This article is mentioned in:

Introduction

Clear cell renal cell carcinoma (ccRCC) is the most common histological subtype of kidney cancer in adults (70–75% of cases) (1). Approximately 30% of early-stage patients develop metastases after surgery for localized disease (2). This carcinoma is histologically characterized by the presence of clear and abundant cytoplasm (2,3). The aggressive characteristics of renal clear cell carcinoma interfere in the efficiency of conventional treatment approaches, such as chemotherapy and radiotherapy (4).

Mutations or deletions associated with tumor suppressor genes may cause the loss or inactivation of these negative regulators. However, the loss of function may be caused by epigenetic changes (5), including acetylation, methylation, phosphorylation, ubiquitinylation, sumoylation, carbonylation and regulation of microRNAs (6). Epigenetic alterations occur at a high frequency, are reversible upon treatment with pharmacological agents and arise at defined regions within a gene (7,8). These features make epigenetics a very attractive field for cancer detection and several studies have shown multiple genes with different types of epigenetic alterations in tumor tissue (9–11).

Analyzing the gene expression and regulatory mechanisms in tumor samples has shown great efficiency in providing new information on signaling pathway activation of several processes, such as evasion of apoptosis, angiogenesis and metastatic potential.

The occludin (OCLN) gene, responsible for producing the protein occludin, is known to be present in all the tight junctions (TJs). TJ formation and expression has been described as a hallmark of epithelial cells and their subsequent loss from cancer cells tends to be considered as a direct effect of epithelial-mesenchymal transition (EMT) (12). Furthermore, occludin is structurally connected to the TGF-β receptor contributing to essential changes in EMT cell phenotype (13).

Growth arrest-specific 1 (GAS1) gene is associated with cell arrest and the protein is linked to the outer cell membrane by a glycosyl phosphatydylinositol. GAS1 protein binds to Hedgehog ligands for the potentiation of Hedgehog signaling (14).

Solute carrier family 2 member 1 (SLC2A1) gene encodes the solute carrier family 2 facilitated glucose transporter member 1 protein essential for cellular energy metabolism pathways (15–17). This protein is able to develop bidirectional flow of glucose according to the gradient of the substrate. The overexpression of the SLC2A1 protein has been found to be involved in increased uptake of glucose in human breast adenocarcinoma cell lines (18).

In the present study, we selected genes described in the literature related to cellular differentiation and proliferation. We analyzed their expression by quantitative PCR (qPCR) and immunohistochemistry, and examined the possible epigenetic mechanisms that regulate genes in ccRCC samples and cell lines. OCLN and GAS1 were underexpressed in ccRCC and miR-122 and miR-34a may regulate their expression in this type of cancer. Furthermore, we showed by qPCR and immunohistochemistry that SLC2A1 was significantly overexpressed in ccRCC samples. These genes may be involved in the molecular biology and development of the disease and analysis of them may contribute to insights for new prognoses, treatments and understanding this type of tumor.

Materials and methods

ccRCC samples

The samples were collected from 101 patients diagnosed with ccRCC, including 45 fresh samples of ccRCC, 56 paraffin-embedded samples of ccRCC, 2 fresh-frozen histologically healthy renal tissue and 24 paraffin-embedded samples of histologically healthy renal tissues, all of which were confirmed by pathologists (Tables I and II). The samples were obtained from the Hospital de Base at Faculdade de Medicina de São José do Rio Preto (Sao Paulo, Brazil). The use of patient-derived material was approved by the Research Ethics Committee of the IBILCE-UNESP and FAMERP (Sao Paulo, Brazil) and written consent was obtained from all the patients. Tissues were obtained from patients undergoing tumor resection surgery, and a diagnosis of ccRCC was determined post-operatively using histopathology. The samples were classified according to the criteria provided by the International Union against Cancer (19).

Table I. Epidemiological and clinicopathological characteristics of the patients diagnosed with ccRCC (fresh samples).

Table I.

Epidemiological and clinicopathological characteristics of the patients diagnosed with ccRCC (fresh samples).

