A lipidomic analysis of CSF from HIV-infected subjects identifies metabolites that predict change in cognitive performance.

Type: Poster
Title: A lipidomic analysis of CSF from HIV-infected subjects identifies metabolites that predict change in cognitive performance
Authors: Haughey N, Mielke M, Bandaru V, Sacktor N, Grant I, Letendre S, Fox H, Chang L, Wozna V, Pardo C, McArthur J
Date: 03-05-2012
Abstract:Background: Approximately 50% of people infected with HIV will develop neurocognitive impairment (NI). A biomarker that identifies those likely to decline or improve in cognitive performance (DCP or ICP) would provide the opportunity for intervention at the earliest stages of NI, and would be useful as a surrogate marker to gage the effectiveness of neuroprotective therapeutics. Methods: We used a lipidomic approach to analyze CSF from 348 subjects (291 HIV+; 57 HIV-) collected from multiple sites including: Johns Hopkins, CHARTER, University of Hawaii and Puerto Rico. Baseline and follow-up CSF (within approximately 1 year) were available for the majority (69.8%) of subjects, and most (85.3%) were taking an ARV regimen at baseline. Initial screening studies lead us to focus on sphingomyelin (SM), ceramide and sterol species as bioactive metabolites that are altered in the setting of HIV-infection in accordance with NI. Multinomial logistic regressions were used to examine the relationship between lipid levels, DCP and ICP. Results: There were no cross sectional associations at baseline between any of the lipid or sterol species measured and NI. Lower SM/ceramide ratios for C24:1 at baseline predicted DCP (p=0.047). The predictive potential of this ratio appeared to be driven by increases of ceramide. For sterols, we found that subjects with lower levels of multiple cholesterol esters (CE) were more likely to exhibit DCP (p=0.046-0.007 depending on the exact CE species). None of the SM or ceramide species predicted ICP. However, subjects with higher baseline levels of two triglyceride species (POS, p = 0.005, and POL, p=0.006) were more likely to exhibit ICP. We next determined if levels of these metabolites changed over time, in association with DCP or ICP. In subjects with DCP, there was evidence for progressive disruptions in multiple long-chain species of ceramides. In those subjects who exhibited ICP, initially elevated triglyceride species were lower at follow-up, suggesting a normalization of triglyceride metabolism. Conclusions: These findings show that early changes in sphingolipid and sterol metabolism predict changes in cognitive performance. These data identify pathways that may be targets for neuroprotective therapy, and the metabolites of these pathways may provide useful biomarkers to identify subjects at increased risk for NI.