Gene expression analysis by quantitative qRT-PCR was performed on fresh frozen tissue samples obtained from 26 CRC patients, 24 samples of corresponding normal mucosa sections and 42 colorectal adenoma samples. Cryostat sections were prepared from each specimen using a Microm HM 505E (Zeiss) and upper and lower sections from each cryosection set were evaluated by a pathologist for relative cell type content.

Formalin-fixed and paraffin-embedded tissue from 24 CRC and 12 AD samples was used for DNA isolation and DNA methylation assessment, whereas 14 of the CRC samples with matched normal colonic mucosa sections, along with 10 AD samples, were used for immunohistochemical evaluation.

Ten additional samples of non-neoplastic colonic epithelia obtained from normal tissue areas on the margins of resected colorectal tumors were collected from 10 anonymous patients by scraping the colonic epithelial layers with plastic swab sticks. The samples were used for DNA isolation and served as controls in DNA methylation analysis.

Clinical tissue specimens were collected at the Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology in Warsaw in accordance with local ethics committee approval and individual patients' informed consent. Patient characteristics are presented in Table I.

HCT 116 and HT29 cell lines were cultured in McCoy's 5A medium (Sigma-Aldrich Inc., St. Louis, MO, USA) supplemented with 10% fetal bovine serum. SW480 and COLO-205 cell lines were cultured in RPMI-1640 medium (Sigma-Aldrich Inc.) supplemented with 5% and 10% fetal bovine serum, respectively. The Caco-2 cell line was cultured in MEM medium (Sigma-Aldrich Inc.) supplemented with 20% fetal bovine serum (Gibco). All culture media contained 1% antibiotics (PenStrep, Gibco). Cell culture was performed at 37°C in a humidified 5% CO₂ atmosphere. The cells were grown in culture medium until 50% confluence and subsequently treated with 0, 2.5, 5 and 10 µM 5-aza-2′-deoxycytidine (Abcam, Cambridge, MA, USA) for 3 days. The medium containing 5-aza-dC was refreshed daily.

Relative PTPRH expression was assessed with qRT-PCR with respect to MIQE guidelines (19). Total RNA from tissue samples and cell lines was isolated using RNeasy Mini (Qiagen) and quantified with NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, waltham, MA, USA). RNA samples were checked for quality on the Agilent 2100 Bioanalyzer and samples with RIN >7 were included. Each RNA sample (1 µg) was subjected to reverse transcription with SuperScript II Reverse Transcriptase (Invitrogen) using both random hexamer and oligo dT, according to manufacturer's protocol. qRT-PCR was carried out using a 384-well format and an Applied Biosystems 7900HT Fast Real-time PCR System (Applied Biosystems) with Sensimix SYBR kit (Bioline), according to manufacturers' instructions in a volume of 5 µl. RT-qPCR was performed using the following cycling conditions: 95°C for 10 min followed by 40 cycles of 95°C for 15 sec and 60°C for 45 with subsequent dissociation curve protocol. ACTB (β-actin) was selected as a reference gene after a previous evaluation of 4 potential reference genes: ACTB, UBC, RPLPO, GAPDH with GeNorm software (20). The standard curves based on amplifying known cDNA template concentrations were used for assessment of PCR efficiency and linear dynamic range. Normalized PTPRH expression was calculated using Q-gene software (21). Primer sequences are listed in Table II.

Immunohistochemical staining was performed on 4-µm formalin-fixed, paraffin-embedded tissue sections of CRCs and matched normal mucosa from 14 patients as well as 14 colorectal adenomas using the Envision Detection System (Dako). Sections were deparaffinized with xylene and rehydrated in a series ethanol solutions of decreasing concentration. Heat-induced epitope retrieval was carried out in a Target Retrieval Solution pH 9.0 (Dako) in a 96°C water bath, for 20 min. After cooling the retrieval solutions for 25 min at room temperature, the slides were treated for 5 min with a Blocker of Endogenous Peroxidase (Dako). Slides were incubated with polyclonal antibody (Ab) against PTPRH (PA5-31340, dilution 1:200) (Thermo Fisher Scientific) for 30 min at room temperature and subsequently labelled with the Envision Detection System (Dako). The color reaction product was developed with 3,3′-diaminobenzidine, tetrahydrochloride (Dako) as a substrate, and nuclear contrast was achieved with hematoxylin counterstaining. Staining intensity was assessed by a four-grade scale: 0, lack of expression; 1, weak expression; 2, moderate expression; and 3, strong expression. The stained tissue slides were examined by the pathologist who was blinded to the qRT-PCR results.