Sample	Gender	Age (years)	Alcohol	Smoking	TNM
F1	M	58	No	No	T1N1M1
F2	F	39	No	No	T2N0M0
F3	F	71	No	No	T4N0M1
F4	F	64	No	No	T2N0M0
F5	M	46	No	No	T1bN0M0
F6	M	70	No	No	T2N0M0
F7	M	70	Yes	Yes	T3aN0M0
F8	F	58	Yes	Yes	T3N3M1
F9	M	36	No	No	T1aN0M0
F10	M	48	Yes	No	T2N0M0
F11	M	52	No	No	T3aN0M0
F12	F	62	No	No	T1aN0M0
F13	M	75	No	Yes	T1N0M0
F14	F	49	No	No	T1bN0M0
F15	F	71	No	No	T1N0M0
F16	F	79	No	No	T1aN0M0
F17	F	29	No	No	T2N0M0
F18	M	70	No	No	T2N0M0
F19	M	39	No	Yes	T2N0M0
F20	M	66	Yes	No	T2N0M0
F21	M	77	No	No	T2N0M1
F22	F	23	No	No	T2N0M0
F23	F	53	No	No	T2N0M0
F24	M	69	No	No	T3N0M0
F25	F	60	No	No	T3bN0M1
F26	F	79	No	No	T3aN0M0
F27	F	59	No	No	T2N0M0
F28	F	64	No	No	T1N0M0
F29	M	53	Yes	No	T2N0M0
F30	M	56	Yes	No	T1N1M1
F31	M	54	No	No	T2N0M0
F32	M	52	Yes	No	T1aN0M0
F33	M	67	Yes	No	T1aN0M0
F34	M	41	No	Yes	T1N0M0
F35	F	55	No	No	T3aN2M1
F36	F	52	Yes	Yes	T3aNxMx
F37	F	25	No	Yes	NI
F38	M	56	No	No	T1bN0Mx
F39	F	83	No	No	T1N0M0
F40	M	68	Yes	Yes	NI
F41	F	71	No	Yes	T1bNxMx
F42	M	73	No	Yes	T3NxM1
F43	M	63	Yes	No	T3aN0M0
F44	M	62	Yes	Yes	NI
F45	M	39	Yes	No	T2NxM0

[i] M, male; F, female; NI, no information was obtained. ccRCC, clear cell renal cell carcinoma; TNM, tumor-node-metastasis.

Table II. Epidemiological and clinicopathological characteristics of the patients diagnosed with ccRCC (paraffin-embedded samples).

Table II.

Epidemiological and clinicopathological characteristics of the patients diagnosed with ccRCC (paraffin-embedded samples).

Sample	Gender	Age (years)	Alcohol	Smoking	TNM
P1	M	80	No	No	T4N1M1
P2	M	50	No	No	T2N0M0
P3	M	61	No	No	T3bN0M0
P4	F	69	No	No	T1bN0M0
P5	M	52	No	Yes	T2N0M0
P6	M	88	No	Yes	T2N0M0
P7	M	60	No	No	T1bN0M0
P8	M	58	No	No	T1bN1M1
P9	M	71	No	Yes	T1N0M0
P10	M	56	No	No	T3bN0M1
P11	M	77	No	No	T2N1M1
P12	F	32	Yes	No	T1aN0M0
P13	M	68	No	No	T3bN0M0
P14	M	83	Yes	Yes	T2N0M1
P15	F	65	No	No	T2N0M0
P16	F	54	No	No	T1bN0M0
P17	M	44	NI	NI	NI
P18	M	65	No	No	T3aN0M1
P19	F	72	No	Yes	T2N0M1
P20	F	60	No	No	T2N0M0
P21	M	77	No	No	T1bN0M0
P22	M	52	No	No	T1bN0M0
P23	M	74	No	No	T3aN0M1
P24	M	75	No	No	T2N0M0
P25	M	78	No	No	T2N0M0
P26	F	66	No	No	T2N0M0
P27	M	64	No	No	T1bN0M0
P28	M	67	Yes	Yes	T2N0M0
P29	M	49	No	No	T2N0M0
P30	F	58	No	No	T1bN0M0
P31	M	84	No	Yes	T1bN0M0
P32	F	48	No	Yes	T2N0M0
P33	F	55	No	No	T1bN0M0
P34	F	50	No	No	T2N0M0
P35	F	46	No	No	T1aN0M0
P36	M	43	No	No	T2N0M0
P37	F	54	No	Yes	T3aN0M0
P38	F	86	No	Yes	T3N0M0
P39	M	51	No	No	T2N0M0
P40	M	74	No	No	T2N0M0
P41	F	51	No	Yes	T1aN0M0
P42	M	63	Yes	Yes	T3aN0M0
P43	M	56	No	No	T2N0M0
P44	F	76	No	Yes	T1bNxMx
P45	M	70	No	Yes	T1aNxMx
P46	M	58	No	No	T1bN0M0
P47	M	56	No	No	T1bN0M0
P48	M	76	No	No	T3N0M0
P49	F	61	No	No	NI
P50	M	81	Yes	No	NI
P51	M	76	Yes	Yes	T2N0M0
P52	F	72	No	No	T1N0M0
P53	F	79	No	No	T2N0M0
P54	M	59	No	Yes	T2N0M0
P55	M	46	No	No	T2N0M1
P56	F	89	No	Yes	T1N0M0

