In Vivo hyperpolarized carbon-13 magnetic resonance spectroscopy reveals increased pyruvate carboxylase flux in an insulin-resistant mouse model


  • Potential conflict of interest: Nothing to report.

  • This study was supported by an intramural funding from the A*STAR Biomedical Research Council.


The pathogenesis of type 2 diabetes is characterized by impaired insulin action and increased hepatic glucose production (HGP). Despite the importance of hepatic metabolic aberrations in diabetes development, there is currently no molecular probe that allows measurement of hepatic gluconeogenic pathways in vivo and in a noninvasive manner. In this study, we used hyperpolarized carbon 13 (13C)-labeled pyruvate magnetic resonance spectroscopy (MRS) to determine changes in hepatic gluconeogenesis in a high-fat diet (HFD)-induced mouse model of type 2 diabetes. Compared with mice on chow diet, HFD-fed mice displayed higher levels of oxaloacetate, aspartate, and malate, along with increased 13C label exchange rates between hyperpolarized [1-13C]pyruvate and its downstream metabolites, [1-13C]malate and [1-13C]aspartate. Biochemical assays using liver extract revealed up-regulated malate dehydrogenase activity, but not aspartate transaminase activity, in HFD-fed mice. Moreover, the 13C label exchange rate between [1-13C]pyruvate and [1-13C]aspartate (kpyr->asp) exhibited apparent correlation with gluconeogenic pyruvate carboxylase (PC) activity in hepatocytes. Finally, up-regulated HGP by glucagon stimulation was detected by an increase in aspartate signal and kpyr->asp, whereas HFD mice treated with metformin for 2 weeks displayed lower production of aspartate and malate, as well as reduced kpyr->asp and 13C-label exchange rate between pyruvate and malate, consistent with down-regulated gluconeogenesis. Conclusion: Taken together, we demonstrate that increased PC flux is an important pathway responsible for increased HGP in diabetes development, and that pharmacologically induced metabolic changes specific to the liver can be detected in vivo with a hyperpolarized 13C-biomolecular probe. Hyperpolarized 13C MRS and the determination of metabolite exchange rates may allow longitudinal monitoring of liver function in disease development. (HEPATOLOGY 2013)

The coordinated actions of insulin and glucagon ensure that glucose homeostasis is maintained across a wide range of physiological conditions. In obesity-associated type 2 diabetes, control of glucose metabolism by these two regulatory hormones is impaired, resulting in hepatic insulin resistance and excessive endogenous glucose production.1 To date, it has not been possible to evaluate this metabolic dysfunction in the liver by a noninvasive in vivo method.

Carbon-13 (13C) magnetic resonance spectroscopy (MRS) has been used to study hepatic gluconeogenesis since the 1980s. However, its inherent low sensitivity has largely limited its application to the study of steady-state metabolism in perfused livers with long acquisition times2 and is thus unsuitable for longitudinal studies. The recent development of hyperpolarized 13C MRS addresses this problem by improving the signal-to-noise ratio (SNR) by more than 10,000-fold,3 making it possible to visualize uptake of 13C labeled pyruvate in the liver and its subsequent metabolic conversion catalyzed by specific enzymes in real time.4, 5

In gluconeogenesis, the conversion of pyruvate into phosphoenolpyruvate (PEP) in the liver is accomplished in two enzyme-mediated steps: anaplerosis of pyruvate into oxaloacetate (OAA) catalyzed by pyruvate carboxylase (PC), followed by conversion of OAA into PEP mediated by PEP carboxykinase (PEPCK). PEPCK is commonly considered the control point for liver gluconeogenesis and its overexpression leads to hyperglycemia. However, deletion of PEPCK reduced gluconeogenic flux by only 40%,6 suggesting that PC may play a more-central role in controlling gluconeogenesis.7

In this study, we investigated the effect of insulin resistance on in vivo gluconeogenic flux in high-fat diet (HFD)-fed mice with real-time measurements of hyperpolarized [1-13C]pyruvate anaplerosis into the tricarboxylic acid (TCA) cycle by PC, as well as its transamination into [1-13C]alanine catalyzed by alanine transaminase (ALT). We also evaluated the sensitivity of hyperpolarized 13C MRS in detecting changes in hepatic glucose production upon pharmacological intervention. Such capability may facilitate longitudinal assessment of therapeutic response in diabetic drug development, as well as a better understanding of the mechanism of action of candidate compounds.


