338-5p and anti-421, respectively) in VCaP cells (Supplementary Fig. S1F). As expected, anti-338-5p or anti-421 significantly induced SPINK1 expression in VCaP cells with concomitant increase in cell invasion and migration (Fig. 1E and 1F and Supplementary Fig. S1G and S1H), while there was no change in the endogenous ERG expression (Fig. 1E and Supplementary Fig. S1G). Conversely, we observed that 22RV1 cells stably overexpressing miR-338-5p or miR-421 (22RV1-miR-338-5p and 22RV1-miR-421, respectively) show a significant reduction in SPINK1 expression at both transcript (~80-90%) and protein (Fig. 1G) levels. Since, SPINK1 overexpression has also been implicated in colorectal, lung, pancreatic, and ovarian cancers (20), we sought to examine if SPINK1 is regulated by a similar mechanism in cancers of different cellular/tissue origins. Thus, we determined the status of SPINK1 expression in multiple cancer cell lines (Supplementary Fig. S2A and S2B). Furthermore, SPINK1+ cancer cell lines, namely, colorectal (WiDr), melanoma (SK-MEL-173), pancreatic (CAPAN-1) and prostate (22RV1) upon transfecting with mimics for miR-338-5p or miR-421 showed a significant decrease in SPINK1 expression both at transcript and protein levels (Supplementary Fig. S2C and S2D). This provides irrevocable evidence that these two miRNAs modulate the expression of SPINK1 transcript irrespective of the tissue background. Furthermore, to ascertain whether decrease in oncogenic properties is indeed due to miR-338/-421 mediated reduction in SPINK1 expression, a rescue cell migration assay using human recombinant SPINK1 (rSPINK1) was performed. As expected, miR-338 and miR-421 overexpressing 22RV1 cells show decrease in cell migration, while adding rSPINK1 to these miRNAs overexpressing cells rescued the invasive phenotype, indicating that miR-338/-421 mediated effects are indeed due to decrease in SPINK1 expression (Supplementary Fig. S2E).
Ectopic expression of miR-338-5p and miR-421 attenuate SPINK1-mediated oncogenesis
SPINK1 overexpression is known to contribute to cell proliferation, invasion, motility and distant metastases (10,12,21). Hence, to understand the functional relevance of miR-338-5p/ -421, we examined 22RV1-miR-338-5p and 22RV1-miR-421 stable cells for any change in their oncogenic properties. Both 22RV1-miR-338-5p (C1 and C2) and 22RV1-miR-421 (pooled and C1) cells showed a significant decrease in cell proliferation compared to control (22RV1-CTL) cells (Fig. 2A). Similarly, reduced invasive properties of 22RV1-miR-338-5p and 22RV1-miR-421 cells were noted (~40% and 60% respectively) (Fig. 2B). While, only a modest decrease in cell proliferation and invasion was observed in pooled 22RV1-miR-338-5p cells (Supplementary Fig. S3A and S3C). To assess neoplastic transformation, soft agar colony formation assay was performed, where both 22RV1-miR-338-5p and 22RV1-miR-421 cells exhibited marked reduction (~60% and ~80% respectively) in number and size of the colonies (Fig. 2C). Likewise, 22RV1-miR-338-5p and 22RV1-miR-421 cells demonstrate significantly lower numbers (~70% and ~60% respectively) of dense foci (Fig. 2D). Next, to examine whether overexpression of these miRNAs in benign immortalized prostate epithelial RWPE-1 cells show any phenotypic change, we performed cell-based functional assays. As expected, no significant change in cell proliferation or migration was observed in cells transfected with miR-338 mimic, while miR-421 mimic shows a marginal decrease in proliferation and migration (Supplementary Fig. S3D). Further, to examine the effect of these miRNAs in a SPINK1-independent context, we established stable miR-338-5p/-421 overexpressing prostate cancer PC3 cells and carried-out functional assays. Surprisingly, a significant decrease in cell proliferation, migration and foci formation was observed in miRNAs overexpressing PC3 cells (Supplementary Fig. S3E and S3G), suggesting that these miRNAs perhaps also target key regulators involved in cell division and motility. To demonstrate that miR-338-5p/-421 modulate SPINK1 expression and attenuate SPINK1-mediated oncogenicity irrespective of the tissue background, we performed functional assays using colorectal carcinoma WiDr cells (SPINK1+) stably overexpressing these miRNAs. As anticipated, a significant decrease in the oncogenic potential of the miR-338-5p/-421 overexpressing WiDr cells was observed (Supplementary Fig. S3H and S3I).
To examine tumorigenic potential of 22RV1-miR-338-5p and 22RV1-miR-421 cells in vivo, chick chorioallantoic membrane (CAM) assay was performed, and relative number of intravasated cancer cells was analyzed. Consistent with in vitro results, 22RV1-miR-338-5p and 22RV1-miR-421 cells showed significant reduction in the number of intravasated cells compared to control (Supplementary Fig. S4A and S4B). Likewise, a significant reduction in the tumor weight was recorded in the groups implanted with 22RV1-miR-338-5p and 22RV1-miR-421 cells (Fig. 2E). To evaluate distant metastases, lungs and liver excised from the chick-embryos were characterized for the metastasized cancer cells. The groups implanted with miRNAs overexpressing cells revealed ~80% reduction in cancer cell metastases to lungs (Fig. 2F), while no sign of liver metastases was observed in either group. Further, tumor xenograft experiment was recapitulated in immunodeficient NOD/SCID mice (n=8 per group) by subcutaneously implanting 22RV1-miR-338-5p, 22RV1-miR-421 and control 22RV1 cells into flank region, and trend of tumor growth was recorded. A significant reduction in the tumor burden was observed in the mice bearing miR-338-5p and miR-421 overexpressing xenografts as compared to control (~70% and 85% reduction respectively) (Fig. 2, G-H). To examine spontaneous metastases, lung, liver and bone marrow specimens were excised from the xenografted mice, and genomic DNA was quantified for the presence of human specific Alu-sequences. A significant decrease (~85% for miR-338-5p and ~90% for miR-421) in cancer cell metastases was observed in the group implanted with miRNAs overexpressing cells (Fig. 2I). Similar to CAM assay, cancer cells failed to metastasize to murine liver (data not shown). Furthermore, significant drop (~50%) in Ki-67-positive cells in the miRNAs overexpressing xenografts confirms that tumor regression was indeed due to decline in cell proliferation (Fig. 2J). Taken together, our findings indicate that miR-338-5p/-421 downregulate the expression of SPINK1 and abrogate SPINK1-mediated oncogenic properties and tumorigenesis.
MiR-338-5p and miR-421 exhibit functional pleiotropy by regulating diverse biological processes
To explore critical biological pathways involved in the tumor-suppressive properties rendered by miR-338-5p/ -421 in SPINK1+ cancers, we determined global gene expression profiles of miRNAs overexpressing 22RV1 cells. Our analysis revealed 2,801 and 2,979 genes significantly dysregulated in 22RV1-miR-338-5p and 22RV1-miR-421 cells respectively relative to control, when filtered by log2 fold change of 0.6, FDR<0.05 and P<0.05 (Supplementary Table S2 and S3). Remarkably, ~22% (704 genes) of the downregulated and ~15% (506 genes) of the upregulated transcripts show an overlap in miR-338-5p and miR-421 overexpressing cells (90% confidence interval) (Fig. 3A), indicating that these two miRNAs regulate a significant number of common gene sets and cellular processes. To examine biological processes commonly regulated by miR-338-5p/-421, we employed DAVID (Database for Annotation, Visualization and Integrated Discovery) and GSEA (Gene set enrichment analysis). Most of the downregulated genes were associated with DNA doublestrand break repair by homologous recombination, cell cycle regulation including G2/M-phase transition, stem-cell maintenance, histone methylation and negative regulation of cell-cell adhesion. Whereas, genes involved in negative regulation of gene expression or epigenetics, intrinsic apoptotic signaling pathways, negative regulation of metabolic process and cell cycle were significantly upregulated (Fig. 3B, Supplementary Table S4 and S5). Moreover, GSEA also revealed enrichment of gene signatures associated with oncogenic pathways and cancer hallmarks. Conversely, 22RV1-CTL cells showed significant enrichment of genes involved in sustaining proliferative signaling (EGFR and MEK/ERK) and cell cycle regulators (E2F targets and G2/M transition). While, positive enrichment for tumor suppressive p53 signaling was found in miRNA overexpressing cells as compared to control (Fig. 3C), indicating its role in reduced oncogenicity. Additionally, an overlapping network of pathways using Enrichment map revealed regulation of cell-cycle phase transition and DNA repair pathways (overlap coefficient=0.8, P<0.001, FDR=0.01), as one of the significantly enriched pathways for both miRNAs (Supplementary Fig. S5).