Cell pellets were resuspended in ice-cold RIPA buffer, incubated for 30 min in 4°C and centrifuged at 12500 rpm for 20 min at 4°C. Samples were resolved using SDS-PAge and electrotransferred to polyvinylidene fluoride membranes (EMD Millipore). PTPRH was detected with polyclonal anti-PTPRH Ab (Pierce PA5-31340 PTPRH Rabbit polyclonal Ab, Pierce; Thermo Fisher Scientific), and a secondary Ab conjugated to HRP (#31460, Pierce; Thermo Fisher Scientific). Detection of β-actin with mouse HRP conjugated mAb (8H10D10) (Cell Signalling, #12262) served as control. Detection was performed by the enhanced chemiluminescence method (Pierce; Thermo Fisher Scientific).

DNA methylation levels of two 5′ PTPRH regions [−190 to −253 upstream and +57 to +147 downstream from transcription start site (TSS)] were measured in CRC cell lines as well as in 14 CRC, 14 AD and 10 normal mucosa samples (obtained from swabs of normal colon epithelium) using the pyrosequencing assay according to the published protocol (22). DNA was isolated with the QIAamp DNA Mini kit (Qiagen) and bisulfite converted using the EpiTect kit (Qiagen), according to manufacturer's instructions. The PCR reaction was performed in a 30 µl mixture containing 1X PCR buffer, 2 mM MgCl₂, 0.25 mM dNTPs, 0.2 µM of each primer, 0.5 units of FastStart DNA Polymerase (Roche Applied Science) under the following conditions: 94°C for 3 min, with subsequent 40 cycles of 30 sec at 94°C, 30 sec at the annealing 52°C temperature and 30 sec at 72°C and final elongation for 7 min at 72°C. The primer sequences are listed in Table II.

The PCR products were purified and analyzed using the PyroMark Q24 System (Biotage AB, Uppsala, Sweden), according to manufacturer's instructions. For every particular sample the average methylation level of all CpGs within each analyzed region was calculated and used for comparing the sample groups.

The cells were cross-linked for 10 min at room temperature by adding formaldehyde directly to culture medium (final concentration 1%), then formaldehyde was quenched at room temperature for 5 min with glycine at 125 mM concentration. Fixed cells were suspended in 100 µl of hypotonic buffer A with NP-40 detergent (10 mM HEPES pH 7.9, 2 mM MgCl₂, 2 mM KCl, NP-40 0.5%) supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific; 78441) and incubated on ice for 5 min followed by centrifugation at 2000 rpm for 5 min in 4°C. Next, nuclei pellets were resuspended in lysis buffer (12.5 mM Tris-HCl, pH 8.0; 2.5 mM EDTA; 0.25% SDS) containing protease and phosphatase inhibitors (Thermo Fisher Scientific; 78441). Chromatin was sheared in a Bioruptor Plus (Diagenode) using the following protocol: 30 sec on-off cycles for 10 min at high intensity. ChIP assays were done using the Matrix-ChIP on polypropylene plates (23). ChIP DNA data were expressed as percent of input DNA, or as the ratio of modified histone to total histone H3 (24). The following antibodies were used in ChIP assay: non-specific rabbit IgG (I-1000, Vector Laboratories), Pol2 (4H8) (Santa Cruz; sc-47701), Histone H3 (Abcam, ab1791), H3K4me3 (Diagenode; pAb-003-050), H3K27me3 (Millipore, 07-449). PCR primers for PTPRH promoter and 3′ regions (exon 20) were used as well as primers for two control promoter regions of transcriptionally active (ACTB) and inactive (HBB) genes. Primer sequences are listed in Table II.

Gene expression values and DNA methylation levels were compared using the two-sided Mann-whitney U test. A significance threshold level of α=0.05 was adopted. Data were analyzed and visualized by GraphPad Prism (GraphPad Software). Microarray data stored in the Gene Expression Omnibus (GEO) were analyzed using GEO2R, being an interactive web tool integrated into the GEO service (25).

We searched for tumor specific expression changes of all genes encoding different RPTPs in our previously reported microarray transcriptome data for normal and neoplastic colorectal tissues (18). We found that PTPRH mRNA had the most reduced expression in CRC of all genes encoding RPTPs with a fold change (FC) of expression at −5.32 (Table III).

Consistent with this are the comparisons of PTPRH expression in normal colonic samples and CRC in 5 different datasets from the GEO database (26–29) that uniformly showed a significant decrease of PTPRH expression in cancer sections (GSE40967 logFC=−2.18; GSE25070: logFC=−1.4; GSE44076 logFC=−2.79; GSE21510 logFC=−2.97; GSE32323 logFC=−2.4).