ALT, alanine transaminase; AST, aspartate transaminase; 13C, carbon-13; ChREBP, carbohydrate response element-binding protein; FAS, fatty acid synthase; G-6-Pase, glucose-6-phosphatase; HFD, high-fat diet; HGP, hepatic glucose production; IHTG, intrahepatic triglyceride; IPGTT, intraperitoneal glucose tolerance test; ITT, insulin tolerance test; IV, intravenously; kpyr->ala, 13C-label exchange rate between pyruvate and alanine; kpyr->asp, 13C-label exchange rate between pyruvate and aspartate; kpyr->lac, 13C-label exchange rate between pyruvate and lactate; kpyr->mal, 13C-label exchange rate between pyruvate and malate; kpyr->oaa, 13C-label exchange rate between pyruvate and oxaloacate; LDH, lactate dehydrogenase; MDH, malate dehydrogenase; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy; NAFLD, nonalcoholic fatty liver disease; OAA, oxaloacetate; PC, pyruvate carboxylase; PDH, pyruvate dehydrogenase; PEP, phosphoenolpyruvate; PEPCK, phosphoenolpyruvate carboxykinase; SEM, standard error of the mean; SNR, signal-to-noise ratio; SREBP-1c, steroid regulatory element-binding protein; TCA, tricarboxylic acid.

Materials and Methods

Animal Welfare.

Starting at weaning age of 3 weeks, male C57/Bl6 mice received either a standard chow diet with 6.0% (w/w) fat, 47.0% (w/w) carbohydrates, and 18.0% (w/w) protein, with metabolizable energy of 3.1 kcal/g (Harlan Teklad, Madison, WI), or an HFD containing 34.9% (w/w) fat, 26.3% (w/w) carbohydrates, and 26.2% (w/w) protein, with metabolizable energy of 5.2 kcal/g (Research Diets, Inc., New Brunswick, NJ), for 24 weeks. Mice on these two diets are referred to as Chow-fed and HFD, respectively, in this article. All animals were fasted 24 hours before examination. All procedures involving animals were approved by the A-STAR Institutional Animal Care and Use Committee (nos. 080351 and 090428).

Animal Handling.

Anesthesia was induced with 2.0% isoflurane mixed with medical air. Body temperature was maintained at 37°C. A catheter was inserted into the tail vein for intravenous (IV) administration of the hyperpolarized 13C pyruvate inside the magnetic resonance imaging (MRI) system. Mice were subsequently sacrificed for measurement of plasma metabolites and liver extraction.

Glucose and Insulin Tolerance Tests.

The intraperitoneal glucose tolerance test (IPGTT) and insulin tolerance test (ITT) were conducted as described previously.8 Details are available in the Supporting Materials.

Body-Composition Measurement.

Body compositions were measured with an EchoMRI 100 (Echo Medical Systems, Houston, TX). Body fat mass and body lean mass were measured within 1 minute.

In Vivo Proton MRS Determination of Fat Content in Hepatic Steatosis.

Proton MRS measurements were performed with a 7-T preclinical MRI system (ClinScan; Bruker BioSpin MRI GmbH, Ettlingen, Germany). The voxel of interest was a 64-mm3 volume placed in the right hepatic lobe of the liver, with care taken to ensure exclusion of major blood vessels. Intrahepatic triglyceride (IHTG) content was expressed as a percentage of the fat signal peak area with reference to the combined signal with water9 (see Supporting Materials for more details).

In Vivo Measurements of Hepatic Metabolism With Hyperpolarized 13C.

