AI-guided discovery of a brain-penetrant LILRB4 inhibitor active in microglial and 5xFAD models of Alzheimer's disease.
TL;DR
In human induced pluripotent stem cell-derived microglia, compound 4 reduced SHP1/2 phosphorylation, attenuated NF-κB activation and IL-1β secretion, and restored Aβ uptake without compromising cell viability, and the compound exhibited oral bioavailability and brain penetration.
Abstract
The inhibitory microglial receptor LILRB4 (ILT3) suppresses amyloid-β (Aβ) clearance in Alzheimer's disease (AD) through ApoE-dependent signaling. Here, we report an independent artificial intelligence-guided approach for discovering small-molecule inhibitors of the LILRB4-ApoE interaction. Ultralarge-scale screening of approximately 500 million compounds identified chemically distinct molecules that bind LILRB4 with nanomolar affinity and inhibit ApoE engagement, as validated using orthogonal biophysical and biochemical assays. Structural modeling and mutational analysis defined an interdomain pocket and key residues associated with ligand recognition. In human induced pluripotent stem cell-derived microglia, compound 4 reduced SHP1/2 phosphorylation, attenuated NF-κB activation and IL-1β secretion, and restored Aβ uptake without compromising cell viability. The compound exhibited oral bioavailability and brain penetration. In male and female 5xFAD mice, once-daily oral administration improved Y-maze performance and reduced regional Aβ42 levels, inflammatory cytokines, and CD86-positive microglia. These findings establish retrieval-based artificial intelligence as a complementary route to LILRB4 ligand discovery and identify compound 4 as a pharmacologically differentiated candidate for further optimization.