MedicineLakex

 

In vivo hyperpolarized carbon13 magnetic resonance spectroscopy reveals increased pyruvate carboxylase flux in an insulinresistant mouse model

In Vivo Hyperpolarized Carbon-13 Magnetic Resonance Spectroscopy Reveals Increased Pyruvate Carboxylase Flux in an Insulin-Resistant Mouse Model Philip Lee,1,2 Waifook Leong,1,3 Trish Tan,1,3 Miangkee Lim,1 Weiping Han,1,3,4 and George K. Radda1 The pathogenesis of type 2 diabetes is characterized by impaired insulin action andincreased hepatic glucose production (HGP). Despite the importance of hepatic metabolicaberrations in diabetes development, there is currently no molecular probe that allows mea-surement of hepatic gluconeogenic pathways in vivo and in a noninvasive manner. In thisstudy, we used hyperpolarized carbon 13 (13C)-labeled pyruvate magnetic resonance spec-troscopy (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 micedisplayed higher levels of oxaloacetate, aspartate, and malate, along with increased 13C labelexchange rates between hyperpolarized [1-13C]pyruvate and its downstream metabolites,[1-13C]malate and [1-13C]aspartate. Biochemical assays using liver extract revealed up-regu-lated malate dehydrogenase activity, but not aspartate transaminase activity, in HFD-fedmice. Moreover, the 13C label exchange rate between [1-13C]pyruvate and [1-13C]aspartate(kpyr->asp) exhibited apparent correlation with gluconeogenic pyruvate carboxylase (PC) ac-tivity in hepatocytes. Finally, up-regulated HGP by glucagon stimulation was detected byan increase in aspartate signal and kpyr->asp, whereas HFD mice treated with metformin for2 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 glu-coneogenesis. Conclusion: Taken together, we demonstrate that increased PC flux is animportant pathway responsible for increased HGP in diabetes development, and that phar-macologically induced metabolic changes specific to the liver can be detected in vivo with ahyperpolarized 13C-biomolecular probe. Hyperpolarized 13C MRS and the determinationof metabolite exchange rates may allow longitudinal monitoring of liver function in diseasedevelopment. (HEPATOLOGY 2013;57:515-524) Thecoordinatedactionsofinsulinandglucagon has not been possible to evaluate this metabolic dys- ensure that glucose homeostasis is maintained function in the liver by a noninvasive in vivo method.
across a wide range of physiological conditions.
Carbon-13 (13C) magnetic resonance spectroscopy In obesity-associated type 2 diabetes, control of glucose (MRS) has been used to study hepatic gluconeogenesis metabolism by these two regulatory hormones is since the 1980s. However, its inherent low sensitivity impaired, resulting in hepatic insulin resistance and ex- has largely limited its application to the study of cessive endogenous glucose production.1 To date, it steady-state metabolism in perfused livers with long Abbreviations: 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, intraperitonealglucose tolerance test; ITT, insulin tolerance test; IV, intravenously; kpyr->ala, 13C-label exchange rate between pyruvate and alanine; kpyr->asp, 13C-label exchangerate between pyruvate and aspartate; kpyr->lac, 13C-label exchange rate between pyruvate and lactate; kpyr->mal, 13C-label exchange rate between pyruvate andmalate; kpyr->oaa, 13C-label exchange rate between pyruvate and oxaloacate; LDH, lactate dehydrogenase; MDH, malate dehydrogenase; MRI, magnetic resonanceimaging; MRS, magnetic resonance spectroscopy; NAFLD, nonalcoholic fatty liver disease; OAA, oxaloacetate; PC, pyruvate carboxylase; PDH, pyruvatedehydrogenase; 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.
From the 1Singapore Bioimaging Consortium, Singapore; 2Clinical Imaging Research Centre, National University of Singapore, Center for Translation Medicine, Singapore; 3Metabolism in Human Diseases, Institute of Molecular and Cell Biology, Singapore; and 4Department of Biochemistry, Yong Loo Lin School of Medicine,National University of Singapore, Singapore.
Received April 16, 2012; accepted August 8, 2012.
This study was supported by an intramural funding from the A*STAR Biomedical Research Council.
