Given the prime importance of UDP-glucuronosyltransferase (UGT) 2B15 and UGT2B17 in inactivating testosterone and dihydrotestosterone, control of their expression and activity in the prostate is essential for androgen signaling homeostasis in this organ. Although several studies provide evidence of transcriptional control of UGT2B15 and UGT2B17 by various endogenous and exogenous compounds, potential post-transcriptional regulation of UGT2B15 and UGT2B17 by microRNAs (miRs) in prostate cancer cells has not been examined. The present study identified a putative miR-376c target site in the 3′-untranslated regions (UTRs) of both UGT2B15 and UGT2B17 mRNAs. In accordance with the possibility that this miRNA negatively regulates UGT2B15 and UGT2B17 expression, there is an inverse correlation in the levels of miR-376c and UGT2B15/UGT2B17 mRNAs in prostate cancer cell lines versus normal prostate tissue. In LNCaP cells, transfection of miR-376c mimics inhibited the glucuronidations of testosterone, 4-methylumbelliferone (a substrate of UGT2B15), and androsterone (a substrate of UGT2B17). miR-376c reduced both UGT2B15 and UGT2B17 mRNA and protein levels and the activity of luciferase reporters containing UGT2B15 or UGT2B17 3′-UTRs. This microRNA-mediated repression was significantly abrogated by mutating the miR-376c binding site in the 3′-UTRs of both UGTs. Collectively, these data indicate that the expression of UGT2B15 and UGT2B17 is negatively regulated by the binding of miR-376c to the 3′-UTRs of UGT2B15 and UGT2B17 in prostate cancer cells. This represents the first evidence for post-transcriptional regulation of UGT2B15 and UGT2B17 by miRNAs in prostate cancer cells and may have importance in regulating androgen receptor signaling.
The glucuronidation of lipophilic compounds results in products that are generally more water soluble, thus promoting their detoxification and clearance (Mackenzie et al., 1997). In humans, glucuronidation is primarily carried out by members of the UDP-glucuronosyltransferase (UGT) 1A, UGT2A, and UGT2B subfamilies (Mackenzie et al., 2005). The UGT1A subfamily has nine functional enzymes (UGT1A1 and UGT1A3–UGT1A10), which are encoded by a single locus (2q37) through splicing of an isoform-specific exon 1 to a set of shared exons 2–5, whereas the UGT2A and UGT2B subfamilies have 10 functional enzymes (UGT2A1, UGT2A2, UGT2A3, UGT2B4, UGT2B7, UGT2B10, UGT2B11, UGT2B15, UGT2B17, and UGT2B28), which are encoded by individual genes, except for UGT2A1 and UGT2A2, which have an isoform-specific exon 1 and shared exons 2–6. These two subfamilies form a gene cluster at 4q13 (Mackenzie et al., 1997, 2005).
Most UGTs are expressed in the liver, where glucuronidation activity plays a major role in systemic detoxification and clearance of small lipophilic compounds. However, UGTs are also expressed in a variety of extrahepatic tissues (e.g., breast, prostate, lung, kidney, and colon), and thus, have a critical role in maintaining intratissular homeostasis and the signaling activity of biologic active endogenous and exogenous molecules, which are primarily inactivated through glucuronidation (Nishimura and Naito, 2006; Nakamura et al., 2008; Izukawa et al., 2009; Ohno and Nakajin, 2009; Court, 2010; Court et al., 2012; Schaefer et al., 2012). For example, the prostate depends on androgen signaling for normal development and function. However, excessive androgen action promotes prostate cancer development and progression (Heinlein and Chang, 2004). As the active androgens testosterone and dihydrotestosterone are mainly inactivated by UGT2B15 and UGT2B17 in the prostate, the expression of these UGTs is a major determinant of androgen effects on this organ (Turgeon et al., 2001; Chouinard et al., 2007). Hence, transcriptional and possibly post-transcriptional regulation of intraprostatic UGT2B15 and UGT2B17 expression controls androgen homeostasis and signaling activity in the prostate, and disruption of this regulation may impact the risk of developing androgen-sensitive prostate diseases, such as prostate cancer. Indeed, a number of studies have investigated the transcriptional regulation of UGT2B15 and UGT2B17 and have shown that these two UGTs are negatively regulated in androgen receptor–positive prostate cancer cell lines by various endogenous and exogenous compounds, including androgens, cytokines, growth factors, chenodeoxylcholic acid, and calcitrol (the active metabolite of vitamin D) (Guillemette et al., 1996, 1997; Levesque et al., 1998; Chouinard et al., 2006; Bao et al., 2008; Kaeding et al., 2008a,b). Of particular importance, the androgen-mediated downregulation of UGT2B15 and UGT2B17 allows androgens to modulate their own signaling activity through a feedback regulatory loop within prostate cancer cells (Chouinard et al., 2008). Despite this importance, little is known about the mechanisms controlling UGT2B15 and UGT2B17 expression at the post-transcriptional level in prostate cancer.
