Testis specific Y-like 5: gene expression, methylation and implications for drug sensitivity in prostate carcinoma

Demethylating agent 5-aza-2′-deoxycytidine (Decitabine, DT) was from (Sigma Chemical Company, St Louis, MO). Antibodies used were rabbit anti-TSPYL5 (Immunoblot), rabbit anti- CDKN1A (Thr-145) (Santa Cruz Biotechnology. Santa Cruz, CA), rabbit anti-TSPYL5 (Sigma, Immunohistochemistry), rabbit anti-AKT, mouse anti-DNMT3B (Novus Biologicals, Littleton, CO), anti-DNMT1, anti-PTEN, anti-β-actin, anti-Histone-H3, anti- p-CDKN1A (T-145), anti-pAKT (Ser- 473) (rabbit), including secondary HRP-conjugated anti-rabbit and mouse (Cell Signaling Technology, Danvers, MA). Chemotherapy drugs paclitaxel (px) and docetaxel (dtx) were procured from the local veterinary pharmacy.

The PC cell lines, DU145, LNCaP and non-tumorigenic (NT) prostate epithelial cells RWPE-1 were purchased from ATCC (Manassas, VA). All of the carcinoma cells were maintained in custom RPMI or DMEM/F12 media with 10% FBS and Gentamycin. The RWPE-1 cells were maintained in a keratinocyte serum free media with growth factor supplements. The cells were tested routinely for mycoplasma contamination with the MycoAlert luciferase kit (Lonza, Allendale, NJ). Archival formalin fixed paraffin embedded (FFPE) tumor specimens from normal, benign or prostate carcinoma patients were obtained from the Pathology department at the University of Missouri Hospital after institutional IRB approval.

Total RNA from prostate carcinoma cells (DU145 and LNCaP), epithelial cells (RWPE-1) and FFPE prostate tissues was extracted using RNeasy or RNeasy FFPE kits (Qiagen, Valencia, CA), respectively. cDNA was generated from total RNA using a cDNA synthesis kit (Bio-rad, Hercules, CA). PCR was performed with TSPYL5 primers. β-actin was used as a housekeeping gene. The PCR conditions were as follows: denaturation at 98 °C for 1 min, followed by 28 cycles at 95 °C for 30 s, 55 °C for 30 s and 70 °C for 30 s, with a final extension at 70 °C for 8 min. The amplified PCR products were analyzed by 2% agarose gel electrophoresis containing Gel Red (Biotium, Hayward, CA). Quantitative real-time PCR (qRT-PCR), was performed with CFX Connect and a Sybr Green reaction (Biorad). The following primers were used for TSPYL5 PCR: Forward, 5′-TGGGCCCTTCTACTGGTGAACTTT-3′; Reverse, 5′- TCACCTGGAGCCACAGCATAATGA-3′. The mRNA expression in tissues was analyzed and the relative cumulative density was calculated by measuring area under curve (AUC) for each sample using an image processing and analysis program (Image J, NIH). Percentage average was obtained for each group and an arbitrary number of 1 was assigned for highest percentage group and subsequent groups were assigned numbers relative to 1 for graphical representation.

The genomic DNA isolated from PC cells using DNeasy Blood & Tissue Kit or tumor tissues using QIAamp DNA FFPE Tissue Kit (Qiagen) was bisulfite-modified with EZ-DNA Methylation-Gold Kit (Zymo Research, Irvine, CA) according to manufacturer’s instructions. The bisulfite reaction was carried out with 500 ng genomic DNA. Bisulfite converted DNA samples were stored at −20 °C until further use.

