Hamilton Se-Hwee Oh

We found that a subset of plasma proteins likely originate predominantly from one organ. We leveraged that organ-protein mapping with machine-learning models to build organ age estimators, or “aging clocks”, for 11 major organs and found that organs age at different rates within and across people, and that an accelerated organ predicts disease in that same organ. In roughly 50,000 UK Biobank participants, we found that accumulating aged organs raised mortality risk step by step, while youthful brain and immune profiles uniquely tracked with longevity.

Amyloid and tau biomarkers can detect which pathology is present but do not reliably predict the rate of cognitive decline in Alzheimer’s disease. Across several major Alzheimer's cohorts, we discovered that synaptic proteins in cerebrospinal fluid were the strongest correlates of cognitive decline, and the ratio between two synaptic proteins, YWHAG and NPTX2, proved a simple robust predictor of decline independent of pathology. We also built plasma proteomics models of cognitive decline in Alzheimer’s as surrogates for cerebrospinal fluid YWHAG:NPTX2.

T cells had been largely overlooked in neurodegeneration relative to neurons and microglia. Using combined single-cell RNA and T-cell receptor sequencing of patient cerebrospinal fluid, we found clonally expanded cytotoxic CD8 T cells in Alzheimer's disease, clonally expanded CD4+ T cells in Lewy body dementia, and rising inflammatory monocytes with normal aging.