Derepression of CLDN3 and CLDN4 during ovarian tumorigenesis is associated with loss of repressive histone modifications

Abstract

Unlike epigenetic silencing of tumor suppressor genes, the role of epigenetic derepression of cancer-promoting genes or oncogenes in carcinogenesis remains less well understood. The tight junction proteins claudin-3 and claudin-4 are frequently overexpressed in ovarian cancer and their overexpression was previously reported to promote the migration and invasion of ovarian epithelial cells. Here, we show that the expression of claudin-3 and claudin-4 is repressed in ovarian epithelial cells in association with promoter 'bivalent' histone modifications, containing both the activating trimethylated histone H3 lysine 4 (H3K4me3) mark and the repressive mark of trimethylated histone H3 lysine 27 (H3K27me3). During ovarian tumorigenesis, derepression of CLDN3 and CLDN4 expression correlates with loss of H3K27me3 in addition to trimethylated histone H4 lysine 20 (H4K20me3), another repressive histone modification. Although CLDN4 repression was accompanied by both DNA hypermethylation and repressive histone modifications, DNA methylation was not required for CLDN3 repression in immortalized ovarian epithelial cells. Moreover, activation of both CLDN3 and CLDN4 in ovarian cancer cells was associated with simultaneous changes in multiple histone modifications, whereas H3K27me3 loss alone was insufficient for their derepression. CLDN4 repression was robustly reversed by combined treatment targeting both DNA demethylation and histone acetylation. Our study strongly suggests that in addition to the well-known chromatin-associated silencing of tumor suppressor genes, epigenetic derepression by the conversely related loss of repressive chromatin modifications also contributes to ovarian tumorigenesis via activation of cancer-promoting genes or candidate oncogenes.

Fig. 1.

DNA methylation in the CLDN3 and CLDN4 promoters in ovarian cell lines and NOSE cells. (A) CLDN4 and (B) CLDN3 DNA methylation status by MSP analysis in ovarian cell lines and NOSE cells. The location of CpG islands and regions for MSP and bisulfite sequencing PCR (BSP) are indicated in the genomic sequence of CLDN3 and CLDN4 (top). The transcript and protein levels for CLDN3 and CLDN4 were measured by quantitative real-time reverse transcription–PCR (qRT–PCR) and western blot, respectively (middle). The qRT–PCR values are represented as mean ± SD. DNA methylation status was analyzed by MSP and quantitative MSP (qMSP) (bottom). Unmethylated (UM) and methylated (M) bands are shown and the quantitative methylation levels by qMSP are represented as percentage of methylated reference (PMR, %). Human leukocytes (h.L), methylated in vitro by SssI treatment (h.L.M), were used as a positive control. (C) CLDN4 and (D) CLDN3 DNA methylation status by bisulfite sequencing in ovarian cell lines. CpG sites analyzed by BSP in the CLDN4 (21 CpG sites) and CLDN3 promoter region (47 CpG sites) are indicated. At least 10 clones were sequenced and each row represents the methylation pattern of an individual cloned PCR product (filled square, methylated CpG sites; open square, unmethylated CpG sites).

Fig. 2.

DNA methylation status in the CLDN3 and CLDN4 promoters in ovarian FFPE tissues. (A) CLDN3 DNA methylation status in ovarian FFPE tissues. Claudin-3 expression in ovarian tissues was examined by immunohistochemistry and expression levels were scored on a scale of 0–3 (upper panel). Immunohistochemistry (IHC) = 0 indicates negative expression and a value >0 was regarded as positive expression. The methylation level (y-axis) was calculated as the mean percentage of methylated reference (PMR) value of two to six repeated quantitative MSP (qMSP) assays (lower panel). A PMR >0 was regarded as methylated. A horizontal line within the box indicates the median value. Outliers and extremes are indicated as open circles in the box plots. Differences in methylation levels between two groups of ovarian FFPE tissues were analyzed by the non-parametric Mann–Whitney test. A P-value < 0.05 (two-sided) was regarded as statistically significant. Statistical analyses were carried out using SPSS Version 12. (B) Relationship of DNA methylation with claudin-3 expression in ovarian carcinoma tissues. The methylation level was also compared between claudin-3-expressing (IHC = 1, 2, 3) and non-claudin-3-expressing tissues (IHC = 0) in ovarian carcinoma FFPE tissues. (C) CLDN4 DNA methylation status in ovarian FFPE tissues. Claudin-4 expression was determined by immunohistochemistry (upper panel) and DNA methylation level by qMSP (lower panel).

