Reciprocal regulation of PCGEM1 and miR-145 promote proliferation of LNCaP prostate cancer cells

Prostate cancer gene expression marker 1 (PCGEM1) is a long non-coding RNA (lncRNA) overexpressed in prostate cancer (PCa) cells that promotes PCa initiation and progression, and protects against chemotherapy-induced apoptosis. The microRNA miR-145 functions as a tumor suppressor in PCa. We speculate that reciprocal regulation of PCGEM1 and miR-145 promote proliferation of LNCaP prostate cancer cells. To test this hypothesis, the interaction between PCGEM1 and miR-145 was examined using a luciferase reporter assay. Expression levels were selectively altered in LNCaP cells and noncancerous RWPE-1 prostate cells by transfection of miR-145 or small interfering RNA sequences against (siRNA) PCGEM1. Relative expression levels were detected by RT-PCR, tumor cell growth and early apoptosis by the MTT assay and flow cytometry, respectively, and tumor cell migration and invasion properties by transwell assays. The effect of siRNA PCGEM1 and miR-145 transfection on prostate cancer growth in vivo was examined in the (nu/nu) mouse model. PCGEM1 and miR-145 exhibited reciprocal regulation; downregulation of PCGEM1 expression in LNCaP cells increased expression of miR-145, while overexpression of miR-145 decreased PCGEM1 expression. Transfection of the miR-145 expression vector and siRNA PCGEM1 inhibited tumor cell proliferation, migration, and invasion, and induced early apoptosis both in vitro. In contrast, there was no effect on RWPE-1 cells. We demonstrate a reciprocal negative control relationship between PCGEM1 and miR-145 that regulates both LNCaP cell proliferation and nu/nu PCa tumor growth. The results also identify PCGEM1 and associated regulators as possible targets for PCa therapy.


Background
Long non-coding RNAs (lncRNAs) are untranslated transcripts longer than 200 nucleotides baring many of the structural characteristics of mRNAs, including a polyA tail, 5′-capping, and a promoter structure, but no conserved open reading frame [1][2][3][4][5][6]. Many lncRNAs are expressed at specific times and in specific tissues during development, and exhibit a variety of slicing patterns. It has been proposed that lncRNAs are involved in the epigenetic regulation of coding genes, and thus exert a powerful effect on a number of physiological and pathological processes, including the pathogenesis of many human cancers [7][8][9][10][11].
Given the structural similarly with mRNAs, lncRNAs may be another important member of the non-coding RNA family [18]. The interaction between lncRNAs and miRs has been linked to the invasion and metastasis of tumors [19]. For example, the miR-29a epigenetically modulated expression of the lncRNA MEG3 in hepatocellular carcinoma (HCC) through promoter hypermethylation [20]. Loss of miR-31 expression in triple-negative breast cancer (TNBC) lines is attributed to hypermethylation of its promoter-associated CpG islan. MicroRNA-31 anchors the novel lncRNA LOC554202 and adjusts its transcriptional activity [21]. Moreover, the lncRNA HULC can inhibit the expression of the tumorigenic miR-372 [22].

Design and construction of eukaryotic expression vector for hsa-miR-145
The mature hsa-miR-145 sequence (5′-GUCCAGUUU CCCAGGAAUCCCU-3′) is available from the miRNA Registry (MIMATOOOO437). To prevent formation of a termination signal, TTGGCCACTGACT was selected as the region in a miR expression vector template. The sequence TGCT was added to the 5′ positive-sense strand template of the miR expression vector and GTCC to the 5′ antisense strand template. Further, a nonspecific sequence was designed and sent to Shanghai GenePharma Co, Ltd. for synthesis. The assay was according to previously described [35]. The eukaryotic expression vector plasmid targeting hsa-miR-145 was named pmiR-145.