[i] M, male; F, female; NI, no information was obtained. ccRCC, clear cell renal cell carcinoma; TNM, tumor-node-metastasis.

Cell lines and treatments

The 786-O (primary clear cell adenocarcinoma) cell line was obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) and HaCaT (an immortal human keratinocyte cell line) was kindly provided by Luisa Lina Villa (Department of Radiology, Center of Translational Oncology Investigation, São Paulo State Cancer Institute, Universidade de São Paulo, Brazil). 786-O cells were cultured in RPMI-1640 medium, and HaCaT cells were cultured in Dulbeccos modified Eagles medium (DMEM) (both from Gibco by Life Technologies, Grand Island, NY, USA). The media were supplemented with 10% fetal bovine serum (Cultilab, Sao Paulo, Brazil), 100 U/ml of penicillin and 100 µg/ml of streptomycin (both from Invitrogen, Grand Island, NY, USA). Cell cultures were grown at 37°C in a 5% CO2 atmosphere. 786-O cells were plated into 6-well plates, cultured for 24 h, and treated for 72 h with fresh 5, 2 or 1 µM of 5-Aza-dC (Sigma-Aldrich, St. Louis, MO, USA) dissolved in phosphate-buffered saline. Due to its chemical instability, fresh medium with 5-Aza-dC was added every 24 h. For the trichostatin A (TSA) experiments, the cells were seeded in 6-well plates, cultured for 24 h, and treated for 24 h with fresh 1,000, 500, 300, 200 or 100 nM of TSA (Sigma-Aldrich) dissolved in dimethyl sulfoxide (DMSO). To assess the combined effect of both drugs, we performed the co-treatment of cells with 5 µM of 5-Aza-dC and 1,000 nM of TSA using the following protocol: 5-Aza-dC was added for 72 h, after which it was removed and TSA was added for an additional 24 h. All experiments were performed in triplicate.

Cytotoxicity assay

786-O cells were seeded into 96-well plates at a density of 103 cells for 5-Aza-dC treatment, 104 cells for TSA and 5×103 cells for co-treatment. After the 5-Aza-dC and TSA treatments, 1 mg/ml of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma-Aldrich) was added to each well and incubated for 30 min at 37°C. Then, the MTT was removed and 100 µl of 100% DMSO (Sigma-Aldrich) was added to each well and the absorbance was measured at 562 nm. Each experiment was performed in triplicate and in 2 independent assays.

RNA and DNA extraction

Genomic DNA and total RNA extraction were performed using TRIzol® reagent (Life Technologies) using both cell lines, the 45 fresh samples of ccRCC and 2 fresh-frozen histologically healthy renal tissue samples, following the manufacturer's instructions. The DNA and RNA quality was verified by electrophoresis through an agarose gel and stained with ethidium bromide. For quantitative real-time PCR, 5 µg of total RNA from each sample was used to synthesize cDNA with a High-Capacity cDNA Archive kit (Applied Biosystems, Foster City, CA, USA), according to the manufacturer's instructions. The cDNA quality was ascertained by PCR using the housekeeping gene β-actin. The PCR products were analyzed by electrophoresis on a 1% agarose gel and stained with ethidium bromide.

