September 06, 2012
Chromatin proteins, primarily histones, are a great way to control what parts of a cell’s DNA are accessible to its machinery. These proteins coat the DNA and are marked up in certain ways to indicate how available a piece of DNA should be. A methyl group here, an acetyl group there and a cell “knows” where the genes are that it is supposed to read!
Of course this structure needs to be maintained or a cell might start to misread parts of its DNA as starting points of genes. Then RNA polymerase II (RNAPII), the enzyme responsible for reading most protein-coding genes, would start making RNA from the wrong parts of the DNA, wreaking havoc in a cell.
One place where maintaining chromatin structure might be especially tricky is within the coding parts of genes. It is easy to imagine RNAPII barreling down the DNA, knocking the proteins aside like pins in a bowling alley. But it doesn’t. For the most part the chromatin structure stays the same and survives the onslaught of an elongating RNAPII.
Two key marks for keeping histones in place are the trimethylation of lysine 36 of histone H3 (H3K36me3) that is mediated by Set2p, and a general deacetylation of histone H4 that is mediated by the Rpd3S histone deacetylase complex. We know this because loss of either complex causes an increase in H4 acetylation and transcription starts from within genes.
In a recent study in Nature Structural & Molecular Biology, Smolle and coworkers identified two key components that help chromatin resist an elongating RNAPII in the yeast S. cerevisiae. The first, called the Isw1b complex, binds H3K36me3 and the second, the Chd1 protein, binds RNAPII itself. That these two were involved wasn’t surprising since previous work had suggested they helped prevent histone exchange at certain genes.
What makes this work unique is that the researchers showed the global importance of these proteins in the process and were able to tease out some of the fine details of what is going on at the molecular level. They used electrophoretic mobility shift assays to show that Isw1b bound the trimethylated form of H3 via its Ioc4p subunit and used chromosome immunoprecipitation coupled to microarrays (ChIP-chip) to show that Isw1b localized to the middle of genes in vivo. They also showed that when Set2p was removed, the localization disappeared (presumably because of the loss of the trimethylation of lysine 36). They clearly demonstrated that Isw1b is found primarily in the middle of genes.
While these results indicate that the Ioc4p-containing Isw1b complex is moored to the middle of genes via its interaction with H3K36me3, it does not establish what it is doing there. For this the researchers knocked out Isw1b and Chd1 and showed via genome tiling arrays a global increase in cryptic transcription starts. The DNA in the middle of genes was now being used inappropriately by RNAPII as starting points for transcription. Further investigation with Isw1b and Chd1 knockouts showed an increase in chromosome exchange and an increase in acetylated H4 in the middle of genes.
Whew. So it appears that Isw1b and Chd1 inhibit inappropriate starts of transcription by keeping hypoacetylated histones in place over the parts of a gene that are read. They are two of the key players in maintaining the right chromatin structure over genes. They help keep RNAPII from railroading histones aside as it elongates, thus protecting the cell from inappropriate transcription starts.
by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics
Categories: Research Spotlight