HEPATOLOGY, February 2013 acquisition times2 and is thus unsuitable for longitudi- before examination. All procedures involving animals nal studies. The recent development of hyperpolarized were approved by the A-STAR Institutional Animal 13C MRS addresses this problem by improving the sig- Care and Use Committee (nos. 080351 and 090428).
nal-to-noise ratio (SNR) by more than 10,000-fold,3 Animal Handling. Anesthesia was induced with making it possible to visualize uptake of 13C labeled 2.0% isoflurane mixed with medical air. Body temper- pyruvate in the liver and its subsequent metabolic con- ature was maintained at 37
C. A catheter was inserted version catalyzed by specific enzymes in real time.4,5 into the tail vein for intravenous (IV) administration In gluconeogenesis, the conversion of pyruvate into of the hyperpolarized 13C pyruvate inside the magnetic phosphoenolpyruvate (PEP) in the liver is accomplished resonance imaging (MRI) system. Mice were subse- in two enzyme-mediated steps: anaplerosis of pyruvate quently sacrificed for measurement of plasma metabo- into oxaloacetate (OAA) catalyzed by pyruvate carboxyl- lites and liver extraction.
ase (PC), followed by conversion of OAA into PEP Glucose and Insulin Tolerance Tests. The intraperi- mediated by PEP carboxykinase (PEPCK). PEPCK is toneal glucose tolerance test (IPGTT) and insulin toler- commonly considered the control point for liver gluco- ance test (ITT) were conducted as described previously.8 neogenesis and its overexpression leads to hyperglyce- Details are available in the Supporting Materials.
mia. However, deletion of PEPCK reduced gluconeo- Body-Composition Measurement. Body composi- genic flux by only 40%,6 suggesting that PC may play a tions were measured with an EchoMRI 100 (Echo more-central role in controlling gluconeogenesis.7 Medical Systems, Houston, TX). Body fat mass and In this study, we investigated the effect of insulin body lean mass were measured within 1 minute.
resistance on in vivo gluconeogenic flux in high-fat In Vivo Proton MRS Determination of Fat Con- diet (HFD)-fed mice with real-time measurements of tent in Hepatic Steatosis. Proton MRS measurements hyperpolarized [1-13C]pyruvate anaplerosis into the tri- were performed with a 7-T preclinical MRI system carboxylic acid (TCA) cycle by PC, as well as its trans- (ClinScan; Bruker BioSpin MRI GmbH, Ettlingen, amination into [1-13C]alanine catalyzed by alanine Germany). The voxel of interest was a 64-mm3 vol- transaminase (ALT). We also evaluated the sensitivity ume placed in the right hepatic lobe of the liver, with of hyperpolarized 13C MRS in detecting changes in care taken to ensure exclusion of major blood vessels.
hepatic glucose production upon pharmacological Intrahepatic triglyceride (IHTG) content was expressed intervention. Such capability may facilitate longitudi- as a percentage of the fat signal peak area with refer- nal assessment of therapeutic response in diabetic drug ence to the combined signal with water9 (see Support- development, as well as a better understanding of the ing Materials for more details).
mechanism of action of candidate compounds.
In Vivo Measurements of Hepatic Metabolism With Hyperpolarized 13C. [1-13C]pyruvic acid (40mg; Cambridge Isotope Laboratories, Cambridge, MA), Materials and Methods mixed with 15 mM of trityl radical (OXO63; GE Animal Welfare. Starting at weaning age of 3 Healthcare, Amersham, UK) and a trace amount of Dot- weeks, male C57/Bl6 mice received either a standard arem (Guerbet, Birmingham UK), was polarized and dis- chow diet with 6.0% (w/w) fat, 47.0% (w/w) carbohy- solved in a hyperpolarizer (Oxford Instruments, Oxford, drates, and 18.0% (w/w) protein, with metabolizable energy of 3.1 kcal/g (Harlan Teklad, Madison, WI), or [1-13C]pyruvate (0.5 mmol/kg body weight) was injected an HFD containing 34.9% (w/w) fat, 26.3% (w/w) IV over 3 seconds, and 60 individual liver spectra were carbohydrates, and 26.2% (w/w) protein, with metab- acquired over 1 minute in a 9.4-T preclinical MRI scan- olizable energy of 5.2 kcal/g (Research Diets, Inc., ner. MRS quantification and analysis protocols are New Brunswick, NJ), for 24 weeks. Mice on these two described in further detail in the Supporting Materials.
diets are referred to as Chow-fed and HFD, respec- Glucagon and Metformin Stimulation. Chow-fed tively, in this article. All animals were fasted 24 hours mice were anesthetized and IV injected with glucagon Address reprint requests to: Philip Lee, Ph.D., Singapore Bioimaging Consortium, 11 Biopolis Way, #02-01 Helios, 138667 Singapore. E-mail: philip_lee@sbic.
a-star.edu.sg; fax: þ65 64789957.
Copyright V C 2012 by the American Association for the Study of Liver Diseases.
View this article online at wileyonlinelibrary.com.
DOI 10.1002/hep.26028Potential conflict of interest: Nothing to report.
Additional Supporting Information may be found in the online version of this article.
