Introduction

Prostate cancer (PCa) remains one of the major medical burdens in males. The estimated number of new cases and deaths from PCa in the United States in 2017 is 161,360 and 26,730, respectively [1]. The incidence rate of PCa is lowest in Asian countries and in India it is the sixth most commonly diagnosed cancer among men [2]. Although prostate-specific antigen (PSA) detection in serum has facilitated the early detection of prostate cancer, there are limitations to this as elevated serum PSA is not specific to the malignant disease and it gives a high false-positive and a false-negative rate of approximately 15% [3]. In addition, PSA as a prognostic marker also has detrimental effects as screening detects indolent tumors along with those that can become life threatening over a period of time [4]. Moreover, PCa patients treated for localized prostate cancer show relapse within 5 years [5]. This lack of sensitivity and specificity in serum PSA level calls for a better PCa biomarker. Thus, miRNA profiling in the serum of PCa patients can lead to identification of putative biomarkers for PCa and can be a tool to establish patients with various stages of PCa and, therefore, can be treated accordingly.

MicroRNAs (miRNAs) are a class of non-protein coding endogenous 19–20 nucleotides long small single-stranded RNAs which were originally discovered in Caenorhabditis elegans lin four locus and 7 years later in mammals let seven was discovered [6]. These are evolutionarily conserved in nature [7]. MiRNAs are found to have a regulatory role and they regulate around 1–5% of the human genome and at least 30% of the protein coding gene [8]. They negatively regulate target gene expression at the post-transcriptional level either by degrading the target mRNA or inhibiting translation of the mRNA into functional proteins [9]. MiRNA is found to regulate vital functions of the cell including apoptosis, proliferation, cell cycle, differentiation, stem cell maintenance and metabolism [10]. The database miRBase reports around 2000 miRNA discovered in the human genome and it is believed that 30% of human genome is under miRNA regulation [11].

Various miRNAs reported have been shown to regulate the hallmarks of cancer and dysregulation of miRNA expression profile is associated with numerous human cancers, including lung, brain, liver, colon, breast, leukemia and prostate cancer [12,13,14]. Conceptually these miRNAs may function as tumor suppressors and oncogenes [12] depending on the target tumor suppressor gene or oncogenes, respectively. For example, miR-15 and miR-16 act as tumor suppressors by targeting anti-apoptotic gene B cell lymphoma 2 (BCL2) mRNA and miR-17-19 acts as oncogene by targeting two tumor suppressor genes phosphatidylinositol-3,4,5,-triphose-3-phosphatase (PTEN) and retinoblastoma like protein-2 (RB2) [13]. Since miRNAs acts as both tumor suppressors and oncogenes and the fact that miRNAs are diferentially expressed in cancer [14] as compared to normal tissue indicates that miRNAs can be potential clinical targets for cancer therapy.

The development of minimal invasive tests for the detection and monitoring of malignancies can greatly reduce the burden of cancer [15]. One such approach is miRNA as blood-based (circulating) cancer biomarkers as they can be readily detected in small volume of samples using specific and sensitive quantification using real time PCR [15]. MiRNAs can be isolated from the most body fluids, including serum, plasma, urine, saliva, breast milk, tears and semen [16]. They are highly stable in circulation as they are resistant to RNase degradation due to their short sequence [17, 18]. The packaging of miRNAs in lipid vesicles, bound by RNA-binding proteins and associated with high-density lipoprotein, has a wider relevance as miRNA can be transferred between individuals orally [16]. Moreover, being highly conserved between species, it allows the use of animal models for pre-clinical studies [16]. Looking at the redundant role of miRNA, it can be expected to have differential expression pattern in different stages of cancer depending on the requirement of cells and the relationship of this differential expression pattern of miRNAs with PCa hallmarks can have great therapeutic application.

Although several studies have investigated the differential expression pattern of miRNAs in PCa and their association with clinicopathological parameters, unique signatures that can be used as biomarker to detect PCa, its prognosis, therapy selection and response are missing. Thus, in this review, we have identified the most consistently reported dysregulated microRNAs in prostate cancer and have tried to evaluate the genes being regulated by these four microRNAs viz., miRNA-141, miRNA-375, miRNA-221 and miRNA-21 in regulating prostate cancer hallmarks.

MicroRNA biogenesis

The biogenesis of miRNA is a cascade of events which comprises three stages: miRNA transcription, miRNA maturation and finally RISC complex formation (Fig. 1). The first stage, miRNA transcription, initiates in the nucleus with transcription of miRNA by RNA polymerase II [19, 20] which generates 5′ capped and 3′ polyadenylated primary transcripts (pri-miRNA) of variable lengths (around 1–3 kb) [21]. The second stage, miRNA maturation, is catalyzed by a ribonuclease (RNase III), called Drosha which is a large protein of ~ 160 kDa and its cofactor DiGeorge syndrome critical region gene 8 (DGCR8) together forms the microprocessor complex (500–650 kDa) and leads to the cleavage of pri-miRNA into precursor-miRNA (pre-miRNA), a sterm-loop structure, comprising of ~ 70 nucleotides [22,23,24]. Pre-miRNA is then exported to the cytoplasm by a nuclear export factor called exportin-5 [25], where another RNase III, Dicer along with TAR (HIV) RNA-binding protein cleaves it into an RNA duplex of ~ 22 nucleotide, which is a double-stranded miRNA–mRNA duplex of the mature miRNA and its complementary strand [26, 27]. The mature miRNA then associates with argonaute proteins to form an miRNA-protein complex called RNA-induced silencing complex (RISC) or RNA interference (RNAi) complex and the complementary strand gets degraded [28]. Depending on the complementarity between the “seed” sequence of the miRNA and the “seed-match” sequence of target mRNA, the target mRNA is either degraded by Ago1/2 or leads to translational repression [29].