Since MAPK signaling pathways involving a series of protein kinase cascades play a critical role in the regulation of cell proliferation, we examined the phosphorylation status of MEK (pMEK) and ERK (pERK), as a read-out of this pathway. In agreement with our in-silico analysis, a significant decrease in pMEK and pERK was observed in 22RV1-miR-338-5p and 22RV1-miR-421 cells (Fig. 3D). E2F transcription factors are known to interact with phosphorylated retinoblastoma, and positively regulate genes involved in S-phase entry and DNA synthesis (22), thus we next examined the E2F1 level in miRNAs overexpressing cells, surprisingly a notable decrease in E2F1 was observed (Fig. 3D). Further, a significant decrease in the expression of genes involved in G1/S transition such as cyclin E2 (CCNE2), cyclin A2 (CCNA2) and cyclin-dependent kinase (CDK1 and CDK6), including mini-chromosome maintenance (MCM3 and MCM10), required for the initiation of eukaryotic replication machinery was recorded (Fig. 3E). Thus, these findings corroborate with previous literature that during DNA damage, CDKs being cell-cycle regulators crosstalk with the checkpoint activation network to temporarily halt the cell-cycle progression and promote DNA repair (23). Intriguingly, presence of putative miR-338-5p/miR-421 binding sites on the 3’UTRs of these cell cycle regulators (Supplementary Table S6) further support that these targets could be directly controlled by these miRNAs. Next, to validate that miR-338-5p/-421 overexpression leads to S-phase arrest, 22RV1 cells transfected with miR-338-5p or miR-421 mimics were subjected to cell cycle analysis, a significant increase in the S-phase arrested cells was noted (Supplementary Fig. S6A). To delineate that this increase in the S-phase cells is indeed due to cell-cycle arrest and not because of DNA replication, BrdU-7AAD-based cell cycle analysis was performed, which revealed a significant decrease in the percentage of BrdU incorporated cells in S-phase (Fig. 3F). Next, we performed Annexin V-PE staining to examine if overexpression of miR-338-5p/-421 lead to apoptosis, a marginal increase in the early apoptotic cells was evident in miR-338-5p/-421 mimics transfected cells (Supplementary Fig. S6B). Taken together, our findings strongly indicate that miR-338-5p/-421 overexpression led to S-phase arrest, thus elucidating the mechanism for reduced cell proliferation and dramatic regression in tumor growth.
MiR-338-5p and miR-421 regulate oncogenic long non-coding RNA MALAT1 post-transcriptionally
MiRNAs are known to regulate the expression of multiple coding as well as long noncoding RNAs (lncRNAs). Thus, we examined lncRNAs which are specifically dysregulated in 22RV1-miR-338-5p and 22RV1-miR-421 cells, gene expression profiling revealed several deregulated lncRNAs including MALAT1 which is associated with metastatic cancers (24) (Fig. 4A). To ascertain the association between MALAT1 and miRNAs, we analyzed MALAT1 expression in 22RV1-miR-338-5p and 22RV1-miR-421 cells, a significant decrease in the MALAT1 transcript levels were observed in miRNAs overexpressing cells as compared to control (Fig. 4B). Further, miRNAs target prediction indicates putative miR-338-5p/ -421 binding sites on the MATAT1 transcript (Fig. 4C), thus MATAT1 region harboring miR-338-5p (+7233-7421 bp) or miR-421 (+6501-6708 bp) binding sites was cloned in the luciferase reporter plasmid (MALAT1-Luc-338 and MALAT1-Luc-421 respectively). Subsequently, HEK-293T cells co-transfected with MALAT1-Luc-338 or MALAT1-Luc-421 and respective miRNA mimics, showed a significant reduction in the luciferase reporter activity (Fig. 4C). To ascertain that miR-338-5p and miR-421 target MALAT1 post-transcriptionally, VCaP cells transfected with anti-338-5p or anti-421 were characterized for MALAT1 expression, as expected a significant increase in MALAT1 and SPINK1 expression was observed (Fig. 4D).
Differential expression of MALAT1 has been associated with G1/S and G2/M transitions of the cell cycle (25). Thus, we sought to understand the inference of miR338-5p/miR-421/MALAT1 axis in cell cycle regulation. Putative binding sites of E2F1, an important regulator of G1/S phase transition, were found within critical region (+50bp to -700bp) of the MATAT1 promoter (26). To explore the functional interplay between E2F1 and MALAT1, we silenced E2F1 in VCaP cells, which resulted in significant decrease in MALAT1 expression (Fig. 4E). Further, significant enrichment of E2F1 over input on the MALAT1 promoter was confirmed by chromatin immunoprecipitation (ChIP) assay (Fig. 4F), indicating involvement of E2F1 in the regulation of MALAT1. Taken together, we demonstrated that miR-338-5p/-421 regulate the expression of two oncogenic drivers, SPINK1 and MALAT1, and possibly mediate S-phase cell cycle arrest in MALAT1- and E2F1-dependent manner (Fig. 4G). Thus, our findings elucidate one of the plausible mechanisms involved in conferring aggressive phenotype of SPINK1+ subtype, typically associated with higher Gleason score and metastases.
Ectopic expression of miR-338-5p and miR-421 suppresses Epithelial-to-Mesenchymal Transition (EMT) and stemness
Association between EMT and cancer stem cells (CSCs) has been well-established, indicating that a subpopulation of neoplastic cells, which harbor self-renewal capacity and pluripotency, are associated with highly metastatic and drug-resistant cancers (27). Since miR-338-5p/-421 overexpression in 22RV1 cells show regression in tumor burden and metastases (Fig. 2, F-I), we evaluated our microarray data for the genes involved in EMT and stemness, and noted a marked decrease in their expression (Fig. 5A) including the EMT-inducing transcription factors (28) namely, SNAI1 (SNAIL), SNAI2 (SLUG), and TWIST1 (Fig. 5, B-C). Since, SNAIL and SLUG are known to negatively regulate CDH1 (E-Cadherin) (29), an epithelial marker involved in cell-cell adhesion, we next examined E-Cadherin expression. Interestingly, miR-338-5p/ -421 overexpressing cells show a prominent increase in the membrane localization of E-Cadherin, while a significant decrease in the expression of vimentin, a mesenchymal marker was observed (Fig. 5D).
In addition, the expression of genes associated with cancer stem cell-like properties were examined in 22RV1-miR-338-5p and 22RV1-miR-421 cells. Strikingly, the expression of well-known pluripotency markers, such as AURKA, SOX9 and OCT-4, and stem-cell surface markers EPCAM, CD117 (c-Kit), and ABCG2, an ATP-binding cassette transporter, were markedly downregulated in miRNAs overexpressing cells (Fig. 5, E-F). Moreover, a subpopulation (CD117+/ABCG2+) of 22RV1 cells, known as prostate carcinoma-initiating stemlike cells, exhibits stemness and multi-drug resistance (30). Having confirmed that miR-338-5p/-421 downregulate expression of ABCG2 and c-Kit, we next examined the efflux of Hoechst dye via ABC-transporters in the absence or presence of verapamil, a competitive inhibitor for ABC transporters (31). As expected, 22RV1-miR-338-5p and 22RV1-miR-421 cells show a significant reduction (~91% and 89% respectively) in the side population (SP) cells involved in Hoechst dye efflux (Fig. 5G). Efflux assay performed in the presence of verapamil show substantial reduction in the SP cells due to inhibition of Hoechst efflux in both control and miRNAs overexpressing cells (Fig. 5G). Further, to confirm that overexpression of these microRNAs lead to decrease in CSC-like properties, prostatosphere assay, a surrogate model for testing enhanced stem cell-like properties was performed. As expected, 22RV1-miR-338-5p and 22RV1-miR-421 cells showed a significant decrease in the size and prostatosphere formation efficiency (Fig. 5, H-I). Moreover, prostatospheres formed by miRNAs overexpressing cells exhibit a significant reduction in the expression of genes implicated in cancer cells self-renewal and stemness (Fig. 5J). Intriguingly, miR-338-5p and miR-421 putative binding sites on the 3’UTR of EPCAM, c-Kit, SOX9, SOX2 and ABCG2 were also noticed (Supplementary Table S6), suggesting a possible mechanism involved in the downregulation of these genes.
