Improved accuracy and precision of gas chromatography/mass spectrometry measurements for metabolic tracers

Abstract

The use of stable-isotope tracer methodology to study substrate metabolic kinetics requires accurate measurement of the tracer to tracee ratio (TTR), often by gas chromatography/mass spectrometry (GC/MS). Many approaches for measurement of the TTR by GC/MS do not use standards of known isotopic enrichment to control for variability in instrument response. In addition, most GC/MS applications exhibit some degree of concentration dependency whereby the measured ion abundance ratio varies with the quantity of sample analyzed, thereby placing a limitation on the accuracy of isotopic enrichment standard curves unless the quantities of standards and samples analyzed are closely matched. We document the degree to which day-to-day variability can affect the instrument response for several GC/MS analyses of metabolic tracers when isotopic enrichment standards are not used to control for variable instrument response. Furthermore, we report a new approach that incorporates concentration dependencies within a standard curve to improve the accuracy and precision of TTR measurements over a range of sample quantities analyzed. The new approach was applied to plasma samples obtained from experimental protocols performed in human subjects with three commonly used tracers: 2H 2-palmitate, 15N 2-urea, and 13C-leucine. Variability in the day-to-day instrument response was 84% and 26% for 2H 2-palmitate and 15N 2-urea, respectively; in addition, up to 10% variability due to concentration dependency was noted for these applications. The new approach virtually eliminated these sources of variability. After controlling for concentration dependency, a threefold reduction in the standard error was noted when the enrichment of 13C-leucine measured by electron-impact (EI) ionization GC/MS was correlated against negative chemical ionization (NCI) GC/MS. These data demonstrate that our new approach decreases the errors in TTR determination caused by variations in instrument response and concentration dependency. This approach is generically applicable, and can improve the accuracy and precision of TTR determinations for most GC/MS analyses.