Selection of genes

We analyzed the differential expression of the following genes by qPCR: E-cadherin (CDH1), N-cadherin (CDH2), claudin 1 (CLDN1), cathepsin B (CTSB), platelet-activating factor acetylhydrolase 2 (PAFAH2), growth arrest-specific 1 (GAS1), laminin β1 (LAMβ1), occludin (OCLN), phosphoglycerate dehydrogenase (PHGDH), solute carrier family 2 member 1 (SLC2A1), transforming growth factor β1 (TGFβ1), tight junction protein 1 (TJP1) and vimentin (VIM). The genes were selected according to their role in cellular differentiation and proliferation.

Quantitative real-time polymerase chain reaction

The qPCR amplification was performed with Power SYBR-Green and an ABI 7300 Real-Time PCR System (Applied Biosystems, Warrington, UK). In brief, the reaction mixture (20 µl of total volume) contained 25 ng of cDNA, gene-specific forward and reverse primers and 10 µl of 2X Quantitative SYBR-Green PCR Master Mix. The samples were tested in triplicate. The relative expression of each specific gene was calculated using the following previously published (20) formula: R = (E target)ΔCt target (control - sample)/(E endogenous ΔCt endogenous (control - sample). A 2-fold-change was set as the cut-off. Gene expression was analyzed in 45 ccRCC samples, in the 786-O cells after treatment and the HaCaT cell line. Two healthy renal fresh-frozen tissue samples were used as the healthy reference (control group). A non-treated 786-O cell line was used as a reference sample and GAPDH was selected as the endogenous control gene.

Bisulfite sequencing

Genomic DNA was subjected to a sodium bisulfite treatment to modify unmethylated cytosine to uracil, as described by Calmon et al (21). Bisulfite-treated DNA was amplified by PCR, using primers designed to amplify CpG-rich regions located at the 5′ region of OCLN (−438 to +1232, relative to the transcription start site (TSS) and encompassing 164 CpG dinucleotides] and GAS1 [-485 to +1518, relative to the TSS and encompassing 269 CpG dinucleotides). The amplified products were cloned using a CloneJET PCR Cloning kit (Thermo Scientific, Carlsbad, CA, USA). Five positive clones from the 786-O cell line before and after treatment with 5-Aza-dC, one ccRCC sample and one healthy renal sample were sequenced using a BigDye Terminator v3.1 Cycle Sequencing kit and an ABI3130XL sequencer, according to the manufacturer's instructions (Applied Biosystems). The methylation percentage for each sample was calculated as the proportion of unconverted CpG dinucleotides among all of the CpG dinucleotides analyzed from the 5 positive clones.

Chromatin immunoprecipitation (ChIP)

ChIP was performed using a ChIP-IT® High Sensitivity kit according to the manufacturers instructions (Active Motif, Carlsbad, CA, USA). The 786-O and HaCaT cells were cross-linked with 1% formaldehyde for 10 min at room temperature and nuclei were isolated. Isolated nuclei were sonicated on ice to break the chromatin into 8,200–1,200 bp fragments. Soluble chromatin was used in immunoprecipitation with H3K4me3, H3K27me3, H3K4me2 and H3K9ac antibodies (Thermo Scientific, Rockford, IL, USA). IgG was used as a negative control (Abcam, Cambridge, MA, USA) and immune complexes were absorbed with protein G magnetic beads. After reversing the cross-linking and treating with proteinase K, the immunoprecipitated DNA was analyzed on a gene-specific basis using qPCR. The quality control ChIP was performed using a ChIP-IT® qPCR Analysis kit (Active Motif) with Human Negative Control Primer Set 1 and Human Positive Control Primer Set GAPDH-2, according to the manufacturers instructions.

MicroRNA analysis

Two microRNA databases, miRBase (www.mirbase.org) and TargetScanHuman (www.targetscan.org), were used to select microRNAs that target the OCLN and GAS1 mRNA. For microRNA expression analysis, total RNA was reverse transcribed into cDNA using miR-34a, miR-122 and U48-specific steam-loop primers. Expression of mature miR-34a and miR-122 was quantified as described in the qPCR methods. The microRNA expression was analyzed in 30 ccRCC samples and in the cell lines 786-O and HaCaT. Two healthy renal fresh-frozen tissue samples were used as a healthy reference. Each sample was run in triplicate, and the miRNA expression levels were normalized to U48.