Epigenetic regulators, such as ten-eleven-translocation (TET) family member, TET1, converts 5’-methylcytosine (5mC) to 5’-hydroxymethylcytosine (5hmC), are well-known to induce pluripotency and maintain self-renewal capacity (32). Thus, we analyzed the expression of TET family members in miRNAs overexpressing cells; strikingly a significant decrease in TET1 was observed (Fig. 5K). Since, ABCG2 and c-Kit, which are implicated in drug-resistance, were downregulated in miRNAs overexpressing cells, thus sensitivity of these cells to chemotherapeutic drug was evaluated. Interestingly, 22RV1-miR-338-5p and 22RV1-miR-421 cells show enhanced sensitivity to doxorubicin as compared to control (Supplementary Fig. S6C). Collectively, miR-338-5p/-421 downregulate the expression of genes implicated in multiple oncogenic pathways namely EMT, stemness and drug resistance, signifying that these two tumor suppressor miRNAs could represent a novel approach for integrative cancer therapy (Fig. 5L).
EZH2-mediated transcriptional repression of miR-338 and miR-421 drives SPINK1-positive prostate cancer
Aberrant transcriptional regulation, genomic loss or epigenetic silencing are well-known mechanisms involved in miRNAs deregulation (33,34). Since SPINK1+ PCa patients exhibit reduced expression of miR-338-5p/-421, we sought to decipher the mechanism involved in miRNAs silencing. EZH2, being a member of polycomb group protein play critical role in epigenetic gene silencing by promoting H3K27me3 marks. Thus, we interrogated Memorial Sloan Kettering Cancer Center (MSKCC) patients’ cohort using cBioPortal (http://cbioportal.org) for any plausible association between SPINK1 and EZH2 expression. Interestingly, most of the SPINK1+ specimens comprising Gleason scores 3 and 4 show concordance with EZH2 expression (Fig. 6A). Further, TCGA-PRAD patients harboring higher expression of EZH2 show increased levels of SPINK1, MALAT1, and decreased expression of miR-338-5p/-421 as compared to EZH2-low patients (Fig. 6B). To further confirm the concordance between these two oncogenes, we subsequently evaluated SPINK1 and EZH2 status by performing immunohistochemistry (IHC) and RNA in situ hybridization (RNA-ISH) respectively on the paraffin embedded tissue microarrays (TMAs) comprising a total of 238 PCa specimens. Interestingly, of the 238 PCa specimens evaluated, 21% (50 cases) were positive for SPINK1 expression, and 88% (44 cases) of these SPINK1+ specimens show positive staining for EZH2 (Fig. 6C). While, 75% (141 cases) of the SPINK1-negative (SPINK1−) patients show EZH2 expression as well, notably 71% of these SPINK1−/EZH2+ cases exhibit lowest EZH2 intensity (score 1). Conversely, trend shown in Fig. 6C depicts about ~50% of the SPINK1+/EZH2+ patients fall into low EZH2 expression group (score 1), ~36% in medium EZH2 (score 2), and ~14% in high EZH2 range (score 3 and 4), indicating a significant association between SPINK1+ status and EZH2 expression (Fig. 6C; χ2=13.66; P=0.008). Thus, in corroboration to previous reports (4,14), our data suggests a more pronounced role of epigenetic alterations in ETS-fusion negative cases. Although, 6 of the 50 SPINK1+ cases failed to show any expression of EZH2, pointing that an alternative mechanism may be involved in SPINK1 regulation or possibly miRNA-338/-421 genomic deletion could be a cause in such cases. Additionally, in another independent PCa cohort (GSE35988), increased expression of SPINK1 and EZH2 was observed in localized as well as metastatic specimens, while MALAT1 was specifically upregulated in metastatic cases as compared to benign or localized (Supplementary Fig. S7A), indicating a plausible interplay of these oncogenic drivers in disease progression.
To investigate whether epigenetic silencing of these miRNAs is mediated by EZH2, we screened the promoters of miR-338, miR-421 and FTX (miR-421 host gene) for the putative transcription factor binding sites and identified MYC and MAX (Myc-Associated Factor X) elements within ~2 kb upstream of Transcription Start Site (TSS). MYC is known to form a repressive complex with EZH2 and HDACs, and downregulate multiple tumor suppressive miRNAs, which in turn target PRC2-interacting partners (35). In addition, EZH2-silenced DU145 cells miRNA expression data (GSE26996) indicates an increase in the expression of numerous EZH2-regulated miRNAs including miR-338 and miR-421 (Supplementary Fig. S7B). We therefore examined the promoters of miR-338, miR-421 and FTX for the recruitment of EZH2, interestingly a significant enrichment of EZH2 over input was observed on the promoters of miR-338 and FTX (Fig. 6D). No enrichment on miR-421 promoter was observed (Supplementary Fig. S7C), indicating that the host gene FTX promoter regulates the expression of this intronic miRNA. Next, to confirm EZH2-mediated methyltransferase activity, we sought to identify H3K27me3 marks on these promoters, a remarkable enrichment of H3K27me3 marks on the miR-338 and FTX promoters were noted relative to IgG control (Fig. 6D and Supplementary Fig. S7C), confirming the role of EZH2 mediated epigenetic silencing of miRNA-338-5p/-421.
Comprehensive GSEA analysis revealed that miRNA-338/-421 overexpressing cells show an enrichment for EZH2 interacting partners, including PRC2 members (36) and EZH2 regulated genes (37,38) (Fig. 6E and Supplementary Fig. S7D), indicating that these two miRNAs in turn regulate EZH2 partners and their target genes. Thus, we next examined the putative binding of miRNA-338-5p/-421 on the 3’UTR of the PRC2 members, interestingly both miRNAs show negative mirSVR binding score (Supplementary Table S6). Moreover, a significant decrease in the transcript levels of EZH2, and its interacting partners SUZ12, RBBP4, RBBP7 and MTF2 were observed in miRNA-338-5p/-421 overexpressing cells (Supplementary Fig. S7E and Fig. 6F). Subsequently, we checked for EZH2 recruitment and H3K27me3 histone methylation marks on the promoters of miR-338 and FTX in stable 22RV1-miR-338-5p, 22RV1-miR-421 and control cells. As expected a significant decrease in the EZH2 occupancy and H3K27me3 repressive marks were observed on the promoter regions of miR-338 and FTX in miRNA overexpressing cells (Fig. 6G). Since, MALAT1 is known to interact with EZH2 and facilitates its recruitment on its target genes (18), thus to confirm this interaction we performed RNA immunoprecipitation (RIP) assay in 22RV1 cells. Interestingly, immune-complex pulled down by EZH2 antibody show ~22-folds enrichment of MALAT1 as compared to IgG control (Fig. 6H), indicating that MALAT1 directly binds to EZH2, might promote its occupancy, and H3K27me3 repressive marks at miR-338-5p/-421 promoters leading to epigenetic silencing. Collectively, our data also indicate that overexpression of miR-338-5p/-421 downregulates EZH2 expression and its interacting members, leading to impaired histone methyltransferase activity of PRC2 and reduced SPINK1-mediated oncogenicity, thereby establishing a double-negative feedback loop.
Since inhibitors for chromatin modifiers are known to erase epigenetic marks, we tested 3-Deazaneplanocin A (DZNep), an inhibitor of the histone methyltransferase; 2’-deoxy-5-azacytidine (5-Aza), a DNA methyltransferase (DNMT) inhibitor and Trichostatin A (TSA), a HDAC inhibitor in 22RV1 cells and examined the expression of miR-338-5p/-421. Treatment with TSA, DZNep, 5-Aza alone or a combination of DZNep and TSA in 22RV1 cells showed a modest increase in miR-338-5p/-421 expression, while 5-Aza and TSA together resulted in ~9-fold increase (Fig. 7A). Furthermore, 5-Aza and TSA combination results in significant increase in miRNAs expression accompanied with a notable decrease (~60-80%) in SPINK1 levels (Fig. 7B). Since, 3’-arm of miR-338 (miR-338-3p) is known to negatively regulate Apoptosis Associated Tyrosine Kinase (AATK) expression (39), likewise a significant reduction in the AATK expression was noticed in our study (Fig. 7B). Furthermore, a deletion construct of FTX show decreased expression of miR-374/-421 cluster (40). In line with this, a significant increase in the FTX and miR-421 expression was reported upon 5-Aza and TSA combinatorial treatment, signifying the importance of host gene FTX in the regulation of miR-421 (Fig. 7B).
