Mechanisms of RNF168 nucleosome recognition and ubiquitylation

Abstract

RNF168 plays a central role in the DNA damage response (DDR) by ubiquitylating histone H2A at K13 and K15. These modifications direct BRCA1-BARD1 and 53BP1 foci formation in chromatin, essential for cell-cycle-dependent DNA double-strand break (DSB) repair pathway selection. The mechanism by which RNF168 catalyzes the targeted accumulation of H2A ubiquitin conjugates to form repair foci around DSBs remains unclear. Here, using cryoelectron microscopy (cryo-EM), nuclear magnetic resonance (NMR) spectroscopy, and functional assays, we provide a molecular description of the reaction cycle and dynamics of RNF168 as it modifies the nucleosome and recognizes its ubiquitylation products. We demonstrate an interaction of a canonical ubiquitin-binding domain within full-length RNF168, which not only engages ubiquitin but also the nucleosome surface, clarifying how such site-specific ubiquitin recognition propels a signal amplification loop. Beyond offering mechanistic insights into a key DDR protein, our study aids in understanding site specificity in both generating and interpreting chromatin ubiquitylation.

Keywords: 53BP1; BRCA1-BARD1; DNA damage response; NMR spectroscopy; RNF168; UbcH5c; X-ray crystallography; chromatin; cryo-EM; nucleosome; ubiquitin ligase.

Figure 1.. Interactions of RNF168 with the ubiquitylated NCP

(A) Top: Domain structure of RNF168. Bottom: Schematics of nucleosomal H2A K13 and K15 ubiquitylation catalyzed by RNF168, illustrating signal amplification by recognition of H2AK13ub or H2AK15ub. (B and C) Fluorescence polarization binding curves of fluorescently-labeled NCP or NCP carrying H2AK13ub and H2AK15ub (NCP^ub) with full-length RNF168 (RNF168^FL) and RNF168 MIU2-LRM (aa 430–481) (RNF168^ML) (B), and with WT RNF168^ML and mutants (C). Data are mean ± s.d. for each data point (n = 3 technical replicates). K_d values are indicated. ND means not determined. (D) Cryo-EM reconstruction (left and middle) and structural model (right) of the RNF168^FL-NCP^ub complex displayed in two orientations. While the complex was assembled with RNF168^FL, the only density detected was for RNF168 MIU2-LRM. (E) Close-up view of polar interactions (yellow dashes) between RNF168 LRM R477 and indicated acidic patch residues in H2A, and hydrophobic interactions between LRM Y474 and indicated residues in H2A and H2B. (F) Close-up view of RNF168 MIU2-LRM interactions with the NCP and ubiquitin linked to H2A K15, highlighting canonical interaction of ubiquitin with MIU2 and proximity of MIU2 to the NCP surface. See also Figures S1 and S2.

Figure 2.. Structure and activity of RNF168^R-UbcH5c with the NCP

(A) Fluorescence polarization binding curves of RNF168^R, WT and mutants, with fluorescently-labeled NCP. Data are mean ± s.d. for each data point (n = 3 technical replicates). K_d values are indicated. ND means not determined. (B) Left: Quantification of the NCP or H2A-H2B (WT or mutants) ubiquitylation from n = 3 independent experiments. Bar graphs show the mean and s.d. for each data point. P values were calculated using a two-sample two-tailed Student t-test; *P < 0.05, **P < 0.01, ***P < 0.001, NS means not significant. Right: Representative SDS-PAGE gels of ubiquitylation assays of the NCP, WT H2A-H2B, H2A^K13S-H2B and H2A^K15S-H2B using UBA1, UbcH5c and RNF168^R. (C) Cryo-EM reconstruction (left and middle) and structural model (right) of RNF168^R-UbcH5c-NCP complex displayed in two orientations. Locations of H2A K13 and K15 (yellow spheres) and active site C85 of UbcH5c (red sphere) are circled in red. (D–G) Close-up views of the interactions between RNF168^R and H2A-H2B (D and E), RNF168^R and UbcH5c (F), and active site of UbcH5c and H2A-H2B (G). Relevant side chains, hydrogen bonds (yellow dashed lines) and Zn⁺² (purple spheres) are highlighted. (H) Ubiquitylation reaction profiles of H2A^K13S-H2B and H2A^{E61A/E64A/D91A}-H2B^E114A catalyzed by WT RNF168^R and mutants. Reactions were probed by monitoring the NMR signal intensity changes of unreacted ¹⁵N-labeled ubiquitin. See also Figures S3–S6.

Figure 3.. Cryo-EM characterization of RNF168^R-UbcH5c~Ub-NCP and RNF168^R-UbcH5c-Ub^B-NCP

(A) Cryo-EM reconstruction (left and middle) and structural model (right) of RNF168^R-UbcH5c~Ub-NCP complex displayed in two orientations. Locations of H2A K13 and C15 (yellow spheres) and active site C85 of UbcH5c (red sphere) are circled in red. (B) Self-ubiquitylation of UbcH5c at indicated lysine residues detected by mass spectrometry. An abundance ratio for each site was determined by comparing the number of fragments containing the specific lysine with ubiquitylation to the total number of fragments containing the same lysine, both with and without ubiquitylation. (C) Proposed mechanism for UbcH5c self-ubiquitylation at K144 based on the crystal structure of UbcH5b~Ub self-assembly mediated by ubiquitin backside interaction (PDB 3A33). K144 in red is proximal to the UbcH5c active site. (D) Cryo-EM reconstruction (left and middle) and structural model (right) of RNF168^R-UbcH5c-Ub^B-NCP complex displayed in two orientations. Zn⁺² ions are purple spheres. See also Figures S6–S10.

