BGPT: Paper Review: Structural basis of specific H2A K13/K15 ubiquitination by RNF168

Fuel Your Discoveries

Quick Answer Copied

Mechanistic core

RNF168’s RING domain binds the nucleosome acidic patch via an Arg-rich helix, which positions the E2~Ub catalytic geometry so that ubiquitin is placed on H2A K15 (and can accommodate K13). The paper builds this from NMR (dimer + nucleosome), XL-MS, targeted mutagenesis, and an integrative HADDOCK-based model, and contrasts RNF168’s binding orientation with PRC1’s RING1B/BMI1 mode.

Long Answer

Paper Review

Structural basis of specific H2A K13/K15 ubiquitination by RNF168

DOI: 10.1038/s41467-019-09756-z

What the paper claims (mechanistic wiring)

Binding site: RNF168’s RING domain binds directly to the nucleosome H2A/H2B acidic patch using an Arg-rich helix.
Active geometry: Integrative modeling places the E2 catalytic center close to H2A K15 and proposes that nucleosome sterics promote a closed, active E2~Ub conformation compatible with E2~Ub contacts.
Specificity logic: RNF168 uses a nucleosome binding mode and orientation distinct from PRC1’s RING1B/BMI1; a targeted nucleosome mutation (H2B E110A in their numbering) can suppress RNF168-dependent ubiquitination while leaving RING1B-mediated ubiquitination largely unaffected.

Evidence map (from data types to mechanism)

Normalization is qualitative; the “moderate” label for docking reflects that docking/ensembles are model-based and depend on restraints, even though the paper validates the model with mutagenesis and crosslinking constraints.

Detailed critique of the main mechanistic claims

1) RNF168 RING binds the acidic patch (direct evidence)

NMR on H2A/H2B dimers: The paper reports NMR changes (peak intensity changes and chemical shift perturbations/relaxation dispersion effects) that cluster around residues in/near the acidic patch, and interprets line broadening/dispersion as evidence of binding for specific residues (e.g., V45/L103/T116 region in their H2B mapping).
NMR on nucleosomes (methyl-TROSY): Using ILV-labeled nucleosomes, they report fast-exchange CSPs for methyl groups in/around the acidic patch, with an inferred complex lifetime upper limit on the order of ~1 ms.
Mutational functional coupling: Their Arg-rich helix residues are tested by site-directed mutagenesis; for example, R63A reduces nucleosome binding/ubiquitination activity, and R67/68A reduces ubiquitination strongly while altering binding mode.

Skeptical notes / blind spots: NMR indicates direct binding and dynamics, but (as the authors also note) chemical-shift perturbations alone do not fully specify orientation; the paper therefore uses XL-MS to resolve orientation. This is a strength, but it still leaves open whether additional transient binding modes exist at other acidic patch-adjacent surfaces under different conditions.

2) XL-MS provides an orientation constraint

The paper reports an intermolecular crosslink between RNF168 K46 and H2B K105 observed across replicates, and uses this as a single unambiguous distance restraint in HADDOCK docking.
They further report that predicted additional crosslinks were not all observed, and interpret this as plausibility/formation probability differences under restraint geometry and dynamics.

3) E2 positioning and K15 specificity: strong model, but not a fully atomic “for free” structure

The paper describes a ternary modeling pipeline: select a conserved E3–E2~Ub structure, superimpose onto the RNF168–nucleosome complex, then dock UbcH5c while enforcing an E2 activation “linchpin” hydrogen bond (R55 ↔ E2-Q92 backbone in RNF168).
They report a best model where the E2 active site is close to H2A K15 (~4.2 Å Sγ–Nζ distance) and discuss that K13 is dynamic and thus not always present in modeled electron density.
They argue specificity against nearby H2B ubiquitination sites by comparing distances (e.g., reported larger distances to H2B K120/K125 in their modeled solutions).

