Histones undergo a range of post-translational modifications, including acetylation, methylation, phosphorylation, ubiquitination and sumoylation. The idea that histone modifications, particularly acetylation and methylation, have fundamental roles in controlling gene transcription is now so firmly established that it is perhaps surprising to realize that this only became apparent during the past 5–10 years.
Histones had been known to be acetylated and methylated for a long time, and the idea of a function in transcription was considered as early as the 1960s. For example, in their 1964 paper, Allfrey, Faulkner and Mirsky demonstrated that histone acetylation can reduce their efficacy as inhibitors of transcription, and thought that this implied "a dynamic and reversible mechanism for activation as well as repression of RNA synthesis." In subsequent decades, enzymes that could acetylate histones were biochemically characterized and cloned. However, their relevance for regulating gene transcription was largely ignored.
In 1996, two papers — published within a month of each other — showed that histone acetylases and deacetylases were, in fact, well-known transcriptional regulators. These studies provided the first clear connection between histone acetylation and transcriptional regulation. Both papers were hailed as breakthroughs and, from then on, understanding how histone modifications control chromatin structure and gene expression became a significant area of enquiry. In the first paper, Allis and colleagues set out to clone a histone acetyltransferase (HAT) from Tetrahymena thermophila. The cloned HAT turned out to have a high level of sequence relatedness to the yeast Gcn5, a transcriptional activator, with almost 80% sequence identity in some domains. Consistent with the sequence homology, the authors showed that recombinant Gcn5 had HAT activity in vitro. In the second study, Schreiber and colleagues purified mammalian histone deacetylases using an inhibitor as an affinity matrix. This approach, combined with the microsequencing of purified proteins, yielded a surprise — one of the proteins had 60% sequence identity to yeast Rpd3, a characterized transcriptional repressor. The two antagonistic enzymatic activities had the opposite functional effects — activation or repression — on transcription.
So what about histone methylation? By 2000, several histone acetylases and deacetylases had been identified, but a functional link between histone methylation and chromatin structure or gene transcription remained elusive. Genetic screens in fruitflies and fission yeast had shown that the fruitfly suppressor of variegation 3-9 (Su(var)3-9) and fission yeast clr4 were important for establishing and propagating heterochromatin — a higher-order chromatin structure that is repressive for transcription. Work on the mammalian homologues had extended this link, but a mechanistic understanding of how these proteins controlled heterochromatin formation was lacking. In 2000, work from Jenuwein and colleagues showed that the mammalian homologue of Su(var)3-9 was a histone lysine methyltransferase that selectively methylated histone H3 at Lys9.
Together with earlier studies establishing a role for histone tails in transcriptional regulation (see Milestone 16), these three studies were pivotal for focusing the attention of the field on histone modifications and their importance as epigenetic markers.
