Fig 1: Confirming the labeling of STARD7 and STING.A, AlphaFold protein structure of mouse STARD7 highlighting sites of labeling (blue and purple) and the canonical lipid-binding pocket (light purple) (UniProt: Q66JV0, AF-Q66JV0-F1-v4). B, In-gel fluorescence of HEK293T cells transiently overexpressing FLAG-tagged GFP or STARD7 treated with various probes tested (10 μM, 1 h). Data presented are from n = 2 independent experiments. C, Overlapping SEE-CITE labeled proteins by 4c (100 μM)/ 5c (100 μM) in ‘Fig. 5b’, 3 (4 μM)/ 5c (20 μM) in ‘Extended Data Fig. 7C’ and 5c (10 μM)/ 5c (100 μM) in ‘H’ with 5c-enriched proteins from protein-directed AfBPP in ‘Extended Data Fig. 6F’ in K562 cells. Treatment time is 1 h. D, Mapping sites of labeling for STING by 5c (PDB: 6NT5 for STING dimer). E, Scatter plots of STING-modified peptides in HEK293T cells transiently overexpressing STING treated with 2a (20 μM, 1 h) or 5c (100 μM, 1 h). Data are average log2(H/L) values (n = 2 biological replicates) including singletons imputed using H/L ratio of 20. Colored dots are shown for STING labeled peptides. F, Immunoblotting to validate of STING-HA expression as a dimer (~75 kDa) and a monomer (~37 kDa) in HEK293T cells treated with 2a (20 μM, 1 h) or 5c (100 μM, 1 h). G, In-gel fluorescence analysis of the same labeled proteome in ‘F’. The mock sample was prepared separately. H, Waterfall plot using average log2(H/L) values (n = 3 biological replicates) and the standard error of the mean (SEM) values for peptides modified by 5c (10 μM or 100 μM, 1 h) in K562 cells excluding singletons. Colored dots are shown for COX5A (red)-, FECH (yellow)- and RTN4 (blue)- labeled peptides. I, AlphaFold protein structure of human RTN4 with probe-labeled peptide harboring C1101 residue highlighted (AlphaFold ID: AF-Q9NQC3-F1). For ‘F,G’, data presented are from n = 1 independent experiment. All MS data can be found in Supplementary Table 9. Source data
Fig 2: SEE-CITE and functional studies confirm activity and labelling sites for asciminib binders.a, General SEE-CITE-based AfBPP workflow. b, Scatter plots of peptides labelled by 4c (100 μM, 1 h) and 5c (100 μM, 1 h) in K562 cells. Data are average log2(H/L) values (n = 2 biological replicates) including singletons imputed using H/L ratio of 20. Coloured dots indicate probe-modified peptides for selected high-value targets. c, Heatmap comparing average log2(H/L) values for SEE-CITE-identified sites of labelling for selected targets in b and those in K562 cells labelled with 5c (20 μM, 1 h) or 3 (4 μM, 1 h). d,e, In-gel fluorescence analyses of HEK293T cells overexpressing wild-type (WT) RTN4, mutant C1101S-RTN4 labelled with 5c (100 μM, 1 h) (d) or in a competitive manner with asciminib (0.5 h) and 5b (1 μM, 0.5 h) (e) with GFP-HEK293T cell line as a control. f, A schematic representation of NOGO66–RTNR interaction, assayed with AP fusion proteins. g,h, Quantification of Nogo66–RTNR interaction via AP binding assays in HEK293T cells stably overexpressing RTNR incubated with AP fusion proteins pre-treated with asciminib (10 μM, 15 min) (g) or overexpressing GFP or RTNR incubated with AP, wild-type (WT) Nogo66 or C1101W-Nogo66 proteins (h). i, Docking of asciminib into the SEE-CITE COX5A labelling region (T79-Y82). j–l, In-gel fluorescence analyses of HEK293T cells transiently overexpressing FLAG-tagged GFP, wild-type COX5A or mutant COX5A in a competitive manner with asciminib and/or dasatinib (1 h) and 5b (j) or 5c (k) (10 μM, 1 h) or in a UV-dependent manner with 5c (10 μM, 1 h) (l). m, Immunoblot analysis of HEK293T cells transiently overexpressing FLAG-tagged GFP or COX5A (WT and mutants). n, FLAG immunoprecipitation (IP) with HEK293T transiently overexpressing FLAG-tagged GFP or COX5A (WT and mutants). o, State 3 respiration in plasma membrane-permeabilized HeLa cells stably overexpressing GFP–HA or COX5A–HA with or without asciminib treatment (10 μM, 4 h). Cells were offered pyruvate/malate as respiratory substrates. p, A schematic representation depicting effects of asciminib on COX5A and mitochondria functions. For g, h and o, data presented are mean ± s.d., with P values calculated using two-tailed unpaired and paired Student’s t-tests, respectively, using n = 3 biological replicates (g and h) and n = 6 biological replicates (o) (from left to right: P = 0.651571 and 0.000313 (g); P = 0.019546, 0.402216, 0.018749, 0.000398, 0.000959 and 0.000051 (h); P = 0.0461, 0.180008, 0.270314 and 0.007380 (o)) with *P < 0.05, **P < 0.01,***P < 0.001, and NS (not significant) P > 0.05. For d, e and j–m, data are presented for n = 3 independent experiments. NS, not significant; WT, wild type; OE, overexpression. All MS data can be found in Supplementary Table 9. Illustrations in a, f and p created in BioRender; Ngo, C. https://biorender.com/bxsja9h (2026).Source data
Fig 3: Characterization of COX5A as a bona fide off-target candidate of asciminib.A, Structural representation of COX5A-labeled sites (G(T)79, Y80 and D81) in close proximity with a defined lipid-binding pocket with phosphotidylsulfocholine (PSC) ligand (PDB: 1V54 for bovine heart cytochrome C oxidase structure). B-C, In-gel fluorescence ‘B’ and immunoblot analysis ‘C’ of HEK293T cells transiently overexpressing FLAG-tagged GFP or COX5A treated with 5b (1 μM, 10 μM, 1 h) in ‘A’ or asciminib/ dasatinib (10 μM, 1 h) in ‘B’ in a UV-dependent manner. D, AP-MS analysis of HEK293T cells transiently overexpressing FLAG-tagged GFP, wildtype (WT) and mutant COX5A, showing protein intensities (n = 3 biological replicates). Data presented are mean ± SD. E, Co-immunoprecipitation of COX5A-HA overexpressing HEK293T stable cells that were transiently transfected with COX5A-FLAG, treated with asciminib or dasatinib (10 μM, 1 h). F, Mean ratios of assembled complex IV to Complex II assessed by Blue Native Polyacrylamide Gel Electrophoresis (BNGE) for HEK293T cells treated with asciminib or dasatinib (10 μM, 6 h). Quantitative assembly blots were shown in ‘Supplementary Fig. 7F’. Data presented are mean ± SD with significance values between experimental conditions calculated using paired student’s t-test (n = 6 biological replicates; from left to right, p = 0.008397 and 0.336080) with *p < 0.05, **p < 0.01,***p < 0.001, and NS (not significant) p > 0.05. For ‘B,C,E’, data presented are representative n = 3 (B,C) and n = 2 (E) independent experiments. All MS data can be found in Supplementary Table 9. Source data
Supplier Page from DNASU for COX5A (Homo sapiens) in pDONR221 (Gateway donor/master vector)