MicroRNAs (termed miRNA or miR hereafter) have recently emerged as major post-transcriptional regulators of gene expression (Filipowicz et al., 2008). It was estimated that over 60% of all protein-coding genes are regulated by miRNAs in mammals (Lewis et al., 2003; Friedman et al., 2009). miRNAs are small (generally 21–25 nt), noncoding, and single-stranded RNAs. Each miRNA has a unique seed sequence, corresponding to nucleotides at positions 2 to 7 from its 5′ terminus, which determines its specificity (Grimson et al., 2007). miRNAs promote mRNA degradation and/or inhibit translation through their binding to miRNA target sites, which are generally located within the 3′-untranslated regions (UTRs) of target mRNAs (Hu and Coller, 2012). A Watson-Crick pairing between an miRNA seed sequence and its seed site (nucleotides at positions 2–7 from the 3′ terminus of an miRNA target site) is essential for miRNA targeting (Grimson et al., 2007; Friedman et al., 2009). By searching for the presence of individual miRNA seed site sequences, bioinformatic programs have been frequently used to predict potential miRNA target sites throughout the human genome (Lewis et al., 2005; Grimson et al., 2007; Friedman et al., 2009). Using this strategy, the present study identified a potential miR-376c target site in the 3′-UTRs of both UGT2B15 and UGT2B17. We further demonstrated that this site mediates the downregulation of both UGT2B15 and UGT2B17 expression by miR-376c. This represents the first evidence for post-transcriptional regulation of UGT2B15 and UGT2B17 by miRNAs in prostate cancer cells.
Materials and Methods
miRNA Mimics and Human Prostate Tissue Total RNA.
Synthetic mimics of hsa-miR-376c (MI0000776) or a negative control miRNA (termed miR-neg) were purchased from Shanghai GenePharma (Shanghai, China). Three total RNA samples of normal prostate tissues were purchased from Ambion (Dallas, TX) and Applied Biosystems (Foster City, CA) (RNA pooled from three Caucasian males; catalogue number AM6000), Life Technologies (Carlsbad, CA) (RNA from Caucasian male), and Clontech (Mountain View, CA) (RNA pooled from 12 Caucasian males), which were termed as prostate tissue 1, 2, and 3, respectively. The expression levels of miR-376c, UGT2B15, and UGT2B17 in these tissue RNA samples were quantified as described below. Five human cell lines, including the prostate cancer LNCaP, VCaP, and Du145 cell lines, the liver cancer HepG2 cell line, and the human embryonic kidney 293 cell line were purchased from American Type Culture Collection (Manassas, VA).
Cell Transfection, RNA Extraction, and Reverse-Transcriptase Quantitative Real-Time Polymerase Chain Reaction.