Fig. 1
figure 1

Schematic representation of microRNA biogenesis

Dysregulated microRNAs in prostate cancer

Circulating miRNAs being abundant in blood, stable, a non-invasive approach for cancer detection and the findings that human blood comprises stably expressed miRNAs [30] have drawn the attention of various studies investigating the potential of miRNAs as blood-based biomarkers. An increasing number of miRNA expression studies investigating differential expression of miRNAs as potential diagnostic, prognostic and predictive tools have been reported, suggesting the potential of miRNAs as novel biomarkers for prostate cancer therapy (summarized in Table 1). The first report on miRNA profiling in prostate cancer was published in 2007 [31]. This study comprised both in vitro and in vivo samples from benign prostatic hyperplasia (BPH) and prostate cancer patients. The expression of 319 miRNAs was analyzed in 6 prostate cancer cell lines, 9 prostate cancer xenografts samples, and 13 clinical prostate tissue samples (4 BPH, 5 untreated prostate carcinomas, and 4 hormone refractory prostate carcinomas) [31]. Expression of only 128 miRNAs (40%) was detectable in array hybridization which was further validated by dot blot hybridization as well as qRT-PCR [31]. Between BPH and carcinoma samples, 51 miRNAs were found to be differentially expressed, 37 miRNAs were found to be downregulated in carcinoma samples and 14 were found to be upregulated [31].

Table 1 Differential expression profile of microRNAs in PCa

As summarized in Table 1, initial studies have reported various miRNAs as differentially expressed in different stages of prostate cancer; however, data are inconsistent and no overlapping expression pattern has been found. The high variability among the data reported by various groups could be because of various factors such as sample size, sample type and screening methodology. Moreover, not all the differentially expressed miRNAs were validated. By validating clinically relevant miRNA in prostate cancer, one can come up with miRNA targets that have clinical application. Thus, identifying consistently reported differentially expressed microRNAs can widen up new horizons for PCa management. In this review, we have identified a panel of four most consistently reported, differentially expressed miRNAs in prostate cancer. These miRNAs: miR-141, miR-375, miR-221 and miR-21 depicted similar trend in multiple studies with different study structure and exhibited consistent significance.

The first report on miR-141 as a potential diagnostic marker was reported in the year 2008 by Mitchell et al. [32]. The authors validated six miRNAs including miR-141 in the serum samples of a case control cohort of 25 metastatic PCa patients with 25 age-matched healthy controls [32]. Of all the miRNAs miR-141 showed the greatest differential expression (46-fold overexpressed) in the patient sample and could differentiate between advanced metastatic PCa cases and healthy controls with a specificity of 60% and sensitivity of 100% [32]. Later in the year 2011, contradictory results were reported by Yaman Agaoglu et al., who investigated the expression pattern of three miRNAs, miR-21, miR-141 and miR-221 and found that there was no difference in the expression level of miR-141 when localized, local advance and metastatic PCa patient’s plasma samples were compared with healthy controls [33]. However, the expression, miR-21 and miR-221, could differentiate between patients and healthy controls [33]. Also, among the three miRNAs investigated miR-141 was reported to be the strongest discriminator of metastatic PCa from localized/local advanced disease [33]. Another study was carried out in the same year by Brase et al., where they identified 69 miRNAs to be upregulated in the serum of metastatic PCa patients [34]. MiR-141 along with other two miRNAs (miR-375 and miR-200b) showed the highest correlation with clinical parameters [34]. Followed by this in 2012, Selth et al. used transgenic adenocarcinoma of mouse prostate (TRAMP) mouse model of the prostate to discover miRNAs associated with PCa [35]. They identified eight miRNAs based on their serum levels and human homologs were further validated in the sera of 25 human PCa patients with metastatic CRPC and 25 healthy controls [35]. The authors showed that miR-141 along with other three miRNAs (miR-298, miR-346 and miR-375) were consistently upregulated in PCa patients with metastatic CRPC compared to healthy controls [35]. Another study in the year 2012 by Nguyen et al. showed that the expression of miR-141 along with miR-375 and miR-378* could discriminate between metastatic PCa patients and low-risk localized patients [36]. Also the expression of miR-141 was significantly higher in prostate tumor samples compared to normal prostate tissue [36]. Taken together, these studies indicated miR-141 as a marker of metastatic prostate cancer which could differentiate between healthy controls, primary PCa patients and metastatic CRPC patients with high specificity and sensitivity and also correlate with PCa aggressiveness.

MiR-375 was first identified as pancreatic islet-specific miRNA that regulates glucose-induced insulin secretion in murine embryonic β cell line MIN6 [37]. However, various miRNA expression profiling studies revealed dysregulated expression of miR-375 in various malignancies for instance, hepatocellular carcinoma [38], gastric cancer [39], head and neck cancer [40], esophageal carcinoma [41], melanoma and glioma [42]. MiR-375 as a biomarker for PCa was first reported in the year 2011 by Brase et al. [34]. The authors identified 69 miRNAs which were upregulated in the serum samples of metastatic PCa patients compared to primary PCa patients [34]. Three miRNAs including miR-375 showed the highest correlation with tumor stage and Gleason score [34]. Also, miR-375 expression showed considerable association with lymph-node metastasis; however, it could not discriminate between high-risk PCa patients (Gleason score 8) and intermediate risk PCa patients (Gleason score 7) [34]. MiR-375 also showed higher expression in prostate tumor tissues compared to normal epithelium [34]. A similar study by Bryant et al. in the year 2012 reported differential expression of 16 miRNAs including miR-375, in plasma samples from 16 metastatic PCa patients compared to 55 localized PCa patients [43]. Expression of three miRNAs including miR-375 along with miR-200b could discriminate between metastatic PCa patients and localized PCa patients [43]. Also the expression of miR-375 was found to show correlation with metastatic CRPC in a study by Selth et al. [35]. Similar results were reported by another group in the year 2013 (Nguyen et al.) where they showed that the expression of miR-375 was found to be upregulated in serum from CRPC patients compared to serum from localized PCa patients [43]. Another study in the year 2015 highlighted the potential of miR-375 as prognostic biomarker of PCa where the authors reported significant association of elevated miR-375 levels with shorter overall survival (mortality rate approximately 80%). Thus, miR-375 expression correlates with clinicopathological parameters and can act as a prognostic biomarker of PCa [44].