Grouping and cell transfection
The experimental culture groups included 1) untransfected LNCaP and RWPE-1 cells (control groups), 2) cells transfected with pmiR-145 or miR-145 mimics (1.6 μg/ml and 50 nM, respectively), 3) cells transfected with the scrambled nucleotide sequence and empty vector (negative control or NC groups, 50 nM), 4) cells transfected with a miRNA inhibitor (NI group, 100 nM), 5) a negative control for NI (NCI group, 50 nM), 6) cells transfected with siRNA PCGEM1 sequence (siRNA PCGEM1 group, 50 nM). Cells in log phase growth were seeded on 6-well culture plates (2 × 10 5 cells/well) and transfected when the cell fusion rate reached 70%. The DNA Lipofectamine 2000 or RNA Lipofectamine 2000 compound was added according to the manufacturer's instructions (Invitrogen). After 6 h, the transfection medium was discarded. Cells were washed with serumfree RPMI 1640 and then cultured in RPMI 1640 supplemented with 10% FBS.

Luciferase reporter assay
The whole mRNA sequences of the PCGEM1 gene were obtained by PCR amplification and cloned separately into multiple cloning sites of the psi-CHECKTM-2 luciferase miRNA expression reporter vector. HEK293T cells were transfected with miR-145 mimic, miR-145 inhibitor, a control miRNA, a miRNA inhibitor control, or empty plasmid using Lipofectamine 2000 according to the manufacturer's instructions. Nucleotide-substitution mutation analysis was carried out using direct oligomer synthesis of PCGEM1 sequences. All constructs were verified by sequencing. Luciferase activity was measured using the dual luciferase reporter assay system kit (Promega Co, Madison, WI, USA) according to the manufacturer's instructions on a Tecan M200 luminescence reader.

MTT assay
The viability of LNCaP and RWPE-1 cells was determined by MTT assay. Briefly, cells at 5 × 10 4 /ml were transfected with siRNA PCGEM1 (siRNA PCGEM1 groups, 50 nM), empty plasmid and scramble sequence (negative control group, 1.6 μg/ml), or pmiR-145 (pmiR-145 group, 1.6 μg/ml) in the presence of Lipofectamine 2000 and serum-free RPMI 1640 media for 6 h. Cells were plated in 96-well plates in medium containing 10% FBS for another 24, 48, or 72 h. MTT stock solution (20 μl, 5 mg ⁄ ml) was added to each well for a final MTT concentration of 0.45 mg⁄ ml and the plate was incubated for 4 h at 37°C. Media was then removed and dimethylsulfoxide (DMSO) (150 μl added to dissolve the blue formazan crystals (the product of MTT conversion by viable cells) at room temperature for 30 min. The relative change in viable cell number was estimated by absorbance at 570 nm on a Bio-Rad microtiter plate reader (Hercules, CA, USA).

Flow cytometry assay
LNCaP and RWPE-1 cells were seeded at 1.0 × 10 6 /ml in 24-well plates (Costar) and transfected in 500 μl media/ well siRNA PCGEM1 (siRNA PCGEM1 groups, 50 nM), empty plasmid and scramble sequence (negative control group, 1.6 μg/ml), or pmiR-145 (pmiR-145 group, 1.6 μg/ml) by Lipofectamine 2000 reagent (Invitrogen) in serum-free RPMI 1640 for 6 h. After transfection, 500 μl of the appropriate growth medium containing 20% FBS were added to each well. Cells were incubated for another 48 h then harvested, washed twice with PBS, fixed with 70% ethanol, and treated with RNase A (1 mg/ml). Finally, the cells were double-stained with FITC-conjugated annexin-V and propidium iodide (PI) solution (50 μg/ml). For each sample, data from approximately 10000 cells were recorded in the list mode on logarithmic scales. Apoptosis and necrosis were analyzed by quadrant statistics on double negative, annexin-Vpositive/PI-negative, annexin-V-negative/PI-positive, and double-positive cells.