LNCaP cells were maintained in RPMI 1640 medium (Gibco/Life Technologies, Grand Island, NY) supplemented with 5% (v/v) fetal bovine serum at 37°C and 5% CO2. For transfection, cells were precultured for 32 hours in phenol red–free RPMI 1640 medium (Invitrogen, Carlsbad, CA) supplemented with 5% dextran-coated charcoal–stripped fetal bovine serum and then plated into six-well plates at approximately 1×106 cells per well. After overnight culture, cells were transfected in triplicate with an miR mimic (miR-376c or miR-neg) at 30 nM using 8 µl of Lipofectamine 2000 (Invitrogen) per well. Twenty-four hours post-transfection, total RNA was extracted from transfected cells using TRIzol reagent according to the manufacturer’s protocol (Invitrogen). Quantitation of target genes (i.e., UGT2B15, UGT2B17, β-actin, and GAPDH) using reverse-transcriptase quantitative real-time polymerase chain reactions (RT-qPCRs) was conducted as previously reported (Hu and Mackenzie, 2009, 2010; Hu et al., 2010), unless otherwise specified. Quantification of miRNAs miR-376c and miR-U6 small nuclear-2 RNA (termed RNU6-2) was performed as recently reported by Balcells et al. (2011). Briefly, RNA samples were treated with DNase I (Invitrogen) and then polyadenylated using poly(A) polymerase (New England Biolabs, Ipswich, MA) to generate a poly(A) tail at the 3′-termini of the miRNAs. The poly(A)-tailed miRNAs were converted to cDNAs using Superscript III reverse transcriptase (Invitrogen) and the reported primer (5′-CAGGTCCAGTTTTTTTTTTTTTTTVN-3′) (Balcells et al., 2011). RT-qPCR was performed using a Corbett Rotor-Gene 3000 (Corbett Research, Mortlake, NSW, Australia) in a 15-μl reaction mixture containing 2 μl of cDNA (∼40 ng), 7.5 µl of 2× GoTaq Syber Green Master Mix (Promega, Madison, WI), and a pair of miRNA-specific forward and reverse primers (at a final concentration of 600 nM each). The PCR reaction was initially incubated at 95°C for 10 minutes, followed by 40 cycles of 95°C for 10 seconds and 60°C for 40 seconds. At the end of the 40th cycle, the temperature was ramped from 60 to 94°C to produce a melting curve. The expression level of miR-376c was presented relative to that (set as a value of 1) of RNU6-2 (a housekeeping miRNA) using the 2–ΔΔCT method (Livak and Schmittgen, 2001). The miRNA-specific primers used are listed in Table 1.
For the quantification of primary miRNA (termed pri-miR) 376c and RNU6-2 in human prostate tissue cDNA, quantitative real-time PCR was performed using a RotorGene 3000 (Corbett Research) in a 20-μl reaction containing 3 μl of cDNA sample, 10 μl of GoTaqqPCR master mix (Promega), and a pair of either pri-miR-376c– or RNU6-2–specific forward and reverse primers (at a final concentration of 500 nM each). The pri-miR–specific primers used are listed in Table 1. The PCR reaction was initially incubated at 95°C for 5 minutes, followed by 40 cycles of 95°C for 10 seconds, 57°C for 15 seconds, and 72°C for 20 seconds. At the end of the 40th cycle, the temperature was ramped from 55 to 95°C to produce a melting curve. The expression level of pri-miR-376c was presented relative to that of RNU6-2 using the 2–ΔΔCT method as mentioned above.
Generation of Luciferase Reporter Constructs and Mutagenesis.
The UGT2B15 mRNA (NM_001076) contains a 513-base pair (bp) 3′-UTR, whereas the UGT2B17 mRNA (NM_001077) contains a 463-bp 3′-UTR. A 454-bp UGT2B15 3′-UTR region between the stop code (TAG) and the poly(A) tail was amplified from a commercial genomic DNA sample (Roche Diagnostic, Indianapolis, IN) by Phusion hot-start high-fidelity DNA polymerase (Thermo Fisher Scientific, Pittsburgh, PA) and subsequently cloned into the XbaI restriction site of the pGL3-promoter vector (Promega), generating the reporter construct pGL3/2B15/UTR. Similarly, a 431-bp UGT2B17 3′-UTR region was amplified and subsequently cloned into the XbaI site of the pGL3-promoter vector generating the reporter construct pGL3/2B17/UTR. Of note, the XbaI site is located immediately downstream of the luciferase coding sequence in the pGL3-promoter vector. Three primers used for cloning were a common forward primer (5′-CCGCTCTAGATTATATCAAAAGCCTGAAGT-3′), the UGT2B15-specific reverse primer (5′-CCGGTCTAGAGACACTTTATTTTCAGATCC-3′), and the UGT2B17-specific reverse primer (5′-CCGCTCTAGAGAAGATTTCATTGGCAAAAT-3′). Using the pGL3/2B15/UTR construct as a template and the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA), the miR-376c seed site (5′-UCUAUGUC-3′), which is complementary to the miR-376c seed sequence (3′-AGAUACAA-5′), was changed to 5′-CGGUUGUC-3′, producing the mutated construct pGL3/2B15/UTR/miR-376c/MT. Similarly, the miR-376c seed site (5′-UCUAUGUC-3′) that is complementary to the miR-376c seed sequence (3′-AGAUACAA-5′) was mutated to 5′-CGACUGUC-3′ in the pGL3/2B17/UTR construct to make the mutated construct pGL3/2B17/UTR/miR-376c/MT. The identities of wild-type and mutated constructs were confirmed by DNA sequencing. The sequences of the primers used for mutagenesis are given in Table 1.
