Mapping the dynamics of epigenetic adaptation during heterochromatin misregulation.

A classical and well-established mechanism that enables cells to adapt to new and adverse conditions is the acquisition of beneficial genetic mutations. Much less is known about epigenetic mechanisms that allow cells to develop novel and adaptive phenotypes without altering their genetic blueprint. It has been recently proposed that histone modifications, such as heterochromatin-defining H3K9 methylation (H3K9me), normally reserved to maintain genome integrity, can be redistributed across the genome to establish new and potentially adaptive phenotypes. To uncover the dynamics of this process, we developed a precision engineered genetic approach to trigger H3K9me redistribution on-demand in fission yeast. This enabled us to trace genome-scale RNA and chromatin changes over time prior to and during adaptation in long-term continuous cultures. Establishing adaptive H3K9me occurs over remarkably slow time-scales relative to the initiating stress. During this time, we captured dynamic H3K9me redistribution events ultimately leading to cells converging on an optimal adaptive solution. Upon removal of stress, cells relax to new transcriptional and chromatin states rather than revert to their initial (ground) state, establishing a tunable memory for a future adaptive epigenetic response. Collectively, our tools uncover the slow kinetics of epigenetic adaptation that allow cells to search for and heritably encode adaptive solutions, with implications for drug resistance and response to infection.

Identification of a novel, tunable interface in the S.pombe HP1 protein, Swi6, that underpins epigenetic inheritance.

HP1 proteins bind dynamically to H3K9 methylation and are essential for establishing and maintaining transcriptionally silent epigenetic states, known as heterochromatin. HP1 proteins can dimerize, forming a binding interface that interacts with and recruits diverse chromatin-associated factors. HP1 proteins rapidly evolve through sequence changes and gene duplications, but the extent of variation required to achieve functional specialization is unknown. To investigate how changes in amino acid sequence impact epigenetic inheritance, we performed a targeted mutagenesis screen of the dimerization and protein interaction domain of the S.pombe HP1 homolog Swi6. We discovered that substitutions mapping to an auxiliary motif in Swi6 outside the dimerization interface can lead to complete functional divergence. Specifically, we identified point mutations at a single amino acid residue that resulted in either persistent gain or loss of function in epigenetic inheritance without affecting heterochromatin establishment. These substitutions increase Swi6 chromatin occupancy in vivo and alter Swi6-protein interactions that selectively affect H3K9me inheritance. Based on our findings, we propose that relatively minor changes in Swi6 amino acid composition can lead to profound changes in epigenetic inheritance, underscoring the remarkable plasticity associated with HP1 proteins and their ability to evolve new functions.

Uncoupling the distinct functions of HP1 proteins during heterochromatin establishment and maintenance.

H3K9 methylation (H3K9me) marks transcriptionally silent genomic regions called heterochromatin. HP1 proteins are required to establish and maintain heterochromatin. HP1 proteins bind to H3K9me, recruit factors that promote heterochromatin formation, and oligomerize to form phase-separated condensates. We do not understand how these different HP1 properties are involved in establishing and maintaining transcriptional silencing. Here, we demonstrate that the S. pombe HP1 homolog, Swi6, can be completely bypassed to establish silencing at ectopic and endogenous loci when an H3K4 methyltransferase, Set1, and an H3K14 acetyltransferase, Mst2, are deleted. Deleting Set1 and Mst2 enhances Clr4 enzymatic activity, leading to higher H3K9me levels and spreading. In contrast, Swi6 and its capacity to oligomerize were indispensable during epigenetic maintenance. Our results demonstrate the role of HP1 proteins in regulating histone modification crosstalk during establishment and identify a genetically separable function in maintaining epigenetic memory.

Uncoupling the distinct functions of HP1 proteins during heterochromatin establishment and maintenance.

H3K9 methylation (H3K9me) marks transcriptionally silent genomic regions called heterochromatin. HP1 proteins are required to establish and maintain heterochromatin. HP1 proteins bind to H3K9me, recruit factors that promote heterochromatin formation, and oligomerize to form phase-separated condensates. We do not understand how HP1 protein binding to heterochromatin establishes and maintains transcriptional silencing. Here, we demonstrate that the S.pombe HP1 homolog, Swi6, can be completely bypassed to establish silencing at ectopic and endogenous loci when an H3K4 methyltransferase, Set1 and an H3K14 acetyltransferase, Mst2 are deleted. Deleting Set1 and Mst2 enhances Clr4 enzymatic activity, leading to higher H3K9me levels and spreading. In contrast, Swi6 and its capacity to oligomerize were indispensable during epigenetic maintenance. Our results demonstrate the role of HP1 proteins in regulating histone modification crosstalk during establishment and identifies a genetically separable function in maintaining epigenetic memory.