Methyl specific PCR (MSP) was performed in PC cells as well as FFPE tumor tissues using bisulfite-converted DNA with primers designed to include two CpG dinucleotides in each forward and reverse primer. Two sets of primers (CpG island) were designed; one, for methylated sequence (which retains CpG complementarity); 5′-GAGGTTATAGTTTAGGGGGAGTTG-3′; R- 5′- CCAAACAACACAAATACAAACTAAC-3′. For unmethylated sequences (complimentary to TpG sequence), the primers F- 5′-GAGAAATTTGTTGAGATTTAAAGTGA-3′; R- 5′CCATCACAAAAAAACATAATA-CACC-3′ were used. The presence of a methylated band in PCR is indicative of methylation in the original sequence [18]. Primers were designed using MethPrimer program [19]. The MSP and unmethylated sequence (USP) PCR bands in tissues upon gel electrophoresis (2% agarose) were analyzed for AUC using the Image J program. The percent methylation for each sample was calculated using AUC of methylated A (M) and unmethylated bands A (U) as follows: Percentage = A (M) ×100/A (M) + A (U).

Pyrosequencing (PSQ) of genomic DNA to quantitate the methylation of individual cytosine residues was performed as described earlier [20]. PSQ is a fast, reliable and quantitative method for analysis of CpG methylation [21]. Methylation analysis of DU145, LNCaP and RWPE-1 cells was performed with TSPYL5 specific primers (CpG island shores) which consisted of a forward (5′- AGAGAAAGTAAAGGTGGATGTTATAATGT-3′), biotinylated reverse (5′-Biosg/ATACTTCCATCCCTTACTATATAACCTA-3′) and sequencing primers (5′-AAAGGAGGTGTTGATAT-3′) designed for a TSPYL5 promoter sequence, followed by DNA sequencing in a Pyro Mark ID system by employing the Pyro Gold reagents kit (Life Technologies, Grand Island, NY). The primers were designed using a PSQ assay design program. The average degree of methylation at four CpG sites was analyzed using Pyro Mark ID software and results are depicted as percentage methylation.

IHC studies were performed as described previously [22] to identify the protein expression levels and cellular localization of TSPYL5 in non-malignant and malignant FFPE human prostate tissues using intelliPATH FLX (Biocare Medical). The analyzed tissue specimens included core tissue from patients with prostate adenocarcinoma (Gleason scores (GS) ranging from 6 to 9), normal and benign prostate tissues. Human testis tissue was used as positive control to detect TSPYL5 protein expression.

Immunoreactivity was scored by a board-certified pathologist (ME) in at least five random fields at 400× magnification in each section and the intensity of protein staining was scored on a 0–3+ scale (0 = no staining, 1 + = weak staining, 2 + = moderate staining, and 3 + = strong staining). The percentage of cells staining positive was scored on 1 - 4 scale (1 = 0–25% positive PC cells, 2 = 26–50% positive cells, 3 = 51–75% positive cells, and 4 = 76–100% positive cells). Composite score (CS) (0–12) was obtained by multiplying the staining intensity and percent of immunoreactive cells. Statistical significance was evaluated by the Mann–Whitney test. P < 0.05 was considered significant. H & E staining was performed according to standard procedures described in literature. Grading is assigned according to 2005 International Society of Urological Pathology Consensus Statement on Gleason Grading of Prostate Cancer (Epstein JI, Allsbrook WC Jr, Amin MB, Egevad LL; ISUP Grading Committee. The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma [23].

For overexpression of TSPYL5, LNCaP cells were plated in 6 well plates (0.3-1 ×10⁶/well) and allowed to grow to 70–80% confluency at 37 °C. The mammalian expression vector TSPYL5/pCMV6-AN-GFP (PV-TSPYL5) or pCMV6-AN-GFP (PV) (Origene, Rockville, MD) with N-terminal tGFP tag was transiently transfected into LNCaP cells using Lipofectamine 3000 (ThermoFisher Scientific, Walthem, MA) according to the manufacturer’s protocol and were allowed to grow for 72 h, harvested and subsequently used for further studies.

A cell viability assay was performed as described previously [24] using a WST-1 assay (Roche Applied Science, Indianapolis, IN) with or without 10 nM of chemotherapy drugs px or dtx. The results are expressed as percent viable cells after respective analysis. All experiments were performed in triplicate.

