Prospective cohort studies and meta-analyses examining the relationship between HDL-cholesterol (C) and stroke risk are discordant and question the value of HDL-C as a marker for stroke risk prediction. However, the relationship between other metrics of HDL, function and proteome, and stroke remain unexplored. We investigated changes in HDL protein composition and function -cholesterol efflux capacity (CEC)- after acute ischemic stroke, and their relationship to long-term functional and neuronal recovery after stroke. Plasma samples were collected from healthy controls (N = 35) and from stroke patients either 24 (early, N = 35) or 96 hours (late, N = 20) after stroke onset. ABCA1 mediated CEC was measured using murine macrophages. Stroke recovery was assessed at 3 months after stroke event using the Modified Rankin Scores (MRS) and the NIH Stroke Scale (NIHSS). Proteomic analysis of HDL by parallel reaction monitoring indicated a distinct time-dependent remodeling after stroke. 15 proteins were significantly altered following stroke, with 6 proteins significantly changing between 24 and 96 h post stroke. Inflammation-related proteins (SAA1, SAA2, PON1) were increased at both time points. Macrophage CEC, consistent with inflammatory remodeling of HDL proteome, was reduced by 50% (P<0.0001) in both early and late post-stroke samples compared to the controls. Changes in 6 post-stroke significantly correlated with stroke recovery scores (P<0.05). Further, the multiple linear regression model adjusting for baseline stroke severity confirmed that these proteins that predict the stroke recovery. Changes in HDL associated proteins within the first 96 hours post stroke could be used as markers to predict functional stroke recovery.

A small portion of HDL digest was pooled within experimental groups and analyzed by data-dependent acquisition. DDA raw files were searched with SEQUEST HT against a human Ensembl protein database using Proteome Discoverer. Results files were used to create a spectral library for subsequent PRM analysis. 0.25 ug of individual patient HDL digest was analyzed by PRM method for 95 peptide precursors for 42 proteins. Fragment peak area was normalized to total ion current and N15 APOA1 peptides.

HDL was isolated from patient plasma by sequential ultracentrifugation. HDL protein, with [N15] APOA1 protein, was reduced, alkylated, and digested by trypsin. Digested peptides were dried down and resuspended in 5% acetonitrile, 0.1% formic acid.