Comparison of derivatization and chromatographic methods for GC–MS analysis of amino acid enantiomers in physiological samples

Abstract

GC–MS analysis of fluorinated and non-fluorinated chloroformate and anhydride derivatives of amino acid (AA) enantiomers on two different chiral columns was compared for the direct quantification of free l- and d-AAs in human serum and urine in a single analytical run. Best sensitivity was achieved with pentafluoropropionic anhydride/heptafluorobutanol derivatives separated on a Chirasil- l-Val column. However, the occurrence of racemization during derivatization precluded accurate quantification of AA enantiomers. Derivatization with methyl chloroformate/methanol and separation on an Rt-γDEXsa column did not exhibit racemization and yielded ten baseline separated racemates of proteinogenic AAs with resolution values greater than 2.4. However, protein and peptide hydrolysis occurred in serum and urine during the highly exothermal derivatization reaction under alkaline conditions. Removing serum proteins by precipitation before derivatization and performing the reaction at neutral pH enabled the determination of accurate free AA enantiomer concentrations. Accuracy of quantification was validated by an established nonchiral GC–MS method for AA analysis. Reliable quantification was achieved using stable-isotope labeled l-AAs as internal standards. Limits of detection (LOD) and lower limits of quantification (LLOQ) for the d-AAs were in the range of 3.2–446 nM and 0.031–1.95 μM, respectively. Relative standard deviations ( N = 6) for the measurement of AAs in urine and serum ranged from 0.49–11.10% to 0.70–3.87%, respectively. The method was applied to the analysis of urine from 19 patients with renal insufficiency. In comparison to healthy probands, D-ratios of Ala, Val, Pro, Thr, Asp, and Asn were significantly increased.