Fig. 3.

Histone modifications in the CLDN3 and CLDN4 promoters in ovarian cancer cell lines. (A) Histone modifications in the CLDN3 and (B) CLDN4 promoters. Location of the regions analyzed by the ChIP assay is indicated (top). Three regions of CLDN3 (ChIP I to ChIP III) and two regions of CLDN4 (ChIP I to ChIP II) were analyzed for histone modifications by ChIP assays (left). Quantitative enrichment of histone modifications in the CLDN3 and CLDN4 promoters was also assessed by quantitative ChIP (qChIP) assays using Taqman probe (right). y-axis represents the fold enrichment relative to IgG control. Gene expression levels of CLDN3 and CLDN4 in each cell line are indicated under the name of cell line.

Fig. 4.

Effect of the loss of H3K27me3 on CLDN3 and CLDN4 expression in ovarian cancer cells. (A) Effect of EZH2 knockdown by EZH2 siRNA on CLDN3 and CLDN4 expression. Knockdown of EZH2 transcripts (EZH2 var1 and EZH2 var2) and the decrease in protein levels of H3K27me3 were assessed in TOV-112D cells 72 h after siRNA treatment (100 nM). The effect of EZH2 knockdown on H3K27me3 enrichment in the CLDN3 and CLDN4 promoters was also evaluated by quantitative ChIP in TOV-112D cells. *P < 0.05, Mann–Whitney test. CLDN3 and CLDN4 messenger RNA (mRNA) levels after EZH2 siRNA treatment were compared with the levels after control siRNA treatment. (B) Effect of LY294002 on CLDN3 and CLDN4 expression. CLDN3 and CLDN4 expression levels were determined by quantitative real-time reverse transcription–PCR (qRT–PCR) in TOV-112D cells after a 24 h treatment with 20 μM LY294002. (C) CLDN3 and CLDN4 expression in TOV-K27R cells. The level of H3K27me3 in TOV-112D, TOV-K27R and TOV-WtH3 cells was assessed by western blotting and the mRNA levels of CLDN3 and CLDN4 were assessed by qRT–PCR.

Fig. 5.

Effects of DZNep, 5-aza-dC and TSA treatment on CLDN3 and CLDN4 expression and epigenetic modifications in ovarian cell lines. (A) Changes in CLDN3 expression in ovarian cells treated with 5 μM DZNep (D), 200 nM TSA (T) or combination of DZNep/TSA. CLDN3 induction by quantitative real-time reverse transcription–PCR is represented as the fold change (y-axis), calculated relative to no treatment (NT). (B) Effect of pharmacological treatments on Polycomb repressive complex-2 (EZH2, EED and SUZ12) protein levels. (C) ChIP analysis of the CLDN3 promoter in TOV-112D and OV-90 cells untreated or treated with DZNep/TSA. Fold enrichment was measured by quantitative ChIP. *P < 0.05. (D) Effect of DZNep, 5-aza-dC and/or TSA on the expression, methylation status and histone modifications of CLDN4 in TOV-112D cells. Changes in CLDN4 expression (top), DNA methylation (middle) and histone modifications (bottom) after DZNep, TSA, 5-aza-dC (A, 5μM) or their combined treatments are represented, respectively. *P < 0.05.