Migration and invasion assay
Cells were transfected with siRNA PCGEM1 (siRNA PCGEM1 groups, 50 nM), empty plasmid and scramble sequence (negative control group, 1.6 μg/ml), or pmiR-145 (pmiR-145 group, 1.6 μg/ml) by Lipofectamine 2000 reagent in serum-free RPMI 1640 for 6 h. One day after transfection, 1 × 10 5 cells were collected, resuspend in 100 μl basal medium, and transferred to the transwell chamber. A 600 μl volume of complete medium was added to the well and the chamber inserted. Plates were incubated at 37°C for 48 h. Remaining cells were swabbed from the top transwell membrane filter and the chamber submerged in 4% paraformaldehyde for 20 min. Cells in the well were stained with crystal violet for 10 min, washed in PBS buffer, then counted by light microscope to determine transwell migration.
The transwell assay was performed as above except with Matrigel in the wells of 24-well plates. The Matrigel was first incubated in pre-chilled basal medium (40 μl) at 37°C for 2 h. The excess medium was discarded, 100 μl basal medium added to the well with the Matrigel and 600 μl to the chamber, following by incubation at 37°C overnight. Transfected cells (1 × 10 5 ) were resuspend in 100 μl basal medium and transfer to the transwell chamber. Then, 600 μl complete medium was added to the well and the chamber inserted. The plates were incubated at 37°C for 24 or 48 h. Cells were stained with crystal violet for 10 min, washed with the PBS and counted using an inverted microscope.

In vivo treatment
BalB/c (nu/nu) mice from the Animal Center of Guangzhou Province (Guangdong, China) received subcutaneous injections of 2 × 10 6 LNCaP cells into each axilla area. When xenograft tumors became palpable (about 0.1 mm 3 ), mice were randomly divided into the control group receiving PBS injection (100 μl), siRNA PCGEM1 (500 nM), negative control (plasmid, scramble sequence 16 μg), PmiR-145 (16 μg), with 6 mice each group. There was no difference in baseline tumor size between the groups. Tumor volume was calculated every 3 days according to the formula v = ab2π/6, where "a" is the maximum tumor diameter and "b" the minimum diameter. After treatment for 20 d, mice were euthanized and tumors were dissected and weighed.

Data analysis
All results are the averages of at least three independent experiments from separately treated and transfected cultures. Data are expressed as the mean ± SD. Statistical comparisons were made by one-way analysis of variance (ANOVA). P < 0.05 was considered to indicate a statistically significant difference.

miR-145 regulates PCGEM1 expression by binding to the PCGEM1
ReRNA, an lncRNA target miR prediction software, predicted miR-145 binding sites at 983 bp and 1004 bp in the PCGEM1 ( Figure 1C). The full-length sequence of PCGEM1 (1643 bp) was cloned downstream of the luciferase gene in the psiCHECK carrier to construct the psiCHECK-2-PCGEM1 carrier. Co-transfection of HEK293T cells with miR-145 mimics was performed to detect the binding site on the PCGEM1 (983-1004 bp). The co-transfection of HEK293T cells with miR-145 mimics and psiCHECK-2-PCGEM1 also significantly inhibited luciferase activity (P <0.05) ( Figure 1A); however, co-transfection of HEK293T cells with miR-145 mimics and empty psiCHECK-2 had little effect on the activity of luciferase (P > 0.05) ( Figure 1B). We concluded that miR-145 can regulate the expression of PCGEM1 by directly binding to target sites within the PCGEM1 sequence.