Furthermore, EZH2 is also known to interact with DNMTs, thus enabling chromatin remodeling and DNA methylation (41). Hence, we next examined the presence of methylated CpG marks on the promoters of miR-338-5p and FTX. Interestingly, methylated DNA immunoprecipitation (MeDIP) revealed locus-specific enrichment in the 5mC levels over 5hmC on these regulatory regions (Fig. 7C). To ascertain the presence of DNA methylation marks we performed bisulfite sequencing using PCa cell lines, a relative increase in the methylated CpG sites on miR-338 and FTX promoters was observed in 22RV1 cells (SPINK1-positive) as compared to VCaP (ERG-positive) cells (Fig. 7D). No significant difference in the methylated CpG sites on the AATK and miR-421 promoters was observed (Supplementary Fig. S7F and S7G). To understand clinical relevance, bisulfite sequencing was carried out on SPINK1-positive (n=5) and ERG fusion positive (n=5) PCa patients’ specimens. Interestingly, all SPINK1-postive specimens exhibit increased methylation marks on the promoters of miR-338 and FTX as compared to ERG positive (Fig. 7D). Taken together, our results strongly indicate that epigenetic machinery comprising of EZH2 and its interacting partners play a critical role in the epigenetic silencing of miRNA-338-5p and miR-421 in SPINK1+ subtype, which in turn reaffirms its silencing by a positive feedback loop.
Discussion
In this study, we unraveled the underlying molecular mechanism involved in the overexpression of SPINK1 exclusively in ETS-fusion negative PCa. Our study provides a molecular basis for SPINK1 overexpression, brought about by the epigenetic repression of key post-transcriptional negative regulators of SPINK1 namely, miR-338-5p and miR-421. We demonstrated the tumor suppressive roles of miR-338-5p/-421, which exhibits functional anticancer pleiotropy in SPINK1+ subtype, by attenuating oncogenic properties, tumor growth and metastases in murine model. Conversely, miR-421 has also been reported to be a potential oncogenic miRNA in multiple cancers (42,43). However, in corroboration with our findings, a recent report suggested tumor suppressive role of miR-421 in prostate cancer (44). We also established that miR-338-5p/-421 overexpressing cells display perturbed cell-cycle machinery triggered by dysregulated cyclins and CDKs, subsequently leading to S-phase arrest. It has been shown that miRNAs targeting multiple cyclins/CDKs are more effective than the FDA-approved CDK4/6 inhibitor in triple-negative breast cancer (45), thus supporting our findings that replenishing these miRNAs may prove advantageous in SPINK1+ cancers. Moreover, besides SPINK1, miR-338-5p/-421 also targets MALAT1, which could possibly result in downregulation of E2F1, thus contributing to arrest in G1/S transition, although in-depth molecular mechanisms behind this remain to be elucidated.
Emerging evidences suggest a complex interaction between EMT and CSCs during cancer progression, and in developing resistance towards anti-cancer drugs. Previous studies have implicated the role of several miRNAs, such as miR-200 family, miR-205 and miR-34a (46,47) in regulating the expression of genes involved in metastases, stemness and drug resistance. Furthermore, miR-338 exhibits tumor suppressive role, and inhibits EMT by targeting ZEB2 (48) and PREX2a (49) in gastric cancer. Here, we identified miR-338-5p/-421 as critical regulators of EMT-inducing transcription factors and -associated markers, which in turn led to decreased stem-cells like features. Moreover, CSCs are known to express ABC transporters, which efflux the chemotherapeutic drugs during resistance (50). Remarkably, miR-338-5p/-421 overexpression shows decreased expression of ABCG2 and c-KIT, consequently a significant drop in the drug-resistant side population, indicating that these two miRNAs are highly effective in conferring drug-sensitivity and reducing the therapy-resistant CSCs. Collectively, our findings provide a solid foundation for qualifying these miRNAs as an adjuvant therapy for the SPINK1+ as well as other drug resistant malignancies.
Numerous lncRNAs have been reported to be dysregulated in prostate cancer (51,52), for instance, PCGEM1 and PRNCR1 are highly expressed in aggressive PCa and enhance the AR-mediated gene activation program (53). Furthermore, MALAT1 renders cell cycle arrest in G1/S transition and mitotic phase by modulating the expression of E2F1 and B-MYB (25). Similarly, nuclear lncRNAs often interact with the components of chromatin-remodeling complexes such as EZH2, SUZ12, CBX7, CoREST and JARID1C, and mediate gene silencing or activation by modulating their activity (52,54). For example, MALAT1, a known nuclear lncRNA interacts with EZH2, facilitates its occupancy and the H3K27me3 activity on the PRC2 target genes (18), in corroboration to this, we also found that MALAT1 interacts with EZH2 and might facilitate its recruitment on miRNAs regulatory regions. Hence, we propose a molecular model for the functional interplay involving SPINK1, MALAT1 and miR-338-5p/-421, wherein MALAT1 facilitates recruitment of EZH2, which acts as an epigenetic switch and by its histone methyltransferase activity establishes H3K27me3 repressive marks on the promoters of miR-338 and FTX, a miR-421 host gene (Fig. 7E). This finding was further strengthened by a recent TCGA study (4), wherein a subset of PCa patients’ harboring SPOP-mutation/CHD1-deletion exhibits elevated DNA methylation levels accompanied with frequent events of SPINK1 overexpression. Recently, a new subtype of ETS-fusion-negative tumors has been defined by frequent mutations in the epigenetic regulators and chromatin remodelers (55). Yet another study, using genome-wide methylated DNA-immunoprecipitation sequencing revealed higher number of methylation events in TMPRSS2-ERG fusion-negative as compared to normal and TMPRSS2–ERG fusion-positive PCa specimens (14), thus collectively, these independent findings reaffirm the critical role of epigenetic pathways engaged in the pathogenesis of SPINK1+ subtype.
Interestingly, increased methylated regions in the ETS-fusion negative patients have been attributed to hypermethylation of miR-26a, a post-transcriptional regulator of EZH2 (14). Thus, given the central role played by EZH2 and the epigenetic mechanism involved in ETS-fusion negative cases, our findings rationalize the role of EZH2-mediated epigenetic regulation of miR-338-5p and miR-421 in SPINK1+/ETS-negative subtype. In consonance with this, overexpression of miR-338-5p/-421 also results in decreased Tet1 expression. Converging lines of evidences suggest dual role of Tet1 in promoting transcription of pluripotency factors as well as recruitment of PRC2 on the CpG rich promoters (56). Taken together, miR-338-5p/-421 mediated decrease in Tet1 expression might possibily contribute in reduced stemness and drug-resistance. We also conjecture that decrease in Tet1 expression may result in reduced PRC2 occupancy on the miRNA promoters, diminish epigenetic silencing marks, and consequently downregulate their targets including SPINK1.
Currently, there is no effective therapeutic intervention for SPINK1+/ETS-negative PCa as well as for other SPINK1+ malignancies, although use of monoclonal EGFR antibody has been suggested (57). Nevertheless, outcome of the phase I/II clinical trials using cetuximab (58) and small molecules inhibitors for EGFR has been largely unsuccessful (59,60). For instance, in a phase Ib/IIa clinical trial using cetuximab and doxorubicin combination therapy, only a fraction of CRPC patients (~8%) showed >50% PSA decline (58), revealing its limited effectiveness. Owing to the pleiotropic anti-cancer effects exhibited by miRNA-338-5p/-421, we propose microRNA-replacement therapy as one of the potential therapeutic approaches for SPINK1+ cancers; nonetheless in-vivo delivery methods and stability are some of the major challenges for successful translation into the clinic (61). While not restricted to this, the present study also suggests alternative avenues for the treatment of SPINK1+ malignancies, for instance inhibitors against DNMTs, HDACs or EZH2, several of which are already in clinical trials (62,63), or selective inhibition of MALAT1 by using antisense oligonucleotides. Conclusively, we moved the field forward by addressing an important question that how SPINK1 is aberrantly overexpressed in ETS-fusion negative PCa, and stratification of patients based on SPINK1-positive and miRNA-338-5p/-421-low criteria could improve therapeutic modalities and overall management strategies.
METHODS
Animals
For mice xenograft studies, we used five to six weeks old NOD.CB17-Prkdcscid/J (NOD/SCID) male mice (Jackson Laboratory) randomized into three groups (N=8 for each experimental condition) before implanting the cells. Mice were anesthetized using a cocktail of ketamine/xylazine (50 and 5 mg/kg respectively, via intraperitoneal route) and were subcutaneously implanted with 22RV1-CTL, 22RV1-miR-338-5p or 22RV1-miR-421 cells (2×106) suspended in 100μl of saline with 20% Matrigel into the dorsal both flank sides of the mice. A blinded assessment of tumor growth was conducted twice a week using digital Vernier’s calipers, and tumor volumes were calculated using the formula (π/6) (L × W2), (L=length; W=width). Spontaneous metastasis to lungs and bone marrow of the xenografted mice was analyzed by performing qPCR using primers specific for human specific Alu-sequences as mentioned in the Supplementary Table S7. All procedures involving mice were approved by the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA) and conform to all regulatory standards of the Institutional Animal Ethics Committee of the Indian Institute of Technology, Kanpur.