Luciferase Reporter Assays.
LNCaP cells were plated in 96-well plates at approximately 1.25 × 105 cells per well. After overnight culture, cells were transfected in triplicate with the internal control pRL-null vector (5 ng in each well), a luciferase reporter (100 ng in each well), and mimics of one miRNA at 30 nM, as indicated in the legend of Fig. 2. Twenty-four hours after transfection, cells were harvested and assayed for firefly and Renilla luciferase activities using the Dual-Luciferase Reporter Assay System (Promega) as reported (Hu and Mackenzie, 2010). The firefly luciferase activity was normalized to the Renilla activity and then presented relative to that of the empty pGL3-promoter vector (set as a value of 100%).
Testosterone and Androsterone Glucuronidation Assays.
LNCaP cells were transfected with miRNA mimics (miR-376c or miR-neg) at 30 nM. Seventy-two hours later, whole cell lysates were prepared in TE buffer (10 nM Tris-HCl and 1 mM EDTA, pH 7.6) and protein concentrations of the lysates were determined using the Bradford protein assay, according to the manufacturer’s protocol (Bio-Rad, Hercules, CA). For testosterone glucuronidation assays, duplicate 100-μl reactions of each sample containing 100 mM potassium phosphate, pH 7.5, 4 mM MgCl2, 1440 pM [14C]testosterone (NEN, Boston, MA), 25 μg lysate protein, and 2 mM UDP-glucuronic acid (Sigma-Aldrich, St. Louis, MO) were incubated at 37°C in a shaking water bath for 2 hours. Reactions were terminated by the addition of 300 μl of chloroform. The reaction mixture was vortexed and then centrifuged at 11,000g for 5 minutes. One third of the supernatant (66 μl) from each sample was further extracted with an equal volume of 100% ethanol and centrifuged at 11,000g for 3 minutes. Sixty microliters of the supernatant of each sample was spotted onto silica gel thin layer chromatography plates (Uniplate; Analtech, Newark, DE) and chromatographed for 1 hour in a solvent of choloform/methanol/water/acetic acid (65:25:4:2). The plates were placed on Kodak storage phosphor screens (Amersham Biosciences, Piscataway, NJ) for 72 hours and then imaged using the Typhoon 9000 Imager (GE Healthcare, Giles, UK). Band intensity was measured using ImageQuant version 5.2 (GE Healthcare).
For androsterone glucuronidation assays, 100-μl reactions were prepared containing 100 mM potassium phosphate, pH 7.5, 4 mM MgCl2, 200 μM androsterone (Sigma-Aldrich), 25 μg cell lysate protein, 200 μM UDP-glucuronic acid (Sigma-Aldrich), and 319 pmol of [14C]UDP-glucuronic acid (PerkinElmer, Boston, MA) and incubated at 37°C in a shaking water bath for 2 hours. Reactions were terminated by the addition of 200 μl of ethanol. The reaction mixture was vortexed and then centrifuged at 11,000g for 3 minutes. Aliquots of 100 µl from each reaction were spotted onto silica gel thin layer chromatography plates and chromatographed for 1 hour as described above. The dried plates were then imaged and quantified as described above. To demonstrate that androsterone is selective for UGT2B17, human UGT2B15 and UGT2B17 supersomes were used as controls in the assay (In Vitro Technologies, Noble Park North, VIC, Australia).