[1-13C]pyruvic acid (40 mg; Cambridge Isotope Laboratories, Cambridge, MA), mixed with 15 mM of trityl radical (OXO63; GE Healthcare, Amersham, UK) and a trace amount of Dotarem (Guerbet, Birmingham UK), was polarized and dissolved in a hyperpolarizer (Oxford Instruments, Oxford, UK), as described previously.10 Hyperpolarized [1-13C]pyruvate (0.5 mmol/kg body weight) was injected IV over 3 seconds, and 60 individual liver spectra were acquired over 1 minute in a 9.4-T preclinical MRI scanner. MRS quantification and analysis protocols are described in further detail in the Supporting Materials.

Glucagon and Metformin Stimulation.

Chow-fed mice were anesthetized and IV injected with glucagon (20 μg/kg), followed by hyperpolarized 13C MRS measurements 10 minutes later. This interval was chosen specifically to coincide with the maximal increase in blood glucose level (Supporting Fig. 1). HFD mice were treated with metformin (200 mg/kg, once-daily) for 2 weeks. In vivo measurements of hepatic metabolism were performed before and after treatment.

Ex Vivo Biochemical Enzyme Activity Assays.

For each enzyme-activity assay, 100 mg of liver tissue was homogenized in 200 μL of ice-cold 100-mM Tris-HCl buffer, then centrifuged for 10 minutes at 13,000×g to remove insoluble material. All assays were based on continuous spectrophotometric rate determination. Details are available in the Supporting Materials.

Statistical Analysis.

All statistical analysis was performed with the Graphpad Prism software package (GrapPad Software, Inc., La Jolla, CA). Data were presented as means ± standard error of the mean (SEM). Statistical significance in hyperpolarized 13C metabolite signal ratios and ex vivo hepatic enzyme activity comparisons were assessed by using a two-tailed unpaired Student t test. For the correlations between 13C-exchange rates and ex vivo enzyme activities, Pearson's product moment was computed, after which the two-tailed Student t test was used to test for statistical significance. For the IPGTT, ITT glucose blood tests, and insulin serum test, two-way analysis of variance, followed by Bonferroni's post-tests, were used. The significance limit was set at P < 0.05.


Mice on Prolonged HFD Develop Hepatic Steatosis, Glucose Intolerance, and Insulin Resistance.

We first defined the pathophysiological effect of prolonged HFD feeding. HFD-fed mice developed hepatic steatosis, with more than an 8-fold higher IHTG level than Chow-fed mice (Table 1). HFD mice were also hyperglycemic and hyperinsulinemic. Hematoxylin and eosin and Oil Red O histology revealed massive lipid deposits in the hepatocytes of HFD-fed mice (Supporting Fig. S2). Other physical characteristics included a higher body weight and body fat composition, together with a lower body lean content. IPGTT revealed impaired glucose tolerance in HFD mice, as evidenced by delayed glucose clearance at 45, 60, 90, and 120 minutes after infusion (Fig. 1A,B). In addition, there was simultaneous compensatory increase in insulin secretion (Fig. 1C,D). ITT revealed a reduced blood glucose decrease in HFD mice, compared to Chow-fed mice (Fig. 1E,F), indicative of insulin resistance in HFD mice. Together, these results show that HFD mice were glucose intolerant, insulin resistant, hyperglycemic, and hyperinsulinemic, clear indications of pre–type 2 diabetes.11

Table 1. Metabolic Characteristics of HFD-fed mice* (N = 7 in each group)
CharacteristicsChowHFDP Values
  • *

    N = 7 in each group.

Body weight, g36.6 ± 0.956.0 ± 0.9<0.01
Body fat, %14.2 ± 1.939.9 ± 0.8<0.01
Body lean, %81.2 ± 1.658.7 ± 1.5<0.01
Intrahepatic TG, %5.0 ± 0.542.7 ± 5.1<0.01
Fasting glucose, mmol/L6.56 ± 0.429.25 ± 0.13<0.01
Fasting insulin, ng/mL0.37 ± 0.013.94 ± 0.95<0.01
Fed glucagon, pg/mL189.6 ± 12.3266.6 ± 27.1<0.05
Plasma TG, mmol/L1.72 ± 0.042.06 ± 0.05<0.01
Figure 1.

