Metabolism of Glycerol, Glucose, and Lactate in the Citric Acid Cycle Prior to Incorporation into Hepatic Acylglycerols*

Background: The contribution of glyceroneogenesis to hepatic acylglycerol synthesis is controversial. Results: Exogenous glucose and glycerol contribute to the glycerol backbone of acylglycerols through both direct and indirect pathways. Conclusion: The citric acid cycle plays a major role in acylglycerol synthesis. Significance: A method is presented that measures the direct and indirect contributions to the glycerol backbone by 13C NMR. During hepatic lipogenesis, the glycerol backbone of acylglycerols originates from one of three sources: glucose, glycerol, or substrates passing through the citric acid cycle via glyceroneogenesis. The relative contribution of each substrate source to glycerol in rat liver acylglycerols was determined using 13C-enriched substrates and NMR. Animals received a fixed mixture of glucose, glycerol, and lactate; one group received [U-13C6]glucose, another received [U-13C3]glycerol, and the third received [U-13C3]lactate. After 3 h, the livers were harvested to extract fats, and the glycerol moiety from hydrolyzed acylglycerols was analyzed by 13C NMR. In either fed or fasted animals, glucose and glycerol provided the majority of the glycerol backbone carbons, whereas the contribution of lactate was small. In fed animals, glucose contributed >50% of the total newly synthesized glycerol backbone, and 35% of this contribution occurred after glucose had passed through the citric acid cycle. By comparison, the glycerol contribution was ∼40%, and of this, 17% of the exogenous glycerol passed first through the cycle. In fasted animals, exogenous glycerol became the major contributor to acylglycerols. The contribution from exogenous lactate did increase in fasted animals, but its overall contribution remained small. The contributions of glucose and glycerol that had passed through the citric acid cycle first increased in fasted animals from 35 to 71% for glucose and from 17 to 24% for glycerol. Thus, a substantial fraction from both substrate sources passed through the cycle prior to incorporation into the glycerol moiety of acylglycerols in the liver.

It is well established that the glycerol moiety of triglycerides and other acylglycerols in adipose tissue can be derived directly from glucose (1,2). The role of the citric acid cycle in conversion of pyruvate or equivalent molecules to glycerol for production of acylglycerols was demonstrated in studies of adipose tissue more than 40 years ago (3,4); this process is termed glyceroneogenesis. In the past decade, attention has turned to the sources of the glycerol moiety of acylglycerols in liver (5,6). Unlike adipose tissue, liver has the capacity to phosphorylate free glycerol via glycerol kinase to yield glycerol 3-phosphate (G3P), 2 which then becomes esterified with fatty acids. Thus, it is now generally accepted that the carbons in the glycerol backbone of acylglycerols synthesized in the liver are derived from three potential sources: glucose via glycolysis to the level of the triose phosphates, glycerol via glycerol kinase, or glyceroneogenesis from pyruvate and intermediates of the citric acid cycle (see Fig. 1A).
Recently, studies with labeled water ( 2 H 2 O or 3 H 2 O) have been adapted for assessment of the relative contributions of the various sources to the glycerol backbone. In the presence of labeled water, the number of hydrogen atoms ( 2 H or 3 H) incorporated into glycerol will differ depending on the source of the glycerol moiety (5,6). Several studies using the water tracer method found that glyceroneogenesis contributed significantly to the glycerol moiety in liver triglycerides (7)(8)(9). Glyceroneogenesis was defined as the synthesis of G3P from precursors other than glycerol or glucose, including pyruvate, lactate, alanine, and intermediates of the citric acid cycle (9). According to this definition, conversion of glucose to pyruvate followed by carboxylation to oxaloacetate and subsequent decarboxylation to phosphoenolpyruvate and metabolism back to the glycerol moiety would not be considered glyceroneogenesis. This pathway, illustrated in Fig. 1B, would be considered an indirect pathway from conversion of glucose carbons to the glycerol backbone after passing through the citric acid cycle. The total contribution of glucose to hepatic triglyceride-glycerol, defined as the sum of direct and indirect pathways, has been reported to be modest, ϳ11-28% of total triglyceride-glycerol depending upon the nutritional state. In contrast to the total contribution from glucose, the contribution from glyceroneogenesis was reported to be much larger (ϳ60%) and to be independent of nutritional state (9). Glyceroneogenesis was also reported to be increased in hepatic lipogenesis in the setting of type 2 diabetes as determined using the labeled water tracer method (10).