Human Prostate Cancer Specimens
All prostate cancer (PCa) specimens used in this study were procured from King George’s Medical University, Lucknow, India. Clinical specimens were collected after obtaining written informed consent from the patients and Institutional Review Board approvals from the King George’s Medical University, Lucknow and Indian Institute of Technology, Kanpur, India. A total of 20 PCa specimens were selected for this study based on the SPINK1 and TMPRSS2-ERG status, confirmed by qPCR, immunohistochemistry and Fluorescent in situ hybridization (FISH) for gene rearrangement (64). The PCa specimens used in this study were collected from men who underwent needle core biopsies and transurethral resection of the prostate (TURP) to relieve obstructive symptoms from locally advanced disease between year 2014 and 2016. None of the patients received preoperative radiation or androgen deprivation therapy. All patients included in this study were of Indian descent residing in the northern part of India and were de-identified. Prostate cancer tissue microarrays (TMA) specimens (n=238) were obtained from Dept. of Pathology, Henry Ford Health System, Detroit, Michigan, USA, after getting written informed consent from the patients and approval from Institutional Review Board. TMAs were stained for SPINK1 and EZH2 by performing immunohistochemistry (IHC) and RNA in situ hybridization (RNA-ISH) respectively.
Cancer Cell Lines and authentication
Prostate cancer cell lines (22RV1, VCaP and PC3), Colorectal (WiDr), Pancreatic (CAPAN-1), Melanoma (SK-MEL-173), prostate epithelial cells (RWPE-1) and human embryonic kidney 293T cells (HEK293T) were procured from the American Type Culture Collection (ATCC) and were maintained using ATCC recommended medium supplemented with 10% fetal bovine serum and Gibco Penicillin-Streptomycin (Thermo-Fisher). Cell lines were cultured in CO2 incubator (Thermo-Fisher) supplied with 5% CO2 at 37°C temperature.
To ensure the identity, short tandem repeat (STR) profiling of all cell lines were performed at the Lifecode Technologies Private Limited, Bangalore and DNA Forensics Laboratory, New Delhi. The profiles were compared with reference STR genotypes available at ATCC, DSMZ-German Collection of Microorganisms and Cell Cultures, and Biosample databases to authenticate the identity and check for any cross contamination. All cell lines were routinely tested for Mycoplasma contamination using PlasmoTest mycoplasma detection kit (InvivoGen).
Transfection of microRNA mimetics, antagomiRs and Small interfering RNA
Synthetic mimics and antagomiRs for the miR-338-5p and miR-421, and negative controls (Exiqon) were transfected using Lipofectamine RNAiMAX (Invitrogen) with a final concentration of 30pmol. The cells were seeded at 40% confluency and were transfected with respective miRNA mimics next day, followed by a second transfection after 24 hours. Subsequently, cells were processed for quantitative analysis and functional assays. Same transfection protocol was followed for On-Targetplus small interfering RNA (siRNA) for SPINK1 (J-019724-07, GE Dharmacon) and E2F1 (J-003259, GE Dharmacon).
Real-Time Quantitative PCR
Briefly, total RNA was extracted using miRNeasy Mini Kit (Qiagen) for miRNA related experiments or else TRIzol (Ambion), and 1μg of RNA with good integrity was reverse transcribed into cDNA using SuperScript III (Invitrogen) in the presence of random primers (Invitrogen). For Real Time Quantitative PCR (qPCR) all reactions were performed in triplicates using SYBR Green Master Mix (Applied Biosystems). The relative expression of the target gene was calculated for each sample by using the ΔΔCt method as described before (12,21). Sequences for all the primer sets used in this study are listed in the Supplementary Table S7.
TaqMan microRNA Assay
Total RNA including miRNA fraction was isolated using miReasy RNA extraction kit (Qiagen). QPCR for the miRNA stem-loop was performed using target-specific stem-loop reverse transcription primers using TaqMan microRNA reverse transcription kit (Thermo Fisher), followed by Taqman assays (Applied Biosystems) following manufacturer’s instructions on the Step OnePlus Real Time PCR System (Applied Biosystems). Relative expression of the target miR-338-5p, miR-421, and miR-876-5p (Applied Biosystems Assays IDs: 4427975, 4427975, 4427975 respectively) was normalized to RNUB6 (Assay ID: 4427975).
MicroRNA 3’UTR SPINK1 and MALAT1 luciferase reporter assay
A Firefly/Renilla Dual-Luciferase reporter vector pEZX-MT01 (GeneCopoeia) was used for cloning full length SPINK1 3’UTR wild type, and mutant with altered residues in the binding sites of miR-338-5p and miR-421. Similarly, MALAT1 wild type (250 bp) harboring miR-338-5p and miR-421 binding sites was also cloned into the same vector. Cells were seeded in a 24-well plate at 30-40% confluency, and co-transfected with 30pmol of miRNA mimics along with 25ng of pEZX-MT01 constructs using lipofectamine RNAiMax (Invitrogen). Luciferase assay was performed using Dual-Glo luciferase assay (Promega) 24 hours after the second transfection. Firefly luciferase activity was normalized to Renilla luciferase activity for each individual sample.
Cell proliferation, invasion and migration assays
For cell proliferation assay, cells were seeded in 12-wells culture plates (10,000 cells/well). At the indicated time points cells were trypsinized and counted on the Z-Series Coulter counter (Beckman Coulter). Cell invasion assays were performed using Transwell Boyden chambers of 8μm pore size (Corning) (21,65). Briefly, RPMI-1640 media supplemented with 20% FBS was added to the lower compartment, and 100,000 cells in serum-free media were added onto Transwell insert coated with Matrigel (BD Biosciences). After 24 hours incubation at 37°C with 5% CO2, the non-invading cells and Matrigel were gently removed with a cotton swab from the Transwell inserts. Invasive cells located on the lower side of the inserts were fixed in formaldehyde (4% in PBS) and stained with crystal violet (0.5% w/v). Images of the representative field were taken on the Axio Observer Z1 microscope (Zeiss). The invaded cells were quantified by de-staining with 10% (v/v) acetic acid in distilled H2O, and absorbance of de-staining solution was measured at 550nm. Same protocol was followed for cell migration assay, except no Matrigel was coated on the inserts.
Foci formation assay
For foci formation assay, cells (2×103) were plated in six-well culture dishes in cell line specific recommended culture media supplemented with 5% heat-inactivated fetal bovine serum (Invitrogen) and incubated at 37°C, media was changed every third day. The assay was terminated after 3 weeks and cells were fixed in formaldehyde (4% in PBS) and stained with crystal violet solution (0.05% w/v). Representative images were taken on the Axio Observer Z1 microscope (Carl Zeiss).
Soft agar colony assay
For anchorage-independent growth assay, soft agar plates were prepared by pouring 2ml of 0.6% low melting-point agarose (Sigma) dissolved in RPMI-1640 medium in 6-well dishes, after polymerization, second layer containing 2ml of 0.3% agar in RPMI-1640 medium supplemented with 10% FBS, and stable 22RV1-CTL, 22RV1-miR-338-5p and 22RV1-miR-421 cells (~1.5×104) resuspended were poured on the top of the first layer. Soft agar assay plates were incubated at 37°C for 20 days, and colonies greater than 40μm in size were counted.
Chick Chorioallantoic Membrane (CAM) assay
The chick embryo CAM assay was performed as explained previously (66). Briefly, fertilized eggs were incubated in a humidified incubator at 38°C for 10 days. The CAM was released by applying low pressure to the hole over the air sac and the shell was cut to make a square 1cm2 windows. Two million cells (22RV1-miR-338-5p, 22RV1-miR-421 or 22RV1-CTL) were implanted near the allantoic vein onto the CAM in 10 days post-fertilized eggs. The windows were subsequently sealed and the eggs were incubated at 38°C. For intravasation experiments, genomic DNA from lower CAM was isolated using Phenol/chloroform method and presence of the tumor cells was quantified by performing quantitative human Alu-specific PCR. The upper CAMs were isolated, fixed and immunostained for human-specific cytokeratin-18 as previously described (67). To assess tumor growth and metastasis, the assay was terminated on 18th day post-implantation and extra-embryonic tumor mass were excised and weighed. For metastases, embryonic lungs and liver was harvested, genomic DNA was isolated and subjected to quantitative human Alu-specific PCR as mentioned previously (66). Briefly, the standard curve for the Alu-specific PCR was prepared using different dilutions of genomic DNA (stock concentration 60ng/μl) isolated from 22RV1 cells along with 500ng of chicken DNA spiked in all the standards. About ~5ng of DNA template was used from each of these standards for the Alu-specific PCR. Further, the presence of Alu repeats in the human genomic DNA (cancer cells metastasized) isolated from embryonic lungs and liver was evaluated by performing Alu specific PCR using primers as listed in Supplementary Table S7.
