Combination of cerebrospinal fluid VAMP-2 and core AD biomarkers reflects episodic memory deficits in sporadic Alzheimer’s disease

Background: Synapse loss is an early event that precedes neuronal death and symptom onset and is considered the best neuropathological correlate of cognitive decline in Alzheimer’s disease (AD). Vesicle-associated membrane protein 2 (VAMP-2) has emerged as a promising biomarker of AD-related synapse degeneration in cerebrospinal fluid (CSF). The aim of this study was to explore the CSF profile of VAMP-2 across the AD continuum in relation to core AD biomarkers, other synaptic proteins, neurogranin (Ng) and synaptosomal-associated Protein-25 kDa (SNAP-25) and cognitive performance. Methods: : We developed a digital immunoassay on the Single Molecule Array platform to quantify VAMP-2 in CSF and used existing immunoassays to quantify Ng, SNAP-25 and core CSF AD biomarkers. The clinical study included 62 cognitively unimpaired AD biomarker negative subjects and 152 participants across the AD continuum from the SPIN cohort (Sant Pau Initiative on Neurodegeneration). Global cognition was measured by the Mini-Mental State Examination (MMSE) and episodic memory by the Free and Cued Selective Reminding test (FCSRT). Statistical methods included χ 2 tests, spearman correlation, and regression analyses. For stepwise regression, age-at-analysis and years of education were forced into the model. Entry of sex, APOE ε4 status, CSF Aβ 1-40 , Aβ 1-42 , Aβ 42:40 ratio, t-tau, p-tau, VAMP-2, SNAP-25 and Ng was conditional on their lowering the Akaike Information Criterion. Results: : The VAMP-2 assay had a good analytical performance (repeatability 8.9%, intermediate precision 10.3%). Assay antibodies detected native VAMP-2 protein in human brain homogenates. CSF concentrations of VAMP-2, neurogranin and SNAP-25 were lower in preclinical AD stage 1 and higher at later AD stages compared to controls and were associated with core AD biomarkers, particularly total tau (adj. r 2 =0.62 to 0.78, p<0.001). The optimal model for association with episodic memory (adj. r 2 =0.56, p<0.001) was observed in participants on the AD continuum and included age (p=0.001), years of education (p=0.001), t-tau (p<0.001), Aβ 42:40 ratio (p=0.02) and VAMP-2 (p<0.001). The same model was the best predictor of MMSE (adj. r 2 =0.27, p<0.001). Conclusions: Our novel digital immunoassay accurately measures VAMP-2 changes in CSF, which serve as a surrogate measure of episodic memory in sporadic AD in combination with core AD biomarkers.