Single-cell sequencing analysis and machine learning model reveal aberrant TIM-3 expression in microglia during Alzheimer’s disease progression

BACKGROUND: Alzheimer’s disease (AD) is characterized by progressive neurodegeneration and neuroinflammation, with microglial dysfunction playing a pivotal role in its pathogenesis. The specific function of the immune checkpoint molecule TIM-3 (encoded by HAVCR2) in microglia remains unclear. METHODS: We performed an integrated multi-omics analysis combining bulk and single-cell RNA sequencing datasets from AD patients. Bioinformatics approaches, including gene set enrichment, cell-cell communication, and pathway analysis. Furthermore, machine learning algorithms were used to stratify AD patients into molecular subtypes based on HAVCR2 expression. Finally, network pharmacology analysis was conducted to identify potential therapeutic compounds targeting TIM-3 + microglia. RESULTS: We identified a novel TIM-3 + microglial subset that drives pathogenesis through a dual mechanism: (i) sustained neuroinflammation via TNF-α/NF-κB and IL-6/JAK/STAT3 signaling, and (ii) impaired amyloid-β clearance linked to PI3K-AKT and FcγR pathways. TIM-3 + microglia displayed strong interactions with known AD risk genes such as INPP5D and TREM2. Clinically, machine learning classified AD patients into two molecular subtypes: a “hyper-inflammatory” subtype and a “metabolically dysregulated” subtype, revealing distinct pathological drivers. Network pharmacology predicted tretinoin (ATRA) as a candidate drug targeting the CD86–PTPRC–ITGAX axis in TIM-3 + microglia. CONCLUSIONS: This study reveals the immunopathological role of TIM-3 + microglia in AD and highlights their contribution to neuroinflammation and impaired protein clearance. The identification of molecular subtypes and a potential drug candidate (ATRA) provides a framework for precision medicine strategies targeting microglial heterogeneity in Alzheimer’s disease.