Nitric Oxide Regulates Gene Expression in Cancers by Controlling Histone Posttranslational Modifications

Abstract

Altered nitric oxide (•NO) metabolism underlies cancer pathology, but mechanisms explaining many •NO-associated phenotypes remain unclear. We have found that cellular exposure to •NO changes histone posttranslational modifications (PTM) by directly inhibiting the catalytic activity of JmjC-domain containing histone demethylases. Herein, we describe how •NO exposure links modulation of histone PTMs to gene expression changes that promote oncogenesis. Through high-resolution mass spectrometry, we generated an extensive map of •NO-mediated histone PTM changes at 15 critical lysine residues on the core histones H3 and H4. Concomitant microarray analysis demonstrated that exposure to physiologic •NO resulted in the differential expression of over 6,500 genes in breast cancer cells. Measurements of the association of H3K9me2 and H3K9ac across genomic loci revealed that differential distribution of these particular PTMs correlated with changes in the level of expression of numerous oncogenes, consistent with epigenetic code. Our results establish that •NO functions as an epigenetic regulator of gene expression mediated by changes in histone PTMs.

Figure 1. Nitric oxide changes histone posttranslational modifications at numerous sites on core histones

All cells were treated with the •NO-donor DETA/NO (500 μM) and histones were isolated for PTM analysis. (A, B) Relative changes in PTMs at key sites on the core histones H3 and H4 measured by high-resolution mass spectrometry. MDA-MB-231 cells were cultured in the presence or absence of •NO and histones were isolated at the indicated time points (0, 24, 48 h). The “wash” column shows data collected from cells that were treated with •NO for 24 h, followed by 24 h of incubation in •NO-free media (histones were collected 48 h after initial treatment). Significant differences and standard error of mean are provided in Fig. S1 and Table S1. (C) Temporal changes at H3K9me2/ac in response to •NO. MDA-MB-231 cells were exposed to •NO for 24 h, followed by 24 h of incubation in •NO-free media. At the indicated time points, histones were isolated and immunoblots were conducted for H3K9me2/ac. (D) •NO-mediated changes in H3K9me2/ac in 9 different cell types. All cells were exposed to •NO for 24 hours, histones were extracted and immunoblotted for H3K9me2/ac.

Figure 2. Nitric oxide induces changes in gene expression and cellular phenotype

(A) List of the 10 most up- and downregulated genes in response to •NO. MDA-MB-231 cells were treated with 500 μM DETA/NO for 24 hours. mRNA was extracted and prepared for hybridization onto GeneChip® PrimeView™ Human Gene Expression Arrays (Affymetrix). Microarray data was analyzed to determine differential gene expression in the presence or absence of •NO. (B) Real-time measurement of cell proliferation (+/− •NO) using the xCELLigence® DP system. MDA-MB-231 cells were seeded in E-plates containing 10% serum and allowed to adhere for 12 h before the addition of •NO (250μM DETA/NO). Cell proliferation was measured for 12 hours following •NO treatment. (C) Real-time measurement of cell migration (+/− •NO) using the xCELLigence® DP system. MDA-MB-231 cells were seeded in CIM plates containing 10% serum as a migratory stimulant. Cell migration was measured for 12 hours following addition of •NO (250μM DETA/NO).

Figure 3. Nitric oxide results in differential patterns of H3K9ac and H3K9me2 distribution across genomic loci

MDA-MB-231 cells were treated with •NO (500 μM DETA/NO) for 24 hours; H3K9me2/ac ChIP-DNA was isolated using validated antibodies and sequenced to appropriate depths. MACS2 (narrow peak caller) and SICER (broad peak caller) algorithms were used for peak calling within H3K9ac and H3K9me2 enriched regions respectively. (A, B) Heatmaps represent the distribution of H3K9ac in untreated controls. Histograms represent differences in the average H3K9ac enrichment profiles around the TSS and TSE in control and •NO-exposed cells. (C) Number of genes associated with H3K9ac peaks that occur within +/− 5 kb of TSS in control and •NO-treated cells. (D) Representative heatmaps demonstrating the distribution patterns of H3K9ac across genomic loci at TSS and TSE following •NO exposure. Example genes associated with •NO-induced increases in H3K9ac around the TSS are highlighted by green boxes. (E, F) Heatmaps represent the distribution of H3K9me2 in untreated controls. Histograms represent differences in the average H3K9me2 enrichment profiles around the TSS and TSE in control and •NO-exposed cells. (G) Number of genes associated with H3K9me2 peaks that occur within +/− 200 kb of TSS in control and •NO-treated cells. (H) Representative heatmaps demonstrating the distribution patterns of H3K9me2 across genomic loci at TSS and TSE following •NO exposure. Example genes associated with •NO-induced increases in H3K9me2 around the TSE are highlighted by green boxes.

Figure 4. Genes upregulated by •NO gain promoter H3K9ac enrichment

(A) Visualization of increased H3K9ac around gene promoters following •NO exposure using Integrative Genomics Viewer (IGV). Bar plots represent expression changes of corresponding mRNA transcripts as measured by GeneChip® PrimeView™ Human Gene Expression Arrays (Affymetrix). (B) Motif analysis from the JASPAR core database using CLOVER within the same gene set revealed multiple Ets-1 transcription factor binding sites around H3K9ac-associated promoters emergent following •NO treatment. Histogram represents average H3K9ac enrichment around potential Ets-1 binding sites. The “dip” at 0 represents Ets-1 binding and nucleosome exclusion. Right panel is a magnified version of the left panel. (C) Ets-1 binding motif sequence.

Figure 5. •NO-induced gene expression correlates to changes in H3K9me2 around their promoters

(A, B, D) Changes in the expression levels of three •NO-regulated cancer genes measured using GeneChip® PrimeView™ Human Gene Expression Arrays (Affymetrix). (C, E) Visualization of H3K9me2 enrichment at the candidate gene loci using Integrative Genomics Viewer (IGV).