HP1 dimerization but not LLPS drives the condensation and segregation of H3K9me-marked chromatin

ABSTRACT Heterochromatin protein 1 (HP1) is a conserved chromatin-associated factor implicated in the establishment and maintenance of H3K9me-marked heterochromatin, potentially through phase separation–mediated condensation. Whether HP1 in molecular terms works primarily via dimerization or liquid–liquid phase separation (LLPS) is unresolved. Using the C. elegans HP1 orthologue HPL-2 and a combined in vitro–in vivo approach, we systematically dissected the molecular determinants of HPL-2 function in heterochromatin condensation. Through specific mutants, we demonstrate that HPL-2 dimerization, but not LLPS, is essential for condensing H3K9me chromatin arrays in vitro and for maintaining H3K9me heterochromatin foci in C. elegans embryos. We further show that HPL-2 dimerization is sufficient to mediate segregation of H3K9me from unmodified chromatin arrays in vitro , generating biphasic condensates reminiscent of cellular heterochromatin domains. Surprisingly, HPL-2 mutants defective in condensation cause only minor transcriptional changes at canonical heterochromatin loci, implying that HP1-dependent heterochromatin foci and gene silencing are not tightly coupled in vivo . Nonetheless, these mutant C. elegans exhibit profound physiological and developmental defects. Our findings establish dimerization as the principal molecular mechanism of HP1-driven H3K9me-chromatin condensation, elucidate the auxiliary role of LLPS, and reveal the uncoupling between HP1-dependent heterochromatin and transcriptional regulation. GRAPHICAL ABSTRACT