Tissue array construction and immunohistochemistry

Tissue microarrays were constructed using a tissue microarrayer (Beecher Instruments, Silver Spring, MD, USA). Briefly, the slides were reviewed by a pathologist, and areas containing each category were annotated on the hematoxylin and eosin (H&E) slides. Fifty-six 1.0 mm in diameter cylindrical tissue cores from the ccRCC tumor samples and 24 from the tumor margins were taken from the corresponding regions of the paraffin blocks and transplanted into a recipient paraffin block.

For immunohistochemical staining, all of the paraffin-embedded sections were cut into 4-µm in thickness and the endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 30 min. Antigen retrieval was performed using sodium citrate buffer (pH 6,0) for 20 min at 96°C in a steam pan. The sections were then incubated with SLC2A1 polyclonal primary antibody (1:200) (Abcam) diluted in 1% BSA at 4°C overnight. After incubation with a primary antibody, the sections were incubated with a biotinylated secondary antibody (Santa Cruz Biotechnology, Paso Robles, CA, USA) and then exposed to a streptavidin complex (HRP ready-to-use; DakoCytomation, Carpinteria, CA, USA). Positive staining was identified using DAB substrate color detection (Dako, Cambridge, UK). The sections were counterstained with hematoxylin.

The SLC2A1 densitometric analyses were conducted with an Axioskop II microscope using the Axiovision™ software (both from Zeiss, Oberkochen, Germany). For these analyses, 20 different points were analyzed from 3 different fields of each tumor fragment and the average immunoreactivity intensity [arbitrary units (AU)] was recorded.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 5 Software (GraphPad Software Inc., San Diego, CA, USA). The relative expression levels detected by qPCR in ccRCC were transformed into natural logarithms. The Wilcoxon signed-rank test was applied to compare the gene expression levels in tumor and healthy tissues. The Mann-Whitney U test and Wilcoxon signed-ranks tests were used to compare the protein expression levels detected by immunohistochemistry.

Results

Evaluation of EMT gene expression in ccRCC

To evaluate gene expression in 45 ccRCC fresh samples and the 786-O cell line, we performed qPCR for the following target genes: CDH1, CDH2, CLDN1, CTSB, PAFAH2, GAS1, LAMβ1, OCLN, PHGDH, SLC2A1, TGFβ1, TJP1 and VIM.

CDH2, CTSB, GAS1, LAMβ1, OCLN, PHGDH, PAFAH2 and TJP1 were significantly downregulated in the ccRCC samples when compared to the reference samples. SLC2A1 was significantly overexpressed in tumor tissues when compared to healthy renal tissues. CDH1, TGFβ1, VIM and CLDN1 were not differentially expressed (Fig. 1).

Figure 1. Gene expression profile in ccRCC tissues. mRNA expression of the selected genes using qPCR in 45 ccRCC fresh samples and in the 786-O cell line. The results are shown as the fold-change in the expression relative to the healthy sample (*p<0.001; Wilcoxon signed-rank test).

CDH1, CTSB, GAS1 and OCLN were significantly downregulated in 786-O cells. SLC2A1 and VIM were significantly overexpressed in 786-O cells. CDH2, CLDN1, LAMβ1, PAFAH2, PHGDH, TGFβ1 and TJP1 were not differentially expressed (Fig. 1).

OCLN and GAS1 were chosen for epigenetic analysis due to the presence of CpG islands in their promoter regions and the observed downregulation in both the ccRCC samples and 786-O cells. SLC2A1 was selected for immunohistochemical analysis since it was overexpressed in both the ccRCC samples and 786-O cells.

