APC/C-mediated ubiquitylation of extranucleosomal histone complexes lacking canonical degrons

Abstract

Non-degradative histone ubiquitylation plays a myriad of well-defined roles in the regulation of gene expression and choreographing DNA damage repair pathways. In contrast, the contributions of degradative histone ubiquitylation on genomic processes has remained elusive. Recently, the APC/C has been shown to ubiquitylate histones to regulate gene expression in pluripotent cells, but the molecular mechanism is unclear. Here we show that despite directly binding to the nucleosome through subunit APC3, the APC/C is unable to ubiquitylate nucleosomal histones. In contrast, extranucleosomal H2A/H2B and H3/H4 complexes are broadly ubiquitylated by the APC/C in an unexpected manner. Using a combination of cryo-electron microscopy (cryo-EM) and biophysical and enzymatic assays, we demonstrate that APC8 and histone tails direct APC/C-mediated polyubiquitylation of core histones in the absence of traditional APC/C substrate degron sequences. Taken together, our work implicates APC/C-nucleosome tethering in the degradation of diverse chromatin-associated proteins and extranucleosomal histones for the regulation of transcription and the cell cycle and for preventing toxicity due to excess histone levels.

Fig. 1. APC3 binds nucleosome acidic patch through HMGN-like nucleosome binding motif.

a Electrostatic surface of nucleosome (PDB 3AFA), left, and zoomed view of acidic patch showing residues mutated in acidic patch interaction screen, right. b Profile plots of APC/C subunits and HMGN1 for triplicate wild-type and mutant nucleosome affinity proteomics experiments (data from). Individual quantitated peptides shown in grey and weighted averages in red or blue. ∆AP: H2A E61A, E64S, N68A, D72S, N89A, D90A, E91S. Additional profile plots shown in Supplementary Fig. 1. c Pulldowns from HEK293T nuclear lysates using reconstituted, wild-type or mutant biotinylated nucleosomes followed by western blot for APC2, APC5 or APC/C coactivator, CDC20. Pulldown-westerns performed one time. d Schematic of nucleosome label transfer crosslinking experiment with H2A E91C linked to Mts-Atf-Biotin reagent, left. Photocrosslinking followed by TCEP addition, allows transfer of biotin to crosslinked protein. Simultaneous blot for biotin and APC3 consistent with photocrosslinking to APC3. Photocrosslinking performed two times with equivalent results. e Sequence alignment of proposed APC3 nucleosome-binding loop and HMGN proteins showing partial conservation of HMGN nucleosome binding motif. Source data are provided as a Source Data file.

Fig. 2. APC/C ubiquitylates extranucleosomal histones in absence of canonical degron.

a APC/C-CDC20-mediated ubiquitylation assays using indicated nucleosomes, H3/H4 tetramers, or H2A/H2B dimers assembled with one fluorescently labeled histone indicated with asterisk. Cyclin B^N* included as control APC/C substrate. Reactions were visualized by fluorescent scanning of SDS-PAGE gels. Representative images of n = 3 independent experiments. b Time course of H2B ubiquitylation by APC/C-CDC20 using H2A/H2B* dimer alone or bound to Nap1 histone chaperone as a substrate. Representative image of n = 4 independent experiments. c APC/C-dependent ubiquitylation assays with indicated substrates and HSL1 D-box and/or HSL1 KEN-box containing peptide competitors. Representative images of n = 2 independent experiments. d APC/C-mediated ubiquitylation assays with H3*/H4 tetramer substrates titrated with unlabeled H3/H4, Securin, or Cyclin B^N competitor substrates. Representative image of n = 3 independent experiments. Source data are provided as a Source Data file.

Fig. 3. APC/C ubiquitylates histone tails and folded domains.

a APC/C-mediated ubiquitylation assays using indicated tailless histone substrates visualized by fluorescent scanning of SDS-PAGE gels. Representative images of n = 2 experiments. b Schematic of mass spectrometry identified ubiquitylation sites (open circles) mapped onto histone structures with tails and folded histone regions and the position of all lysines annotated. Source data are provided as a Source Data file.

Fig. 4. Cryo-EM models of APC/C-CDC20-UBE2C-X-H2A/H2B and -X-H3/H4 complexes.

a Schematic of UBE2C-X-H2A/H2B crosslinked E2-X-substrate complex, top left. UBE2C C114 crosslinked to H2B K11C mutant using BMOE. Representative image of Coomassie-stained gel of reconstituted APC/C-CDC20-UBE2C-X-H2A/H2B complex, bottom left. Cryo-EM map low pass filtered at 10 Å resolution to visualize density in APC/C central cavity, middle, and proposed model overlaid with map, right, of APC/C-CDC20-UBE2C-X-H2A/H2B complex. b Schematic of UBE2C-X-H3/H4 crosslinked E2-X-substrate complex, top left. UBE2C C114 crosslinked to H3 K27C mutant using BMOE. Gel of reconstituted APC/C-CDC20-UBE2C-X-H3/H4 complex, bottom left. Cryo-EM map low pass filtered at 10 Å resolution to visualize density in APC/C central cavity, middle, and proposed model overlaid with map, right, of APC/C-CDC20-UBE2C-X-H3/H4 complex. One H3/H4 heterodimer is modeled. Gels is (a, b) completed once. Source data are provided as a Source Data file.

Fig. 5. APC8 contributes to coactivator-independent histone complex ubiquitylation.

a Representative 2D classes of APC/C-CDC20-UBE2C-X-H2A/H2B and -X-H3/H4 showing absence and presence of APC/C dimers, respectively. b Representative mass photometry histograms of APC/C alone or in complex with 320 nM H2A/H2B dimers or H3/H4 tetramers with and without N-terminal tails, demonstrating H3/H4-mediated APC/C dimerization. Additional replicates and concentrations are shown in Supplementary Fig. 9a. c APC/C-dependent ubiquitylation assays of indicated histone complexes in the presence of coactivator CDC20 and/or the activating UBE2S^CTP. Representative images of n = 3 experiments. d Ubiquitylation assays of indicated histone complexes using APC/C platform (Plat), APC Plat + APC8, APC/C or APC/C-CDC20. Representative images of n = 3 experiments. Assays is (c, d) visualized by fluorescent scanning of SDS-PAGE gels. Source data are provided as a Source Data file.

Fig. 6. H2A disordered tails are required for APC8-dependent H2B ubiquitylation.

a Schematic of APC/C architecture highlighting distance between APC8 and UBE2C active site cysteine, C114. b APC/C-mediated ubiquitylation assays using H2A/H2B dimers with one histone fluorescently labeled and the other histone tailless visualized by fluorescent scanning of SDS-PAGE gels. Representative image of n = 3 experiments. c Ubiquitylation assays of H2B in the context of H2A tail truncations using APC/C platform (Plat), APC Plat + APC8, APC/C or APC/C-CDC20. Representative image of n = 3 experiments. d Model of chromatin-tethered APC/C polyubiquitylating histones mobilized by polymerases or chromatin remodeling complexes, left, or DNA or nucleosome bound proteins, right. Source data are provided as a Source Data file.