These observations are not consistent with the conventional concept that glycolysis to the level of trioses is the major source for G3P needed for fatty acid esterification. The conclusion that glyceroneogenesis provided most of the glycerol backbone in hepatic acylglycerol production was based on a technique using 14 C-labeled glucose in combination with tritiated water ( 3 H 2 O) to quantify the contribution of glucose to the glycerol moiety via the citric acid cycle (9). This indirect contribution of [U-14 C 6 ]glucose via lactate was estimated based on the appearance of triglyceride-[2,3-14 C 2 ]glycerol. However, this labeling pattern is not the only isotopomer produced by [U-14 C 6 ]glucose via lactate/pyruvate; triglyceride-[1,2-14 C 2 ]glycerol plus triglyceride-[U-14 C 3 ]glycerol may also be generated during passage through the citric acid cycle. Additional triglyceride-[1,2-14 C 2 ]glycerol formation may be possible because [2,3-14 C 2 ]G3P generated from the citric acid cycle is in the equilibrium of glycerol, a symmetric molecule, producing [1,2-14 C 2 ]G3P. This approach (9) is important because it recognizes that labeled water tracers cannot distinguish glyceroneogenesis from glucose metabolism to the glycerol backbone via the citric acid cycle. However, this method may underestimate the contribution of glucose to the glycerol backbone through the indirect pathway because not all possible glycerol isotopomers are considered.
In addition to glucose, glycerol via glycerol kinase has been believed to be a significant source of the glycerol backbone of acylglycerols in the liver. Nonetheless, the contribution of free glycerol to the glycerol backbone is often not measured or is reported as a minor contribution by use of the water tracer method (8,9). Like glucose, free glycerol could conceivably contribute to the glycerol backbone via cycling through the citric acid cycle. Free glycerol is in equilibrium with triose phosphates and can be metabolized to pyruvate, oxaloacetate, phosphoenolpyruvate, and gluconeogenesis. However, the contribution of free glycerol to the glycerol backbone after metabolism in the citric acid cycle (see Fig. 1C) has not been considered previously.
Because fatty liver, defined as overproduction and storage of hepatic triglycerides, is a major and growing clinical problem (11), it is important to understand the contribution of each nutritional source of carbon to the glycerol backbone. Furthermore, it is important to develop a simple method using stable isotopes to quantify these pathways because studies with 3 H and 14 C are not acceptable for patients. Deuterated water ( 2 H 2 O) can be given to humans (8,10), but as noted, glyceroneogenesis may be overestimated, and the glucose contribution may be underestimated using this technique. Here, we explored an alternative approach to examine the sources of glycerol in hepatic acylglycerols of whole animals using one of three 13 C-enriched substrates: [U-13 C 6 ]glucose, [U-13 C 3 ]glycerol, or [U-13 C 3 ]lactate. 13 C NMR analysis of glycerol hydrolyzed from liver fats enabled us to measure the independent contribution of each substrate to the glycerol backbone and also distinguishes between the direct versus indirect pathway contributions of glucose or glycerol to the glycerol moiety in the livers of whole animals. This study demonstrates that glucose and glycerol are indeed the main sources of the glycerol backbone but that a significant portion of this contribution occurs after metabolism of glucose or glycerol to the level of pyruvate, followed by carboxylation to oxaloacetate and subsequent synthesis to the glycerol backbone. Because the indirect pathway is detected as glyceroneogenesis by labeled water methods, the results from these earlier studies should be reinterpreted considering the possibility of cycling involved in glucose and glycerol contributions.