Protein was extracted from whole cell lysates using the M-PER mammalian protein extraction reagent (Thermo Scientific), and the concentrations were estimated by the Bradford method. Equal amounts of protein (35 μg) were loaded on to the gel. Subsequently, the proteins were blotted on to a nitrocellulose membrane. The membrane was probed separately with primary antibodies for TSPYL5, CDKN1A, and AKT including P-CDKN1A (Thr 145), pAKT (S-473), β-actin, histone H3, PTEN, DNMT-1 and DNMT3b. Following incubation with the primary antibody at 4 °C overnight, the membrane was incubated with a horseradish peroxidase-labeled secondary antibody and visualized with Luminate Forte Western HRP substrate (Millipore, Billerica, MA). The blot was imaged in a Kodak imaging station (Carestream Health). The protein band ratios were calculated from the protein band intensities obtained using Image J program.

TSPYL5 gene expression was analyzed in triplicate in PC cells (DU145, and LNCaP) and non-tumor (NT) epithelial cells (RWPE-1) by qRT-PCR analysis. qRT-PCR analysis indicated variable TSPYL5 mRNA expression in the cells tested (Fig. 1a). While the TSPYL5 mRNA expression was not significant between RWPE-1 and LNCaP (P ≥ 0.05), the expression was significantly low in DU145 cells (P = 0.02). Total cell lysates were analyzed for TSPYL5 protein expression by immunoblot analysis. As anticipated, the protein expression was very insignificant in DU145 but low to moderate in LNCaP and RWPE-1 cells (Fig. 1b), respectively. The difference in TSPYL5 protein expression was evaluated based on protein band intensities (Fig. 1c). The decrease in TSPYL5 protein expression was found to be highly significant between RWPE-1 and DU145 (P = 0.001) while moderate difference was observed between RWPE-1 and LNCaP, (P = 0.04) cells.

Due to the differential expression of mRNA in various cells used in this study, we analyzed the TSPYL5 gene methylation status in all the cells. First, we treated endogenously absent or low TSPYL5 expressing PC cells DU145 and LNCaP with DT. DT was effective in strong induction of TSPYL5 mRNA in LNCaP (P = 0.001) and DU145 (P = 0.0021) cells compared to wild type (WT) counterparts (Fig. 2a), suggesting that TSPYL5 gene is a target primarily for aberrant methylation. Next, in order to analyze the presence of DNA methylating enzymes DNMT1 and DNMT3B, an RT- PCR was performed to observe the variation in mRNA of these enzymes across the cells tested. As shown (Fig. 2b), all the cells had mRNA expression of these enzymes. Interestingly, it also was evident in DU145 cells where endogenous TSPYL5 expression was very low, both DNMT1 and DNMT3b mRNA expression was relatively high compared to the other cells, in which one or the other of the enzyme mRNA expression was low. Varying DNMT1 and DNMT3B protein expression was also observed in the nuclear fraction of the cells. Histone–H3 was used as a housekeeping protein.

We further investigated the TSPYL5 gene methylation status by MSP analysis. The MSP primers were designed within the chromosomal 8 regions (97,277,582-97,277,700) of the TSPYL5 gene (CpG islands, Fig. 3a) and sodium bisulfite-modified genomic DNA was used as a template. MSP results revealed a differential methylation pattern among the cells (Fig. 3b). TSPYL5 gene exhibited decreased methylated band intensity in the following order: DU145 > LNCaP > RWPE-1. An intense methylation band was observed in DU145 (P =0.001) and LNCaP (P = 0.012) cells compared to RWPE-1 (Fig. 3c). While DU145 cells had no unmethylated band, the LNCaP cells had a very dim unmethylated band compared to RWPE-1 cells, which had an intermediate intensity unmethylated band.

