Discovery of CRBN-Dependent WEE1 Molecular Glue Degraders from a Multicomponent Combinatorial Library

Abstract

Small molecules promoting protein-protein interactions produce a range of therapeutic outcomes. Molecular glue degraders exemplify this concept due to their compact drug-like structures and ability to engage targets without reliance on existing cognate ligands. While cereblon molecular glue degraders containing glutarimide scaffolds have been approved for treatment of multiple myeloma and acute myeloid leukemia, the design of new therapeutically relevant monovalent degraders remains challenging. We report here an approach to glutarimide-containing molecular glue synthesis using multicomponent reactions as a central modular core-forming step. Screening the resulting library identified HRZ-1 derivatives that target casein kinase 1 α (CK1α) and Wee-like protein kinase (WEE1). Further medicinal chemistry efforts led to identification of selective monovalent WEE1 degraders that provide a potential starting point for the eventual development of a selective chemical degrader probe. The structure of the hit WEE1 degrader complex with CRBN-DDB1 and WEE1 provides a model of the protein-protein interface and ideas to rationalize the observed kinase selectivity. Our findings suggest that modular synthetic routes combined with in-depth structural characterization give access to selective molecular glue degraders and expansion of the CRBN-degradable proteome.

Figure 1.

Expanding the chemical space of CRBN molecular glue degraders. (A) Cartoon representation of CRBN-dependent protein degrader mechanism of action. (B) Schematic design of the multicomponent reaction-based library with incorporation of glutarimide moiety (above) and utilization of the GBB reaction for 1-step HRZ-1 scaffold synthesis. (C) Primary MCR library hits of the antiproliferative phenotypic screening. MOLT-4 cells were incubated with a compound, dimethylsulfoxide (DMSO) as a negative control, or CC-90009 GSPT1 degrader as a positive control at the indicated range of concentrations (n = 3). After 3 days, cell proliferation was evaluated by the Cell-Titer Glo assay.

Figure 2.

WEE1 and CK1α (also known as CSNK1A1) are degraded by compounds 1–3 in a CRBN E3 ligase-dependent manner. (A) Whole cell proteomics analysis of compound 1. (B) Western blot of WEE1 and CK1α in wildtype and CRBN −/− MOLT-4 cells treated for 5 h with 1 μM compounds 1–3. (C) Western WEE1 and CK1α degradation under in MOLT4 cells pretreated with MLN4924 (1 μM) for 1 h, followed by treatment with compound 1 and 2 for 5 h.

Figure 3.

WEE1 degradation selectivity validation. Selective degradation of WEE1 by compounds 6 and 13 were confirmed using HiBiT lytic assay after 5 h incubation. Results of the assay in Jurkat HiBiT-CK1α (A) and HiBiT-WEE1 (B) cells. Compound 10 was used as a positive control in both experiments. Mean HiBiT signal is presented for three replicates (error bars – ± SD). Compound 6 (C) and 13 (D) were incubated with MOLT-4 cells for 5 h and the whole cell proteomics analysis was performed.

Figure 4.

Structural characterization of WEE1 recruitment to CRBN facilitated by compound 10. (A) TR-FRET assay showing biochemical ternary complex induction by the hit compounds at a range of concentrations. (B) Mutation of the G-loop glycine leads to rescue of a HiBiT-WEE1 KD degradation by compound 10. (C) Density map of the DDB1–CRBN–compound 10–WEE1 complex visualized by cryo-EM (left) and magnified WEE1–CRBN interaction region with two possible compound binding modes (right). (D) Comparison of the binding modes of compound 10 with lenalidomide and SJ3149 in the CRBN–kinase interaction pocket.

Figure 5.

Characterization of the HRZ-1 scaffold derivatives CK1α/WEE1 degradation and antiproliferative effect in a range of cancer cell lines. (A–E) Western blots showing consistent and preferential degradation of WEE1 by selective degraders 6 and 13. CC90009 and compound 10 were used as negative and positive controls, respectively. (F,G) Heat maps demonstrating effect of compounds 1–14 on the cancer cells viability using EC₅₀ (compound concentration leading to half-maximal cell death) and the maximum relative cell death achieved at 10 μM treatment. Evaluation of the cellular viability effect of gemcitabine in the presence of (H) selective WEE1 molecular glue and (I) PROTAC degraders. The cell lines tested are indicated above the graphs.