Impaired glucose tolerance, increased insulin secretion, and reduced insulin sensitivity in HFD mice. (A) HFD mice exhibited glucose intolerance, compared to Chow-diet fed mice. Mice were 27 weeks old. N = 8 for each group. (B) Glucose AUC (area under curve) calculated based on data in (A). (C) Plasma insulin levels in HFD (N = 3) and Chow-diet fed mice (N = 10) in IPGTT. (D) Glucose-induced insulin secretion calculated by integrating the AUC after baseline subtraction based on data in (C). (E) ITT revealed that HFD mice (N = 4) exhibited lower insulin sensitivity than Chow-diet fed control (N = 10). (F) Insulin-induced glucose level in plasma calculated by integrating AAC (area above curve) based on data in (E). Data are presented as mean ± SEM. *P < 0.05; **P < 0.01.

Elevated Carbohydrate Anaplerosis in Mice With Hepatic Steatosis as Detected by Hyperpolarized 13C MRS.

We next examined whether metabolic changes in gluconeogenesis could be detected in vivo with hyperpolarized [1-13C]pyruvate. Pyruvate is at a major metabolic junction and generates four metabolite intermediates, each catalyzed by a distinct enzyme or enzyme complex: lactate by LDH (lactate dehydrogenase); alanine by ALT; acetyl-coA by PDHC (pyruvate dehydrogenase complex); and oxaloacetate by PC (pyruvate carboxylase). Because of the abundance of LDH and ALT in the liver, rapid 13C label exchange from [1-13C]pyruvate to [1-13C]lactate and [1-13C]alanine rendered the lactate and alanine the two largest metabolite peaks in the MRS spectrum (Fig. 2A). PDH flux could be assessed by the changes in [1-13C]bicarbonate levels (Fig. 2A,B). The anaplerotic role of pyruvate was observed by its conversion into OAA, a vital intermediate metabolite involved in gluconeogenesis and oxidative phosphorylation. [1-13C]OAA can be rapidly converted into [1-13C]phosphoenolpyruvate, [1-13C]malate, [1-13C]aspartate, and [6-13C]citrate, catalyzed by PEPCK, malate dehydrogenase (MDH), aspartate transaminase (AST), and citrate synthase, respectively. In the MRS spectra, we were able to detect [1-13C]malate and [1-13C]aspartate peaks, consistent with observations in the perfused mouse liver.4 Because the conversion of OAA to malate and aspartate are reversible reactions, there is 13C label exchange between these three metabolites. In addition, reversible dehydration of [1-13C]malate to [1-13C]fumarate, catalyzed by fumarase, resulted in the repositioning of the 13C label between the C1 and C4 positions of fumarate. This, in effect, gave rise to [4-13C]malate, [4-13C]aspartate, and [4-13C]OAA peaks.4, 12 A representative time course displaying the progression of metabolite signals is shown in Fig. 2B. These results show that the major downstream pathways of pyruvate can be monitored with hyperpolarized [1-13C]pyruvate.

Figure 2.

Elevated pyruvate anaplerosis in mice with hepatic steatosis is detected by hyperpolarized 13C MRS. (A) Representation of the in vivo measured hyperpolarized 13C spectra in fatty liver. Metabolites that were detected include [1-13C]lactate (183.0 ppm), [1-13C]malate (181.5 ppm), [4-13C]malate (180.2 ppm), [1-13C]pyruvate hydrate (179.1 ppm), [4-13C]aspartate (177.8 ppm), [1-13C]alanine (176.4 ppm), [4-13C]OAA (175.6 ppm), [1-13C]aspartate (175.0 ppm), [1-13C]pyruvate (170.8 ppm), and [1-13C]bicarbonate (160.8 ppm). (B) Representative time course depicting the simultaneous production of downstream metabolites upon infusion of hyperpolarized [1-13C]pyruvate. (C) Normalized metabolite signal comparison between mice fed on Chow and HFD. (D) Corresponding 13C label exchange rates (N = 8 for each group). Data are presented as mean ± SEM. *P < 0.05; **P < 0.01. Values are displayed in Supporting Table 1.