DNA methylation and histone acetylation analysis

The methylation status of CpG islands at the 5′ region of the OCLN and GAS1 genes was investigated by bisulfite sequencing in ccRCC samples, healthy renal tissue and 786-O cells before and after 5-Aza-dC treatment. The bisulfite sequencing was performed in the −438 to +1232 region of OCLN and the −485 to +1518 region of GAS1. The methylation levels of OCLN and GAS1 were not significantly reduced in 786-O cells before and after 5 µM of 5-Aza-dC treatment (24.39–22.19% for OCLN and 22.5–22.23% for GAS1). The methylation level of the genes compared between the ccRCC samples and the healthy renal samples did not significantly differ, 4–1% for OCLN and 1.8–0% for GAS1 (Fig. 2).

Figure 2. Methylation levels in OCLN and GAS1 gene promoters. The transcription initiation site is represented by +1. An expanded view shows the position of CpG islands and the region analyzed by bisulfite sequencing. Each row represents one sequenced clone, and open circles and filled circles represent unmethylated and methylated CpG dinucleotides, respectively. (A) Bisulfite sequencing of OCLN and GAS1 before and after 5-Aza-dC treatment. (B) Bisulfite sequencing of OCLN and GAS1 in healthy and tumor samples.

An MTT assay was performed to detect the possible effects of 5-Aza-dC and TSA treatment on 786-O cell viability. All of the drug concentrations tested showed viability >75%, including the co-treatment (Fig. 3).

Figure 3. 786-O cell viability after 5-Aza-dC and TSA treatment. Cell viability of 786-O cells after 5-Aza-dC, TSA and co-treatment with 5-Aza-dC and TSA.

The cell growth was inhibited in a dose-dependent manner by 5-Aza-dC treatment in 786-O cells. The expression of OCLN mRNA was higher in the cells treated with 5 and 2 µM of 5-Aza-dC than in the untreated cells (Fig. 4A). There was no difference in the GAS1 mRNA expression after treatment with any 5-Aza-dC concentration (Fig. 4A).

Figure 4. OCLN and GAS1 expression levels after 5-Aza-dC and TSA treatment. (A) mRNA expression of OCLN and GAS1 genes in the 786-O cell line after 5-Aza-dC treatment. (B) mRNA expression of OCLN and GAS1 genes in the 786-O cell line after TSA treatment, (C) mRNA expression of OCLN and GAS1 in the 786-O cell line after co-treatment with 5-Aza-dC and TSA (*p<0.001; Wilcoxon signed-rank test).

Similar to the 5-Aza-dC treatment, TSA treatment showed a better response at the highest concentration (1,000 nM). The OCLN gene was upregulated in the 786-O cells after 1,000 nM TSA treatment when compared to the untreated cells (Fig. 4B). Other TSA concentrations were not able to significantly upregulate OCLN expression. The GAS1 expression levels were unchanged after TSA treatment in 786-O cells when compared to the untreated cells (Fig. 4B).

The 5-Aza-dC (5 µM) and TSA (1,000 nM) co-treatment efficiently upregulated OCLN expression in 786-O cells. However, GAS1 gene expression was not altered after co-treatment (Fig. 4C).

ChIP

Since there was alteration in the OCLN gene expression after 5-Aza-dC treatment without any difference in the DNA methylation level in its promoter region and it is known that 5-Aza-dC treatment may alter both DNA and histone methylation (22), we investigated whether the loss of OCLN expression in 786-O cells could be due to histone modifications. Regarding the GAS1 gene, we could not observe any change in its expression after the 5-Aza-dC and TSA treatments, and therefore, we also investigated if histone modifications could contribute to GAS1 supression in the absence of DNA hypermethylation or histone deacetylation.

In both cell lines, ChIP followed by qPCR analysis showed that all of the histone modifications were reduced or absent in the promoter regions of OCLN and GAS1. It was not possible to differentiate the relative occupancy of the marks studied between the healthy and tumor cell lines. The Human Negative Control Primer Set 1 and Human Positive Control Primer Set GAPDH-2 served as negative and positive controls, respectively (Fig. 5).

Figure 5. Histone modifications in OCLN and GAS1 gene promoters. ChIP analyses of the levels of H3K4me3, H3K27me3, H3K4me2 and H3K9ac in the OCLN and GAS1 promoter regions. The values are represented as fold-change over the input.

microRNA analysis

We performed stem-loop reverse transcription PCR (RT-qPCR) to quantify miR-122 and miR-34a expression in the 786-O and HaCaT cell lines. The relative expression of miR-122 and miR-34a in the 786-O cells was higher than in the HaCaT cells (Fig. 6). Both microRNAs were significantly overexpressed in tumor tissues when compared to healthy renal tissues. These results suggest that upregulation of miR-122 and miR-34a may be, at least in part, involved in the downregulation of OCLN and GAS1 in ccRCC.