MiR-221 is encoded in tandem located on the X chromosome (Xp11.30) in human, mouse and rat and is highly conserved in vertebrates [45]. MiR-221 is reported to act as an oncomir in various epithelial cancers including prostate cancer [46,47,48,49]. MiR-221 as a biomarker of PCa was first reported in the year 2011 by Agaoglu et al. [33]. The authors reported miR-221, along with miR-21, to be elevated in the plasma samples of PCa metastatic patients compared to healthy controls [33]. Also the expression of miR-221 was found to be significantly higher in metastatic PCa cases compared to localized/local advanced PCa [33]. Yet in another study (Selth, 2012) miR-221 in combination with miR-20a, miR-21 and miR-145 could discriminate between intermediate or high-risk PCa patients compared to low-risk PCa as categorized by cancer of the prostate risk assessment (CAPRA) score and D’Amico’s criteria [35]. However, there was no significant association between independent expression of miR-221 and PCa aggressiveness by either CAPRA or D’Amino score [35]. Collectively, these studies show that the expression of miR-221 is associated with metastatic PCa and can discriminate between different stages of PCa.

MiR-21 is located in chromosome 17q23.2 in human and is evolutionarily conserved across vertebrate species [50]. The expression of miR-21 is found to be overexpressed in various human tumors and cancer cell lines [51,52,53], including prostate tumors and it is reported to be an oncogene targeting various tumor suppressor genes [54, 55]. As mentioned above, miR-21 along with miR-221 was reported by Agaoglu et al. in 2011 to be upregulated in metastatic PCa patients compared to healthy controls and could discriminate between metastatic and localized/local advanced PCa cases [33]. Zhang et al. in the same year showed that the expression of miR-21 is upregulated in androgen-dependent PCa (ADPC) and castration-resistant PCa (CRPC) group of patients compared to localized PCa and BPH [56]. Also, serum levels of miR-21 corresponded to that of PSA levels in ADPC and CRPC patients [56]. Moreover, patients who were resistant to chemotherapy had an elevated level of miR-21 compared to responsive group of patients [56]. However, expression of miR-21 could not discriminate between localized PCa and BPH [56]. In a study by Shen et al., expression of miR-21 along with miR-20a was significantly associated with CAPRA score as well as with clinicopathological variables of PCa [57]. Thus, miR-21 expression correlates with CAPRA score and PSA levels and can work as a diagnostic and prognostic marker in PCa.

The molecular basis for dysregulated expression of microRNAs in prostate cancer

The molecular mechanism underlying the dysregulated expression of microRNAs in prostate cancer includes change in copy number of microRNAs [62], epigenetic modifications (DNA methylation and histone modification) [63], upregulated expression of Dicer [64], mutations in the stem regions of pre-miRNAs [65], single nucleotide polymorphisms (SNPs) [66] and androgen receptor (AR) regulated mechanisms [67]. Copy number alterations of miRNAs and their regulatory genes are highly prevalent in cancer and may account partly for the dysregulated microRNA expression profile in many cancers including prostate cancer [62]. It has been seen that in epithelial cancers there is a high frequency of copy number alterations in microRNA containing regions of the genome which correlates with the respective microRNA expression [62]. This copy number alterations correlates with miRNA transcript expression; however, the same does not hold true for many miRNAs. Also microRNA regulation is controlled by transcriptional regulation of the host gene [68]. Since half of the microRNA genes are located in introns of protein coding gene [69], these are more susceptible to epigenetic silencing by aberrant methylation of the CpG island located in the 5′UTR of the host gene [68]. Moreover, as half of the human promoter regions comprises CpG islands, thus, differential promoter methylation may lead to aberrant expression of microRNAs [70]. Additionally, there exists a bidirectional cross-control mechanism between microRNAs and epigenetic machinery in which microRNA expression can be regulated by the epigenetic system and in return the components of epigenetic machinery are modulated by microRNAs [71]. For instance, miR-29 expression showed inverse correlation and downregulated the expression of DNA methyltransferases (DNMT-3A and -3B) in lung cancer (55). Furthermore, enforced expression of miR-29 leads to decreased global DNA methylation with simultaneous restoration of expression of tumor suppressor genes. Besides these, the position of microRNA in the genome (fragile sites, genomic regions of loss heterozygosity, or cancer-associated regions) [72] and mutations in the microRNA processing machinery also contributes to the microRNA deregulation in cancer [73]. For instance, miR-142 located on chromosome 17 but also found in the breakdown junction of a t(8,17) translocation causes an aggressive B cell leukemia by upregulation of a translocated c-Myc gene which is under the control of an upstream miR-142 promoter [74]. The upregulated expression of Dicer along with other components (EIF2C2, EIF2C1, XPO5, MOV10, HSPCA and TNRC6B) of miR machinery also contributes towards deregulated expression of microRNAs and also correlates with the aggressiveness associated with PCa metastasis and Gleason score [64]. Human Dicer is located in the subtelomeric region of chromosome 14 (14q32.13) and any genomic instability at 14q32 induces Dicer upregulation [64]. In addition, single nucleotide polymorphisms (SNPs) within precursor-microRNAs (pre-miRNAs) also affect microRNA expression levels [75]. For instance, in papillary thyroid carcinoma, an SNP in the pre-miR-146a is reported to decrease mature miRNA expression [76]. Another aspect of regulating the microRNA expression in PCa is androgen-induced androgen-receptor (AR) which binds to the promoter region of microRNA and leads to its over expression followed by androgen-dependent (AD) cell growth and castration resistance in PCa [67].