We next examined the metabolic changes in gluconeogensis in HFD mice. When compared to control mice, the ratios of [1-13C]malate/tCarbon, [4-13C]OAA/tCarbon, [1-13C]aspartate/tCarbon, and [1-13C]alanine/tCarbon were significantly larger in fatty livers of HFD-fed mice (Fig. 2C), whereas no significant change in [1-13C]lactate/tCarbon and [1-13C]bicar/tCarbon ratios was observed, suggesting higher PC and ALT activities, but not LDH or PDH activities. Consistently, the rate of 13C label exchange between [1-13C]pyruvate and [1-13C]alanine, kpyr->ala, was increased by more than 70% in HFD-fed mice, whereas no significant change was detected in its exchange with [1-13C]lactate, kpyr->lac (Fig. 2D). The exchange rate between [1-13C]pyruvate and [1-13C]malate, kpyr->mal, which is dependent on both PC and MDH enzyme catalysis, increased significantly in fatty liver. The exchange between [1-13C]pyruvate and [1-13C]aspartate, kpyr->asp, mediated by both PC and AST enzymes, was also elevated in the steatotic liver. Exchange rate between [1-13C]pyruvate and [4-13C]OAA, kpyr->oaa, showed a more than 3-fold increase in the HFD group, relative to Chow-fed mice. The corresponding time courses over 60 seconds illustrated the relatively faster production of these four-carbon metabolites in the steatotic liver (Supporting Fig. S3). Together, the flux measurements showing increased PC activity in HFD-fed mouse livers suggest that PC may be a central player in enhanced gluconeogenesis in the prediabetic stage.

Increased Malate, Aspartate, and Alanine 13C Metabolite Signals in Fatty Liver.

With the observation that [1-13C]malate and [1-13C]aspartate signals were significantly increased in fatty liver, we next sought to understand the mechanism underlying the changes. Because each pathway involves two mediating enzymes, PC/MDH and PC/AST, respectively, it is essential to distinguish each enzyme's contribution to the 13C metabolite signal. Ex vivo enzyme-activity assays of liver extracts obtained from both HFD- and Chow-fed mice revealed a significant up-regulation of PC activity in fatty liver (Fig. 3A). However, there was no apparent increase in AST activity (Fig. 3B), indicating that the larger [1-13C]aspartate signal was primarily the result of increased PC activity. Hepatic MDH activity, on the other hand, was up-regulated in diabetic mice (Fig. 3C). Therefore, the higher [1-13C]malate signal could be attributed to a combination of increased PC and MDH activities. This combined effect probably led to increased 13C label exchange between OAA, malate, and fumarate, thus contributing to an elevated [4-13C]OAA signal (Fig. 2C). These results further support the critical role of PC in gluconeogensis in the prediabetic stage.

Figure 3.

Up-regulation of gluconeogenic fluxes in fatty liver is accompanied by increased intracellular hepatic enzyme activities. (A) Increased anaplerotic influx of pyruvate was reflected in the elevated enzyme activity of PC. (B) No significant change was observed in AST activity. (C) MDH activity was elevated in steatotic hepatocytes. (D) Augmented gluconeogenic enzyme PEPCK activity corroborated the state of increased gluconeogenesis in the insulin-resistant liver. (E) Abnormal liver function in diabetes was evident in the increased ALT flux (N = 8 for each group). Data are presented as mean ± SEM. *P < 0.05; **P < 0.01.

Another key enzyme in gluconeogenesis, PEPCK, was concomitantly up-regulated in the insulin-resistant liver (Fig. 3D). This further corroborated the observation that elevated pyruvate anaplerosis was required to support the increased hepatic glucose production in diabetic mice. The higher exchange rate between [1-13C]pyruvate and [1-13C]alanine indicated faster transamination, which was confirmed in the biochemical ALT activity assay (Fig. 3E).

Hyperpolarized 13C Metabolic Fluxes as Potential Biomarkers of Liver Function.