H&E and immunostaining of tumor xenografts
Tumor tissues excised from the xenografted mice were fixed in 10% buffered formalin overnight, followed by dehydration using increasing concentration of ethanol. Subsequently, tumor specimens were embedded in paraffin and serially sectioned at 3μm thickness using microtome (Leica) as described earlier (21). Briefly, tissue sections were deparaffinized and dehydrated/rehydrated using standard protocol, followed by antigen-retrieval in the citrate buffer (pH 6.0) for 10 minutes at 100°C. Endogenous peroxidase activity was quenched using 3% hydrogen peroxide for 5 minutes. Sections were then blocked with 10% goat serum and probed with anti-mouse Ki-67 (1:400, CST, 9449S) at 4°C overnight, followed by secondary horseradish peroxidase (HRP)-conjugated antibody (DAKO), and HRP activity was detected using DAB (3, 3 -diaminobenzidine) peroxidase (HRP) substrate kit (DAKO). Quantification of IHC staining was performed in a blindfolded manner. The numbers of cells positive for Ki-67 staining were manually counted from ten random histological sections for each mouse.
Gene expression array analysis
For global gene expression profiling, total RNA was isolated from stable 22RV1-miR-338-5p, 22RV1-miR-421 and 22RV1-CTL cells as described earlier and subjected to Agilent Whole Human Genome Oligo Microarray profiling (dual color) using Agilent platform (8×60K format) according to the manufacturer’s protocol. A total of three microarray hybridizations were performed using each stable miRNA overexpressing cell line samples against control cells. Microarray data was normalized by following locally weighted linear regression (also known as Lowess) (68), and data was normalized using GeneSpringGX software for the raw data files. Differentially regulated genes were clustered using hierarchical clustering based on Pearson coefficient correlation algorithm to identify significant gene expression patterns. Further, for multiple hypotheses testing adjustments were applied using Benjamini and Hochberg procedure to calculate the FDR-corrected P-values (with FDR< 0.05) for the differentially expressed genes. Data was filtered to include only features with significant differential expression (log2 fold change greater than 0.6 or less than -0.6, P< 0.05) i.e. ~1.6-fold average over- or under-expressed genes, were then used for the enrichment of biological processes using DAVID bioinformatics platform. Further, enrichment of the biological pathways, and molecular signatures that were enriched upon miRNA overexpression with respect to control were analyzed using Gene set enrichment analysis (GSEA). A network based enrichment of critical miR-338-5p and miR-421 overlapping biological pathways was generated using Enrichment Map (69), a plug-in for Cytoscape network visualization software (http://baderlab.org/Software/EnrichmentMap/). The heatmap.2 function of R package ‘gplots’ was used to create the heat maps.
Western Blot analysis
Cell lysates were prepared in radioimmunoprecipitation assay (RIPA) lysis buffer, supplemented with complete protease (Roche) and phosphatase inhibitors mixture (Calbiochem). Protein samples were prepared in 1X SDS sample loading buffer; size fractionated on the SDS-PAGE and transferred onto a Polyvinylidene Difluoride membrane (PVDF) membrane (GE Healthcare). The PVDF membrane was then incubated for 1 hour at room temperature in blocking buffer [Tris-buffered saline, 0.1% Tween (TBS-T), 5% non-fat dry milk], and were incubated overnight at 4°C with the following primary antibodies: antiphosphor -MEK or -ERK rabbit (1:1000, 9121S or 4377S) or total-MEK or –ERK (1:1000, 9126S or 4695S), anti-E2F1 rabbit (1:1000, 3742S), anti-TET1 rabbit (1:2000, ab121587) and anti-β-Actin rabbit (1:3000, 4970S). Subsequently, blots were incubated with horseradish peroxidase-conjugated secondary anti-mouse or anti-rabbit antibody (1:5000, Jackson ImmunoResearch Laboratories) for 2 hours at room temperature, and were washed with 1X TBS-T buffer, and the signals were visualized by enhanced chemiluminescence system (GE Healthcare) as described by the manufacturer.
Immunofluorescence analysis
Cells were grown on the glass coverslips and fixed in 4% para-formaldehyde, washed with 1X PBS, permeabilized with 0.3% Triton X-100 in PBS for 10 min and blocked with 5% goat serum in 0.1% Triton X-100 in 1X PBS for 2 hours at room temperature. Subsequently, cells were incubated with the following primary antibodies: SPINK1 (1:100, Abnova, H00006690-M01), E-cadherin (1:400, CST, 3195S), N-cadherin (1:400, Abcam ab98952), Slug (1:50, CST, 9585S), Snail (1:50, CST, 3895S), c-Kit or CD117 (1:400, CST, 3308S), SOX-9 (1:400, Merck millipore, AB5535, a kind gift from Dr. A. Bandyopadhyay, IITK), TET1 (1:500, Abcam, ab121587). Subsequently, cells were incubated with Alexa Fluor-488 conjugated secondary anti-mouse or anti-rabbit antibodies (1:600, CST, 4412 or 4408). The coverslips with the stained cells were mounted on the slides using Vectashield with DAPI (Vector laboratories). Images were captured on the Axio Observer Z1 microscope (Carl Zeiss) equipped with high-resolution CCD camera or at LSM780LNO Carl Zeiss Confocal microscope.
Prostatosphere Assay
Briefly, prostate cancer cells (10000 cells/ml) were cultured in suspension in low-adherence plate using serum-free DMEM-F12 (Invitrogen) supplemented with B27 (1:50, Invitrogen), 20 ng/ml EGF (Invitrogen), 20 ng/ml FGF (Invitrogen) and Penicillin-Streptomycin (Thermo-Fisher Scientific) as previously described (70). Small population of cells which formed prostatospheres were collected by gentle centrifugation and were mechanically dissociated into single cells suspension and then passaged for several generations (4-5 passages at an interval of 10-12 days) following similar culture conditions and were assessed for the sphere initiation/forming efficiency or self-renewal capacity. Spheres larger than 50μm in diameter were counted and plotted as percent sphere-forming efficiency. Representative images of the sphere were taken on phase contrast mode on the Axio Observer Z1 microscope (Carl Zeiss). Total RNA was isolated at day 6 and 12 as described before and was subjected to qPCR for detecting the expression of cancer stem cell like markers.
Flow cytometry
For BrdU-7AAD cell cycle analysis, 22RV1 cells were transfected with mimics for miR-338-5p, miR-421 or control for 2 consecutive days, followed by BrdU pulse labelling for 2 hours after second transfection. Subsequently, cells were stained with anti-BrdU antibody conjugated to fluorescein isothiocyanate (FITC) (BrdU-7AAD flow kit, BD Biosciences), and then with 7-Aminoactinomycin D (7-AAD), following manufacturer’s instructions. Samples were subjected to FACS (BD FACS Calibur) excited at 488nm. Forward Scatter (FSC) and Side Scatter (SSC) parameters were adjusted to gate the population of interest. The 7-AAD signals were recorded on the linear scale while BrdU-FITC signals were recorded on the logarithmic scale.
For Hoechst side population assay, 22RV1-CTL, 22RV1-miR-338-5p or 22RV1-miR-421 cells were stained with Hoechst 33342 (5 μg/ml), in the presence or absence of an ABC efflux pump inhibitor, verapamil (used as a negative control) followed by incubating samples at 37°C in water bath for 2 hours with periodic agitation. Later, cells were centrifuged at 2000 rpm for 5 min at 4°C and resuspended in cold 1X PBS and washed twice. Samples were kept at 4°C until analysis. Propidium iodide (PI) was added at a concentration of 5μg/ml to exclude dead cells. For detection of side population (SP), Hoechst blue and red signals were acquired using a 460/50 and 670/30 nm band-pass filters respectively. While 7-AAD was excited at 488 nm and its emission was measured in logarithmic scale through a band pass filter of 670/30. Since, Hoechst Red signals are comparatively lower than that of Blue, a relatively higher laser power was used, and an optimal resolution of the SP cells was found using 30-35mW of power with the UV laser. A dim tail of SP cells enriched faction was gated using a dot plot displaying Hoechst Blue and Hoechst Red scatter. A minimum of 100,000 live cell events were acquired to resolve the SP cell population in each sample.