4-Methylumbelliferone Glucuronidation Assay.
LNCaP cells were transfected with miRNA mimics (miR-376c or miR-neg) at 30 nM, and 72 hours later, whole cell lysates were prepared and the protein concentrations of the lysates were determined as described above. Two hundred microliter reactions of each sample containing 100 mM potassium phosphate, pH 7.4, 4 mM MgCl2, 400 μM 4-methylumbelliferone (4-MU) (Sigma-Aldrich), and 225 μg lysate protein were preincubated at 37°C for 5 minutes in a shaking water bath. After the preincubation, the reactions were initiated by the addition of 5 mM UDP-glucuronic acid and further incubated for 2 hours. Reactions were terminated by the addition of 2 μl of 70% perchloric acid (Chem-Supply, Gillman, SA, Australia) and kept on ice for 10 minutes. Samples were centrifuged at 5000g for 10 minutes at 4°C and a 60-μl aliquot of the supernatant was analyzed by high-performance liquid chromatography as previously described (Uchaipichat et al., 2004). 4-MU glucuronide formation was calculated using a standard curve of 4-MU glucuronide (Sigma-Aldrich). To demonstrate that 4-MU is selective for UGT2B15, human UGT2B15 and UGT2B17 supersomes were used as controls in the assay (In Vitro Technologies).
Statistical analysis was performed with a two-tailed Student’s t test using GraphPad Prism 6 (GraphPad Software, La Jolla, CA). A P value of less than 0.05 was considered statistically significant.
Putative miR-376c Target Site Is Present in the UGT2B15 and UGT2B17 3′-UTRs.
TargetScan is an online well verified bioinformatic program that predicts miRNA target sites in 3′-UTRs of target mRNAs using parameters, such as seed pairing, 3′-pairing, and local AU content (Lewis et al., 2005; Grimson et al., 2007; Friedman et al., 2009). Using TargetScan 6.2 software, we identified one putative miR-376c target site in the 3′-UTRs of both UGT2B15 and UGT2B17 mRNAs. As shown in Fig. 1, the UGT2B15 and UGT2B17 miR-376c target sites are highly conserved in both position and sequence, with only a C/U mismatch at position 19 and a G deletion at position 17 in the UGT2B15 miR-376c target site. Moreover, both target sites are conserved between humans and chimpanzees, as revealed by TargetScan 6.2 (http://www.targetscan.org/) (data not shown). TargetScan classifies predicted miRNA target sites into four types: 6mer (a pairing to the 2–7 nt miRNA seed), 7mer-m8 (a pairing to the 2–7 nt miRNA seed and nucleotide 8), 7mer-A1 (a pairing to the 2–7 nt miRNA seed plus an A at nucleotide 1), and 8mer (a pairing to the 2–7 nt miRNA seed and nucleotide 8 plus an A at nucleotide 1). Studies have shown the following hierarchy of miRNA site efficacy: 8mer > 7mer-8m > 7mer-A1 > 6mer (Grimson et al., 2007; Nielsen et al., 2007). According to this nomenclature, both the UGT2B15 and UGT2B17 miR-376c target sites are 7mer-8m sites. In addition to seed pairing, a Watson-Crick pairing to the 3′ proportion of the miRNA (3′-pairing) at specific positions (such as nucleotides 12–17) has been shown to enhance the efficacy of miRNA targeting (Grimson et al., 2007). A 3′-pairing to the predicted miRNA is found at nucleotides 12–16 in the UGT2B15 miR-376c target site and nucleotides 12–18 in the UGT2B17 miR-376c site.
UGT2B15 and UGT2B17 mRNA Levels Are Inversely Correlated to the Levels of miR-376c in Prostate Cell Lines and Tissues.
