Profiling of urinary amino-carboxylic metabolites by in-situ heptafluorobutyl chloroformate mediated sample preparation and gas chromatography–mass spectrometry

•Concurrent 1,1,1,2,2,3,3-heptafluorobutyl chloroformate mediated metabolite derivatization and liquid-liquid microextraction (LLME) for urinary GC–MS analysis in-situ described.•Reaction products of 153 diagnostic urinary metabolites treated with HFBCF systematically investigated.•Sample preparation protocol tested for calibration and validation of 132 urinary metabolites.•112 amino-carboxylic metabolites directly quantified in a 25 μL urine sample aliquots.•The described method proved suitable for urinary GC–MS profiling.

A novel 1,1,1,2,2,3,3-heptafluorobutyl chloroformate reagent (HFBCF) was examined for in-situ derivatization of amino-carboxylic metabolites in human urine. The arising reaction products exhibit greatly reduced polarity which facilitates combining the derivatization and liquid-liquid microextraction (LLME) from an aqueous urine into an isooctane phase and immediate gas chromatographic–mas spectrometric analysis (GC–MS). The sample preparation protocol is simple, proceeds without an alcohol excess and provides cleaner extracts than other urinary GC–MS based methods. Moreover, thiol metabolites bound in disulfide bonds can be released by reduction with tris(3-hydroxypropyl)phosphine (THP) prior to the developed derivatization and LLME step. In order to evaluate potential of the novel method for GC–MS metabolomics, reaction products of 153 urinary metabolites with HFBCF, particularly those possessing amino and carboxyl groups (56 amino acids and their conjugates, 84 organic acids, 9 biogenic amines, 4 other polar analytes) and two internal standards were investigated in detail by GC–MS and liquid chromatography-mass spectrometry (LC–MS). One hundred and twenty metabolites (78%) yielded a single product, 25 (16%) and 2 metabolites (2-methylcitrate, citrate) generated two and more derivatives. From the examined set, analytically applicable products of 5 metabolites were not detected; the derivatives of 3 metabolites were only suitable for LC–MS analysis. Electron ionization (EI) of the examined analytes contained characteristic, diagnostic ions enabling to distinguish related and isomeric structures. The new method was validated for 132 metabolites using two internal standards in artificial urine and with special attention to potential disease biomarker candidates. The developed sample preparation protocol was finally evaluated by means of a certified organic acid standard mixture in urine and by GC–MS analysis of 100 morning urines obtained from healthy patients (50 males and 50 females), where 112 physiological metabolites were quantified in a 25 μL sample aliquot. The quantification data for the set were satisfactory, most metabolites were found within the range reported in the reference human metabolome (HMDB) database and literature. The reported results suggest that the described method has been a novel promising tool for targeted GC–MS based metabolomic analysis in urine.