Methionine metabolism influences genomic architecture and gene expression through H3K4me3 peak width

Abstract

Nutrition and metabolism are known to influence chromatin biology and epigenetics through post-translational modifications, yet how this interaction influences genomic architecture and connects to gene expression is unknown. Here we consider, as a model, the metabolically-driven dynamics of H3K4me3, a histone methylation mark that is known to encode information about active transcription, cell identity, and tumor suppression. We analyze the genome-wide changes in H3K4me3 and gene expression in response to alterations in methionine availability in both normal mouse physiology and human cancer cells. Surprisingly, we find that the location of H3K4me3 peaks is largely preserved under methionine restriction, while the response of H3K4me3 peak width encodes almost all aspects of H3K4me3 biology including changes in expression levels, and the presence of cell identity and cancer-associated genes. These findings may reveal general principles for how nutrient availability modulates specific aspects of chromatin dynamics to mediate biological function.

Fig. 1

MR reduces H3K4me3 but maintains its genomic distribution. a MR experiment and ChIP-seq data analysis pipeline in human cancer cells HCT116. b Number of H3K4me3 peaks called under high (100 μM) and low (3 μM) methionine conditions in human cancer cells. c Total length of shared and unique peak regions for high and low methionine conditions in human cancer cells. d Definition of peak height, area, and width. e Density scatter plots comparing H3K4me3 peak heights, areas, and widths between high and low methionine conditions in human cancer cells. Colors represent for dot density. Dashed lines show identical x and y coordinates. Spearman’s rank correlation coefficients between the high and low methionine conditions and linear regression coefficients are shown in the table at bottom. f MR experiments scheme in vivo. Mouse image used with permission from Microsoft. g Density scatter plots comparing H3K4me3 peak heights, areas, and widths in mouse liver between high (0.84% w/w) and low (0.12% w/w) methionine diets

Fig. 2

H3K4me3 width dynamics encode biological information. a Heat map showing Spearman’s rank correlation coefficients among fold changes of H3K4me3 peak heights, areas, and widths under MR in human cancer cells. The upper right part shows density scatter plots comparing the corresponding log2(fold change) values. Colors of dots in the scatter plots indicate density of dots. b Same as in a but for mouse liver. c Distribution of fold changes (defined by the values in low methionine condition divided by the corresponding values in high methionine condition) in peak width. The arrow at the bottom denotes the change from sensitive (more reduction in peak width) to robust (less reduction in peak width). d Number of pathways significantly enriched (GSEA FDR Q-value < 1e-5) in peaks with different dynamics under MR in human cancer cells. e Same as in d but for mouse liver. f Examples of annotated pathways in each category. g Annotation of 298 MSigDB pathways enriched in H3K4me3 peaks with sensitive width in human cancer cells. h Representative cancer-related pathways enriched in H3K4me3 peaks with sensitive width in human cancer cells. i Annotation of 48 MSigDB pathways enriched in H3K4me3 peaks with robust width in mouse liver. j Representative liver-specific pathways enriched in H3K4me3 peaks with robust width in mouse liver

Fig. 3

H3K4me3 width dynamics encode cell type-specific TF binding. a Distribution of fold changes (defined by the values in low methionine condition divided by the corresponding values in high methionine condition) in peak width in all peaks (gray), top 500 peaks with sensitive width (orange) and top 500 peaks with robust width (blue) in human cancer cells. b Framework for the TF binding motif enrichment analysis. c Distributions of TF binding motif enrichment Q-values in H3K4me3 peak sets with different dynamics under MR in human cancer cells. Box limits are the 25th and 75th percentiles, center lines are medians, and the whiskers are the minimal and maximal values. d Same as in c but for mouse liver. e Top 10 TF binding motifs enriched in 500 H3K4me3 peaks with sensitive width in human cancer cells. f Same as in e but for robust width in mouse liver. g Number of oncogenes and tumor suppressors in top 10 TFs enriched in H3K4me3 peaks with sensitive width in human cancer cells and those with robust width in mouse liver

Fig. 4

H3K4me3 width dynamics predict differential gene expression. a Framework of RNA-seq data analysis. Mouse image used with permission from Microsoft. b Spearman correlation between peak size descriptors and gene expression levels in human cancer cells. c Same as in b but for mouse liver. d Expression levels of genes associated with different H3K4me3 dynamics under high methionine conditions in human cancer cells. The gene set with sensitive width which has been demonstrated to associate with cell type-specific biological functions and TF binding is highlighted. Box limits are the 25th and 75th percentiles, center lines are medians, and the whiskers are the minimal and maximal values. e Same as in d but for mouse liver. f Differential gene expression in human cancer cells under MR. g Same as in f but for mouse liver. h Fraction of differentially expressed genes in genes with or without H3K4me3 in human cancer cells. i Same as in h but for mouse liver. j Spearman’s rank correlation coefficients between H3K4me3 changes and gene expression changes under MR in human cancer cells. k Same as in j but for mouse liver