H2Bub1 is preferentially localized in transcribed regions and is reduced in exons. (A) Mean H2Bub1/H2B ratios—deduced from ChIP-seq in NT2 cells (y-axis)—of genes grouped according to their expression level. Gene bodies were scaled to fit in 100 equally sized bins from their transcription start site (TSS) to their 3′ end (x-axis). A moving average of five was applied. (B, left) Pie chart depicting the fraction of H2Bub1/H2B peaks (Methods) in each of six gene segments ([UTR] untranslated region; [stop] translation stop codon). (Right) Relative enrichment of H2Bub1/H2B peaks across the gene segments corrected for their genomic size. (*) χ2 test P-value < 0.05; (**) P-value < 10−6. (C) Mean H2Bub1/H2B ratio in NT2 cells (y-axis) of first, second, and third introns of expressed genes (gray, blue, and purple, respectively). Introns were scaled to fit in 100 equally sized bins (x-axis). A moving average of 20 was applied. (D) Mean H2Bub1/H2B ratio in NT2 cells (y-axis) of internal exons (exon), exonic composition regions (ECRs), and pseudo exons (pseudo; red, blue and gray, respectively) in expressed genes. Seventy-five base pairs from each region end and the 250 bp adjacent to them were plotted (x-axis). A moving average of 20 was applied. (E) Mean H2Bub1/H2B ratio in NT2 cells (y-axis) of multiexonic expressed genes longer than 20 kb, grouped according to the number of exons in their first 10 kb; 1.1 kb upstream of the TSS to 10 kb downstream from it was plotted (x-axis). A moving average of 700 was applied. (F) Distribution of H2Bub1 ChIP-seq reads along the CSDE1 and LPHN2 genes (y-axis). Gene architecture is shown for each gene from start to end (x-axis), with thick lines corresponding to exons and thin lines corresponding to introns. (a.u.) Arbitrary units.