We next determined the potential of hyperpolarized 13C metabolic signals as relevant diagnostic biomarkers of liver dysfunction in the diabetic state by examining the relationship between measured in vivo hyperpolarized 13C exchange and actual hepatic enzyme activity. First, the exchange rate between hyperpolarized [1-13C]pyruvate and [1-13C]alanine, kpyr->ala, correlated with ex vivo hepatic ALT activity (Fig. 4A). Because ALT is a key regulatory enzyme in pyruvate recycling and urea production, in vivo kpyr->ala could be used as an important biomarker of liver dysfunction. Second, faster 13C label exchange between [1-13C]pyruvate and [1-13C]aspartate, kpyr->asp, correlated well with higher PC activity in hepatocytes (Fig. 4B). Therefore, kpyr->asp could be a potential biomarker to reflect in vivo gluconeogenic flux in the liver. Together, these results demonstrate that hyperpolarized 13C metabolic signals may be used as relevant diagnostic biomarkers of liver dysfunction, such as in diabetes.

Figure 4.

Correlation between PC and ALT enzyme activities and the corresponding in vivo hyperpolarized 13C metabolic flux biomarkers. Faster [1-13C]pyruvate to (A) [1-13C]alanine, and (B) [1-13C]aspartate exchanges correlate to higher intracellular ALT and PC activities, respectively.

Detecting Changes in Liver Metabolism Induced by Glucagon and Metformin With Hyperpolarized 13C MRS.

To assess the detection sensitivity of hyperpolarized 13C MRS on changes in liver metabolism, we first examined glucagon-induced glucose production in Chow-fed animals. Higher aspartate, bicarbonate, and OAA signals were recorded 10 minutes after IV glucagon injection (Fig. 5A). The corresponding 13C-label exchanges rates (kpyr->asp, kpyr->bic, and kpyr->oaa) were also significantly increased (Fig. 5B). Elevated kpyr->asp and kpyr->oaa are signatures of enhanced hepatic gluconeogenesis (see above), whereas higher kpyr->bic indicates up-regulated pyruvate dehydrogenase (PDH) activity.13 Conversely, metformin treatment successfully reduced hepatic gluconeogenesis, as evidenced by the significantly lower malate and aspartate signals, and the abatement of their corresponding exchange rates, kpyr->mal and kpyr->asp (Fig. 5C,D). Blood glucose level was decreased by 24% as well (Supporting Table 3). These results show that hyperpolarized 13C MRS appears to be sufficiently sensitive for measurement of induced metabolic changes in the liver.

Figure 5.

Hyperpolarized 13C MRS detects changes in hepatic gluconeogenesis upon pharmacological intervention. (A) Significantly higher [1-13C]aspartate and [4-13C]OAA signals and (B) exchange rates were detected 10 minutes after glucagon stimulation. (C) Significantly lower [1-13C]malate and [1-13C]aspartate signals and (D) exchange rates were detected in metformin-treated mice (N = 5 for each group). Data are presented as mean ± SEM. *P < 0.05; **P < 0.01. Values are displayed in Supporting Tables 2 and 3.


Novelty in Measuring Liver Metabolism In Vivo.

Although it is recognized that glucose homeostasis maintained by the tissue trio (muscle, liver, and fat) is disturbed in diabetes,14 it has not been possible to detect and measure the underlying hepatic metabolic aberrations noninvasively in real time. In this study, we demonstrate, for the first time, the novel use of hyperpolarized 13C MRS to quantify and assess enzyme fluxes specific to the liver in a type 2 diabetes mouse model in vivo. By measuring gluconeogenic fluxes, we identify PC and that its downstream MDH activities are up-regulated and suggest the PC pathway as a critical component in the development of hyperglycemia and diabetes.

Heightened Gluconeogenesis in Diabetes Drives Pyruvate Anaplerosis.

