Effects of extraction solvent and sample volume on metabolomic and lipidomic analyses of the human vitreous body

Metabolomic and lipidomic profiling of the human vitreous body is challenged by limited sample availability and a lack of standardized procedures. The objective of this study was to assess the effects of extraction solvent and sample volume on the analytical coverage and repeatability of metabolomic and lipidomic profiling of a single pooled sample of human vitreous body using ultra-high-performance liquid chromatography-high-resolution mass spectrometry (UHPLC-HRMS) and proton nuclear magnetic resonance (1H-NMR) spectroscopy. Vitreous samples were obtained from 19 adults undergoing primary vitrectomy for rhegmatogenous retinal detachment, epiretinal membrane, or macular hole. All samples were homogenized into a single collective pool. Metabolites were extracted using acetonitrile or methanol from 15 to 50 µL samples, while lipids were extracted using Folch, isopropanol, or methyl tert-butyl ether methods from 15 to 125 µL samples. Analytical performance was assessed by molecular coverage and reproducibility (coefficient of variation [CV] < 30%). With methanol extraction, 183–205 metabolites met the CV < 30% criterion across the tested volumes, including 205 metabolites at 25 µL (mean CV, 11.87%), whereas 139–178 metabolites met this criterion with acetonitrile. At low volumes (15–25 µL), 143–184 lipid species met the CV < 30% criteria with Folch extraction, with mean CVs of 16.12–16.21%; 204 lipid species met this criterion at 125 µL. Across native, filtered, and Folch-extracted vitreous preparations, 1H-NMR quantified up to 30 metabolites, with mean CVs of retained spectral buckets below 7%. Extraction solvent and sample volume influence analytical coverage and repeatability in human vitreous metabo-lipidomics, with methanol providing the highest metabolite coverage and the Folch method offering favorable lipid coverage while enabling complementary 1H-NMR analysis at low sample volumes.