MicroRNAs and androgen receptor signaling

Prostate cancer cells depend on androgen for its growth and survival [77]. Thus, patients with metastatic prostate cancer are often treated with drugs that block androgen production [78]. These drugs are provided with antiandrogens such as analogs of luteinizing hormone releasing hormones that work as androgen receptor antagonist [79]. Studies carried out by Chen et al. have shown that androgen works in a paracrine manner in normal prostate cells where binding of androgen receptor to its ligand leads to its dimerization and, thereby, it regulates specific genes in the stromal and epithelial prostate cells, regulating proliferation and survival of epithelial cells [78]. However, in prostate cancer cells the paracrine pathway gets converted to the autocrine pathway where androgen directly stimulates proliferation of malignant prostate cells [80]. The postulated mechanisms underlying the reason behind relapse to an unresponsive hormone refractory stage can be divided into four categories, the first category being the mutations such as amplifications or point mutations of the androgen receptor gene which alters the response of the receptor and sensitizes the cancerous cells even to low concentration of androgens [81], also the antagonists start behaving like agonists; the 2nd category includes cases where activation of androgen receptor takes place independently as a consequence of cross talk between AR signaling pathway and other pathways such as epidermal growth factor receptor, Akt pathway, mitogen-activated protein kinase (MAPK) signaling induced by oncogenes [82]; however, the kinases and substrates involved in this mechanism are unknown; the 3rd category includes alterations in the equilibrium between coactivators and corepressors [83] and the 4th category comprises alternating pathways that bypass androgen receptor signaling and progression of the disease which is independent of androgen receptor [84, 85].

Thus, it shows that AR signaling pathways play a crucial role in PCa progression and, hence, the microRNAs that regulate this signaling pathway are of utmost importance. Therefore, in the present review we have compiled the role played by this panel of miRNAs in apoptosis in PCa.

MicroRNA-141 and androgen receptor

Waltering et al. 2011 showed that very few microRNAs are androgen regulated in both cell lines and xenografts of which miR-141 was upregulated in both PCa cases and castration-resistant prostate cancer (CRPC) compared to BPH [86]. The authors analyzed the microRNA expression in LNCaP-derived models, VCaP cell lines, 13 castrated prostate cancer (PCa) xenografts and clinical samples of untreated PCa patients and CRPC, using microarray and q-RT-PCR [86]. They also studied the functional significance of miR-141 in PCa by overexpressing and suppressing miR-141 in cell lines [86]. The forced expression of miR-141 increased the proliferation of LNCaP cell lines in low concentration of dihydrotestosterone (DHT) and the suppression of miR-141 by anti-miR-141 reduced LNCaP proliferation in low androgen medium, indicating the role of miR-141 in enhancing the growth of CRPC cells in the androgen depleted environment [86].

Another study by Xiao et al. [87] showed the correlation of Shp (small heterodimer protein) expression with miR-141 and subsequently its effect on AR signaling pathway in malignant and non-malignant PCa cells. The Shp mRNA and protein were found to be downregulated in PCa cell lines compared to non-malignant human prostate epithelial cell line (RWPE-1), which was confirmed by real-time PCR, western blot and immunofluorescent staining [87]. The authors also found an upregulated expression of miR-141 in PCa cell lines compared to control. Using miR-141 precursor and anti-miR-141, the authors showed that transfection with anti-miR-141 downregulated the expression of Shp protein in RWPE-1 cell line in a dose-dependent manner and an increase in Shp protein was seen in PCa cell lines after treatment with anti-miR-141 [87]. This suggests that Shp is a target of miR-141 [87]. Thus, the above study showed that Shp, which is a known corepressor and metabolic regulator, is a target of miR-141 and the downregulation of Shp, induced by upregulated miR-141, transcriptionally regulates androgen receptor genes in prostate cells indicating the importance of miR-141 in prostate cancer progression.

MicroRNA-375 and androgen receptor

MiR-375 acts as a tumor suppressor in various cancers viz., esophageal squamous cell carcinoma [88], oral squamous cell carcinoma [89], pancreatic cancer [90], squamous cervical cancer [91], gastric carcinomas [39], head and neck squamous cell carcinomas [92], and melanoma [93]. However, in hormone-dependent cancers, namely prostate cancer and breast cancer, miR-375 is found to be overexpressed and thus is hypothesized to exert an oncogene function [94, 95].

A study by Chu et al. showed that the differential expression pattern of miR-375 is determined by the methylation mediated transcriptional repression of the miR-375 promoter. The AR-positive PCa cells exhibits lower levels of miR-375 and AR-negative PCa cells display a higher level of miR-375 [96]. Androgen receptor negatively regulates the DNA methyltransferases (DNMTs) activity in PCa cells, thereby leading to either hypermethylations or hypomethylation of the miR-375 promoter in AR-negative cells (PC-3) and AR-positive cells (LNCaP), respectively [96]. Also using a demethylating agent such as 5-Aza-dC in AR-positive cells, the lower expression of miR-375 could be reversed, indicating DNA methylation to be a major driving factor in regulating miR-375 expression in prostate cancer cells [96].

The dysregulated expression profile of miR-375 is only relevant to malignant PCa cells such as PC-3 whereas no significant attenuation was observed in benign prostate epithelial cells (RWPE-) [97], indicating a dual role of miR-375 in PCa progression depending on the stage and hormone status [97]. The molecular targets of miR-375 are cyclin D2 (CCND-2) and retinoblastoma 1(RB1) which were seen to be downregulated upon forced expression of miR-375 in PCa cell lines [97].

MicroRNA-375 has been reported to form a complex with miR-93/miR-106b and targets capicua (CIC), which is an HMG box-containing transcriptional repressor [98]. Knock-down of CIC leads to increased expression of cellular retinoic acid binding protein 1 (CRABP1) [99]. CRABP1 is known to have pro-tumorigenic and pro-metastatic activity in mesenchymal tumors [100] and is known to be upregulated in androgen-independent PC-3 cells [101] and CRPC [102].

Another study reported sec23 homolog A, coat complex II component (SEC23A) to be a target of miR-375 [94]. SEC23A works in interaction with N-myc downstream-regulated gene 1 (NDRG1) and the NDRG1 interactions are androgen-regulated driving androgen-dependent PCa to androgen-independent PCa [94]. Further, it has been reported that MHC-I molecules interact with SEC23-24 for their endoplasmic reticulum (ER) to Golgi trafficking, suggesting downregulation of SEC23A by miR-375 which in turn reduces the immunogenicity of PCa cells by reducing the expression of MHC-I [94].