UGT2B15 and UGT2B17 enzymes inactivate testosterone and dihydrotestosterone and are highly expressed in normal and cancerous prostate tissues (Barbier and Belanger, 2008). Hence, these two UGTs are believed to be major determinants of intraprostatic androgen signaling activity (Chouinard et al., 2007). Previous studies have shown that both UGT2B15 and UGT2B17 are highly expressed in androgen receptor–positive prostate cancer cell lines (e.g., LNCaP and VCaP) but are not expressed in androgen receptor–negative prostate cancer cell lines (e.g., PC3 and Du145) (Chouinard et al., 2007; Bao et al., 2008; Hu and Mackenzie, 2010; Hu et al., 2010). We performed RT-qPCR to measure the expression levels of miR-376c, UGT2B15, and UGT2B17 in three androgen receptor–positive prostate cancer cell lines (i.e., LNCaP, VCaP, and DuCaP) and three normal pooled prostate RNA samples from different commercial sources. As illustrated in Fig. 2, our results showed downregulation of miR-376c and upregulation of both UGT2B15 and UGT2B17 in prostate cancer cell lines as compared with normal prostate tissues. This inverse correlation in the expression levels of miR-376c and the two UGTs between cell lines and tissues suggests a negative regulation of UGT2B15 and UGT2B17 by miR-376c in androgen receptor–positive prostate cancer cell lines.
We also measured the levels of the primary miR-376c transcript (pri-miR-376c), UGT2B15, and UGT2B17 in the three prostate cancer cell lines (VCaP, LNCaP, and DuCaP) and 18 normal human prostate tissue samples (Supplemental Fig. 1). Similar to the data shown in Fig. 2, an inverse correlation was observed between pri-miR-376c and the two UGTs in the cancer cell lines as compared with normal prostate tissues.
miR-376c Reduces Both UGT2B15 and UGT2B17 mRNA, Protein, and Glucuronidation Activities in Prostate Cancer LNCaP Cells.
To investigate the potential negative regulation of UGT2B15 and UGT2B17 by miR-376c in androgen receptor–positive prostate cancer cells, we transfected miRNA mimics (miR-376c or miR-neg) into the LNCaP cell line, the most frequently used prostate cancer model cell line (Hu et al., 2014), and then performed RT-qPCR to quantify UGT2B15 and UGT2B17 mRNA levels. As shown in Fig. 3, A and B, both UGT2B15 and UGT2B17 mRNA levels were significantly lower in miR-376c mimic transfected cells as compared with those in miR-neg–transfected cells, supporting a negative regulation of these two UGTs by miR-376c, possibly through targeting the predicted miR-376c site in their 3′-UTRs (Fig. 1). miR-376c did not significantly alter control glyceraldehyde-3-phosphate dehydrogenase mRNA levels (Fig. 3C). We further developed an antibody that recognized both UGT2B15 and UGT2B17 but had no cross-reactivity with three other UGT2B enzymes, namely, UGT2B4, UGT2B7, and UGT2B10 (Supplemental Fig. 2; Supplemental Methods). Western blotting performed with this antibody showed that UGT2B15 and/or UGT2B17 protein levels were significantly reduced in miR-376c mimic-transfected cells as compared with miR-neg–transfected cells (Supplemental Fig. 2). Consistent with this observation, the capacity to glucuronidate testosterone, a substrate of both UGT2B15 and UGT2B17 (Turgeon et al., 2001), was also significantly reduced by miR-376c (Fig. 3D). To confirm the specific downregulation of UGT2B15 and UGT2B17 by miR-376c, we further measured the glucuronidation activities of miR-376C– or neg-miR–transfected LNCaP cell lysates using 4-MU, for which UGT2B15 has an approximately 20-fold higher activity than UGT2B17 (data not shown), and androsterone, a substrate of UGT2B17, but not UGT2B15 (data not shown). As expected, the glucuronidation of 4-MU and androsterone were significantly reduced by miR-376c (Fig. 3, E and F, respectively). The contribution of other UGTs to the glucuronidation of testosterone, 4-MU, and androsterone were negligible as only UGT2B15 and UGT2B17 are expressed to any significant extent in LNCaP cells.
3′-UTRs of UGT2B15 and UGT2B17 Are Direct Targets of miR-376c.
