Su-Lin Zhang

Targeting the FOXP3—T-bet interaction to restore Treg stability in IFN-γ—driven autoimmunity 2254306

Regulatory T cells (Tregs) maintain immune homeostasis through FOXP3-centered transcriptional complexes that tightly control lineage stability and suppressive function. However, how specific FOXP3 mutations disturb this complex and drive pathogenic Treg reprogramming in IPEX syndrome remains unclear. We identified a distinctive mechanism by which the FOXP3 V408M mutation promotes Th1-skewed inflammation and also explored a pharmacological strategy to restore Treg stability. We generated FOXP3 V408M knock-in mice and performed immunophenotyping, transcriptomic, and chromatin conformation analyses to determine how the mutation affects FOXP3—T-bet interaction and Ifng transcription. An AI-driven virtual screening strategy integrating sequence- and structure-based modeling was applied to identify compounds that stabilize FOXP3—T-bet interaction. Functional validation was performed in vitro and in multiple in vivo mouse models. FOXP3 V408M mutation disrupted the FOXP3—T-bet interaction, thereby releasing T-bet from FOXP3-mediated repression and enhancing Ifng transcription. This defect reprogrammed Tregs toward an IFN-γ—producing phenotype that promoted Th1 inflammation. Among the AI-driven screening hits, 430C10 emerged as a first-in-class FOXP3-targeting stabilizer binding an allosteric pocket within the FKH domain. 430C10 reinforced the FOXP3—T-bet interaction and suppressed T-bet—driven IFN-γ production by Tregs. Oral 430C10 treatment markedly alleviated IFN-γ+ Treg—driven inflammation in FOXP3 V408M mice and improved disease outcomes in an acute colitis model under FOXP3 WT settings. Our findings define the FOXP3—T-bet interaction as a tunable checkpoint controlling Treg stability and IFN-γ—driven autoimmunity. Pharmacological stabilization of this interaction with 430C10 provides a proof-of-concept therapeutic strategy for restoring immune homeostasis in IPEX syndrome and related autoimmune diseases. Our research is supported by National Natural Science Foundation of China (82271829, 32130041, 82441047, 82241222); The Innovation Program of Shanghai Municipal Education Commission (21140902900); Noncommunicable Chronic Diseases-National Science and Tech Therapeutic Approaches to Autoimmunity (THER)

Targeting the FOXP3-T-bet interaction to restore Treg stability in IFN-γ-driven autoimmunity.

Development of potent and selective ENPP1 inhibitors based on imidazo[1,2-a]pyrazine and pyrazine scaffolds for cancer immunotherapy.

In this study, new imidazo[1,2-a]pyrazine derivative 15 (YS3074) and pyrazine derivative 17 (YS3375) are identified and optimized as potent and selective ENPP1 inhibitors. Compounds 15 and 17 both exhibited high potency and selectivity toward ENPP1 (IC50 = 4.23 nM and 19.4 nM, respectively) with negligible ENPP2 inhibition (>10 μM) and weak ENPP3 inhibition (>3 μM). Both enhanced cGAMP-induced expression of STING pathway genes (IFNB1, CXCL10, IL6) in a concentration-dependent manner without cytotoxicity, producing stronger and more sustained activation than MV-658. This effect, dependent on the cGAMP-STING axis, was attributed to inhibition of ENPP1-mediated cGAMP hydrolysis. Additionally, compounds 15 and 17 showed no inhibition of the hERG channel (IC50 > 30 μM), indicating a favorable cardiac safety profile. In the CT-26 colorectal cancer model, compounds 15 and 17, administered at 40 mg/kg in combination with an anti-PD-1 antibody, achieved superior tumor growth inhibition (TGI: 73% and 67%, respectively), significantly prolonged survival, and were well tolerated with no observable body weight loss. These results establish 15 and 17 as promising ENPP1 inhibitors with favorable safety profiles, supporting their potential in combination cancer immunotherapy.

Mapping Non-Homologous Pocket Compatibilities to Identify Hidden Drug-Target Relationships: A Pocket Hopping Framework.

Predicting small molecule-protein interactions across nonhomologous proteins remains challenging because shared ligand recognition is often not evident from sequence, fold, or pocket similarity. Here, we introduce pocket hopping, a machine-learning framework that learns residue-level interaction patterns from coligand binding pockets and infers compatibility between nonhomologous pockets for similar chemotypes. Using shared ligands as supervision rather than explicit geometric alignment, pocket hopping identifies pocket relationships that are not readily captured by conventional chemical-, sequence-, or structure-based comparisons. In two case studies, pocket hopping demonstrates broad utility in drug discovery by enabling de novo hit identification and mechanistic interpretation, identifying fedratinib and its analogues as helicase WRN inhibitors. The model also identified the clinical-stage HDAC inhibitor abexinostat as a direct ENPP1 binder and inhibitor, and cellular assays showed enhanced cGAMP-STING signaling under cGAMP stimulation. Together, these results indicate that pocket-level compatibility can complement existing approaches for target identification, hit discovery, and polypharmacology analysis.