Metabolism of singly or multiply 13C‐labeled substrates leads to the production of molecules that contain 13C atoms at various positions. Molecules differing only in the number of isotopic atoms incorporated are referred to as mass isotopomers. The distribution of mass isotopomers of many molecules can be measured by gas chromatography/mass spectrometry after chemical derivatization. Quantification of metabolite mass isotopomer abundance resulting from biological processes necessitates correction of the measured mass isotopomer distribution of the derivatized metabolite for contributions due to naturally occurring isotopes of its elements. This correction must take into account differences in the relative natural abundance distribution of each mass isotopomer (skewing). An IBM‐compatible computer program was developed which (i) calculates the natural abundance mass isotopomer distribution of unlabeled and labeled standards given the molecular formula of the derivatized molecule or fragment ion, and (ii) calculates the natural abundance mass isotopomer distribution of the singly and multiply labeled molecule or fragment via non‐linear fitting to the measured mass isotopomer distribution of the unlabeled molecule or fragment. The output of this program is used to correct measured mass isotopomer distributions for contributions from natural isotope abundances and to verify measured values for theoretical consistency. Differences between predicted and measured unlabeled and 13C‐labeled isotopomer distributions for hydroxamate di‐t‐butyldimethylsilyl (di‐TBDMS) derivatized pyruvate were measured. The program was applied to the mass isotopomer distribution of glucose labeled from [U‐13C3]glycerol and of fatty acids labeled from [U‐13C6]glucose and either [2‐13C2] acetate or [U‐13C2]acetate. In some of these cases, the measured mass isotopomer distributions corrected by the program were different from those corrected by the classical technique. Implications of these differences including those on the calculation of glucose production due to gluconeogenesis in isolated perfused rat liver are discussed.

Mass isotopomer distribution analysis allows studying the synthesis of polymeric biomolecules from 15N, 13C-, or 2H-labeled monomeric units in the presence of unlabeled polymer. The mass isotopomer distribution of the polymer allows calculation of (i) the enrichment of the monomer and (ii) the dilution of the newly synthesized polymer by unlabeled polymer. We tested the conditions of validity of mass isotopomer distribution analysis of glucose labeled from [U-13C3]lactate, [U-13C3]glycerol, and [2-13C]glycerol to calculate the fraction of glucose production derived from gluconeogenesis. Experiments were conducted in perfused rat livers, live rats, and live monkeys. In all cases, [13C]glycerol yielded labeling patterns of glucose that are incompatible with glucose being formed from a single pool of triose phosphates of constant enrichment. We show evidence that variations in the enrichment of triose phosphates result from (i) the large fractional decrease in physiological glycerol concentration in a single pass through the liver and (ii) the release of unlabeled glycerol by the liver, presumably via lipase activity. This zonation of glycerol metabolism in liver results in the calculation of artifactually low contributions of gluconeogenesis to glucose production when the latter is labeled from [13C]glycerol. In contrast, [U-13C3]lactate appears to be a suitable tracer for mass isotopomer distribution analysis of gluconeogenesis in vivo, but not in the perfused liver. In other perfusion experiments with [2H5]glycerol, we showed that the rat liver releases glycerol molecules containing one to four 2H atoms. This indicates the operation of a substrate cycle between extracellular glycerol and liver triose phosphates, where 2H is lost in the reversible reactions catalyzed by alpha-glycerophosphate dehydrogenase, triose-phosphate isomerase, and glycolytic enzymes. This substrate cycle presumably involves alpha-glycerophosphate hydrolysis.

The use of distributions of mass isotopomers in glucose from [U-13C]glycerol to estimate fractional rates of gluconeogenesis was examined. [U-13C]glycerol was infused into normal subjects who ingested acetaminophen and fasted for 60 h. Isotopomer distributions were measured by mass spectrometry in blood glucose and in glucuronic acid from urinary acetaminophen glucuronide. The distributions are incompatible with glucose production solely via gluconeogenesis from a single pool of triose phosphates. Rather, with the assumption of a single enriched triose phosphate pool, the distributions indicate, despite the 60 h of fasting, about as much glucose formation from an unlabeled glucose source as from that pool. Therefore the data indicate cellular heterogeneity in glycerol's metabolism, so that two or more pools with significantly different enrichments were the source of the glucose and glucuronic acid. This heterogeneity is related to much greater concentrations of glycerol in periportal than in pericentral zones of the liver lobule. Beyond evidence for heterogeneity, the findings emphasize a limitation in applying analyses of mass isotopomer distributions to measure polymer biosynthesis in the presence of heterogeneity in the precursor pool.

We conducted an extensive mass isotopomer analysis of citric acid cycle and gluconeogenic metabolites isolated from livers of overnight fasted rats perfused with 4 mM glucose, 0.2 mM octanoate, 1 mM [U-13C3]lactate, and 0.2 mM [U-13C3]pyruvate, in the anterograde or retrograde mode. In both perfusion modes, two distinct isotopomer patterns were observed: (i) those of phosphoenolpyruvate, glucose, malate, and aspartate and (ii) those of citrate, alpha-ketoglutarate, glutamate, and glutamine. Key citric acid cycle parameters and, hence, rates of gluconeogenesis, calculated (Lee, W.-N.P. (1989) J. Biol. Chem. 264, 13002-13004 and Lee, W.-N.P. (1993) J. Biol. Chem. 268, 25522-25526) from our mass isotopomer data did not only vary, but lead to conclusions inconsistent with Lee's citric acid cycle model. Compared to lactate and pyruvate uptake, which sets an upper limit to glucose production, rates of gluconeogenesis calculated (i) with the phosphoenolpyruvate and citrate data were similar, but those calculated (ii) with the glutamate data amounted to only 60%, which is unlikely. All these conclusions are independent of the perfusion modes. We provide evidence that the following processes contribute to the observed labeling discrepancy: (i) the reversibility of the isocitrate dehydrogenase reaction and (ii) an active citrate cleavage pathway for the transfer of the oxaloacetate carbon skeleton from mitochondria to the cytosol. Also, a good fit of our labeling data was obtained with a model of citric acid cycle and gluconeogenesis which we developed to incorporate the above reactions (Fernandez, C.A., and Des Rosiers, C. (1995) J. Biol. Chem. 270, 10037-10042). The following conclusions can be drawn from the calculated reaction rates: (i) about half of the lactate conversion to glucose occurs via the citrate cleavage pathway, (ii) the flux through the reversal of the isocitrate dehydrogenase reaction is almost as fast as that through the citrate synthase reaction, and (iii) the flux through citrate synthase and alpha-ketoglutarate dehydrogenase is 1.6- and 3.2-fold that through pyruvate carboxylase, respectively.