In order to further investigate the extent of methylation in the cells, we analyzed a different chromosomal region within the TSPYL5 gene (Chr 8: 97278367–97278417) for an individual cytosine methylation pattern using PSQ. PSQ quantifies methylation in explicit sequence context, thereby enabling several consecutive CpG sites to be quantified individually in a single assay. We selected the above region (CpG island shore, Fig. 3a) to avoid excessive CG density for design of PSQ primers. Four CpG sites were selected (Position 1–4, Fig. 4a and Additional file 1: Table S1) to analyze the extent of methylation in cytosine residues. Cumulative methylation percentage for individual cell type (Fig. 4b) indicated that the methylation percentage was highest (1.2 fold; P = 0.04) in DU145 cells relative to RWPE-1. The difference in methylation percentage of cytosine residues at the position of interest between LNCaP and RWPE-1 was not statistically significant (P > 0.05). The pyrograms depicting the methylation status of the cytosine residues in the selected sequence for DU145, LNCaP and RWPE-1 cells and the comparative methylation percentage of individual cytosine residues are shown in (Additional file 2: Figure S1).

After identifying an inverse relationship between TSPYL5 mRNA and the presence of methylation in NT and PC cells, we sought to extend the analysis of gene expression and DNA methylation to benign and prostate tumor tissues. TSPYL5 mRNA expression was observed in normal (n = 3) and benign samples (n = 9) (Fig. 5a). In total, 21 tumor samples were analyzed, out of which four samples had a GS of 6, fourteen samples had GS-7 and three samples had GS - 8 or- 9. Tumor tissues with GS ≥ 8 exhibited almost no TSPYL5 expression (Fig. 5b, T19-21). Variable intermediate expression was observed with tumors with a GS-6 or −7 (T1-T18), however, in a few tumor GS-7 samples (T7, T12 and T14) very weak or no TSPYL5 mRNA expression was observed. These three samples had Gleason pattern (4 + 3). The graphical representation of TSPYL5 mRNA expression in different tissues are depicted in (Additional file 3: Figure S2 (a)). Densitometry gel analysis indicated a decrease in the TSPYL5 mRNA expression in tissues with GS-7 (P = 0.012) and GS ≥ 8 (P = 0.001).

For MSP DNA methylation studies in tissues, the genomic DNA was isolated from normal, select benign and tumor samples (n = 3 each) and bisulfite converted before methylation analysis. MSP analysis with CpG island primers, demonstrate that methylation was low in normal and benign samples and lower or intermediate in graded tumor tissues (GS-6 or-7). Unfortunately, we had very little starting material of the tumor tissues with GS =7 (T7, T12, T14) were very little and were unable to assess the methylation analysis in these tissues. However in tissues with GS ≥ 8 increased methylation was observed (Fig. 5c). The cumulative methylation percentage is depicted in (Additional file 3: Figure S2 (b)). Significant methylation was observed between benign and tumor tissues (GS-7; P = 0.047) and GS ≥ 8 (P = 0.032). Furthermore, we analyzed the expression of DNMT’s in high grade advanced PC tumor tissues (Gleason score, GS ≥ 8) by IHC. It appears that predominantly DNMT3b protein was expressed while DNMT1 was found to be relatively low or not detected (Additional file 4: Figure S3).

