(Un)targeted hair metabolomics: Systematic evaluation of different sample pretreatment parameters and application to bleached hair

Abstract

Metabolomic studies have gained increasing interest over the past few years. Hair in particular represents an alternative to common matrices (e.g. blood, urine) for the detection of robust biomarkers: it provides a long-term monitoring of endogenous compounds as they are constantly incorporated during hair growth. For an initial biomarker search, usually a non-hypothesis driven approach (untargeted) is applied. Such an approach could be of great interest for example to identify endogenous markers to objectively differentiate cosmetically untreated from treated hair (e.g. bleaching). However, so far, little is known about the ‘hair metabolome’ and its analysis. Therefore, as a prerequisite, knowledge about presence of endogenous metabolites from different compound classes in hair and a suitable sample preparation for their reliable detection is mandatory. Different decontamination protocols (dichloromethane (DCM)/acetone/H 2 O/acetone and H 2 O/acetone/DCM/acetone), homogenization procedures (pulverization/cutting) and extraction solvent mixtures [acetonitrile (ACN)/H 2 O, ACN/buffer pH 4 and ACN/buffer pH 8.5] were tested. Extracts and corresponding wash water solutions were analyzed by liquid chromatography high-resolution (HR) MS and MS/MS (Sciex 6600 TripleTOF) and GC-HRMS (Thermo Scientific Q Exactive), both with untargeted data acquisition. Cosmetically untreated hair samples of one individual were either left untreated or bleached in vitro for 30 min with a mixture of bleaching oxidase (containing 9% H 2 O 2 ) and bleaching oxidation powder, n = 5 per group. Extracts were analyzed by LC-MS/MS. (Tentative) identification of metabolites was performed using Progenesis Qi and PeakView. The detection of different metabolite classes was possible (e.g. fatty acids, amino acids and derivatives, acylcarnitines) using untargeted hair metabolome analysis. Pulverization of hair samples showed higher homogeneity for most metabolites compared to snippets. Extraction with ACN/H 2 O led to a good coverage of metabolites. Comparing untreated and bleached hair samples, a targeted search for 1H-pyrrole-2,3,5-tricarboxylic acid (PTCA) was conducted as it was recently described as a marker for cosmetic hair treatment [1] . PTCA could be detected and identified with a reference standard. A significant increase was observed after bleaching. Additionally, another metabolite formed during treatment could be found in bleached hair samples and was tentatively identified as 1H-pyrrole-2,3,4,5-tetracarboxylic acid (PTeCA). Establishing a suitable sample preparation for hair metabolomic analyses is a challenging task. However, our study showed reliable metabolite detection from different chemical classes. The identification of PTCA in combination with the identification of an additional metabolite formed through bleaching served as a proof of concept for the developed untargeted workflow. Future studies should focus on the confirmation of these findings with a larger number of hair samples.