To determine whether UGT2B15 and UGT2B17 are direct targets of miR-376c, we cloned the 3′-UTR of either UGT2B15 or UGT2B17 downstream of the luciferase gene in the pGL3-promoter vector and mutated four contiguous bases within the predicted miR-376c seed site sequences to abolish seed pairing as described in Materials and Methods. We cotransfected wild-type or mutated constructs with miRNA mimics (miR-376c or miR-neg) into LNCaP cells and quantified the luciferase activity of each reporter construct. As shown in Fig. 4, B and C, compared with transfection of miR-neg, transfection of the miR-376c mimic into LNCaP cells significantly reduced the luciferase activity of the construct carrying the wild-type 3′-UTR of UGT2B15 (pGL3/2B15/UTR) or UGT2B17 (pGL3/2B17/UTR); however, this reduction was significantly abrogated in the two miR-376c site-mutated constructs (pGL3/2B15/UTR/miR-376c/MT and pGL3/2B17/UTR/miR-376c/MT). We repeated these experiments in a second prostate cancer cell line (DU145) and observed similar results (Supplemental Fig. 3). Taken together, these results demonstrate the functionality of the predicted miR-376c site in prostate cancer cells. As shown in Supplemental Fig. 3, a similar negative regulation of these two UGTs by miR-376c was also observed in the liver cancer HepG2 cell line and the kidney human embryonic kidney 239T cell line, suggesting that this miR-376c site may also be functional in cellular contexts other than prostate cancer cells. Collectively, these results provide evidence that the UGT2B15 and UGT2B17 3′-UTRs are direct targets for miRNA-376c.
UGT2B15 and UGT2B17 are present in the prostate but show elevated expression in androgen receptor–positive prostate cancer cell lines (Fig. 2) (Chouinard et al., 2006; Hu et al., 2010). As both UGTs are responsible for inactivating testosterone and its more active metabolite, dihydrotestosterone, control of their expression is important for maintenance of androgen signaling homeostasis and activity in the prostate (Turgeon et al., 2001; Chouinard et al., 2007, 2008). Previous studies with androgen receptor–positive prostate cell lines (mainly LNCaP cells) have shown regulation of these UGTs at the transcriptional level (Guillemette et al., 1996, 1997; Levesque et al., 1998; Chouinard et al., 2006; Bao et al., 2008; Kaeding et al., 2008a,b). We now show regulation of their expression by a post-transcriptional mechanism involving miR-376c.
The 3′-UTRs of UGT2B15 and UGT2B17 mRNAs are similar in length (513 bp for UGT2B15 and 463 bp for UGT2B17) and exhibit regions of high homology. Indeed, we show in this study that the UGT2B15 and UGT2B17 miR-376c target site is highly conserved in both position and sequence. As mentioned earlier, the miR-376c target site in each UGT 3′-UTR can be designated as a 7mer-8m site, with only a C/U mismatch at position 19 and a G deletion at position 16 in the UGT2B15 site. miRNAs regulate gene expression generally through promoting mRNA degradation and/or inhibiting translation of target mRNAs (Hu and Coller, 2012). Evidence presented in the present study showed that miR-376c reduced UGT2B15 and UGT2B17 mRNA levels, demonstrating their capacity to promote mRNA degradation. Whether these two miRNAs also have a translational inhibitory effect remains to be investigated in future studies. Several human genes have been shown to be direct targets of miR-376c, including growth factor receptor–bound protein 2 in intrahepatic cholangiocarcinoma cells (Iwaki et al., 2013), activin receptor–like kinase 7 in ovarian cancer cells (Ye et al., 2011), and transforming growth factor-α in osteosarcoma cancer cells (Jin et al., 2013).
