Determining the isotopic abundance of a labeled compound by mass spectrometry and how correcting for natural abundance distribution using analogous data from the unlabeled compound leads to a systematic error.

When the isotopic abundance or specific activity of a labeled compound is determined by mass spectrometry (MS), it is necessary to correct the raw MS data to eliminate ion intensity contributions, which arise from the presence of heavy isotopes at natural abundance (e.g., a typical carbon compound contains ~1.1% (13) C per carbon atom). The most common approach is to employ a correction in which the mass-to-charge distribution of the corresponding unlabeled compound is used to subtract the natural abundance contributions from the raw mass-to-charge distribution pattern of the labeled compound. Following this correction, the residual intensities should be due to the presence of the newly introduced labeled atoms only. However, this will only be the case when the natural abundance mass isotopomer distribution of the unlabeled compound is the same as that of the labeled species. Although this may be a good approximation, it cannot be accurate in all cases. The implications of this approximation for the determination of isotopic abundance and specific activity have been examined in practice. Isotopically mixed stable-atom labeled valine batches were produced, and both these and [(14) C6 ]carbamazepine were analyzed by MS to determine the extent of the error introduced by the approach. Our studies revealed that significant errors are possible for small highly-labeled compounds, such as valine, under some circumstances. In the case with [(14) C6 ]carbamazepine, the errors introduced were minor but could be significant for (14) C-labeled compounds with particular isotopic distributions. This source of systematic error can be minimized, although not eliminated, by the selection of an appropriate isotopic correction pattern or by the use of a program that varies the natural abundance distribution throughout the correction.