Figure 6. MicroRNA-122 and microRNA-34a expression are upregulated in ccRCC tissues. miR-122 and miR-34a expression levels in the 786-O cell line compared with the HaCaT cell line and the ccRCC samples compared with healthy renal tissue samples (*p<0.001; Wilcoxon signed-rank test).

Immunohistochemistry

The SLC2A1 protein expression was investigated in ccRCC tissue array by immunohistochemistry.The ccRCC tissues showed a higher expression of the SLC2A1 protein (p<0.0001) relative to the healthy tissues. Fig. 7 shows the SLC2A1 immunoreactivity and the densitometric data.

Figure 7. SLC2A1 is highly expressed in ccRCC tissues compared to healthy samples. Protein detection of (A) SLC2A1 in the ccRCC sample. (B) SLC2A1 in a healthy sample. Scale bars, 20 µm. (C) Histogram of the densitometry of the immunostaining of SLC2A1 in the samples analyzed (*p<0.0001; Mann-Whitney U test).

Discussion

Several prognostic models are used to predict the survival of ccRCC patients, including clinicopathological parameters and protein, genetic and epigenetic biomarkers (23–25). In the present study, we investigated the epigenetic regulatory mechanisms of OCLN and GAS1, genes that may contribute to epithelial-mesenchymal transition (EMT), cellular differentiation and proliferation in renal cell carcinoma.

The OCLN gene is responsible for producing the occludin protein, which is present in all tight junctions. Occludin has been implicated in tight junction permeability and, in particular, to the regulation of size-selective diffusion (26–28). The transmembrane proteins occludin and claudin are present in tight junctions, which are connected to the cytoskeleton via a network of proteins, such as zonula occludens-1 (29). This epithelial junctional complex, along with adherens junctions and desmosomes, is located at the most apical part of the complex (30). Tight junctions determine epithelial cell polarity and undergo dissolution during cancer progression that is accompanied by cytoskeletal changes (31).

Downregulation of OCLN has been observed in liver tumors, breast cancer, endometrial and lung carcinomas (32–35). A study of human cutaneous squamous cell carcinoma observed the loss of OCLN and related this loss to cell-cell adhesion, apoptosis and proliferation (36). OCLN was found to be partially methylated in the promoter and endogenous region of a breast cancer cell line, and treatment with 5-Aza-dC and TSA could induce re-expression of this gene (37). In the present study, we observed hypermethylation in ~20% of the endogenous region, although, treatment with 5-Aza-dC did not change the global methylation profile of the OCLN gene in the ccRCC cell line. However, we demonstrated that the effect of 5-Aza-dC treatment and the synergistic effect of 5-Aza-dC and TSA treatment in 786-O cells were effective in reactivating OCLN gene expression. Thus, 786-O cells showed a methylation profile in the OCLN promoter, despite the suggestion of alternate modes of reactivation based upon OCLN expression. To further investigate whether increased OCLN expression in 786-O cells could be due to chromatin modifications, we performed ChIP for H3K27me3, H3K4me3, H3K4me2 and H3K9ac marks in the OCLN promoter in 786-O cells. ChIP analysis did not show any differences between H3K27me3 silencing marks or changes in H3K4me3, H3K4me2 and H3K9ac activation marks in the OCLN gene promoter region. H3K4me1, H3K4me2 and H3K4me3 marks inform us on the potential prognosis since their increase was found to be inversely correlated with lymph node involvement and distant metastasis in patients with renal cell carcinoma (38).

Another potential regulatory mechanism that may influence OCLN expression is microRNAs. miR-122 has been reported to regulate OCLN expression in intestinal tissue, which contributes to increased intestinal permeability (39). In the present study, miR-122 expression levels were higher in the ccRCC cell line and tumor samples, suggesting that the increase in microRNA expression may contribute to the downregulation of the OCLN gene. miR-122 expression was upregulated in 15 cases of clear cell papillary renal cell carcinoma and in 80 samples of ccRCC (40–42). In breast cancer, high miR-122 levels have been associated with metastasis, likely by increasing nutrient availability in the pre-metastatic niche (43).

