Exploration of the tunability of BRD4 degradation by DCAF16 trans-labelling covalent glues

Abstract

Chemically induced proximity modalities such as targeted protein degradation (TPD) hold promise for expanding the number of proteins that can be manipulated pharmacologically. However, current TPD strategies are often limited to proteins with preexisting ligands. Molecular glues (e.g. glutarimide ligands for CUL4^CRBN), offer the potential to target undruggable proteins. Yet, their rational design is largely unattainable due to the unpredictability of the 'gain-of-function' nature of the glue interaction upon chemical modification of ligands. We recently reported a covalent trans-labelling glue mechanism which we named 'Template-assisted covalent modification', where an electrophile decorated BRD4 inhibitor was effectively delivered to a cysteine residue on DCAF16 due to an electrophile-induced BRD4-DCAF16 interaction. Herein, we report our efforts to evaluate how various electrophilic modifications to the BRD4 binder, JQ1, affect DCAF16 recruitment and subsequent BRD4 degradation efficiency. We discovered a moderate correlation between the electrophile-induced BRD4-DCAF16 ternary complex formation and BRD4 degradation. Moreover, we show that a more solvent-exposed warhead presentation optimally recruits DCAF16 and promotes BRD4 degradation. The diversity of covalent attachments in this class of BRD4 degraders suggests a high tolerance and tunability for the BRD4-DCAF16 interaction. This offers a new avenue for rational glue design by introducing covalent warheads to known binders.

Fig. 1.

A) The chemical structures of the BRD4 inhibitor JQ1 and its related analog GNE011. B) Investigating various cysteine reactive electrophilic handles. C) A schematic depiction of a general covalent trans-labelling event and its relevance to our previous study showing the degradation of BRD4 by covalently trans-labelling DCAF16.

Fig. 2.

A) A schematic depiction of the BRD4_BD2-eGFP and mCherry reporter assay and B) an AlphaScreen assay in the context of BRD4 degraders. C) Dose-dependent BRD4_BD2 degradation by covalent JQ1 analogs as determined by the BRD4_BD2-eGFP and mCherry reporter assay. D) The dose-dependent inhibition of BRD4_BD2 by covalent JQ1 analogs as determined by an AlphaScreen assay. E) The chemical structures of the covalent JQ1 degradation tails included in the covalent SAR. Non-covalent analogs are not included. Note: MMH287, and MMH2 were not tested for AlphaScreen BRD4_BD2.

Fig. 3.

A) A schematic diagram of the BRD4_BD2-DCAF16 TR-FRET assay. B) A 6 h TR-FRET assay for all covalent JQ1 analogs (520 nm/490 nm). C) A graph of TR-FRET maximum ratio vs. time for all covalent JQ1 analogs at 1, 2, 4, 6, and 24 h incubation times. D) A correlational analysis between the TR-FRET Ratio (520 nm/490 nm) and the % BRD4_BD2 degradation for all JQ1 analogs.

Fig. 4.

Cellular activity of the covalent BRD4 degraders. Western blotting analysis of BRD4 degradation by covalent analogs in K562 cells with A) WT and DCAF16 KO, and B) washout after 4 h of treatment (24 h total). C) A schematic depiction of the HiBiT system for quantification of protein abundance. D) Dose-dependent degradation of BRD4 by covalent JQ1 analogs in a BRD4-HiBiT JURKAT cell line. dBET6 is used as a positive control.