MicroRNA-221 and androgen receptor

The expression of miR-221 is upregulated about 6- to 10-fold in LNCaP-derived castration-resistant counterpart androgen-independent (AI LNCaP-abl) cells compared to androgen-dependent LNCaP cells [103]. Upregulating the expression of miR-221 in androgen-dependent (AD), LNCaP cells showed a negative regulation of the AR mediated PSA level and androgen-mediated cell growth; however, it showed no influence on AR expression [103]. Also, transfection of AI LNCaP-abl (CRPC) cells with anti-miR-221 could restore the AR mediated PSA level and androgen-mediated cell growth, indicating the importance of miR-221 in maintaining the CRPC phenotype in PCa [86]. Cyclin-dependent kinase inhibitors, p27/kip1 and p57/kip2, are reported as miR-221 targets [103]. Upregulated expression of miR-221 in PC-3 and LNCaP cell lines and in glioblastoma is associated with upregulated p27/kip1 expression and results in growth inhibition and inability to form colonies in soft agar; however, in AI LNCaP-abl (CRPC) cells upregulated expression of miR-221 had no influence on the p27/kip1 expression, suggesting various other genes which may be involved along with p27/kip1 in maintaining CRPC phenotype in PCa [103].

MiR-221 is also shown to play a role in neuroendocrine (NE) differentiation of prostate cells [104]. NE diferentiation is hypothesized to be a major regulator of CRPC [105], and is associated with various cancerous phenotypes because NE cells do not proliferate; they secretes elevated levels of survival genes viz., Survivin [106] and Bcl-2 [107]; they also secrete certain growth factors and hormones that support the growth of surrounding tumor in a paracrine manner [106]. It has been reported that miR-221 promotes NE differentiation in hormone-sensitive LNCaP cells and sustains the growth of LNCaP cells in androgen-deprived environment [105]. MiR-221 also induces an S-phase arrest in cells [105]. MiR-221 transfection of LNCaP cells is also associated with NSE mRNA upregulation [105].

MiR-221 is shown to have a direct influence on the sensitivity of PCa cells to androgen. The upregulated expression of miR-221 in PCa cell lines is also consistent in human prostate tumor samples. It has been reported that even in the absence of androgen LNCaP cells could grow when expression of miR-221 was upregulated [108]. Also LNCaP with elevated miR-221 expression rescued LNCaP cells from growth arrest at G1 phase of cell cycle due to androgen depletion, promoting androgen-independent (AI) growth [108]. Upregulated expression of miR-221 also leads to differential expression of various AR-responsive genes such as declined expression of PSA [103] along with other AR sensitive genes viz., polycomb protein enhancer of zeste homolog 2 (EZH2) [26] and cell cycle regulatory gene (cdc20) [109]; however, significant change in expression level was observed in AR and promyelocytic leukemia zing finger protein (PLZF) [110]. It is also reported that upregulated expression of miR-221 is associated with downregulation of various miR-221 targets and, thereby, leads to elevated expression of G2, G2/M phase transition and M-phase cyclins, which eventually sustain AI growth of LNCaP cells [108]. HECT domain containing E3 ubiquitin protein ligase 2 (HECTD2) and Ras-related protein Rab-1A (RAB1A) are reported to be two targets of miR-221 and they are negatively regulated by miR-221 in PCa cells and thus they helped in sustaining the AI growth of PCa cells [108].

MicroRNA-21 and androgen receptor

MiR-21 has been shown to be an AR-regulated microRNA that enhances the androgen-dependent growth of prostate cells and develops the CRPC phenotype [111]. In yet another study, it has been shown that p57Kip21 is a target of miR-21 in PCa cells and transfection with miR-21 inhibitors and mimics in MDA-PCa-2b and PC-3 cells upregulated and downregulated p57Kip21 mRNA expression, respectively, thereby hampering the tumor suppressive potential of p57Kip21 [112].

Thus, the above studies show convincing data indicating miR-141, miR-375, miR-221 and miR-21 as makers of metastatic prostate cancer. MiR-141 is globally upregulated in PCa cases irrespective of the stage and hormone status of the cancer, whereas miR-375 is upregulated in AR-negative PCa cells compared to AR-positive PCa cells and the dysregulated expression profile is only relevant to malignant PCa cells indicating a dual role of miR-375 in PCa progression. Similarly, expression of miR-221 and miR-21 is upregulated in AI PCa cell lines compared to AD PCa counterparts and miR-221 also has a role in regulating neuroendocrine (NE) differentiation of prostate cells which is a regulator of CRPC, thus signifying the role of miR-221 in maintaining the CRPC phenotype in PCa.

MicroRNAs and cell proliferation

Cell proliferation is one of the major factors that mark the beginning of cancer and is the central and key process affected in all malignancies [113]. It is marked by loss of balance between cell loss and cell gain followed by invasion and metastasis [114]. Studies have shown that one way by which microRNAs plays a role in the pathogenesis of cancer is by regulating cell proliferation [7, 115]. For instance, it has been reported that miR-122 plays a role in hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) by inhibiting proliferation and growth of malignant tumor cells [116]; miR-19a acts as an oncogene and targets TIA1 (T cell intracytoplasmic antigen) and promotes colorectal cancer proliferation and migration [117]. MiR-144 targets E2F8 (E2F transcription factor 8) gene and inhibits the proliferation of papillary thyroid cancer (PTC), the most common subtype of thyroid cancer [118]; miR-373 suppresses proliferation and invasion in breast carcinoma by inhibiting BCl-2 expression [119]. Similarly in prostate cancer various miRNAs are reported to be tumor suppressor miRs as well as oncomiRs. For instance, miR-181c inhibited cell proliferation in PCa cells viz PC-3 and DU145 by targeting ERK2 (extracellular signal-regulated kinase), a core component of the ERK signaling pathway [120]. Similarly, miR-193a-3p suppressed proliferation in PC-3 and DU145 cells by targeting CyclinD1 and exhibited a significant G1/S phase arrest [121]. MiR-211 also exhibited tumor suppressive effects by inhibiting proliferation of PCa cells by targeting SPARC (secreted protein acidic and rich in cysteine) mRNA [122] which is matricellular glycoprotein and plays a major role in cell proliferation, migration and differentiation, modulating reversible interactions between cells and ECM [123]. Similarly, miR-17 overexpression suppressed LNCaP cells proliferation by downregulating STAT3 expression which is an important transcription factor in the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) signaling pathway that plays a critical role in many biological processes including proliferation [124]. MiR-20b, however, exerted an oncogenic effect in prostate cancer cells by promoting proliferation in VCaP and PC-3 cells by inhibiting PTEN (phosphatase and tensin homolog) expression by directly binding to its 3′-UTR and the proliferating ability of the prostate cancer cells was significantly reduced upon transfection with miR-20b inhibitor [125]. Similarly, miR-671 functions as an oncogene and promotes prostate cell proliferation by inhibiting tumor suppressor gene SOX6 [encoding SRY (sex determining region Y)—box 6] [126]. Thus, miRNAs represent critical regulators of proliferation in PCa and, thus, in the present review we have compiled the reported data associated with the role played by each of these four most dysregulated microRNAs, viz., miR-141, miR-375, miR-221 and miR-21 in PCa proliferation and also made a comprehensive list of target genes that regulates PCa proliferation.

