Histone mimicry in HP1 is required for a conformational switch that regulates assembly of a minimal heterochromatin unit necessary for silencing in vivo

Long-term silencing of large regions of the genome is achieved through the formation of heterochromatin. From yeast to humans, heterochromatin is characterized by two key molecular signatures: (i) di or tri-methylation of lysine 9 of histone H3 (H3K9me2/3), and (ii) heterochromatin protein 1 (HP1). The association of HP1 with H3K9-methylated chromatin drives heterochromatin assembly and spread. Yet, how HP1 assembles on methylated nucleosomal templates and how the HP1-nucleosome complex is regulated are poorly understood. Using S. pombe as a model system, we show that two dimers of the HP1 protein, Swi6, binds to one nucleosome: each dimer contains one chromodomain (CD) that engages one copy of the H3K9-methyl mark, while the other CD is unoccupied. This HP1-nucleosome complex acts as a scaffold for the addition of other HP1 molecules that self-associate through a novel CD-CD interface nucleating from the unoccupied CDs. Chromodomain-mediated polymerization of HP1 on chromatin appears to (1) increase its association with methylated nucleosomes in vitro, (2) bridge neighboring methylated nucleosomes, and (3) increase heterochromatin assembly in vivo. Our data suggests that H3K9-methyl recognition and chromatin coating by HP1 are intrinsic to the fundamental architecture of the HP1-nucleosome complex. But they also raise the question of how methylated chromatin templates HP1 assembly. We found that two key features of heterochromatin, the H3K9me3 and the nucleosomal DNA, promote a conformational change in Swi6 that drives its association with nucleosomes. By binding to methylated nucleosomes, unbound Swi6 dimers switch from an autoinhibited state that is refractory to both methyl mark recognition and higher-order oligomerization to a state that is competent for spreading. Cryo-EM studies of the Swi6-nucleosome complex reveal the architecture of the spreading competent state. In vivo, mutants that disrupt such a switch also result in disruption of heterochromatin. The coupling of a conformational switch in HP1 to the recognition of specific features of methylated chromatin provides a mechanism for how HP1 can specifically target H3K9-methylated chromatin, thus preventing its aberrant spread into euchromatin. Finally, our discovery of these different HP1 conformational states provides a basic starting point for understanding how HP1 can switch between alternative functions in heterochromatin