Animal Studies-The study was approved by the Institutional Animal Care and Use Committee at the University of Texas Southwestern Medical Center. Male Sprague-Dawley rats (300 -350 g) were studied in two different nutritional states. One group had free access to food and water. The other group was fasted for 24 h with free access to water. All animals received an intraperitoneal injection of a mixture of glucose (2 g/kg of body weight), glycerol (0.5 g/kg of body weight), and lactate (0.5 g/kg of body weight) under isoflurane anesthesia. Only one substrate was enriched in 13 C in any given experiment, but all three substrates were present in each experiment. After the injection, rats were placed back into their cage, where they quickly awakened and were allowed free access to water. After 3 h, blood, liver, and skeletal muscle tissues were harvested under sodium pentobarbital anesthesia, and they were further processed for NMR analysis.
Sample Processing for NMR Analysis-Liver tissue (7-8 g) ground to a powder under liquid nitrogen was transferred into a beaker containing CHCl 3 /methanol (2:1, 40 ml). The mixture was stirred for 1 h and filtered using a Whatman filter paper. Deionized water (5 ml) was added, and the mixture was swirled manually for 1 min. The swirled mixture was allowed to settle at room temperature for organic-aqueous layer separation and further centrifuged at a low rpm for clear separation. The upper aqueous layer was aspirated, and the remaining organic layer was dried under a vacuum using a liquid nitrogen trap. The dried residue was dissolved in 4 ml of 1:1 1 N KOH and 90% methanol and incubated for 1 h at 70°C with stirring. After incubation, hexane (8 ml) was added, and the sample was vortexed for 1 min. The mixture was centrifuged at a low rpm to separate the organic-aqueous layers. The upper layer containing fatty acids dissolved in hexane was aspirated. The bottom aqueous layer containing glycerol and glycerol phosphates was eluted through a cation exchange resin (2 ml) with deionized water (15 ml). The eluent was dried and dissolved in 2 H 2 O (160 l) for 13 C NMR acquisition.
Blood, liver, and skeletal muscle tissues were treated with perchloric acid to extract water-soluble components, neutralized with KOH, and centrifuged, and the supernatant was dried. The dried residue was dissolved in 2 H 2 O (160 l) for 13 C NMR acquisition for the analysis of the citric acid cycle intermediates and exchanging pools.
NMR Spectroscopy-All NMR spectra were collected using a Varian INOVA 14.1 T spectrometer (Agilent, Santa Clara, CA) equipped with a 3-mm broadband probe with the observe coil tuned to 13 C (150 MHz). 13 C NMR spectra were collected using a 60°pulse, a 36,765-Hz sweep width, 110,294 data points, and a 1.5-s acquisition time with 1.5-s interpulse delay at 25°C. Proton decoupling was performed using a standard WALTZ-16 pulse sequence. Spectra were averaged over ϳ3000 -7000 scans requiring ϳ3-6 h. All NMR spectra were analyzed using the ACD/Labs PC-based NMR spectral analysis program (Advanced Chemistry Development, Inc., Toronto, Canada).
Statistical Analysis-Data are expressed as means Ϯ S.E. Comparisons between groups were performed using Student's t test. A p value of Ͻ0.05 was considered significant.

Contributions of Exogenous Glucose and Glycerol to the Glycerol Moiety of Acylglycerols via the Citric Acid
Cycle-When a liver exposed to [U-13 C 6 ]glucose or [U-13 C 3 ]glycerol is producing acylglycerols, the appearance of an intact three-carbon [U-13 C 3 ]glycerol backbone in the acylglycerol pool would reflect "direct" formation of G3P from one of these precursors. In contrast, if [U-13 C 6 ]glucose or [U-13 C 3 ]glycerol is first metabolized to pyruvate, oxaloacetate, and the citric acid cycle before forming G3P (Fig. 1, B and C), doubly enriched ([1,2-13 C 2 ]glycerol and [2,3-13 C 2 ]glycerol) and uniformly enriched ([U-13 C 3 ]glycerol) isotopomers would then appear in the acylglycerol pool (Fig. 2). Hence, the appearance of [U-13 C 3 ]glycerol in the acylglycerol pool does not necessarily reflect the direct pathway from [U-13 C 6 ]glucose or [U-13 C 3 ]glycerol. To estimate the contribution of carbon coming solely from the citric acid cycle, [U-13 C 3 ]lactate was included as a third tracer. In this case, any contribution from [U-13 C 3 ]lactate to the glycerol backbone must reflect conversion to [U-13 C 3 ]pyruvate, entry into the citric acid cycle, and exit from the cycle through phosphoenolpyruvate carboxykinase to phosphoenolpyruvate and consequently G3P.