Figure 4.. Cryo-EM structures of post-reaction states where the NCP has been ubiquitylated by RNF168^R-UbcH5c

(A) Cryo-EM reconstruction (left and middle) and structure (right) of post-reaction state 1 (class 1) of the NCP^ub-RNF168^R-UbcH5c complex. Density is visible for RNF168^R, but barely noticeable for UbcH5c. (B) Cryo-EM reconstruction (left and middle) and structure (right) of post-reaction state 2 (class 2) of the NCP^ub-RNF168^R-UbcH5c complex. UbcH5c density matches that in the pre-reaction state in Figure 2C. See also Figure S11.

Figure 5.. Conformational dynamics of RNF168^R-UbcH5c

(A) Overlay of three RNF168^R-UbcH5c structures (from crystal structures of RNF168^R-UbcH5c-H2A-H2B and the cryo-EM structure of RNF168^R-UbcH5c-NCP), aligned with respect to RNF168^R. UbcH5c samples multiple conformations, pivoting by as much as ~55° from the N-terminus of UbcH5c α-helix 1. (B) Top: Representative crystal structure (conformation 1 or open state) of RNF168^R-UbcH5c and cryo-EM structure (conformation 2 or closed state) of RNF168^R-UbcH5c-NCP both from A showing only the RNF168^R-UbcH5c moiety. Selected residues are labeled and approximate distances between RNF168^R and UbcH5c in both conformations are indicated. Bottom: RNF168^R-UbcH5c NMR signals that are affected by the RNF168^R and UbcH5c interaction (see spectral overlay in C (left)) define interfaces consistent with both X-ray and cryo-EM structures. (C) Left: Overlaid ¹H-¹⁵N HSQC spectra of ¹⁵N-labeled RNF168^R-UbcH5c, RNF168^R and UbcH5c. RNF168^R-UbcH5c signals that differ from corresponding signals in isolated RNF168^R and UbcH5c are highlighted by arrows or boxed (and enlarged in B (bottom)). Right: Plot of the chemical shift differences between the spectra of RNF168^R-UbcH5c and RNF168^R (top) and between the spectra of RNF168^R-UbcH5c and UbcH5c (bottom). One and two standard deviations (1σ and 2σ) above the mean chemical shift are indicated. Inset: Surface representation of RNF168^R-UbcH5c. RNF168^R-UbcH5c residues with ≥ 0.1 ppm chemical shift deviation from those in isolated RNF168^R and UbcH5c are colored red. (D) Summary of a 1-μs molecular dynamics (MD) simulations of RNF168^R-UbcH5c bound to H2A-H2B using the cryo-EM structure of RNF168^R-UbcH5c-NCP as starting model. Fifty simulated RNF168^R-UbcH5c models (gray) are overlaid relative to their E3 components. For comparison, RNF168^R-UbcH5c conformations 1 (open state from a crystal structure) and 2 (closed state in the cryo-EM structure) are also included. (E) Determination of the ubiquitylation zone of RNF168^R-UbcH5c using the MD simulations in D. Five-hundred atomic coordinates sampled from the 1-μs MD simulation were used to evaluate changes in atomic positions and predict which residues could be ubiquitylated if they were lysines. Ellipsoids generated using the “measure inertia” function of Chimera were used to approximate the spatial distributions of atoms from selected residues in the MD simulation. The ellipsoids shown are for UbcH5c active site C85 SG (blue), H2A K13 and K15 NZ (yellow), H2A non-lysine residues A12 to S19 CA (orange), and H2B K120 NZ (red). (F) H2A-H2B ubiquitylation by RNF168^R, UbcH5c and UBA1 at indicated H2A lysine residues monitored using SDS-PAGE. See also Figure S12.

Figure 6.. NCP^ub recognition and ubiquitylation by RNF168 and BRCA1-BARD1

(A) Cryo-EM structures of RNF168^R-UbcH5c-NCP and BRCA1^R-BARD1^R-UbcH5c-NCP. The NCPs are oriented similarly to highlight the opposite positioning of UbcH5c by ~180° in the two complexes. (B and C) Close-up views of the interactions between BRCA1^R and H2A-H2B acidic patch (B), and between BARD1^R and H2B C-terminal α-helix (C). (D) Cryo-EM reconstructions (top left and right) and structural models (bottom left and right) illustrating ubiquitin recognition in NCP^ub by RNF168 and BARD1. The side chains of ubiquitin I44 and K63 are shown to highlight the radically different ubiquitin recognition modes. The side chain of D729 in BARD1 forms a polar interaction with K63 (yellow dashed line).

Figure 7.. Reaction cycle of RNF168

Reaction cycle of RNF168. Various modes of interactions of RNF168, RNF168^R-UbcH5c, RNF168^R-UbcH5c~Ub or RNF168^R-UbcH5c-Ub^B with the NCP or NCP^ub are interpreted as different steps or states of the reaction cycle, leading to H2AK13ub and H2AK15ub signal amplification in chromatin.