MicroRNA-141 and proliferation

MiR-141 is reported to have anti-proliferative properties in nasopharyngeal carcinoma, gastric adenocarcinoma cells [127] and hepatocellular carcinoma [128]. In case of PCa miR-141 is shown to interact with kruppel-like-factor-9 (KLF9) and promotes proliferation of PCa cells by upregulating stem cell markers viz., Oct-4, Nanog, SOX-9 and Bmil [103]. Moreover, genes associated with proliferation such as CyclinD1, Cyclin E and c-Myc are reported to be upregulated in miR-mimic transfected PC-3 cells [129]. MiR-141 is also reported to enhance the spheroid forming ability of PC-3 cells when transfected with miR-141 mimics. Thus, miR-141 positively regulates the proliferative and stemness-associated properties in PCa cells [129].

MicroRNA-375 and proliferation

CBX7 (chromobox homolog 7) is reported to be a target of miR-375 [130]. The CBX7 loss is correlated with metastasis in various other cancers such as breast, colon, gastric and pancreatic cancer [131, 132]. Also inhibition of CBX7 by miR-375 leads to abundance of CBX8 in PCa cells and CBX8 is shown to have oncogenic properties in other cancers such as colon [133], esophageal [134] and breast cancer [135]. CBX7 upregulates E-cadherin and knockdown of CBX7 leads to activation of various signaling pathways such as EMT and Wnt/beta-catenin pathway [130]. Thus, increased expression of miR-375 is associated with enhanced proliferation, metastasis and invasion of PCa cells in vitro [130].

MicroRNA-221 and proliferation

MiR-221 positively regulates both cell proliferation and migration and negatively regulates apoptosis of PCa cells [136]. Downregulation of miR-221 is associated with the G0/G1 arrest of cells, indicating a role of miR-221 in cell cycle distribution [136]. Silent information regulator 1 (SIRT1) is one of the putative targets of miR-221 and is reported to act as both oncogene and tumor suppressor depending on the oncogenic pathway specific to the tumor [137]. For instance, in PC-3 cells, increased expression of SIRT1 with downregulated miR-221 expression is associated with inhibition of cell proliferation and migration and increased apoptosis [136], whereas in LNCaP cells, it serves to restrain cell proliferation [138]. Although SIRT1 is not a direct target of miR-221 as shown by the luciferase reporter assay, the biological effects exerted by miR-221 in PCa pathogenesis are in association with SIRT1 [136].

In yet another study, miR-221 has been shown to act as a tumor suppressor, targeting B-cell specific Moloney leukemia virus insertion region homolog 1 (Bmi-1), which is a polycomb ring finger oncogene and downregulation of miR-221 is associated with the promotion of cell proliferation in PCa cells [139].

Another tumor suppressor gene, ARHI, that negatively regulates cell proliferation is reported to be a target of miR-221 and causes cell cycle G0/G1 arrest in PC-3 cells with regulation of genes such as p21, growth arrest and DNA-damage-inducible, alpha (GADD45A) and Hect domain and RLD5 (HERC5) [140]. These genes are reported to function by regulating cell cycle proteins, stimulating DNA excision repair and ISGylation of protein targets, respectively [140].

MicroRNA-21 and proliferation

A recent study has shown that there exists an inverse correlation between miR-21 and phosphatase and tensin homolog deleted on chromosome ten (PTEN) in prostate cancer [141]. PTEN is a tumor suppressor and have both lipid phosphatase and protein phosphatase activity [142] and can inhibit the phosphorylation of downstream PI3K/Akt signaling pathway [141]. Thus, the reduced expression of PTEN is associated with reduced dephosphorylation of PI3K/Akt, enhancing cell proliferation and invasion in prostate cancer cells [141].

Thus, collectively it can be stated that miR-141, miR-375 and miR-21 act as oncogenic miRNAs promoting proliferation of PCa cells and miR-141 also positively regulates the stemness-associated properties of PCa cells. However, there are contradicting data in case miR-221 in PCa cells proliferation and is reported to act both as an oncomiR and a tumor suppressor miRNA regulating PCa cell proliferation. Thus, future studies on comprehensive analysis of miR-221 target genes and regulatory networks will be of great importance to clarify the role of miR-221 in PCa cell proliferation.