Skeptical notes / what would be decisive: The paper’s ternary placement relies on docking + restraints; docking does not guarantee that the sampled conformations correspond to catalytically productive complexes in solution. The paper partially addresses this by validating the RNF168–nucleosome orientation with XL-MS and by testing structural-model-guided mutants that alter ubiquitination outcomes, which increases confidence—but an explicit experimentally resolved ternary structure is still absent (they also report that crystallography of RNF168 RING–nucleosome was unsuccessful).

4) Selective uncoupling from PRC1 (functional separability)

The paper reports an ~180° rotation difference between RNF168 and RING1B binding modes relative to the acidic patch, consistent with target lysines being on opposite sides of the nucleosome with respect to the patch.
They test a nucleosome mutation at the periphery of the RNF168 interface (H2B E110A; E113 in humans) and report RNF168-dependent ubiquitination drops strongly while RING1B-dependent ubiquitination is unaffected (in the assays described).

Blind spot: The separation is demonstrated in vitro with reconstituted nucleosomes and recombinant E3/E2 conditions; real chromatin contains additional factors (e.g., remodelers, histone modifications, competitors) that can change productive encounter rates and effective binding landscapes. The paper itself is explicit about focusing on in vitro mechanistic separation, so extrapolation to in-cell specificity requires careful follow-up.

Reproducibility + data availability (strong points)

NMR data deposition: The paper states NMR data are deposited in BMRB under accession codes 27547, 27791, 27792, and 27786.
Proteomics XL-MS deposition: Crosslinking mass spectrometry data are available via ProteomeXchange under accession PXD012723.
Models deposition: Structural models are reported in PDB-Dev with accession codes PDBDEV_00000028 and PDBDEV_00000029.

Figure-by-figure functional purpose (from the text provided)

Figure	Main contribution	Key residues/constraints mentioned
Fig. 1	RNF168 RING binds acidic patch in H2A/H2B dimers (NMR exchange/CSP clustering)	Acidic patch residues clustered; H2B V45/L103/T116 cited as showing dispersion effects; contrast to I51
Fig. 2	RNF168 RING binds nucleosome acidic patch (methyl-TROSY), and Arg-helix mutation functional tests	CSP examples: H2A L64/L92; H2B V45/L103; R63A reduced nucleosome ubiquitination; XL-MS crosslink support
Fig. 3	Integrative E3–nucleosome model + E2 positioning rationale	Crosslink restraint described as RNF168 K46–H2B K105 in text; E2 proximity to H2A K15 (~4.2 Å in model); R67/68A effects
Fig. 4	RNF168 vs PRC1 uncoupling via nucleosome mutation	H2B E110A (E113 human) selectively suppresses RNF168-mediated ubiquitination without affecting RING1B-mediated ubiquitination (per assay)
Fig. 5	Nucleosome promotes closed active E3–E2~Ub conformations (steric clash argument)	Open vs closed E3–E2~Ub compatibility discussed; clashes vs incompatibility
Fig. 6	Binding-mode comparison electrostatics/surface shapes (RNF168 vs RING1B)	Arg-anchor residues and electrostatic surface differences described; explains binding-mode changes

The table is based only on the provided full-text excerpts/figure descriptions in your input.

Known unknowns & what could overturn the mechanism

E2 identity in vivo: The paper models ternary complexes using UbcH5c as the cognate E2 “in vitro”; in cellular contexts, RNF168 could interact with different or additional E2 partners under regulatory constraints, which could alter which lysines are geometrically favored.
Model uniqueness: Docking ensembles are filtered with SASD criteria and clustering, but multiple conformations could still satisfy restraints; the paper notes that NMR alone doesn’t encode orientation and that the XL-MS crosslink is crucial for orientation, implying that if that restraint is condition-dependent, model conclusions could shift.
Physiological context: The selective “uncoupling” via H2B E110A is shown in reconstituted systems and does not automatically guarantee the same selectivity in chromatin with additional modifications/competitors.

Bespoke BGPT deep dives (click to explore)

Author Reviews (BGPT)

Feedback:

Updated: March 23, 2026