Discovery and Characterization of Small Molecule Drugs to Target Estrogen Receptor Mutants in Breast Cancer
Introduction
The acquisition of estrogen receptor alpha (ERα) gene (ESR1) mutations is a key driver for the development of resistance to current endocrine therapy in breast cancer. Clinical studies have shown that ESR1 mutations are frequently observed in metastatic ER-positive breast cancer patients and are associated with poor survival. It is well appreciated that activating ESR1 somatic mutations, especially Y537S and D538G, can drive estrogen-independent activities. In addition, these ESR1 mutations diminish the potency of the current standard-of-care agents (tamoxifen and fulvestrant) that bind ERα directly. Therefore, it is a critical need to develop next-generation antiestrogens that inhibit ERα mutant signaling in breast cancer to improve patient survival. Here, we search for small molecule inhibitors against ERα mutants Y537S and D538G using DNA-encoded chemical library (DECL) selections.
Methods
ERα ligand binding domain (LBD) proteins corresponding to wild type (WT), Y537S and D538G mutants were expressed in E. coli and purified by Ni-NTA, anion exchange, and size exclusion chromatography. The ability of these purified proteins to bind ligands was tested in biochemical assays to validate their use in DECL selections. We conducted a DNA-encoded chemical library affinity selection using our in-house collection of 6 billion compounds against WT, Y537S and D538G mutants ER ligand binding domain (LBD) proteins in the presence and absence of estradiol. Hits that enriched with these targets were resynthesized off-DNA and tested in biochemical assays. We have performed functional studies with these compounds in wildtype and CRISPR-Cas9 knock-in Y537S or D538G mutant MCF-7 and T47D breast cancer cells.
Results
We have successfully purified microgram amounts of ERα LBD for WT, Y537S, and D538G proteins. The ability of these purified proteins to bind ligands was tested using homogeneous time-resolved fluorescence and fluorescent polarization assays. SRC3 peptide binds to the WT ERα LBD in the presence of estradiol, whereas Y537S and D538G LBDs bind the SRC3 peptide in the absence of estradiol, consistent with these mutants constitutively binding to SRC3. Our multi-billion small molecule collections of DNA-encoded chemical library screens identified several hits in WT and mutant ERα LBDs. To confirm the selection output, we synthesized off-DNA compounds and validated these in biochemical and cell-based studies. Compounds CDD-1274, CDD-1802 and CDD-3209 dramatically decrease the WT and mutant ERα protein levels in many cell types. These compounds greatly reduce protein levels for GREB1, TFF1, c-MYC, E2F1, and survivin in wildtype and mutant breast cancer cells, and they increase levels of p21 and cleaved PARP, a marker for apoptosis in wildtype and mutant ER-positive breast cancer cells but not in ER-negative breast cancer cells.
Conclusion
We have identified potent novel ERα mutant binders using our DNA-encoded chemical library platform. Our compounds are active in biochemical and ERα mutant cell lines, suggesting these molecules are potential chemical probes to explore in in vivo models of breast cancer. In a panel of ER+ breast cancer cell lines were sensitive to CDD-1274, CDD-1802 and CDD-3209, whereas ER-cell lines tested were insensitive. Furthermore, our study rapidly identified novel WT and mutants ERα degrader lead compounds from DNA-encoded chemical library screening with minimal medicinal chemistry efforts.