The fate of [U-13 C 3 ]pyruvate through pyruvate carboxylase versus pyruvate dehydrogenase (i.e. acetyl-CoA) was confirmed by inspecting the labeling patterns of the citric acid cycle intermediates and exchanging pools. Fig. 3 shows the 13 C NMR spectra of liver extracts from three groups of rats given a mixture of glucose, glycerol, and lactate (only one enriched in 13 C  or [2,3-13 C 2 ]glutamine in all spectra provides direct evidence for entry of [U-13 C 3 ]pyruvate largely through pyruvate carboxylase ( Figs. 2A and 3). The presence of a small amount of [4,5-13 C 2 ]glutamate in some spectra reflects the flux of carbons into the cycle via pyruvate dehydrogenase (Figs. 2B and 3).

JOURNAL OF BIOLOGICAL CHEMISTRY 14491
(independent of which substrate is enriched with 13 C), then, the ratio can be used to evaluate the total carbon contribution to the glycerol moiety of acylglycerols coming from the citric acid cycle (the "indirect" pathway) in all other experiments.
As an example, the C2 resonance of glycerol isolated from the liver acylglycerols of a fed rat given [U-13 C 6 ]glucose/glycerol/ lactate showed five resonance components: a singlet (S), a doublet (D), and a triplet (T) (Fig. 5A). The singlet was assumed to arise only from the natural abundance endogenous glycerol backbone, so it was not included in further calculations. In this particular spectrum, D/(D ϩ T) ϭ 22%, whereas T/(D ϩ T) ϭ 78%. The doublet component reflects the sum of [1,2-13 C 2 ]glycerol and [2,3-13 C 2 ]glycerol isotopomers and hence could arise only from the indirect pathway of [U-13 C 6 ]glucose via the citric acid cycle. The triplet component reflects only [U-13 C 3 ]glycerol, but this isotopomer could arise from either the direct or indirect pathway. As noted above, because fed animals given glucose/glycerol/[U-13 C 3 ]lactate showed a constant T/D ratio (36/64 ϭ 0.56) in the glycerol C2 resonance, the fraction of the triplet resulting from the indirect pathway in the [U-13 C 6 ]glucose experiment was then estimated at 12% (22% ϫ 0.56). Consequently, the triplet portion from the direct pathway is 78 Ϫ 12% ϭ 66%, whereas the indirect contribution of [U-13 C 6 ]glucose is 34%. The contributions of all other substrates to total glycerol production were determined similarly and normalized to 100%.
Contributions of Exogenous Substrates to Liver Acylglycerols in Fed Animals-The 13 C NMR spectra of extracts of the aqueous layer obtained after hydrolysis of liver fats show well resolved resonances from glycerol (Fig. 4). The 13 C enrichment

. 13 C enrichments in the glycerol moiety derived from liver acylglycerols estimated by glycerol C1 and C3 resonance analysis (left) or glycerol C2 resonance analysis (right).
A, for fed animals, rats given [U-13 C 6 ]glucose/glycerol/lactate (n ϭ 8) and rats given glucose/[U-13 C 3 ]glycerol/lactate (n ϭ 8) had higher 13 C enrichments compared with rats given glucose/glycerol/[U-13 C 3 ]lactate (n ϭ 7). B, for fasted animals, rats given glucose/[U-13 C 3 ]glycerol/lactate (n ϭ 5) had the highest enrichment, followed by rats given [U-13 C 6 ]glucose/glycerol/lactate (n ϭ 5) and glucose/glycerol/[U-13 C 3 ]lactate (n ϭ 6). The singlet (S) was assumed to reflect the natural abundance level of 13 C. The doublet (D) in the glycerol C1 and C3 resonance represents signals from [1,2-13 C 2 ]glycerol, [2,3-13 C 2 ]glycerol, and [U-13 C 3 ]glycerol. It is easy to appreciate that the contribution of each substrate to the glycerol moiety was sensitive to the nutritional state. In the glycerol C2 resonance, the singlet and the central peak of the triplet (T) are overlapped. The area of the singlet can be estimated by subtracting the triplet contribution assuming a 1:2:1 area ratio for the three components of the triplet. Administration of exogenous 13 C-labeled substrates could produce excess [1-13 C 1 ]glycerol or [3-13 C 1 ]glycerol moiety, which would cause underestimation of the actual enrichments (left). #, p Ͻ 0.05; §, p Ͻ 0.01; ¥, p Ͻ 0.001.

