Novel insights into mechanisms partitioning chromatin states

Eukaryotic chromatin can be divided into two main types: euchromatin with high levels of transcription and heterochromatin, where transcription is repressed. These chromatin types are not static, but show remarkable plasticity during development and adaptation to environmental cues: Heterochromatic genes can get activated, whereas euchromatic genes get repressed. This is orchestrated by a multitude of chromatin-modifying enzymes, which extensively modify the protruding histone tails of the nucleosomes, the fundamental packaging unit of DNA. Such histone modifications serve as binding platforms for a plethora of specific chromatin readers and show distinct characteristics depending on which chromatin state they belong to: Euchromatin is characterized by H3K36me3, H3K4me3 and H2B ubiquitination (H2Bub), whereas heterochromatin is hypoacetylated and carries high levels of H3K9 methylation.
With the development of deep sequencing technologies, the perception of transcriptionally silent heterochromatin changed. It became clear that heterochromatin can also be transcribed,but that these transcripts are quickly recognized and degraded in a chromatin context-dependent manner. This holds true for the fission yeast Schizosaccharomyces pombe (S. pombe), the model organism that I used in my PhD thesis. It has very similar heterochromatic characteristics compared to other eukaryotes, such as H3K9me2 and heterochromatic transcription. Notably, heterochromatin formation in S. pombe involves the RNA interference (RNAi) pathway, which directs de novo heterochromatin formation in a sequence-specific manner via small interfering RNAs (siRNAs). Maintenance and spreading of heterochromatin across nucleosomes is achieved by a self-enforcing feedback loop, wherein heterochromatin is stabilized, expanded, and propagated in a sometimes DNA-sequence independent manner.
Although siRNAs are necessary to maintain constitutive heterochromatin, siRNAs targeting a euchromatic locus do not induce heterochromatin formation. This paradox laid the foundation for my PhD thesis by evoking the question how euchromatic genes are protected from de novo heterochromatin formation by RNAi.
In a first study to address this question, I contributed to the discovery that the Polymerase II-associated factor 1 complex (Paf1C) is a main repressor of RNAi-mediated heterochromatin formation. In Paf1C mutant cells, silencing of the siRNA-targeted euchromatic gene occurred in a stochastic manner, but was maintained efficiently with and without the initial source of siRNAs; a truly epigenetic phenomenon. Yet, initiation of silencing was rather limited, suggesting the existence of additional repressive activities that may function specifically at the initiation step.
I identified the histone acetyltransferase Mst2 to be such a repressive factor. The Mst2 complex represses heterochromatin formation specifically at the initiation step, but does not affect maintenance. This is achieved by the exclusion of Mst2C from heterochromatin due to tethering of Mst2C to H3K36me3-marked nucleosomes, a hallmark of actively transcribed genes. This tethering in turn protects euchromatic genes from silencing. Further, to dissect the mechanism of this protection, I aimed at identifying all potential Mst2 substrates. By employing an acetylomics approach, I discovered that Mst2 also acetylates a specific residue of Brl1. This is exciting, because Brl1 is an enzyme responsible for H2B ubiquitination, a second hallmark of active chromatin. Acetylation of Brl1 increases H2Bub and feeds back to increased transcription activity, which is also marked by increased H3K4me3 levels.
In conclusion, my work led to the discovery of a euchromatic feedback loop, which protects euchromatin from de novo heterochromatin formation. I am fascinated by this finding, because it implies that euchromatic genes can “remember” their active state and actively counteract heterochromatin formation. Furthermore, the interplay of euchromatic and heterochromatic feedback loops creates epigenetic plasticity, which allows cells to keep genes on or off in an almost digital manner. Since the general characteristics of heterochromatin and all identified factors are conserved, I propose that similar feedback loops likely partition chromatin in active and inactive states also in humans and may help promoting epigenetic robustness against environmental challenges and cancer.