Through validation with spectrophotometric assays of liver tissue extracts, we demonstrate that in hepatic steatosis, the larger [1-13C]aspartate metabolite signal may be attributed to a higher PC flux, whereas the increased [1-13C]malate signal is a combined effect of increased PC and MDH activity. Because both [1-13C]aspartate and [1-13C]malate are derived from [1-13C]oxaloacetate through anaplerosis from [1-13C]pyruvate, and a larger [4-13C]oxaloacetate pool is indeed detected simultaneously, we hypothesize that the increase in these downstream metabolite pools is a manifestation of elevated demand for the intermediate OAA, in response to higher anabolic rates of gluconeogenesis and fatty acid synthesis (FAS) (Fig. 6). A few factors may contribute to this phenomenon in fatty liver, as described below.

Figure 6.

Demand for intermediate metabolite OAA drives higher anaplerotic pyruvate carboxylase flux. Anaplerotic flux through PC is augmented in fatty liver as a result of increased demand for the intermediate metabolite, OAA. This phenomenon can be attributed to a series of molecular events triggered by (1) insulin resistance, (2) dominant glucagon effect, and (3) obesity-induced de novo FAS (see text). Metabolites in red boxes can be detected by hyperpolarized 13C MRS.

Insulin resistance.

Insulin insensitivity in the fatty liver is detrimental to the hormone's inhibitory role in gluconeogenesis, primarily through the inactivation of the phosphatidylinositol 3-kinase/serine/threonine kinase–signaling pathway,15 thereby enfeebling the suppression of key gluconeogenic enzymes PEPCK and glucose-6-phosphatase (G-6-Pase) expression.14 In addition, previous studies utilizing radioisotopic analysis also showed that carboxylation of pyruvate into OAA is up-regulated in the diabetic rat liver, concomitant with dramatic increases in PC,16 PEPCK, and G-6-Pase15 expression. These studies corroborate our finding that both PC and PEPCK enzyme activities are increased in the fatty liver, leading to larger 13C-malate, -aspartate, and -OAA signals as well as higher rates of chemical exchange with pyruvate. Indeed, higher hepatic PC activity correlated with increased PEPCK activity (r2 = 0.82; P < 0.0001) (Supporting Fig. 4), further supporting the hypothesis that both PC and PEPCK are important regulators in gluconeogenesis.7

Increased glucagon effect.

In diabetes, pathological alteration of the precise balance between insulin and glucagon action results in excessive hepatic gluconeogenesis and glycogenolysis, both of which induce hyperglycemia. Moreover, inadequate suppression of postprandial glucagon secretion by insulin in the diabetic state causes hyperglucagonemia and evokes elevated HGP, as observed in HFD mice. We previously reported that combined defects in insulin secretion and signaling were not sufficient to cause hyperglycemia in the absence of dysregulated glucagon secretion in a mouse model with deletion of calcium-sensing protein synaptotagmin-7.17 Indeed, glucagon plays a major role in promoting gluconeogenesis in enhancing G-6-Pase activity and PEPCK transcription in the liver, likely through the protein kinase A–signaling cascade mechanism.18 Thereafter, up-regulated gluconeogenesis increases the demand for OAA. In this work, we demonstrated up-regulated PC activity in glucagon-stimulated HGP in Chow-fed animals, as detected in vivo with hyperpolarized 13C MRS, through the biomarker kpyr->asp. Concomitantly, glucagon increases PDH activity.19 This technology appears to possess sufficient sensitivity to detect this phenomenon as well, as evident from the higher kpyr->bic exchange rate. Treatment with a glucagon-receptor antagonist appears to alleviate HGP in the diabetic liver,20 and reducing glucagon signaling is being explored as a potential therapy for diabetes.21 It will be interesting to measure corresponding changes in hepatic metabolism upon therapeutic intervention with a glucagon-receptor antagonist in diabetic animals, and that forms the next phase of our research.

Elevated FAS.

