Justin Shumate

Glucocorticoids (GCs) coordinate immunity, inflammation, and metabolism through allosteric regulation of the glucocorticoid receptor (GR) transcription factor. GCs are indispensable anti-inflammatory drugs yet linking specific ligand–receptor structural states to specific biological outcomes has remained a major barrier to designing safer, more selective therapies. Using structure-based design, we developed selective glucocorticoid receptor modulators (SGRMs) of immune function by extending a steroidal scaffold from the ligand-binding pocket into an adjacent solvent channel. These SGRMs suppressed T cell pro-inflammatory cytokines and promoted differentiation of memory precursor T cells while showing minimal induction of M2 macrophage polarization or T cell checkpoint proteins PD-1 and CTLA-4, all key targets of immunotherapy. Molecular dynamics simulations revealed that solvent-channel substituents function as a lever arm to drive dynamic oscillations in the steroid core, thereby allosterically tuning GR activity states. Systematic perturbation of immune cells with a graded series of ligands enabled a l igand p erturbation with m achine l earning (LPML) framework to map coregulated responses across cell types and identified effector T cell gene networks tightly coupled with immune checkpoint induction. This approach outlines a general strategy for decoding the logic of allosteric drug action, enabling the rational design of SGRMs with tailored immunomodulatory profiles.

Traditional phenotypic drug discovery platforms suffer from poor scalability and/or a lack of mechanistic understanding of discovered probes. We address this by creating Endo-GeneScreen (EGS), a high-throughput platform that identifies small molecules that regulate endogenous protein levels encoded by a preselected target gene within disease-modeling contexts. Two initial screens identify >40 validated small molecules that boost endogenous neuronal Syngap1 levels, a gene that causes a neurodevelopmental disorder when haploinsufficient. EGS assays also accelerate preclinical development of drug candidates and facilitate mode-of-action deconvolution studies of orphaned probes. SR-1815 represents a fully validated proof-of-concept candidate from the platform. It is a previously unknown drug-like small molecule multikinase inhibitor that regulates splicing of Syngap1 transcripts. It restores SynGAP protein abundance to wildtype levels and mitigates major cellular consequences of Syngap1 loss-of-function. Thus, the EGS platform promotes identification and development of small molecules that alter the abundance of disease-linked proteins in a translationally-relevant context.