Fig 2: Structural basis for specific inhibition of UCHL1.a Average distance alignment and schematic representation of domain architecture of human UCH family deubiquitinases. Boundaries of bacterially expressed UCH catalytic domain constructs are given. ULD, UCH37-like domain. b Ubiquitin rhodamine cleavage assay of indicated recombinant UCH DUBs, preincubated with GK13S dilutions for 1 h. Data points are shown as mean ± standard error from N = 2–5 independent experiments, as indicated in the figure. Source data are provided as a Source Data file. c Superposition of UCHL1~GK13S with apo UCHL1 (pdb: 2ETL) and apo UCHL3 (pdb: 1UCH) and close-up view of the binding pocket. Labeled residues are shown as sticks. d Superposition of the equivalent to the compound binding pocket in apo UCHL3 and UCHL3~Ub-VME (pdb: 1XD3). The Ubiquitin C-terminus is shown in cartoon representation. Residues undergoing a conformational change upon Ubiquitin binding are indicated with arrows. e Indicated Flag-UCHL3 constructs with mutations introducing a GK13S binding site were overexpressed in HEK293 cells. Cells were treated with a compound where indicated, and GK13S-bound proteins were visualized by in-gel fluorescence. Uncropped versions of gels and blots are shown in the supplementary information. f Sequence alignment of human Ubiquitin C-terminal Hydrolase (UCH) family members. Secondary structure assignments are based on the UCHL1~GK13S structure. Black arrows indicate residues mutated in e. For a full sequence alignment, see Supplementary Fig. 10a. g Close-up view of the binding pocket of UCHL1~GK13S (left panel) and superposition with full-length UCHL5 (pdb: 3IHR), the catalytic domain of UCHL5 (pdb: 3RII) and UCHL5 in co-complex with an inhibitory fragment of INO80G (pdb: 4UF5) (from left to right).

Fig 3: A proposed model for classical NHEJ. UCHL3 could regulate classical NHEJ by antagonizing Ku80 ubiquitylation in its catalytic activity and phosphorylation dependent manner. See main text for detailed description of the proposed model.

Fig 4: GK13S, but not GK16S, reduces monoubiquitin in U-87 MG cells and thereby phenocopies the effect of a UCHL1 mutant mouse.a Representative microscopy images of U-87 MG cells were taken 72 h after treatment with 1.25 µM of the indicated compounds (including the control Staurosporine, STS). Effects of the compounds on cell growth and viability were assessed using analysis of confluency (bright-field, BF) and apoptosis (propidium iodide staining, PI). See Supplementary Fig. 7a–d for quantitation. b Cellular activity-based protein profiling of intact U-87 MG cells treated with indicated compounds or DMSO for 1 or 24 h. c Schematic representation of overlapping and individual cellular targets of GK13S and GK16S, supporting their application as chemogenomic probes for the investigation of UCHL1. Compare Fig. 2e. d Inhibition of cellular UCHL1. Western blot analysis of endogenous UCHL1 labeled with HA-Ub-VS after treatment of U-87 MG cells with either the indicated compounds or DMSO for 1 or 24 h. e Western blots showing monoubiquitin levels in U-87 MG cells treated with GK13S or GK16S. An additional knockdown of UCHL3 does not aggravate the effect. f Quantification of monoubiquitin intensities upon treatment of U-87 MG cells with DMSO, GK13S, or GK16S (as shown in e). Values correspond to the mean of four or six independent experiments for control or UCHL3 knockdown, respectively. Error values represent the standard error of the mean (s.e.m.). Statistical significance was analyzed using individual one-sample, two-tailed t-tests compared to the mean of “1” as set for the DMSO-treated samples. **p < 0.01 (exact p = 0.0095 for DMSO/GK13S comparison in siScr background and p = 0.0014 for the DMSO/GK13S comparison in the siUCHL3 background); ns, not significant. Uncropped versions of gels and blots are shown in the supplementary information. Source data are provided as a Source Data file.

Fig 5: UCHL3 facilitates classical NHEJ. (A) Schematic representation of a fluorescent reporter assay measuring c-NHEJ efficiency. I-SceI digestion sites and inserted stop codon are indicated by arrow heads and red characters, respectively. (B) U2OS cells transfected with the indicated siRNAs were subjected to c-NHEJ assay. The efficiency of c-NHEJ was normalized to control siRNA-transfected cells and set to 100% (Mean ± SEM, n = 3). (C) U2OS or UCHL3 KO#2 cells transfected with the indicated plasmid coding FLAG (empty vector: EV) or FLAG-UCHL3 were subjected to c-NHEJ assay. The efficiency of c-NHEJ was normalized to EV transfected U2OS cells and set to 100% (Mean ± SEM, n = 3). (D) U2OS cells transfected with the indicated siRNAs were subjected to direct-repeat GFP assay. The efficiency of homology mediated repair was normalized to control siRNA-transfected cells and set to 100% (Mean ± SEM, n = 3). (E,F) U2OS (WT) or UCHL3 KO#1 cells transfected with the indicated siRNAs were processed for immunoblotting analysis (E) or subjected to clonogenic survival assay after IR (F) (Mean ± SEM, n = 3). (G,H) U2OS cells stably expressing FLAG-UCHL3 (wild-type: WT) (G) or catalytically inactive mutant (C95A) (H) were transfected with the indicated siRNAs and subjected to clonogenic survival assay after IR (Mean ± SEM, n = 3). *p < 0.05, ***p < 0.001, N. S.; not significant. Full-length blots are presented in Supplementary Fig. S10.