De novo FAS in the liver is regulated by three known transcription factors: steroid regulatory element-binding protein (SREBP-1c)22; carbohydrate response element-binding protein (ChREBP)23; and peroxisome proliferator-activated receptor gamma.24 Hyperinsulinemia induces hepatic SREBP-1c expression while hyperglycemia stimulates ChREBP activity. These events lead to transcriptional activation of all lipogenic genes including adenosine triphosphate citrate lyase, acetyl-CoA carboxylase, and FAS,25 effectively increasing FAS flux. Because citrate formed in the TCA cycle and shuttled to the cytosol is the primary metabolite required in the production of fatty acids, there is inevitably an increase in demand for this intermediate. Therefore, it is unsurprising that a recent 13C isotopomer study found a 2-fold increase of hepatic TCA cycle flux in patients with nonalcoholic fatty liver disease.26 Because only pyruvate that enters the TCA cycle through PC produces a net increase in cycle intermediates, whereas pyruvate entering through PDH is restricted to energy production only,27 the elevated PC flux and OAA pool observed in diabetic mice must have also catered to the increased FAS demand. This agrees with our recent observation in the hypertrophied heart, in which a larger 13C-citrate signal (from increased pyruvate anaplerosis) was recorded.28 Moreover, detection of citrate pool with hyperpolarized [2-13C]pyruvate substrate has recently been demonstrated to be feasible in the study of myocardial TCA flux29; therefore, similar measurements in the insulin-resistant liver will undoubtedly aid in validating the hypothesis that an enlarged citrate pool supports FAS. Metformin is used clinically to counter elevated FAS and gluconeogenesis in diabetes, primarily through its activation of adenosine-monophosphate–activated protein kinase.30 In this study, we demonstrated that metformin treatment leads to reduced HGP by, at least in part, decreasing PC activity, as well as production of malate and aspartate from pyruvate.

The advent of hyperpolarized 13C MRS has enabled visualization of real-time metabolism in the in vivo mouse liver, in particular, the anaplerosis of pyruvate into the TCA cycle. The distinct patterns in downstream metabolite progression suggest that hyperpolarized 13C MRS is sensitive to subtle differences in metabolic conversions. It is worth noting that LDH-, ALT-, MDH-, and AST-mediated conversions are reversible. Therefore, the appearance of lactate, alanine, malate, aspartate, and OAA peaks resulted from the enzyme-mediated exchange of the hyperpolarized 13C label, in which equilibrium is dependent on the concentrations of both substrate and product, as well as the redox potential. Hence, these metabolite signals reflect the concentration of each metabolite that already exists within the cellular environment and in the plasma, rather than net metabolite production.31 Nonetheless, the validity of hyperpolarized 13C exchange rates to investigate metabolic changes in pathology has been demonstrated in cardiac diseases28, 32 and cancer.33

Biomarkers of Gluconeogenesis: kpyr->asp, kpyr->mal.

Although both malate and aspartate signals were increased in the fatty liver, kpyr->asp, rather than kpyr->mal (data not shown), correlated linearly to PC activity. In addition, kpyr->asp appeared to be more sensitive in detecting glucagon-induced up-regulation in gluconeogenesis because it increased significantly upon glucagon injection. However, with metformin treatment, the changes in gluconeogenesis were large enough to manifest in both kpyr->asp and kpyr->mal. This confounding phenomenon warrants further investigation, in particular, the influence of MDH activity. In addition, from our experience, the metabolite peaks measured in vivo are generally lower in the fed state than fasted, in part because of the abundance of other alternative metabolic substrates in the circulation (e.g., unlabeled pyruvate), which limits the uptake of the infused hyperpolarized 13C-labeled pyruvate. This observation is observed in the lower metabolite peaks in the spectrum of fed mice (Supporting Fig. 3A).

In summary, we demonstrate the application of hyperpolarized 13C MRS in probing metabolic events in the liver and its correlation with enzyme activities. We identify an important role of the PC pathway in the development of hyperglycemia and diabetes. We also demonstrated the capability of this technique to probe changes in hepatic metabolism upon therapeutic intervention, paving the way for longitudinal assessment. The significant correlation between 13C exchange rates and hepatic enzyme activities illustrates the potential of these indices as biomarkers of liver function in diabetes and a wide range of other diseases.


The authors thank Dr. Hongyu Li for technical assistance in the glucose and insulin tolerance tests.