Sources of Glycerol in Hepatic Acylglycerols
MAY 17, 2013 • VOLUME 288 • NUMBER 20 in the glycerol moiety of acylglycerols was estimated using two approaches by analysis of the multiplet areas of the C1 and C3 glycerol resonance at 63.5 ppm and the C2 resonance at 73.0 ppm (Fig. 4). Given the low probability of forming either singly enriched [1-13 C 1 ]glycerol or [3-13 C 1 ]glycerol from any of these substrates, it was assumed that the singlet component corre-sponds to natural abundance levels of 13 C (1.1%) and that the doublet component reflects a combination of [U-13 C 3 ]glycerol, [1,2-13 C 2 ]glycerol, and [2,3-13 C 2 ]glycerol isotopomers derived from the labeled substrate provided in each experiment. With this assumption, the area of the doublet component, normalized to 1.1% 13 C in the singlet component, was 2.30 Ϯ 0.19% in fed rats provided [U-13 C 6 ]glucose/glycerol/lactate, 1.74 Ϯ 0.34% in fed rats given glucose/[U-13 C 3 ]glycerol/lactate, and 0.17 Ϯ 0.03% in fed rats given glucose/glycerol/[U-13 C 3 ]lactate (Fig. 4A, left bar). The 13 C enrichment in the glycerol moiety of acylglycerols did not differ significantly in rats supplied with [U-13 C 6 ]glucose versus [U-13 C 3 ]glycerol (p ϭ 0.17) but was significantly higher compared with rats supplied with [U-13 C 3 ]lactate.
A similar result was found in fed animals when the enrichment was estimated based on the glycerol C2 resonance with the assumption of 1.1% 13 C in the singlet component (Fig. 4A,  right bar). Unlike in the C1 and C3 resonance, the singlet in C2 is not well resolved because of overlap with the central peak of the C2 triplet. Nevertheless, the area of the singlet can be estimated by subtracting the triplet contribution, assuming a 1:2:1 area ratio for the three components of the triplet. Using the area of the C2 singlet in the calculation instead of the C1 and C3 singlet, the contribution of each substrate to the glycerol moiety was slightly higher: 2.30 3 2.94% in rats given [U-13 C 6 ]glucose/ glycerol/lactate, 1.74 3 1.87% in rats given glucose/[U-13 C 3 ]glycerol/lactate, and 0.17 3 0.23% in rats given glucose/ glycerol/[U-13 C 3 ]lactate (Fig. 4A, left graph versus right graph).
Unlike the multiplets contributing to the C1 and C3 resonance, which do not distinguish between the direct versus indirect pathway, the C2 resonance is more informative (Fig. 5). Here, the triplet component largely reflects direct formation of acylglycerol-[U-13 C 3 ]glycerol from either [U-13 C 6 ]glucose or [U-13 C 3 ]glycerol, whereas the doublet component can reflect only the indirect formation of the glycerol moiety after passage of the labeled substrate through the citric acid cycle. The results reported from the glucose/glycerol/[U-13 C 3 ]lactate experiment in fed animals (Fig. 4) showed 64% doublet and 36% triplet in the glycerol C2 resonance. This demonstrates that a small amount of the triplet component also arose from metabolism in the citric acid cycle. This amount was considered in the calculation of the direct versus indirect contribution of [U-13 C 6 ]glucose or [U-13 C 3 ]glycerol as described above. Given this correction, 35 Ϯ 4% of the [U-13 C 6 ]glucose carbons passed through the citric acid cycle prior to formation of the glycerol moiety. Similarly, of the glycerol moiety derived from exogenous [U-13 C 3 ]glycerol, 17 Ϯ 1% passed through the citric acid cycle in fed animals (Fig. 5A).