We have shown in the present study an inverse relationship between the levels of miR-376c and the levels of UGT2B15 and UGT2B17 mRNA in prostate cancer cell lines and normal prostate tissues (Fig. 2). Widespread deregulation of miRNA expression in prostate cancer has been recently reported (Ozen et al., 2008). In particular, miR-376c has been shown to be downregulated in prostate cancer tissues (especially in advanced metastatic and invasive tumors) as compared with normal prostate tissues (Srivastava et al., 2013; Formosa et al., 2014). Given our findings that UGT2B15 and UGT2B17 mRNAs are targets of miR-376c, it seems feasible that miR-376c downregulation might enhance the expression of these two UGTs in prostate cancer in vivo. Consistent with this possibility, it has been recently shown that UGT2B15 and/or UGT2B17 expression is upregulated in metastatic androgen-independent prostate cancer as compared with primary prostate cancer (Stanbrough et al., 2006; Montgomery et al., 2008; Hornberg et al., 2011; Zhang et al., 2011; Mitsiades et al., 2012). A future study could explore this correlation further by obtaining a cohort of matched normal and tumor tissues with defined cancer staging.
microRNA has also been implicated in regulating UGT1A1 expression (Dluzen et al., 2014). All nine functional UGT1A mRNAs share exons 2–5 and thus have the same 3′-UTR sequence and the same set of putative miRNA binding sites. However, a miR-491-3p binding site that is located within this shared UGT1A 3′-UTR has been recently shown to regulate three UGT1As (i.e., UGT1A1, UGT1A3, and UGT1A6) but not the remaining six UGT1As in liver cancer HepG2 and Huh7 cells. This differential hepatic regulation of the UGT1As by miR-491-3p is believed to be attributable to the differing secondary structure of the UGT1A mRNAs (Dluzen et al., 2014). These observations, combined with our findings described herein, provide the first evidence for post-transcriptional regulation of UGTs by miRNAs. As mentioned earlier, all seven functional UGT2Bs are encoded by unique genes and thus each UGT2B mRNA has a unique 3′-UTR sequence. In addition to an miR-376c binding site, the UGT2B15 and UGT2B17 3′-UTRs contain putative binding sites for other miRNAs, such as miR-21, miR-382, miR-375-3p, miR-125b, miR-222, miR-147, miR-1289, miR-376a, miR-3939, miR-409, miR-450b-5p, miR-489, miR-494, miR-548x, and miR-656. However, transfection of miRNA mimics of these miRNAs had no impact on the mRNA levels of either UGT2B15 or UGT2B17 and did not affect the activity of the pGL3-promoter reporter carrying the wild-type UGT2B15 or UGT2B17 3′-UTR in LNCaP cells (data not shown). TargetScan software also predicted putative binding sites for many miRNAs in the 3′-UTRs of the remaining five human UGT2Bs (i.e., UGT2B4, UGT2B7, UGT2B10, UGY2B11, and UGT2B28) (data not shown). Whether these predicted miRNA binding sites are involved in post-transcriptional regulation of these UGT2B genes remains to be determined in future studies.
In summary, the present study demonstrates that UGT2B15 and UGT2B17 mRNAs are direct targets of miR-376c and provides the first evidence for post-transcriptional regulation of UGT2B15 and UGT2B17 by miRNAs in prostate cancer cells. Combined with previously described transcriptional regulation of UGT2B15 and UGT2B17 (Hu et al., 2014), our observations indicate that the expression and hence enzymatic activity of these two androgen-inactivating UGTs are regulated at both the transcriptional and post-transcriptional levels in prostate cancer cells. We anticipate that altered expression of miRNAs, such as miR-376c in the prostate, has a role in fine-tuning intraprostatic androgen signaling homeostasis and activity and hence may impact the risk of developing androgen-sensitive prostate diseases, such as prostate cancer.
Participated in research design: Wijayakumara, Hu, Mackenzie.
Conducted experiments: Wijayakumara, Hu.
Performed data analysis: Wijayakumara, Hu, Mackenzie.
Wrote or contributed to the writing of the manuscript: Wijayakumara, Hu, Mackenzie, Meech, McKinnon.
- Received May 21, 2015.
- Accepted July 9, 2015.
This work was supported by the National Health and Medical Research Council of Australia [ID1020931]. P.I.M. is a Senior Principal Research Fellow of the National Health and Medical Research Council of Australia. D.D.W. is supported by a Flinders Medical Centre Volunteers postgraduate scholarship.
- negative control microRNA
- polymerase chain reaction
- primary miRNA
- U6 small nuclear-2 RNA
- quantitative real-time polymerase chain reaction
- untranslated region
- Copyright © 2015 by The American Society for Pharmacology and Experimental Therapeutics