A Biotin Targeting Chimera (BioTAC) System to Map Small Molecule Interactomes in situ

Abstract

Unbiased chemical biology strategies for direct readout of protein interactome remodelling by small molecules provide advantages over target-focused approaches, including the ability to detect previously unknown targets, and the inclusion of chemical off-compete controls leading to high-confidence identifications. We describe the BioTAC system, a small-molecule guided proximity labelling platform, to rapidly identify both direct and complexed small molecule binding proteins. The BioTAC system overcomes a limitation of current approaches, and supports identification of both inhibitor bound and molecular glue bound complexes.

Figure 1 |. The BioTAC system enables rapid, accurate small molecule target-ID.

A. Schematic depicting the components of the BioTAC system. B. Example molecule, Cpd 1, annotated with key functional groups. C. Immunoblot analysis of BRD4 enrichment following treatment of HEK293 cells transiently transfected with miniTurboFKBP12^F36V with the indicated compounds and 100 µM biotin at the 30 min timepoint. Data representative of n = 2 biologically independent experiments (SI Figure 1J). D. Scatterplot displaying relative FC of streptavidin-enriched protein abundance following treatment of HEK293 cells transiently transfected with miniTurboFKBP12^F36V with the indicated compounds and 100 µM biotin at the 30 min timepoint. Only proteins with P-value < 0.05 in both conditions depicted. Complete datasets in Table S1, plotted individually Figure S3. E. F-test analysis of proteomic data depicted in D., showing significant enrichment of BRD3 and BRD4 in the presence of 1 µM Cpd 1, relative to DMSO and (+)-JQ1 off-compete experiments. F-statistic listed top right, X = scaled abundance ratio.

Figure 2 |. The BioTAC system enables rapid, accurate small molecule interactome-ID.

A. Time-course proteomics of streptavidin-enriched biotinylated proteins isolated from HEK293 cells transiently transfected with miniTurboFKBP12^F36V and treated with the indicated compounds and 100 µM biotin, demonstrating enrichment and competition of known direct targets (30 min) and complexed proteins (60 min, 4 hrs). Only proteins with P-value < 0.05 in both conditions depicted. Complete datasets in Table S1, plotted individually Figure S3. High-confidence hits are defined as those that are enriched > 2-fold in both Cpd 1/ DMSO and Cpd 1 / Cpd 1 + 10x (+)-JQ1, where P < 0.05, plotted upper-right quadrant. B. Fold-change of all reference BRD2, BRD3, and BRD4 interactors (Lambert et. al.) enriched in the Cpd 1 vs. DMSO dataset at the 4 hr time point, showing statistically significant rescue (P < 0.0001, 1-way ANOVA). C. Percent enrichment of reference interactors (Lambert et. al.) in the hits identified at the 4 hr timepoint (dotted line), vs. the percent enrichment of known interactors in 4,000 random protein sets of equivalent size from the DMSO control (gray bars), showing significant enrichment by Cpd 1 BioTAC (> 7 standard deviations away from mean of random chance). D. Gene Ontology analysis, showing significant enrichment of biological processes associated with known BET-protein function in the 4 hr Cpd 1 BioTAC experiment hits.

Figure 3 |. The BioTAC system enables detection of non-degrader molecular glue interactions.

A. Schematic depicting how the BioTAC system can detect molecular glue interactions. B. Chemical structure of Trametinib recruiting bifunctional molecule JWJ-01-280-1/Cpd 2. C. Immunoblot analysis of MEK1 and KSR1 following treatment of HEK293 cells transiently transfected with miniTurboFKBP12^F36V with the indicated compounds and 100 µM biotin at the 4 h timepoint, and streptavidin-based enrichment, showing successful enrichment and competition with trametinib. D. Immunoblot analysis of mKSR1 following treatment of HEK293 cells transiently transfected with miniTurbo-FKBP12^F36V and mKSR1, with the indicated compounds and 100 µM biotin at the 4 h timepoint, and streptavidin-based enrichment, showing successful enrichment and competition with trametinib. In C, D two KSR1 isoforms are observed, produced by alternative splicing. KSR1-L (102 KDa) corresponds to the expected MW of Uniprot Q8IVT5-1 (canonical sequence), KSR1-S (87 KDa) corresponds to the expected MW of variant with residues 1–137 missing, Uniprot Q8IVT5-3, and −4. Our data indicate both isoforms can complex with trametinib-bound MEK1.