The evidence that chromatin is involved in regulating splicing is piling up. But is that just a one-way street? No, says an international team of researchers who now report loads of data suggesting that splicing can return the favor and contribute to histone methylation.
Looking at ChIP experiments in mouse and human genomes, the researchers showed that H3K36me3 levels were low in intronless genes, whereas genes with introns had low levels of the histone mod at the transcription start site, but then the levels increased progressively along the gene according to the amount of transcription. Here are a few other things they found:
- H3K36me3 levels in RNAs from intron-containing genes with three exons were low at the first intron-exon border, rose at the second exon, and rose even higher at the third exon. But when that second exon was spliced out, the gene had much less H3K36me3 at the first intron-exon border, suggesting that splicing influences chromatin mods.
- U2AF1, a gene with introns, had lots of H3K36me3 when undergoing active transcription, but low levels during conditions when transcription declined. However, an intronless gene called Zrsr1 had the same small amount of H3K36me3 regardless of transcription activity. This could mean that when there’s no splicing, transcriptional activation alone isn’t enough to change the mod.
- Inhibiting splicing or transcription caused H3K36me3 levels to decline, but levels rebounded when the researchers removed the inhibitor.
- Their data also suggests that splicing may recruit the HYPB/Setd2 histone methyltransferase to intron-containing genes.
The researchers say their data support the idea that epigenetics and RNA processing are indeed having some “bidirectional” communication.
Check out all of the epigenetic splicing and dicing at Nature Structural & Molecular Biology, July 2011.