The miR-145 and PCGEM1 expression in in the LNCaP cells exhibited reciprocal regulation
To examine possible mutual regulation of PCGEM1 and miR-145, we designed a small interfering RNA sequences to knockdown PCGEM1 expression and detected the expression levels of miR-145. Alternatively, we transfected PmiR-145 into LNCaP cells and then detected the expression levels of PCGEM1. Transfection of the siRNA PCGEM1 downregulated expression of PCGEM1 and resulted in a significant increase in miR-145 expression compared to the NC group and control group (both P < 0.05) (Figure 2).

siRNA-mediated PCGEM1 knockdown and miR-145 overexpression inhibited LNCaP proliferation
To examined the effect of siRNA-mediated PCGEM1 knockdown and concomitant overexpression of miR-145 on PCa cell proliferation, normal prostate cells (RWPE-1) and LNCaP cells were transfected as above and cell numbers estimated by MTT after 24, 48, 72 h. Indeed, both siRNA PCGEM1 and PmiR-145 transfection groups exhibited a significant reduction in LNCaP cell proliferation, while transfection with empty vector and scrambled sequences (negative control group) had no effect compared to untransfected controls ( Figure 3A). In contrast, proliferation of RWPE-1 cell was not changed significantly, even after 72 h ( Figure 3B).

siRNA-mediated PCGEM1 knockdown and miR-145 overexpression induced early apoptosis of LNCaP cell
Cell apoptosis of the above transfection groups was detected by flow cytometry using annexin-V/PI double staining. In total, 25.16% of PCGEM1 knockdown and 23.6% of miR-145-overexpressing LNCaP cells were in early apoptosis ( Figure 4A, C), compared to only 0.63% and 0.82% of RWPE-1 cells, respectively, in these same transfection groups ( Figure 4B, D).

siRNA-mediated PCGEM1 knockdown and miR-145 overexpression inhibits progression of tumor xenografts
LNCaP tumor xenografts were established in Athymic nude mice to evaluate the effects of siRNA PCGEM1 or pmiR-145 on prostate cancer growth in vivo. Compared to the untreated animals, application of siRNA PCGEM1 or pmiR-145 significantly diminished the tumor volume, whereas negtive control group had no effect ( Figure 6). No body weight loss or diarrhea was observed and all animals (treated as well as non-treated) survived. The results shows that shows that reduce the expression of PCGEM1  or over express miR-145 can effectively inhibit prostate cancer growth in vivo.