Investigating Mitotic Inheritance of Histone Modifications Using Tethering Strategies.

The covalent and reversible modification of histones enables cells to establish heritable gene expression patterns without altering their genetic blueprint. Epigenetic mechanisms regulate gene expression in two separate ways: (1) establishment, which depends on sequence-specific DNA- or RNA-binding proteins that recruit histone-modifying enzymes to unique genomic loci, and (2) maintenance, which is sequence-independent and depends on the autonomous propagation of preexisting chromatin states during DNA replication. Only a subset of the vast repertoire of histone modifications in the genome is heritable. Here, we describe a synthetic biology approach to tether histone-modifying enzymes to engineer chromatin states in living cells and evaluate their potential for mitotic inheritance. In S. pombe, fusing the H3K9 methyltransferase, Clr4, to the tetracycline-inducible TetR DNA-binding domain facilitates rapid and reversible control of heterochromatin assembly. We describe a framework to successfully implement an inducible heterochromatin establishment system and evaluate its molecular properties. We anticipate that our innovative genetic strategy will be broadly applicable to the discovery of protein complexes and separation-of-function alleles of heterochromatin-associated factors with unique roles in epigenetic inheritance.

An H3K9 methylation-dependent protein interaction regulates the non-enzymatic functions of a putative histone demethylase.

H3K9 methylation (H3K9me) specifies the establishment and maintenance of transcriptionally silent epigenetic states or heterochromatin. The enzymatic erasure of histone modifications is widely assumed to be the primary mechanism that reverses epigenetic silencing. Here, we reveal an inversion of this paradigm where a putative histone demethylase Epe1 in fission yeast, has a non-enzymatic function that opposes heterochromatin assembly. Mutations within the putative catalytic JmjC domain of Epe1 disrupt its interaction with Swi6HP1 suggesting that this domain might have other functions besides enzymatic activity. The C-terminus of Epe1 directly interacts with Swi6HP1, and H3K9 methylation stimulates this protein-protein interaction in vitro and in vivo. Expressing the Epe1 C-terminus is sufficient to disrupt heterochromatin by outcompeting the histone deacetylase, Clr3 from sites of heterochromatin formation. Our results underscore how histone modifying proteins that resemble enzymes have non-catalytic functions that regulate the assembly of epigenetic complexes in cells.

A cell's identity depends on which of its genes are active. One way for cells to control this process is to change how accessible their genes are to the molecular machinery that switches them on and off. Special proteins called histones determine how accessible genes are by altering how loosely or tightly DNA is packed together. Histones can be modified by enzymes, which are proteins that add or remove specific chemical 'tags'. These tags regulate how accessible genes are and provide cells with a memory of gene activity. For example, a protein found in yeast called Epe1 helps reactivate large groups of genes after cell division, effectively 're-setting' the yeast's genome and eliminating past memories of the genes being inactive. For a long time, Epe1 was thought to do this by removing methyl groups, a 'tag' that indicates a gene is inactive, from histones ­ that is, by acting like an enzyme. However, no direct evidence to support this hypothesis has been found. Raiymbek et al. therefore set out to determine exactly how Epe1 worked, and whether or not it did indeed behave like an enzyme. Initial experiments testing mutant versions of Epe1 in yeast cells showed that the changes expected to stop Epe1 from removing methyl groups instead prevented the protein from 'homing' to the sections of DNA it normally activates. Detailed microscope imaging, using live yeast cells engineered to produce proteins with fluorescent markers, revealed that this inability to 'home' was due to a loss of interaction with Epe1's main partner, a protein called Swi6. This protein recognizes and binds histones that have methyl tags. Swi6 also acts as a docking site for proteins involved in deactivating genes in close proximity to these histones. Further biochemical studies revealed how the interaction between Epe1 and Swi6 can help in gene reactivation. The methyl tag on histones in inactive regions of the genome inadvertently helps Epe1 interact more efficiently with Swi6. Then, Epe1 can simply block every other protein that binds to Swi6 from participating in gene deactivation. This observation contrasts with the prevailing view where the active removal of methyl tags by proteins such as Epe1 switches genes from an inactive to an active state. This work shows for the first time that Epe1 influences the state of the genome through a process that does not involve enzyme activity. In other words, although the protein may 'moonlight' as an enzyme, its main job uses a completely different mechanism. More broadly, these results increase the understanding of the many different ways that gene activity, and ultimately cell identity, can be controlled.