GAS1 is a membrane-associated protein; it is tethered to the membrane through a glycosyl-phosphatidylinositol anchor (44,45). GAS1 contributes to inhibitory signal transduction and interferes with cell proliferation due to the inhibition of DNA synthesis during the G0 to S phase (44,46). The Hedgehog pathway is an essential regulator of carcinogenesis, and the GAS1 protein participates in the Hedgehog signaling pathway through the binding to Hedgehog ligands and enhancing the signaling, thus, indirectly inducing EMT (14,47,48).

Reduced GAS1 expression was found to be positively associated with the survival time in gastric cancer cells (49). The downregulation of the GAS1 gene has been described in colorectal cancer, melanoma and prostate cancer metastasis (50–52).

In the present study, the global methylation profile of GAS1 in ccRCC cells was unchanged before and after 5-Aza-dC treatment. Similarly, there were no gene expression modifications after treatment with TSA or co-treatment with TSA and 5-Aza-dC. Sacilotto et al (53) suggested that GAS1 gene transcription is modulated by chromatin remodeling and histone modifying complex recruitment to the GAS1 promoter in hepatic cells. However, our chromatin immunoprecipitation assays did not show a difference in the H3K27me3 silencing mark, nor were there differences in the H3K4me3, H3K4me2 and H3K9ac activation marks in the GAS1 gene promoter region in ccRCC cells. It is possible that other regulatory mechanisms are altering the expression of GAS1 in ccRCC, such as microRNAs.

GAS1 gene regulation in papillary thyroid carcinoma occurs through miR-34a, leading to cell proliferation and apoptotic suppression (54). miR-34a regulates mesangial proliferation and glomerular hypertrophy by directly inhibiting GAS1 in early diabetic nephropathies (55). In the present study, we measured miR-34a expression levels, which target the GAS1 gene, in ccRCC cell lines and tumor samples and suggest that the increase in this microRNA may contribute to the downregulation of GAS1. miR-34a was upregulated in 8 ccRCC samples and renal carcinoma cell lines, establishing an association between this miR and multiple targets (56,57). miR-34a overexpression supports cell proliferation in the majority of cancers, suggesting an unexpected link between cancer and neuronal and endocrine cell metabolism (58).

The SLC2A1 gene is overexpressed in different types of cancers, including lung, liver and breast (59–61). SLC2A1 facilitates glucose transport and is present in the epithelium and endothelium, contributing to proliferation (59,62,63). In the present study, SLC2A1 was upregulated in ccRCC samples and 786-O cells as determined by both qPCR and immunohistochemistry. Cifuentes et al (64) demonstrated that SLC2A1 and glucose transport are regulated by insulin and therefore may be an important target for conventional pharmaceutical therapy and gene therapy in the treatment of osteosarcoma. The SLC2A1 protein belongs to a class of markers that can detect breast cancer by molecular imaging with more common immunoreactivity in ductal carcinoma in situ than in invasive breast cancer (65,66).

In summary, the silencing of OCLN and GAS1 may contribute to tumor progression due to a possible relationship between these genes and cell-cell adhesion, apoptosis and proliferation. We suggest that the increase in miR-122 and miR-34a expression may contribute to the downregulation of the OCLN and GAS1 genes, respectively. Overexpression of the SLC2A1 gene was higher in ccRCC than that in the non-neoplastic samples. The expression of these genes may be involved in ccRCC progression in addition to other cellular processes that contribute to tumor development. Once these genes have been characterized in ccRCC, molecular genetic tools may be used to explore the biological process involved, thus improving diagnosis, treatment and patient outcomes. More studies should be conducted to assess the overall importance of epigenetic events in the regulation of OCLN and GAS1 expression in ccRCC and in others tumors, as well as its relevancce in the analysis of possible transcription factors that may also contribute to the regulation of the expression of these genes.

Acknowledgements

The present study was supported by the São Paulo State Research Foundation, FAPESP (2012/08853-0), and the National Council for Research, CNPq.

References