For Annexin PI staining, 22RV1 cells were transfected with mimics for miR-338-5p, miR-421 or control miRNA. After 24 hours, cells were washed with cold 1X PBS and were resuspended in 1X Binding Buffer at a concentration of 1×106 cells/ml. Subsequently, 5μl of FITC Annexin V and PI was added and incubated at room temperature for 15min. Subsequently, 400μl of 1X Binding Buffer was added to the samples and were analysed on BD FACSCalibur. Data acquisition was performed on BD FACSCalibur platform, and analysed using FlowJo version 10.7 (TreeStar).
Immunohistochemistry
TMA slides were incubated at 60°C for at least 2 hours. Slides were then placed in EnVision FLEX Target Retrieval Solution, High pH (Agilent DAKO, K800421-2) in a PT Link instrument (Agilent DAKO, PT200) at 750C, heated to 970C for 20 minutes, and then cooled to 750C. Slides were then washed in 1X EnVision FLEX Wash Buffer (Agilent DAKO, K800721-2) for 5 minutes. Slides were then treated with Peroxidazed 1 (Biocare Medical, PX968M) for 5 minutes and Background Punisher (Biocare Medical, BP974L) for 10 minutes with a wash of 1X EnVision FLEX Wash Buffer for 5 minutes after each step. Mouse monoclonal SPINK1 (Novus Biologicals, H00006690-M01) diluted 1:100 in EnVision FLEX Antibody Diluent (Agilent DAKO, K800621-2) was added to each slide, which were then cover slipped with parafilm, placed in a humidifying chamber, and incubated overnight at 40C. The next day, slides were washed in 1X EnVision Wash Buffer for 5 minutes and then incubated in Mach2 Doublestain 1 (Biocare Medical, MRCT523L) for 30 minutes at room temperature in a humidifying chamber. Slides were then rinsed in 1X EnVision Wash Buffer 3 times for 5 minutes each. Slides were then treated with a Ferangi Blue solution (1 drop to 2.5ml buffer; Biocare Medical, FB813S) for 7 minutes. Slides were rinsed 2 times in distilled water, then treated with EnVision FLEX Hematoxylin (Agilent DAKO, K800821-2) for 5 minutes. Slides were rinsed several times in distilled water, immersed in a 0.01% ammonium hydroxide solution, and then rinsed twice in distilled water. Slides were then dried completely. Slides were dipped in xylene approximately 15 times. EcoMount (Biocare Medical, EM897L) was added to each slide, which was then cover slipped.
RNA in situ hybridization
TMA slides were incubated at 60°C for 1 hour. Tissues were then de-paraffinized by immersing in xylene twice for 5 minutes each with periodic agitation. The slides were then immersed in 100% ethanol twice for 3 minutes each with periodic agitation, then air-dried for 5 minutes. Tissues were circled using a pap pen (Vector, H-4000), allowed to dry, and treated with H2O2 for 10 minutes. Slides were rinsed twice in distilled water, and then boiled in 1X Target Retrieval for 15 minutes. Slides were rinsed twice in distilled water, and then treated with Protease Plus for 15 minutes at 40°C in a HybEZ Oven (Advanced Cell Diagnostics, 310010). H2O2, 1X Target Retrieval, and Protease Plus are included in the RNAscope pre-treatment kit (Advanced Cell Diagnostics, 310020). Slides were rinsed twice in distilled water, and then treated with EZH2 probe (Advanced Cell Diagnostics, probe ID: 405491) for 2 hours at 40°C in the HybEZ Oven. Slides were then washed in 1X Wash Buffer (Advanced Cell Diagnostics, 310091) twice for 2 minutes each. Slides were then treated with Amp 1 for 30 minutes, Amp 2 for 15 minutes, Amp 3 for 30 minutes, and Amp 4 for 15 minutes, all at 400C in the HybEZ oven with 2 washes in 1X Wash Buffer for 2 minutes each after each step. Slides were then treated with Amp 5 for 30 minutes and Amp 6 for 15 minutes at room temperature in a humidity chamber with 2 washes in 1X Wash Buffer for 2 minutes each after each step. Red color was developed by adding a 1:60 solution of Fast Red B: Fast Red A to each slide and incubating for 10 minutes. Slides were washed twice in distilled water. Amps 1-6 and Fast Red are included in the RNAscope 2.5 HD Detection Reagents-RED (Advanced Cell Diagnostics, 322360). Slides were then treated with EnVision FLEX Hematoxylin (Agilent DAKO, K800821-2) for 5 minutes. Slides were rinsed several times in distilled water, immersed in a 0.01% ammonium hydroxide solution, and then rinsed twice in distilled water. Slides were then dried completely. Slides were dipped in xylene approximately 15 times. EcoMount (Biocare Medical, EM897L) was added to each slide, which was then cover slipped.
EZH2 and SPINK1 staining Evaluation Criteria
EZH2 expression intensity scoring by RNA-ISH for all the tumor foci was evaluated on the basis of the number of red dots/cell and were graded into five levels ranging from score of 0 to 4 as described previously (71). SPINK1 staining by IHC was used to evaluate SPINK1 positive and negative status of the PCa specimens. Further, an association between SPINK1 and EZH2 expression in patients’ samples was calculated by applying Chi-Squared contingency test (72) on GraphPad Prism.
Chromatin immunoprecipitation
Briefly, cancer cells (~80-90% confluency) were crosslinked with a final concentration of 1% formaldehyde for 10 minutes, followed by quenching with Glycine (125mM) for 10 minutes at room temperature, followed by washing with 1X PBS twice. Next, cell lysis was performed using lysis buffer [1% SDS, 10mM EDTA, 50mM Tris-Cl and protease inhibitor (Roche)] followed by sonication using Bioruptor (Diagenode) to obtain an average length of ~500bp DNA fragments. Chromatin immunoprecipitation (ChIP) assays were carried out using antibodies against E2F1 (CST, 3742), EZH2/KMT6 (Abcam, ab191250), H3K27me3 (CST, 9733) and control rabbit IgG (Invitrogen). Supernatant containing sheared chromatin were incubated at 4°C overnight with 4μg of E2F1 or EZH2 or H3K27me3 and IgG antibodies. Concurrently, the Protein G coated Dynabeads (Invitrogen) were blocked with 100μg/ml BSA (HiMedia) and 500μg/ml sheared salmon sperm DNA (Sigma) and incubated at 4°C overnight. Blocked beads were washed twice with 9:1 dilution buffer: lysis buffer [1% Triton X-100; 150mM NaCl; 2mM EDTA (pH 8.0); 20mM Tris-HCl (pH 8.0) with protease inhibitors] and were incubated with respective antibodies to form antibody-bead conjugates. The antibody-bead conjugates were then washed three times in a low salt wash buffer 1 [1% Triton X-100; 0.1% SDS; 150mM NaCl; 2mM EDTA (pH 8.0); 20mM Tris-HCl (pH 8.0) with protease inhibitors] and once in high salt wash buffer 2 (same as wash buffer 1, except 500mM NaCl). The antibody/protein/DNA complexes were eluted using elution buffer [100mM NaHCO3, 1% SDS, RNaseA and Proteinase K (500μg/ml each)]. DNA was isolated using phenol-chloroform-isoamyl alcohol extraction method, precipitated and washed with 70% ethanol, air-dried, and dissolved in nuclease free water (Ambion). QPCR was performed using appropriate primer sets as listed in Supplementary Table S7.
Methylated DNA Immunoprecipitation (MeDIP)
Genomic DNA was extracted from 22RV1 cells using QIAamp DNA Mini Kit and was sonicated to produce random fragments ranging from 300-1000 bp. About 4μg of fragmented DNA was used for MeDIP assay. The DNA was denatured for 10 min at 95°C and immunoprecipitated with 4μg of monoclonal antibody against 5-mC (Abcam ab10805) or 5-hmC (Abcam ab106918) and IgG (Santa Cruz, sc-2027) in a final volume of 500μl IP buffer (10mM sodium phosphate (pH 7.0), 140 mM NaCl, 0.05% Triton X-100) for 2 hours at 4°C. Dynabeads (40μl) were washed twice with 800 μl PBS-BSA (0.1%) for 5 minutes at room temperature, and then were resuspended in 40μl of 1X IP buffer. The resuspended Dynabeads were added to the samples and incubated for 4-5 hours at 4°C with end-over-end shaking using rotator stirrer. Beads were then collected and washed thrice with 700μl of 1X IP buffer. The beads were treated with proteinase K (500μg/ml) for 3 hours at 50°C, subsequently immunoprecipitated DNA was recovered by phenol-chloroform extraction followed by ethanol precipitation. Real-time PCR reactions were carried out with 40 ng of input DNA and 2μl of the immunoprecipitated DNA following manufacturer’s instructions on the Step OnePlus Real Time PCR System (Applied Biosystems). All reactions were performed in triplicates and the relative fold enrichment of 5-mC over 5-hmC was plotted.
