Simultaneous determination of natural-abundance δ15N values and quantities of individual amino acids in proteins from milk of lactating women and from infant hair using gas chromatography/isotope ratio mass spectrometry.

RATIONALE: In isotope tracer experiments used in nutritional studies, it is frequently desirable both to determine the (15)N/(14)N ratios of target compounds and to quantify these compounds. This report shows how this can be achieved in a single chromatographic run for protein amino acids using an isotope ratio mass spectrometer. METHODS: Protein hydrolysis by acidic digestion was used to release amino acids, which were then derivatized as their N-pivaloyl-O-isopropyl esters. Suitable conditions for sample preparation were established for both hair and milk proteins. The N-pivaloyl-O-isopropyl esters of amino acids were separated by gas chromatography (GC) on a 60 m ZB-WAX column linked via a combustion interface to an isotope ratio mass spectrometer. The (15)N/(14)N ratios were obtained from the m/z 28, 29 and 30 peak intensities and the quantities from the Area All (Vs) integrated peak areas. RESULTS: It is shown from a five-point calibration curve that both parameters can be measured reliably within the range of 1.0 to 2.0 mg/mL for the major amino acids derived from the hydrolysis of human maternal milk or hair samples. The method was validated in terms of inter-day and inter-user repeatability for both parameters for 14 amino acids. The amino acid percentage composition showed a good correlation with literature values. The method was applied to determine the variability in a population of lactating mothers and their infants. CONCLUSIONS: It has been established that δ(15)N values can be simultaneously determined for a complex mixture of amino acids at variable concentrations. It is shown that the percentage composition obtained correlates well with that obtained by calculation from the protein composition or from literature values. This procedure should provide a significant saving in analysis time, especially in those cases where the GC run-time is necessarily long. It allows the satisfactory determination of the variation within a sample population.