27-Hydroxycholesterol, The Estrogen Receptor Modulator, Alters DNA Methylation in Breast Cancer

Abstract

27-hydroxycholesterol (27-HC) is the first known endogenous selective estrogen receptor modulator (SERM), and its elevation from normal levels is closely associated with breast cancer. A plethora of evidence suggests that aberrant epigenetic signatures in breast cancer cells can result in differential responses to various chemotherapeutics and often leads to the development of resistant cancer cells. Such aberrant epigenetic changes are mostly dictated by the microenvironment. The local concentration of oxygen and metabolites in the microenvironment of breast cancer are known to influence the development of breast cancer. Hence, we hypothesized that 27-HC, an oxysterol, which has been shown to induce breast cancer progression via estrogen receptor alpha (ERα) and liver X receptor (LXR) and by modulating immune cells, may also induce epigenetic changes. For deciphering the same, we treated the estrogen receptor-positive cells with 27-HC and identified DNA hypermethylation on a subset of genes by performing DNA bisulfite sequencing. The genes that showed significant DNA hypermethylation were phosphatidylserine synthase 2 (PTDSS2), MIR613, indoleamine 2,3-dioxygenase 1 (IDO1), thyroid hormone receptor alpha (THRA), dystrotelin (DTYN), and mesoderm induction early response 1, family member 3 (MIER). Furthermore, we found that 27-HC weakens the DNMT3B association with the ERα in MCF-7 cells. This study reports that 27-HC induces aberrant DNA methylation changes on the promoters of a subset of genes through modulation of ERα and DNMT3B complexes to induce the local DNA methylation changes, which may dictate drug responses and breast cancer development.

Keywords: 27-hydroxycholesterol; DNA methylation; breast cancer; epigenetics; estrogen receptor alpha; liver X receptor (LXR).

Figure 1

Identification of aberrant DNA methylation signatures: The heatmap represents the hypermethylation in 27-HC treated samples. The heatmaps were generated using beta values obtained from fluorescence intensity from the array scanning and the hypermethylated genes are with beta values greater that 0.5 fold change (treated > 0.5 compared to control) .The genes which showed significant DNA hypermethylation are phosphatidylserine synthase (PTDSS2), mir613, indoleamine 2,3-Dioxygenase 1 (IDO1), thyroid hormone receptor alpha (THRA), Dystrotelin (DTYN), Mesoderm induction early response 1, family member 3 (MIER). The hypermethylated genes are shown in red color. The list of genes was taken for analysis from bisulfite sequencing ( Supplementary Data, S1 )

Figure 2

The heatmap represents the hypomethylation in 27-HC treated samples. The heatmaps were generated using beta values obtained from fluorescence intensity from the array scanning and the genes are considered hypomethylated when beta values are lesser than 0.5 (treated < 0.5 compared to control).

Figure 3

Validation of promoter DNA methylation by bisulfite sequencing: (A) Schematic representation of PTDSS2 promoter region in the chromosome 8, and the boxes contain the sequences of two amplicons from PTDSS2 gene promoter. The CpGs are highlighted in red. In the bisulfite sequencing analysis output from BISMA, the red box indicates methylated CpGs and the blue box indicates the unmethylated CpGs. Increased methylation is observed at the CpG 1, CpG 2, and CpG 3 in the PTDSS2 amplicon 1 of 27-HC-treated cells than DMSO control. There is significantly increased methylation at the amplicon 2 in 27-HC-treated cells than DMSO control. Each row indicates individual clone, and the number at the top indicates the CpG sites of the corresponding amplicons. (B) The bisulfite sequencing analysis for mir613 and location of its position on chromosome 14. The box contains the sequence of amplicon from the mir613 promoter, and the CpG sites are highlighted in red. There is a significant level of hypermethylation at the CpG 2, 3, and 5 in 27-HC-treated cells than DMSO control. Each row indicates individual clone, and the number at the top indicates the CpG sites of the corresponding amplicon.

Figure 4

The 27-hydroxycholesterol (27-HC) weakens the DNMT3B association with estrogen receptor alpha (ERα): Co-immunoprecipitation (Co-IP) of ERα from DMSO- and 27-HC-treated cells has identified that the 27-HC causes a reduction in the association of DNMT3B with the ERα protein. The samples were separated on 12% gel and probed with the corresponding primary antibody. The ERα is used as a loading control. The bar graph represents the intensity of DNMT3B measured using ImageJ. The Co-IP assay was performed in triplicate, and the intensity of DNMT3B and ERα was measured using ImageJ. The normalized DNMT3B signal plotted and the error bar represent the standard error calculated from the biological triplicates. The p values were calculated using paired t-test, and the differences were considered significant at p values < 0.01 (**).

Figure 5

Transcriptional regulation of 27-hydroxycholesterol (27-HC)-mediated DNA hypermethylation: (A) The bar graph represents the qRT-PCR analysis for PTDSS2 and has revealed the downregulation of its expression in 27-HC-treated cells. The GAPDH was used as normalization control, and the error bar represents the standard deviation of three independent replicates. The bottom scheme represents the hypermethylation of PTDSS2 gene promoter leading to transcriptional repression. (B) The qRT-PCR analysis has revealed the biogenesis of miR-613 microRNA in 27-HC-treated cells. The U6 miRNA was used as internal normalization control to measure the relative expression of the miR-613, and the error bar represents the standard deviation of three independent replicates. The scheme at the bottom of the bar graph represents the hypermethylation of mir-613 gene promoters, leading to transcriptional activation and biogenesis of miR-613. The p values were calculated using paired t-test, and the p values < 0.05 were considered significant.

Figure 6

In silico analysis for transcription factor (TF) binding sites on PTDSS2 promoter: (A) The list of predicted TFs for PTDSS2 promoter sequence. The TFs are labeled in the box covering its putative binding sites on the promoter DNA sequence. The hypermethylated CpG sites are highlighted in red. (B) The output file from bisulfite sequencing analysis for PTDSS2 gene promoter using BIMSA software and indicates the percentage of methylation at the top 4 CpG sites in the 27-hydroxycholesterol (27-HC)-treated samples.

Figure 7

The interaction network of hsa-miR-613 with transcription factors (TFs) using Network module of TransMir. Level 1 is predicted, and level 2 is supported by high-throughput experimental data. TFs like ESR1, Sp1, FOXA1, GATA3, FOXM1, and AP1 are reported to be regulated by estrogen or 27-hydroxycholesterol (27-HC).

Figure 8

Analysis of mir-613 network using miRNA-centric network visual analytics platform: (A) Represents the network connecting hsa-mir-613 with the important transcription factors, miRNAs with the main node being NR1H3 (LXRα), SREBPF2, SP1, and ESR1 (ERα). (B) Enriched with function biological process: cell proliferation. (C) Represents the pathway connecting liver X receptor (LXR) and ESR1 (ERα) to hsa-mir-613.