As noted above, the percentages of doubly labeled and uniformly labeled acylglycerols from fasted rats provided glucose/ glycerol/[U-13 C 3 ]lactate were 64 and 36%, respectively. Given the correction using this ratio, the multiplet data of the glycerol C2 resonances derived from fasted animals show that 71 Ϯ 4% of all glucose carbons contributing to the glycerol moiety first passed through the citric acid cycle, whereas 24 Ϯ 1% of all labeled glycerol contributing to the glycerol moiety first passed through the cycle (Fig. 5B).

DISCUSSION
In either fed or fasted animals given a mixture of glucose, glycerol, and lactate, the majority of glycerol in hepatic acylglycerols was derived from glucose and free glycerol. Significant portions of glucose and glycerol contributions occurred after entry into the citric acid cycle, and this fraction was sensitive to the nutritional state. Fasting caused a 2-fold increase in the fraction of acylglycerols derived from glucose via the indirect pathway compared with the fed state. The contribution of lactate to the glycerol moiety was trivial in fed animals, and although it increased somewhat in fasted rats, lactate remained a minor contributor to the glycerol moiety in liver.
Previous studies using the water tracer method to determine the sources of triglyceride-glycerol in liver have noted the possibility of overestimation of glyceroneogenesis as a consequence of metabolism of glucose to pyruvate, followed by synthesis to the glycerol moiety (6). Therefore, Nye et al. (9) complemented the use of 3 H 2 O with the addition of [U-14 C 6 ]glucose to allow correction for the contribution of glucose arising through the citric acid cycle. Triglyceride-glycerol labeled at C2 and C3 was considered in the calculation of cycled glucose. However, other labeling patterns in the glycerol moiety could also arise with passage of glucose carbons through the oxaloacetate pool. [U-14 C 3 ]Pyruvate in liver from glycolysis of [U-14 C 6 ]glucose also results in triglyceride-[U-14 C 3 ]glycerol and triglyceride-[1,2-14 C 2 ]glycerol through the metabolic network involved in the citric acid cycle. Furthermore, [2,3-14 C 2 ]G3P formed from [U-14 C 3 ]pyruvate is in exchange with glycerol, a symmetric molecule, which consequently can become [1,2-14 C 2 ]G3P and eventually triglyceride-[1,2-14 C 2 ]glycerol. The relative amounts of these isotopomers will be sensitive not only to the fraction of glucose carbons entering the cycle via pyruvate carboxylase but also the extent of "backward" scrambling into the symmetric four-carbon intermediates. In this study, we used [U-13 C 3 ]lactate to correct for these pathways, which allowed us to measure the fraction of glucose that passed through the citric acid cycle during glycerol moiety formation. With this correction, the total contribution of glucose to the glycerol moiety was found to be the major source among these three exogenous contributors in fed animals and also an important source in fasted animals, whereas the contribution of exogenous lactate was small in both fed and fasted animals.
We have also shown that the contribution of exogenous glycerol to the glycerol moiety of acylglycerols was important in liver and that glucose was not the only substrate that passed through the cycle. The free glycerol contribution was similar to glucose in fed animals but was almost 2-fold greater than glucose in fasted animals. The contribution of glyceroneogenesis measured by the [U-13 C 3 ]lactate tracer increased in fasted animals, but still, it was the smallest contribution among the sources. Although the indirect contribution of free glycerol was less than that of glucose, it was comparable with the [U-13 C 3 ]lactate contribution under fed (1.87 ϫ 0.17 ϭ 0.32% versus 0.23%) and fasted (3.81 ϫ 0.24 ϭ 0.91% versus 1.31%) conditions. The extensive 13 C labeling in the citric acid cycle intermediates or molecules in exchange with the intermediates confirmed the involvement of the citric acid cycle in the indirect contribution of glucose or free glycerol to the glycerol moiety.