MicroRNAs and epithelial mesenchymal transition (EMT)

Epithelial mesenchymal transition is a developmental program with downregulation of epithelial phenotype and upregulation of mesenchymal characteristics [143]. It plays a major role in the metastasis of tumors of epithelial origin [144]. EMT can be both physiological and pathological and plays a key role in embryonic development and many diseases including cancer [145]. MicroRNAs have been reported to regulate EMT in various cancers, for instance, miR-200 family acts as a suppressor of EMT in various cancer, miR-194 promotes tumorigenesis by positively regulating EMT in colorectal cancer [146]; miR-217 suppressed EMT in gastric cancer by targeting PTPN14 (protein tyrosine phosphatase non-receptor type 14) gene which plays a crucial factor in EMT, metastasis and tumorigenesis [147]; miR-138 acts an a tumor suppressor in breast cancer cells by negatively regulating tumor associated gene Vimentin [148]. In case of prostate cancer, various miRNAs regulated EMT, for instance, miR-200b regulated EMT by increasing epithelial features of PC-3 cells with simultaneous reduction of mesenchymal markers [149]; miR-186 plays a tumor suppressive role in prostate cancer and suppressed EMT by inhibiting Twist1 expression [150]; miR-409-3p/-5p acts as an oncogene and its inhibition reduced the growth of prostate cancer cells with simultaneous induction of MET (mesenchymal–epithelial transition), both in vitro and in vivo [151]; ectopic expression of let-7a downregulated CCR7 (CC chemokines receptor 7) gene expression in PC-3 cells [152]. Mounting evidence supports that miRNAs play a vital role in regulating EMT in prostate cancer and, thus, in the present review we have summarized the role played by each of these four dysregulated microRNAs in regulating EMT in molecular level in prostate cancer from the existing literature.

MicroRNA-141 and EMT

The most potent inducer of EMT, Zeb1, is shown to have a suppressive role on miR-200 family with the most prominent on miR-141 and miR-200c in many cancers such as breast, pancreatic and colorectal cancer [153]. Zeb1 is shown to directly repress transcription of these microRNAs by binding to two conserved sites in their promoter regions [153]. Also, another potent inducer of EMT, TGFβ2, is a putative target of miR-141, exhibiting a regulatory loop between EMT induction and miR-141 expression [153]. However, unlike other members of miR-200 family, which are well known suppressors of EMT, miR-141 is only partial inhibitor of EMT that suppresses Zeb1 and Vimentin but without much effect on other mesenchymal markers such as Zeb2, Snail1, Snail2, Twist and Fibronectin, thus indicating induction of partial MET (mesenchymal–epithelial transition) phenotype [154].

MicroRNA-375 and EMT

Various groups have demonstrated that miR-375 is elevated in prostate cancer [155, 156]; however, there is only a single report showing the association of miR-375 with EMT. The study reported that miR-375 was upregulated in PCa cell lines possessing epithelial phenotype, whereas it was downregulated in cells having mesenchymal phenotypes, indicating miR-375 to be an epithelial marker in prostate cancer cells [157]. The authors have reported that miR-375 acts as a tumor suppressor and also as an inducer of metastasis, in a stage dependent manner [157]. They have also identified yes-associated protein 1 (YAP-1) as the downstream target of miR-375 and have shown that knockdown of YAP-1 in prostate cancer cells leads to downregulation of mesenchymal phenotypes viz., Vimentin and Fibronectin, indicating YAP-1 to be a mesenchymal marker [157]. Additionally, this study also reported that miR-375 is negatively regulated by Zeb-1, which is a key regulator of EMT. Knockdown of Zeb-1 downregulated the expression of miR-375 with the eventual loss of YAP1, indicating YAP1 to be a key downstream target of miR-375 in mediating EMT [157].

MicroRNA-221 and EMT

MiR-221 is a key regulator of EMT in luminal breast cancer cells [158] and there exists a direct correlation between miR-221 expression and E-cadherin repression [159]. Also, it directly targets trichorhinophalangeal 1 (TRPS1), which transcriptionally represses Zeb2 and Dicer, a potent EMT inducer and a key regulator of microRNA maturation, respectively, in pancreatic and breast cancer cells [22, 158]. However, the role of miR-221 in regulation of EMT in prostate cancer is not yet established to the best of our knowledge and more research in the area is needed.

MicroRNA-21 and EMT

Prostate cancer cells are believed to be originated from basal cells but express luminal markers and shows functional properties of basal cells [160,161,162]. The basal cell compartment expresses B cell translocation gene 2 (BTG2) [163] along with ∆Np63α, an isoform of p63 and is responsible for maintaining stemness and controls differentiation during prostate organogenesis [164, 165]. Loss of ∆Np63α is associated with the acquisition of EMT phenotype [166]. Basal protein (BTG2) is a tumor suppressor and its loss is associated with enhanced extracellular signal-regulated kinase 1 (ERK) signaling in prostate cells and is postulated to be associated with the acquisition of EMT phenotype in PCa cells [167]. The expression of miR-21 is upregulated in androgen-independent RWPE-2, PC-3 and DU145, whereas it is downregulated in androgen-dependent LNCaP and 22Rv1 compared to non-neoplastic RWPE-1 cells and prostate epithelial cells (PrEC) [168]; consistently, the BTG2 expression inversely correlates with miR-21 expression in PCa cells [168]. Studies have shown that miR-21 targets BTG2 and leads towards the phenotypic shift of prostatic basal cells towards luminal phenotype with the acquisition of EMT like features and tumorigenecity [167]. Also, the restoration of BTG2 levels in RWPE-2 cells that expresses a higher level of mesenchymal markers could shift the EMT phenotype with downregulation of Vimentin, Fibronectin, cytokeratins 8 and 18 (CK8-18) and with upregulation of p63 [167].

Taken together, the above studies indicate that miR-141 and miR-375 works as inhibitor of EMT targeting various genes involved in the EMT process although miR-141 is reported to be a partial inhibitor of EMT. However, miR-21 is reported to exert oncogenic effects by inducing phenotypic shift in PCa cells, thereby acquiring malignant features. The correlation between miR-221 and EMT has not been assessed to the best of our knowledge. Thus, further studies revealing the function of miR-221 in EMT are required.

