FACT frees Ubp10 from nucleosome sequestration after rapid H2B deubiquitination

Abstract

Chromatin-modifying enzymes must engage their substrate, act on it and release it, but which step limits their turnover is unclear. Using three-color single-molecule fluorescence, we directly visualized the yeast deubiquitinase Ubp10 removing ubiquitin from H2B on individual nucleosomes. Ubp10 cleaved the mark within seconds of binding, without detectable DNA unwrapping, yet it did not let go: ~80% of nucleosomes remained bound after 45 minutes, and unmodified nucleosomes formed similarly long-lived complexes. Ubp10 turnover is therefore limited by release rather than catalysis. The histone chaperone FACT relieved this bottleneck by accelerating Ubp10-nucleosome dissociation. In living yeast, nuclear depletion of the FACT subunit Spt16 increased the slowly diffusing, chromatin-bound fraction of nucleoplasmic Ubp10. These findings identify FACT-dependent release as a likely mechanism for recycling Ubp10 on chromatin. More broadly, a nucleosome can sequester the enzyme acting on it, making enzyme release a step at which chromatin modification can be controlled.