Synaptic dysfunction and glial activation markers throughout aging and early neurodegeneration: a longitudinal CSF biomarker-based study

Background Synaptic homeostasis, maintained by microglia and astroglia, is disrupted throughout aging and early on in neurodegenerative diseases. Our aim was to study the relationship between TREM2-dependent microglial reactivity, astroglial response and synaptic dysfunction in two longitudinal cohorts of cognitively healthy volunteers and determine whether this relationship is influenced by AD core biomarkers. Methods We analyzed cross-sectional and longitudinal associations between cerebrospinal fluid levels of soluble TREM2 (sTREM2), astroglial markers (GFAP, S100B), and synaptic markers (neurogranin, α-synuclein) in cognitively unimpaired participants from the Wisconsin Registry for Alzheimer’s Prevention (WRAP) and the Alzheimer’s and Families (ALFA+) cohort. Biomarkers were quantified using validated immunoassays (NeuroToolKit, Roche), with sTREM2 measured using an in-house MSD-based assay in the WRAP cohort. Linear regression and linear mixed-effects models were used, both unadjusted and adjusted for Aβ42 and p-tau. Subgroup analyses were performed based on AT classification, APOE-ε4 status, and median splits of Aβ42/Aβ40 ratio and p-tau, to capture profiles suggestive of early AD-related neuropathogenesis. Results We found significant cross-sectional associations between sTREM2 and α-synuclein, as well as between sTREM2 and S100B, in subgroups exhibiting AD-related biomarker profiles. Longitudinally, lower baseline neurogranin and α-synuclein and higher S100B predicted greater increases in sTREM2 over time independently of AD-related markers in the WRAP cohort (β = −0.02, p = 0.006; β = −0.02, p = 0.01; β = 0.02, p = 0.03, respectively). In ALFA+, lower baseline α-synuclein also predicted a greater subsequent longitudinal increase in sTREM2, but only among individuals with Aβ42/Aβ40 ratio above the median (β = -0.01, p = 0.05). Notably, higher baseline sTREM2 was associated with a smaller longitudinal increase in neurogranin in both cohorts (β = -0.01, p = 0.03 for WRAP, β = -0.01, p = 0.04 in ALFA+). Conclusions Synaptic dysfunction markers at baseline influence the longitudinal dynamics of CSF sTREM2 independently of AD-pathology related biomarkers throughout aging and earliest stages of neurodegeneration. In turn, higher baseline sTREM2 is associated with more stable neurogranin levels over time. These results suggest an independent interaction between synaptic dysfunction and TREM2-dependent microglial activation throughout aging and early neurodegeneration beyond AD pathology.