(Un)targeted hair metabolomics: first considerations and systematic evaluation on the impact of sample preparation

Abstract

The interest in metabolomic studies has rapidly increased over the past few years. Changes of endogenous compounds are typically detected in plasma or urine. However, the use of hair allows for long-term monitoring of metabolomic changes and has recently started being applied to metabolomic studies. Within the proposed study, we aimed for a systematical investigation of different pre-analytical parameters on detected metabolites from different chemical classes in hair. For this purpose, three different parameters were varied: (1) multi-step decontamination (dichloromethane (DCM), acetone, H2O, acetone; H2O, acetone, DCM, acetone; and H2O, methanol/acetone), (2) homogenization (pulverization vs. cutting into snippets), and (3) extraction (acetonitrile (ACN)/buffer pH 4 vs. ACN/H2O vs. ACN/buffer pH 8.5). To include as many metabolites as possible, samples were analyzed by high-resolution time of flight mass spectrometry coupled to liquid chromatography (HPLC-HRMS) and additionally by gas chromatography high-resolution mass spectrometry (GC-HRMS) followed by untargeted-like data processing, respectively. The application of different decontamination procedures yielded similar results, although pointing to a trend towards increased washing-out effects if protic solvents were used as a first washing step. Pulverization of hair samples was favorable in terms of detected and tentatively identified metabolites. Evaluation of extraction solvents showed differences in extraction yield for the majority of investigated metabolites, yet, a prediction of metabolite extraction according to their pKa values was not possible. Overall, successive decontamination with DCM, acetone, H2O, and acetone; homogenization by pulverization; and extraction with ACN/H2O produced reliable results.