Discussion
Long non-coding RNAs (lncRNAs) are a new class of regulatory RNA [36]. These mRNA-like molecules, which lack significant protein-coding capacity, were once thought to be a part of the genomic "dark matter", but recent studies have implicated lncRNAs in a wide range of biological functions through poorly understood molecular mechanisms [37]. Despite recent insights into how lncRNAs function in such diverse cellular processes as regulation of gene expression and assembly of cellular structures, by and large, the key questions regarding lncRNA mechanisms remain to be answered [38]. The lncRNA Prostate cancer gene expression marker 1 (PCGEM1) is overexpressed in PCa, suggesting roles in proliferation, metastasis, and invasion [39]. In order to reveal the mechanisms regulating expression in PCa, we have predicted PCGEM1 interaction with miR-145 using billogical information (Figure 1C), and futher investigated a possible interaction with the tumor suppressor miR-145. Co-transfection of LNCaP cells with miR-145 mimics or miR-145 inhibitor with psiCHECK-2-PCGEM1 significantly inhibited reporter gene activity but only miR-145 suppressed reported gene expression when transfected with empty psiCHECK-2 ( Figure 1A, B). Thus, miR-145 may regulate PCGEM1 expression by directly binding to target sites within the PCGEM1 sequence.
We then demonstrated a mutual inhibitory control relationship between PCGEM1 and miR-145 by selective siRNA-mediated PCGEM1 knockdown and miR-145 overexpression. Expression of PCGEM1 (locus 2q32) was detected in the androgen receptor-positive cell line LNCaP but not in noncancerous prostate lines or androgenreceptor negative Pca lines [23]. PCGEM1 overexpression in LNCaP cells promotes cell proliferation and a dramatic increase in colony formation, suggesting a role in cell growth regulation [24]. In contrast, miR-145 expression was low in all the prostate cell lines tested (PC3, LNCaP, and DU145) compared to the normal cell line RWPE-1, and in cancerous regions of human prostate tissue compared to adjacent normal prostate tissue [39]. To test the possibility of mutual negative regulation of PCGEM1 and miR-145, we design a small interfering RNA targeting PCGEM1 and a vector for miR-145 overexpression (pmiR-145) and transfected these into LNCaP cells and normal RWEP-1 cells. RT-PCR results showed that knockdown of PCGEM1 in LNCaP cells increased miR-145 expression ( Figure 2A) and that miR-145 overexpression reduced PCGEM1 expression ( Figure 2B). Inhibition of PCGEM1 reduced LNCaP proliferation ( Figure 3A), transwell migration and invasive capacity into Matrigel ( Figures 5A, 6A), and the growth of solid tumors, possibly by promoted early apoptosis ( Figure 4A). However, altering PCGEM1 expression had no significant effect on RWPE-1 cell growth, migration, or invasion (Figures 3, 4, 5, 6B). The proliferation, colony formation, and soft agar growth of liver cancer cells was reduced by inhibiting expression of the lncRNA TUC339 using an siRNA [40], while silencing HULC expression in hepatoma effectively inhibited the growth of liver cancer cells [41]. In contrast, siRNA gene silencing of MEG3 expression promote cell proliferation, whereas overexpression inhibited proliferation and promoted apoptosis [42]. Thus, individual lncRNAs can either promote or inhibit carcinogenesis. Selective knockdown and overexpression of lncRNAs may be feasible strategies to reduce tumor growth in a variety of tissue. Specifically, miR-145 is a well documented tumor suppressor [43][44][45][46], and we successfully constructed an overexpressing vector that suppressed PCa cell growth with no observable effects on noncancerous prostate cells. Similarly, gain-of-function assays revealed that miR-145 transfection inhibited cell proliferation, migration and invasion of PC3 and DU145 PCa cell lines [47].
While PCGEM1 is known to be overexpressed in PCa, it is unknown if overexpression directly causes hyperproliferation and (or) metastasis [24]. Fu [48] found that overexpression of PCGEM1 attenuated doxorubicininduced expression of p53 and p21Waf1/Cip1, and inhibited apoptosis of LNCaP cells. Petrovics [24] revealed that elevated PCGEM1 expression increased cell proliferation and Rb phosphorylation. However, to the best of our knowledge, no study has investigated PCGEM1 regulation by miR-145. An siRNA PCGEM1 inhibited LNCaP cells growth and reduced migration and invasion, likely by raising the expression levels of miR-145.
Indeed, we confirmed direct binding of miR-145 to the PCGEM1 and demonstrated reciprocal regulation of these two transcripts. Moreover, miR-145-mediated suppression Figure 6 Effect of siRNA PCGEM1/pmiR-145 on prostate cancer xenografts. LNCaP tumor xenografts were established in male Athymic nude mice. Animals in the treatment arm received siRNA PCGEM1 (500 nM /kg once daily) or pmiR-145 (16 μg/kg once daily). Control mice received PBS (100 μl/kg once daily). Negative Control mice received empty plasmid or scrambled sequence (16 μg/kg once daily). After 20 daily treatments, tumors injected with siRNA PCGEM1 or pmiR-145 were significant smaller than control or NC group tumors (P △ * < 0.05).
of PCGEM1 suppressed tumor growth in vivo and PCa cell proliferation and invasive capacity in vitro. In turn, Reciprocal regulation of PCGEM1 and miR-145 promote proliferation of LNCaP prostate cancer cells.
In conclusion, our study demonstrates reciprocal negative control of PCGEM1, a tumor-promoting long noncoding RNA, and the tumor suppressor miR-145. This study highlights the interrelationship between two classes of non-coding RNAs. Both downregulation of PCGEM1 or overexpression of the miR-145 reduced the proliferation and invasive capacity of prostate cancer cells in vitro and in vivo.