RNA Immunoprecipitation (RIP) Assay
About ~80% confluent 22RV1 cells were harvested for RIP assay. RIP was performed using EZH2 antibody (Abcam, ab191250). Briefly, the cells were washed twice with ice-cold 1X PBS and scraped in 1X PBS with protease inhibitor (Invitrogen). Next, cells were incubated in RIP buffer [50 mM Tris-HCl (pH 7.9), 0.25 M NaCl, 1% Nonidet P-40 (NP-40), 10 mM EDTA, protease inhibitor cocktail and RNase inhibitor] for 30 min at 4°C. Further, cell lysate was obtained by centrifugation at 12,000 rpm for 10 min at 4°C and was incubated overnight with 4μg of EZH2 antibody or IgG control (Invitrogen). Simultaneously, Protein G coated Dynabeads (Invitrogen) were pre-absorbed with 100μg/ml BSA at 4°C. The pre-absorbed beads were washed thrice with NT2 buffer [50 mM Tris-HCl (pH 7.4), 300 mM NaCl, 1 mM MgCl2, 0.05% Nonidet P-40 (NP40), 1XPIC, RNase inhibitor] and then incubated with RNA-antibody complex to form RNA-antibody-bead precipitates. Further, the beads were washed thrice with NT2 buffer, followed DNase I digestion for 15 min at 37°C and washed twice with NT2 buffer. Co-purified RNA was extracted by Trizol (Invitrogen), followed by cDNA synthesis using the SuperScript kit (Invitrogen). MALAT1 and GAPDH expression was analysed by qPCR as described before.
Bisulfite sequencing
Bisulfite conversion of the genomic DNA was carried out using EpiTect Bisulfite Kit (Qiagen) following manufacturer’s instructions. Briefly, bisulfite converted DNA was used as template for PCR amplification using primers (Macrogen Inc., South Korea) designed using the Methprimer software as listed in Supplementary Table S7. The amplified PCR product was purified using QIAquick PCR purification kit (Qiagen), cloned into pGEM-T Easy Vector (Promega) and transformed into One Shot TOP10 competent cells (Invitrogen). Plasmid DNA was isolated from eight independent colonies and was outsourced for conventional Sanger sequencing at Macrogen Inc., South Korea. The BiQ Analyzer online tool was used to calculate the methylation percentage and to generate the graphical plots.
Data Mining and Computational Analyses
MicroRNA target Prediction by Multiple Programs
MicroRNA prediction programs, namely miRanda, miRMap, PITA and RNA Hybrid were used to predict miRNAs targeting 3’UTR of SPINK1 (Fig. 1A and Supplementary Table S1). The correlation between the expression of the predicted miRNAs and SPINK1 was analysed by employing RNA Sequencing data for the TCGA-PRAD cohort. For Fig. 1A (lower panel), 4C and Supplementary Table S6 miRanda was used to predict the putative binding sites of the miR-338-5p and miR-421 on the 3’UTR of target genes.
Integrative analyses for TCGA-PRAD data
For gene association studies between miRs-338-5p, -421, miR-876-5p, SPINK1, ERG, EZH2 and MALAT1 Illumina HiSeq mRNA and miRNA-Seq data along with clinical information from TCGA-PRAD dataset was downloaded. Overexpression of SPINK1 in PCa exhibits outlier-expression in ~10-15% of the total PCa cases (Tomlins et al., 2008). Thus, to stratify patients with increased expression of SPINK1, we sorted TCGA patients’ samples on the basis of increasing SPINK1 expression (descending order), and divided the dataset into four equal parts by employing Quartile-based normalization method (73), the top 25% of the patients (N=119) corresponding to the upper quartile (QU, log2 (RPM+1)>5.468 or log2 (normalized count+1)>1.892), were assigned as SPINK1 high or SPINK1-positive patient samples and the lower quartile (QL, log2 (RPM+1)<1.124 or log2 (normalized count+1)<-2.611), were considered as SPINK1 low or SPINK1-negative samples. Also, we found about 18 patients with outlier expression of SPINK1 with log2 (RPM+1) of greater than 11.984, which were included in the heat map representation of SPINK1 positive TCGA patients in Figure 1B. No further cut-offs were applied for miR-338-5p, miR-421, miR-876-5p and ERG expression, corresponding expression values (based on SPINK1 cut-off) were considered for these genes for further analysis. Hierarchical Clustering of miRs-338-5p, -421 or 876-5p, SPINK1, and ERG were employed using heatmap.2 of R’s gplot package, which uses Euclidean distance to obtain distance matrix and complete agglomeration method for clustering between each genes and miRNA.
For Kaplan-Meier survival analysis, the survival data included sample type (primary tumors), days to first biochemical recurrence and days to last follow-up for TCGA-PRAD patients was considered. The samples were divided into two groups, higher and lower miRNA expression groups according to the expression level of a miR-338-5p and miR-421 using Cox proportional hazards regression model in R. Next, we performed a 13-year survival analysis of these miRNAs using Kaplan-Meier survival analysis (74) by employing survival package (https://cran.r-project.org/web/packages/survival) in the R environment, and statistical significance was computed using the log-rank test. For clinical relevance of miR-338-5p and miR-421, TCGA-PRAD dataset was analyzed for the association of these miRNAs with clinical parameters such as primary Gleason score, Clinical T score and positive lymph node status. Data analysis was performed by one-way analysis of variance with Tukey’s post hoc test for multiple comparisons, and student’s t-test was applied for comparison between two groups.
The MSKCC cohort (Cancer Cell, 2010) data was retrieved from cBioPortal (http://www.cbioportal.org/) for SPINK1 and EZH2 expression in the prostate cancer patients (N=85), and oncoprints were generated using default parameters (mRNA expression z-score threshold ±2 vs normal). Further, to ascertain possible association between EZH2, SPINK1, MALAT1 and miR-338-5p/-421, TCGA patients’ samples were stratified on the basis of increasing EZH2 expression, and divided the dataset into four equal quartiles, the top 25% of the patients (N=119) corresponding to the upper quartile (QU, log2 (normalized count+1) >7.313), were considered as EZH2 high patient samples and the lower quartile log2 (normalized count+1) <6.36), were considered as EZH2 low samples. The corresponding expression values for SPINK1, MALAT1, miR-338-5p and miR-421 in EZH2 high and EZH2 low groups (without further cut-offs) were considered to association with EZH2 expression (related to Fig. 6B).
Statistical analysis
Statistical significance was determined by either two-tailed Student’s t test for independent samples or one-way Analysis of Variance (ANOVA), otherwise specified. The differences between the experimental groups were considered significant if the P-value of less than 0.05 was obtained. Error bars represent mean ± SEM. All experiments were repeated three times in triplicates.
Data availability
The gene expression microarray data from this study has been submitted to the NCBI Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) under the accession number GSE108558.
Authors’ contributions
V.B., A.Y., and B.A. designed and directed the experimental studies. V.B. and A.Y. performed in vitro cell line-based studies. V.B., A.Y. and S.N. performed the bisulfite sequencing experiments and analysis. V.B., A.Y., R.T. and S.G. performed the gene expression studies, bioinformatics analysis and ChIP assays. A.Y. performed the immunofluorescence and RIP experiments. V.B., A.Y., and B.A. performed statistical analysis and interpreted the data. V.B., A.Y., and B.A. executed the in vivo experiments. A.G. provided PCa patient specimens. S.C., N.G., and N.P. performed immnuohistochemistry and RNA in situ staining on the PCa tissue microarrays. V.B., A.Y., and B.A. wrote the manuscript. B.A. directed the overall project.
Acknowledgements
B.A. is an Intermediate Fellow of the Wellcome Trust/DBT India Alliance. This work is supported by the Wellcome Trust/DBT India Alliance Fellowship [grant number: IA/I(S)/12/2/500635] awarded to B.A. We thank Yuping Zhang, Brendan Veeneman, Mahendra Palecha, Ayush Praveen for their technical support and Anjali Bajpai for critically reading the manuscript. We also thank Jonaki Sen for extending the use of fertilized eggs facility. The IIT Kanpur has filed a patent (IN 201611016564) on the therapeutic applicability of miR-338-5p and miR-421 described in this study in which B.A., V.B. and A.Y. are named as inventors.
Footnotes
Financial support: This work is supported by the Wellcome Trust/ DBT India Alliance grant (IA/I(S)/12/2/500635 to BA).
Conflict of interest: The authors declare no conflicts of interest or disclosures.