Fig 2: UBE3A variants encompass a broad landscape of functional effects.a BAR assay screen of 152 UBE3A variants showing benign (gray), weak loss-of-function (light blue), strong loss-of-function (blue), weak gain-of-function (pink), and strong gain-of-function (red) mutations relative to WT UBE3A. Significance was determined using a One-sample t-test (two-tailed) with Benjamini–Hochberg multiple comparisons correction (FDR = 0.05). Exact numbers of experiments and p-values are provided in Supplementary Data 1. Red, strong gain-of-function; Pink, weak gain-of-function; Gray, no change from WT UBE3A; Light blue, weak loss-of-function; Blue, strong loss-of-function. b Distribution of functional classes for variants tested in our screen. c Solution NMR (PDB: 6U19) of the AZUL domain of UBE3A (green) bound to PSMD4 (gold). Variants tested in our screen are shown in red. d Co-crystal structure (PDB: 4GIZ) of the E6BD domain of UBE3A (green) bound to HPV E6 (gold). Variants tested in our screen are shown in red. e Co-crystal structure (PDB: 1C4Z) of the HECT domain of UBE3A (green) bound to the E2 enzyme UBCH7 (gold). Variants tested in our screen (red) included mutations at the catalytic cysteine (C820) and the E2-binding interface (T656 and F690). f Heat map plot showing BAR assay results from UBE3A sites with multiple variants. Values are normalized to WT UBE3A activity. White shading represents WT UBE3A activity levels, blue shading indicates loss-of-function, and red shading indicates gain-of-function. Scale bar shows the percent change relative to WT UBE3A. Exact numbers of experiments and p-values are provided in Supplementary Data 1, *p < 0.05, **p < 0.005, p < 0.0005, One-sample t-test (two-tailed) with Benjamini–Hochberg multiple comparisons correction (FDR = 0.05).

Fig 3: Phospho-ubiquitinated human Parkin activates Parkin in trans. (A) Overview of experiments to detect in trans activation of Parkin. (B) Preparation of phospho-ubiquitinated Parkin. Recombinant HA-Parkin purified from bacteria was subjected to the ubiquitination assay in the presence of recombinant TcPINK1 WT or KD. Phospho-ubiquitination of Parkin was confirmed by western blot using anti-pUb. (C) MBP-Parkin ubiquitination assay using FLAG-ubiquitin, UBE1 and UbcH7 in the presence of HA-Parkin prepared as in (B). Input; MBP-Parkin without HA-Parkin. (D) Detection of Parkin self-activation by Ub-VS in the presence of phospho-ubiquitinated Parkin. MBP-Parkin was incubated with HA-Parkin [prepared as in (B)] for 10 min. MBP-Parkin~Ub; MBP-Parkin with Ub-VS covalently bound to its catalytic cysteine. Two replications were performed for each experiment in (C) and (D), and the representative results are shown.

Fig 4: In vitro reconstitution of STAM1 ubiquitination by AIP4. A, STAM1 ubiquitination by AIP4 was reconstituted in vitro and performed in the presence of increasing concentrations of βarrestin1. Ubiquitination reactions comprised of E1(Ube1, 42 nM), E2 (UbcH7, 350 nM), E3 (AIP4, 48.5 nM), ubiquitin (11.6 μM), DTT (1 mM), ATP (1 mM) plus STAM1 (42 nM), and varying concentrations of βarr1 (0 nM, 20 nM, 40 nM, 80 nM, 160 nM, or 320 nM) in 40 μl. Reactions were incubated for 90 min at 37 °C and terminated with 40 μl 2× sample buffer. Equal volumes were analyzed by 7% SDS-PAGE and immunoblotting with the indicated antibodies. Polyubiquitinated [Ub(n)] and unmodified STAM1 are indicated. Immunoblots are from one representative experiment. STAM1 ubiquitination was quantified from the STAM (B) and ubiquitin (C) immunoblots using densitometry and shown as line graphs. Data were expressed relative to the signal at 300 nM βarr1 and represent the mean ± S.D. from three independent experiments. Curves were fitted by nonlinear regression, Michaelis–Menten (GraphPad Prism). AIP4, atrophin-interacting protein 4; STAM, signal-transducing adaptor molecule.