Validating PTPRH expression by qRT-PCR on 24 normal colonic mucosa samples, 46 AD and 26 sections of CRC confirmed a stepwise decrease in mRNA expression through the adenoma-carcinoma sequence (Fig. 1). A significant (P<0.0001) downregulation with a FC=−3.4 and FC=−7.6 was observed when normal samples were compared to adenoma and cancer samples, respectively (Fig. 1A).

A subsequent analysis of PTPRH protein expression was performed by immunohistochemical staining of 14 paired CRC/normal mucosa samples and 10 adenomas. Generally, the highest protein expression was observed in the normal mucosa of the colon. The protein was observed in the cytoplasm and membrane of colonic epithelial layer and was clearly polarized. Strong immunoreactivity was observed in cells of the apical epithelium layer and those of the crypt lumen, while in epithelial cells located at the site of lamina muscularis the expression was visibly lower (Fig. 1B). PTPRH expression was lower in each CRC section as compared to corresponding normal mucosa (Fig. 1B). Immunoreactivity was to some extent deceased in adenomas compared to normal tissue, however, this was not so pronounced when CRCs were compared with normal samples (Fig. 1B). In 7 sections of normal mucosa, PTPRH expression was assessed as strong and in 7 sections as moderate, whereas it was scored as moderate, low and undetected in 5, 8 and 1 of CRC samples, respectively. In adenomas, PTPRH expression was high, moderate and low in 1, 7 and 4 samples, respectively.

DNA methylation was assessed in two PTPRH 5′ regions, including the upstream promoter and 5′ UTR/1 exon sequence with pyrosequencing in normal mucosa, AD and CRC samples. Cancer samples demonstrated significantly higher DNA methylation levels for both analyzed 5′ gene regions when compared to normal tissue samples (mean 35.21% and 31% vs. 15.62% and25.35%; P=0.0007 and P=0.00148, respectively). Adenomas and normal samples showed comparable DNA methylation levels (Fig. 2).

PTPRH DNA methylation as well as gene and protein expression were assessed in four colorectal cancer cell lines: SW480, HCT116, COLO205 and HT29. qRT-PCR analysis showed variable expression level of PTPRH in these cell lines, with the highest expression in HT29, moderate in COLO205, low in HCT116 and very low in SW480. The mRNA levels closely correspond to PTPRH protein levels in four cell lines as detected by western blot analysis (Fig. 3A).

Analyzing DNA methylation patterns with pyrosequencing showed that gene/protein expression levels were related to methylation levels. The highest and almost complete DNA methylation was observed in both cell lines with low PTPRH levels; SW480 and HCT116. In contrast, COLO205 and HT29 cells demonstrating higher gene expression presented low DNA methylation (Fig. 3B).

These four cell lines were cultured in the presence of DNA methyltrasferases' inhibitor 5-aza-dC. A stepwise gradient of four concentrations 0, 2.5, 5 or 10 µM of 5-aza-dC was used in culture. The increased PTPRH expression was observed in SW480 and HCT116, but not in the cell lines that lack DNA methylation at the PTPRH (Fig. 3C). The increased expression after 5-aza-dC treatment was also observed at protein level (Fig. 3D). However, the expression in treated SW480 was still low and required long exposure time when detected with western blotting.

Using chromatin immunoprecipitation, we measured the concentration of RNA II polymerase (Pol2) as well as trimethylation levels of both K4 of histone H3 (H3K4m3, the established marker of transcriptionally active chromatin) and K27 of histone H3 (H3K27m3, the marker of inactive chromatin) at the PTPRH promoter and 3′ region of this gene (exon 20) in 4 CRC cell lines. For comparison, the level of immunoprecipitation was also assessed for the ACTB promoter (transcriptionally active, β-actin gene) and HBB promoter (transcriptionally inactive, hemoglobin gene).

The pattern of Pol2 and selected histone posttranslational modifications at the gene promoter corresponded to expression and methylation status. A relatively high Pol2 concentration was observed in the PTPRH promoter region for HT29 and COLO205, i.e. in both cell lines lacking DNA methylation and characterized by high expression. A low Pol2 level was detected at the PTPRH promoter in HCT116 and SW480 that corresponds to the low gene expression (Fig. 4A).

The level of H3K4m3 at the PTPRH promoter, but not 3′ region, also clearly corresponded to expression and methylation status. High levels of this modification were observed in HT29 and COLO205, whereas low levels and no enrichment were seen in HCT116 and SW480 (both with the methylated gene promoter) (Fig. 4B). In contrast, the level of H3K27m3 at the PTPRH promoter showed no enrichment in HT29 and COLO205 and somewhat higher levels in HCT116 and SW480. H3K27m3 levels were high at 3′ gene regions in all the cell lines (Fig. 4C).