MicroRNAs and apoptosis

Apoptosis is programmed cell death with certain morphological changes such as cell shrinkage, membrane blebbing, chromatin condensation and nuclear fragmentation [169]. During the transformation of a normal cell to a malignant one, evasion of cell death is one of the major hallmarks and is accomplished by the impaired balance between pro-apoptotic and anti-apoptotic proteins, reduced caspase activity and disrupted death receptor signaling [170]. Dysregulated microRNA expression is associated with tumorigenesis and microRNAs are reported to act as both pro-apoptotic and anti-apoptotic in various cancers [171]. For instance, miR-491 induces apoptosis in colorectal cancer by regulating BCL-XL (BCL2-like 1 isoform) [172], miR-133a suppresses osteosarcoma progression by inducing apoptosis by targeting BCL-XL and Mcl-1 [173], in pancreatic cancer miR-1284 is reported to induce apoptosis by regulating PI3K/Akt pathway [174], in breast cancer cells miR-125b targets Bak1 and is associated with inhibition of apoptosis in Taxol-resistant cancer cells [175]. Similarly in case of prostate cancer, various miRNAs are reported to regulate apoptosis; for instance, miR-218 plays a tumor suppressive role and induces apoptosis by targeting an oncogene TPD52 (tumor protein D52) which is upregulated in prostate cancer [176]; similarly, miR-466 induced apoptosis in metastatic prostate cancer cells (PC-3 and DU145) with simultaneous induction of G0/G1 cell cycle arrest [177]; miR-143 induced apoptosis in LNCaP cells by inhibiting BCl-2 expression [178]; miR-1180 induced apoptosis in PCa cells by targeting TRAF1 (TNF receptor-associated factor 1) and BAG2 [B cell lymphoma 2 (Bcl 2)-associated athanogene 2] [179]. Thus, we see that there exists a correlation between dysregulated miRNA expression and apoptosis. In the present review, we have discussed the current knowledge about each of these four most dysregulated microRNAs, viz., miR-141, miR-375, miR-221 and miR-21 in apoptosis in PCa.

MicroRNA-141 and apoptosis

MiR-141 is shown to regulate cell death pathway in many cancers. For instance, miR-141 overexpression is associated with apoptosis induction in osteosarcoma by targeting ZEB1 and ZEB2 [180]; in pancreatic cancer, miR-141 acts as an apoptosis inducer by targeting mitogen-activated protein kinase isoform 4 (MAP4K4); however, in case of hepatocellular carcinoma downregulation of miR-141 promotes apoptosis by modulating hepatocyte nuclear factor 3β (HNF-3β) [181]. In prostate cancer, miR-141-3p acts as an inhibitor of apoptosis as shown by downregulation of various apoptosis-related mRNAs such as p21, p27, Bax and Caspase-3 upon miR-141-3p mimic treatment and was upregulated on treatment with miR-141-3p inhibitors [103, 182].

MicroRNA-375 and apoptosis

MiR-375 expression is lowest in RWPE-1 cell line and highest in 22Rv1; however, in PC-3 cell line it shows moderate expression [97]. Forced expression of miR-375 could increase the level of apoptosis in PC-3 cells (metastatic prostate cell line); however, in RWPE-1 cells (a benign prostate cell line) there was no change in the level of apoptosis, thus suggesting that deregulated expression of miR-375 is only relevant in malignant prostate cancer [97]. Also, inhibition of miR-375 expression in 22Rv1 could increase the level of apoptosis, suggesting that miR-375 can have an oncogenic phenotype (in 22Rv1) as well as tumor suppressive phenotype (in PC-3) in the same tumor model [97].

MicroRNA-221 and apoptosis

MiR-221 positively regulates apoptosis in androgen-independent PC-3 and DU145 by increasing the caspase 3/7 activity and simultaneous activation of JAK/STAT pathway by negatively regulating suppressor of cytokine signaling 3 (SOCS3) and interferon regulatory factor 2 (IRF2) and, thus, acts as a tumor suppressor [183]. However, another study revealed that miR-221 negatively regulates apoptosis in PCa cells, both PC-3 and LNCaP with a simultaneous increase in proliferation and decrease in Caspase-3 and Caspase-10 activity by inhibiting TNF-α/CHX-induced apoptosis [19]. It was also observed that knocking down of miR-221 was associated with increased expression of Caspase-10 and sensitizing cells to apoptosis [19]. Thus, it shows that there are contradictory data regarding the role of miR-221 being tumor suppressor or an oncogene in prostate cancer.

MicroRNA-21 and apoptosis

MiR-21 is known to play a role in tumorigenesis in various malignancies [184, 185], including prostate cancer [186] by targeting various genes that are tumor suppressors in nature [187]. One such tumor suppressor gene is FBXO11 (a member of the F-box subfamily lacking a distinct unifying domain) which has been identified as a target of miR-21 by microarray analysis [187]. FBXO11 targets various proteins that play a role in cell cycle control, apoptosis, metastasis, differentiation and also tags proteins for proteosomal degradation by ubiquitination [187]. In prostate cancer models, FBXO11 acts as an inducer of apoptosis and sensitizes DU145 cells to apoptosis [187]. Thus, miR-21 regulates apoptosis in prostate cancer.

Collectively, the above studies revealed that miR-141 and miR-21 act as oncogenic miRNAs in regulating apoptosis in PCa; however, miR-375 acts as both oncogenic and tumor suppressor miRNA in regulating apoptosis in PCa. Also there are contradictory data reported regarding the role played by miR-221 in apoptosis in PCa. Thus, further research exploring the signaling pathways and functional targets of this panel of miRNAs needs to be conducted.

Conclusion

Dysregulated microRNA profile is associated with prostate cancer aggressiveness and affects various critical cellular processes viz., proliferation, apoptosis, EMT and androgen receptor signaling (Table 2). Although various studies in the last decade have tried to study the relationship between microRNAs and PCa, there is high variability in the data reported. In the present review, we have identified a panel of four consistently dysregulated microRNAs in prostate cancer and have prepared a comprehensive list of their experimentally validated targets both in vitro and in vivo. Thus, validation of these promising candidate microRNAs in larger, prospective cohort will not only define their exact role in prostate cancer progression but may also be used as potential prognostic and diagnostic markers in prostate cancer.

Table 2 Genes regulated by the panel of four differentially expressed microRNAs in PCa

The implication of this panel of four microRNAs in PCa management demands uniformity in the study design, sampling method and profiling platform used. Thus, minimizing the ambiguity associated with miRNA profiling studies and developing a gold standard technique might help in near future to bring this panel of miRNAs from bench to bed side.