The indirect contribution observed in this study occurred presumably within the liver itself. However, one cannot exclude the possibility of peripheral metabolism of either glucose or glycerol to lactate, followed by glyceroneogenesis in liver. In the case of rats given glucose/[U-13 C 3 ]glycerol/ lactate, the 13 C enrichment found in blood glucose was only 3% in fed animals and 11% in fasted animals by measured summed enrichments of multiple-labeled glucose isotopomers, including [1,2-13 C 2 ]glucose, [2,3-13 C 2 ]glucose, [1,2,3-13 C 3 ]glucose, [4,[5][6][7][8][9][10][11][12][13] C 2 ]glucose, [5,6-13 C 2 ]glucose, [4,5,6-13 C 3 ]glucose, and [U-13 C 6 ]glucose. In rats given [U-13 C 6 ]glucose/glycerol/lactate, the 13 C enrichment found in blood glucose was 40% in fed animals and 60% in fasted animals. Thus, we further examined the possibility of peripheral lactate contribution to the glycerol moiety of acylglycerols in the livers of fasted rats given [U-13 C 6 ]glucose/glycerol/lactate, which had the highest 13 C enrichments in blood glucose. Fig. 6 shows C2 resonances of lactate from liver, circulating blood, and skeletal muscle of a fasted rat given a mixture of [U-13 C 6 ]glucose, glycerol, and lactate along with the C2 resonance of the glycerol moiety of acylglycerols in liver. The fractions of doublets (produced after cycling) in the glycerol moiety and lactate in liver were much higher than those in either blood or skeletal muscle, indicating that the observed doubly labeled molecules in liver were produced primarily through metabolism in the liver itself. Previously, we observed that skeletal muscle did not produce doubly labeled three-carbon units from [U-13 C 3 ]lactate (12). Although [U-13 C 3 ]lactate in liver could be derived from either glycolysis in liver or peripheral metabolism, the combination of isotopomers found in glycerol isolated from liver acylglycerols was most consistent with involvement of the citric acid cycle in liver.
13 C enrichment in the glycerol moiety of acylglycerols was measured by assuming that the singlet of glycerol C1 and C3 or the singlet of C2 arose from natural abundance 13 C. Administration of exogenous 13 C-labeled substrates could produce excess singlet, which would cause underestimation of the actual enrichments. In this study, the singlet of glycerol C1 and C3 could arise from exogenous 13 C-labeled substrates, whereas the singlet of C2 was essentially only natural abundance 13 C in the glycerol moiety of acylglycerols (Fig. 2). This explains why the enrichment based on the C1 and C3 resonance analysis was consistently lower compared with the enrichment based on the C2 resonance analysis (Fig. 4). Nonetheless, compared with the C2 resonance of the glycerol moiety, the simpler multiplet pattern in the C1 and C3 resonance makes it easy to appreciate the degree of enrichments because the singlet reflects mostly natural abundance 13 C, whereas doublets represent signals from all of the multiply labeled glycerol isotopomers (i.e. [1,2-13 C 2 ]glycerol, [2,3-13 C 2 ]glycerol, and [U-13 C 3 ]glycerol), which cannot arise from natural abundance.
In summary, glucose and glycerol are major contributors to the glycerol moiety of acylglycerols in the livers of both fed rats and fasted rats given a mixture of exogenous glucose, glycerol, and lactate. However, significant fractions of both glucose and glycerol contributions occurred by synthesis of the glycerol moiety after metabolism in the citric acid cycle. Interestingly, the indirect contribution of glucose was Ͼ2-fold greater than the direct contribution in fasted rats. In addition, exogenous glycerol also contributed significantly to the glycerol moiety through the citric acid cycle. 13 C NMR analysis with 13 C-labeled substrates is a powerful tool for the study of sources of the glycerol moiety, distinguishing direct and indirect contributions of glucose and glycerol to the glycerol backbone of acylglycerols in liver. This approach is readily applicable in a clinical setting where acylglycerols (transported by VLDL) released from the liver can be sampled from blood.