IHC analysis of TSPYL5 in human normal prostate and tumor tissues identified protein expression patterns that mirrored the tissue mRNA expression data. A minimum of three tissues were analyzed in each case. Normal human testis tissue (Fig. 6a) was used as positive control for TSPYL5 expression. The testis tissue showed strong membrane, nuclear and cytoplasmic staining (Fig. 6b). In benign prostate tissue, benign acini lined by inner secretory epithelial cells and outer basal cell layer (Fig. 6c), the TSPYL5 expression was prominent in the cytoplasmic membrane (Fig. 6d). The prostate adenocarcinoma specimens with GS-6 (3 + 3), exhibited both nuclear and cytoplasmic staining with a composite score (CS) of 12 (3x4) (Fig. 6f). Prostate adenocarcinoma cases with GS-7 (3 + 4), show moderate cytoplasmic and membrane staining with CS of 8 (2x4) (Fig. 6h). Interestingly, in the tumor specimens with GS-8 (4 + 4) or above the staining intensity was very weak (CS ≤ 1) and is mainly confined to the cell membrane. There is a loss of nuclear and cytoplasmic staining (Fig. 6j). Corresponding H & E stains for testis, benign or prostate tumor tissues were processed in parallel (Fig. 6a, c, e, g and i). The composite scores obtained in tumor tissues with GS-6 or-7 were higher relative to the GS-8. The difference in protein staining intensity between benign and tumor tissues with GS −8 (P = 0.012) and GS-7 (P = 0.032) was significant, while no significant difference was observed in staining intensities between benign and tumor tissues with GS-6 (P > 0.05). Further analysis of tissues with GS 9 (4 + 5) compared to benign tissues (P = 0.008). (Additional file 5: Figure S4). Altogether, TSPYL5 expression diminishes in high grade prostate carcinoma compared to the benign tissue or intermediate grade prostate carcinoma suggesting that TSPYL5 could function as an indicator of disease progression. An overall summary of gene expression, methylation frequency and IHC composite scores are presented in Table 1. The patient tissues used in this study and their assigned Gleason scores are depicted in (Additional file 6: Table S2).

We analyzed the expression of TSPYL5 and other proteins including CDKN1Aand pAKT in, DU145, LNCaP and RWPE-1 cells (Fig. 7a). The protein bands were analyzed (Fig. 7b) by image quantification software as described in the methods. TSPYL5 protein expression was absent in DU145, low in LNCaP and moderate in RWPE-1 cells.

To study the co-expression of other cellular proteins in all the cells, we focused on two important proteins CDKN1A and AKT which participate in cellular proliferation, drug sensitivity and cell survival. Interestingly, CDKN1A expression was low in DU145 cells in which TSPYL5 expression was insignificant. However, high CDKN1A expression was observed in LNCaP compared to RWPE-1 cells in which the endogenous TSPYL5 is either low or moderate, respectively. Phosphorylated CDKN1A (P- CDKN1A, Thr145) was absent in all cell lines. Variable AKT expression was observed in all these cells. Furthermore, LNCaP and RWPE-1 cells had pAKT expression, but no pAKT was observed in DU145. While PTEN (a tumor suppressor protein, TSP) is expressed in both DU145 and RWPE-1 cells in which TSPYL5 expression was absent or moderate, no expression was observed in LNCaP cells. Based on the protein band density (Fig. 7b), the ratios of pAKT/TSPYL5 in different cells were 0.25 (DU145), 18.9 (LNCaP) and 3.46 (RWPE-1). Also, the ratios between CDKN1A/TSPYL5 was 0.14 (DU145), 16.8 (LNCaP) and 3.7 (RWPE-1). These ratios indicate the differences in the relative expression of pAKT and CDKN1A in relation to TSPYL5.

In order to evaluate the drug sensitivity in WT and TSPYL5 overexpressing LNCaP cells, we tested the effect of two standard drugs used for PC treatment, paclitaxel (px) and docetaxel (dtx). Cellular viability in the presence of each drug (10 nM) was tested in WT, cells transfected with vector only (PV) and TSPYL5 overexpressing (PV-TSPYL5) LNCaP cells. Dtx decreased the WT and PV LNCaP cell viability by 40% (P = 0.012, P = 0.014). However, PV-TSPYL5 LNCaP cells exhibited enhanced sensitivity to dtx (62%, P = 0.008) compared to WT cells without the drug (Fig. 8a). Similar results were observed in the presence of px (Additional file 7: Figure S5). Changes in pAKT, CDKN1A and PTEN protein expression in TSPYL5 overexpressing LNCaP cells were analyzed by Western analysis. We noted a reduction in CDKN1A (P = 0.009) in PV-TSPYL5 LNCaP cells which originally had high endogenous CDKN1A. Only a mild decrease (P ≥ 0.05) was observed with pAKT and no changes in PTEN expression (Fig. 8b and c).