HEPATOLOGY, Vol. 57, No. 2, 2013 (20 lg/kg), followed by hyperpolarized 13C MRS Table 1. Metabolic Characteristics of HFD-fed mice* (N 5 7 measurements 10 minutes later. This interval was cho- sen 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 metabo- lism were performed before and after treatment.
Intrahepatic TG, % Fasting glucose, mmol/L Ex Vivo Biochemical Enzyme Activity Assays. For Fasting insulin, ng/mL each enzyme-activity assay, 100 mg of liver tissue was Fed glucagon, pg/mL homogenized in 200 lL of ice-cold 100-mM Tris-HCl Plasma TG, mmol/L buffer, then centrifuged for 10 minutes at 13,000g *N ¼ 7 in each group.
to remove insoluble material. All assays were based oncontinuous spectrophotometric rate determination.
HFD mice were glucose intolerant, insulin resistant, Details are available in the Supporting Materials.
hyperglycemic, and hyperinsulinemic, clear indications Statistical Analysis. All statistical analysis was per- of pre–type 2 diabetes.11 formed with the Graphpad Prism software package Elevated Carbohydrate Anaplerosis in Mice With (GrapPad Software, Inc., La Jolla, CA). Data were pre- Hepatic Steatosis as Detected by Hyperpolarized 13C sented as means 6 standard error of the mean (SEM).
MRS. We next examined whether metabolic changes Statistical significance in hyperpolarized 13C metabolite in gluconeogenesis could be detected in vivo with signal ratios and ex vivo hepatic enzyme activity com- hyperpolarized [1-13C]pyruvate. Pyruvate is at a major parisons were assessed by using a two-tailed unpaired metabolic junction and generates four metabolite inter- Student t test. For the correlations between 13C- mediates, each catalyzed by a distinct enzyme or exchange rates and ex vivo enzyme activities, Pearson's enzyme complex: lactate by LDH (lactate dehydrogen- product moment was computed, after which the two- ase); alanine by ALT; acetyl-coA by PDHC (pyruvate tailed Student t test was used to test for statistical sig- dehydrogenase complex); and oxaloacetate by PC (pyru- nificance. For the IPGTT, ITT glucose blood tests, vate carboxylase). Because of the abundance of LDH and insulin serum test, two-way analysis of variance, and ALT in the liver, rapid 13C label exchange from followed by Bonferroni's post-tests, were used. The sig- [1-13C]pyruvate to [1-13C]lactate and [1-13C]alanine nificance limit was set at P < 0.05.
rendered the lactate and alanine the two largest metabo-lite peaks in the MRS spectrum (Fig. 2A). PDH fluxcould be assessed by the changes in [1-13C]bicarbonate levels (Fig. 2A,B). The anaplerotic role of pyruvate was Mice on Prolonged HFD Develop Hepatic Steato- observed by its conversion into OAA, a vital intermedi- sis, Glucose Intolerance, and Insulin Resistance. We ate metabolite involved in gluconeogenesis and oxidative first defined the pathophysiological effect of prolonged phosphorylation. [1-13C]OAA can be rapidly converted HFD feeding. HFD-fed mice developed hepatic steato- sis, with more than an 8-fold higher IHTG level than [1-13C]aspartate, and [6-13C]citrate, catalyzed by Chow-fed mice (Table 1). HFD mice were also hyper- PEPCK, malate dehydrogenase (MDH), aspartate trans- glycemic and hyperinsulinemic. Hematoxylin and eo- aminase (AST), and citrate synthase, respectively. In the sin and Oil Red O histology revealed massive lipid MRS spectra, we were able to detect [1-13C]malate and deposits in the hepatocytes of HFD-fed mice (Sup- [1-13C]aspartate peaks, consistent with observations in the perfused mouse liver.4 Because the conversion of included a higher body weight and body fat composi- OAA to malate and aspartate are reversible reactions, tion, together with a lower body lean content. IPGTT there is 13C label exchange between these three metabo- revealed impaired glucose tolerance in HFD mice, as evidenced by delayed glucose clearance at 45, 60, 90, [1-13C]malate to [1-13C]fumarate, catalyzed by fuma- and 120 minutes after infusion (Fig. 1A,B). In addi- rase, resulted in the repositioning of the 13C label tion, there was simultaneous compensatory increase in between the C1 and C4 positions of fumarate. This, in insulin secretion (Fig. 1C,D). ITT revealed a reduced effect, gave rise to [4-13C]malate, [4-13C]aspartate, and blood glucose decrease in HFD mice, compared to [4-13C]OAA peaks.4,12 A representative time course dis- Chow-fed mice (Fig. 1E,F), indicative of insulin resist- playing the progression of metabolite signals is shown in ance in HFD mice. Together, these results show that Fig. 2B. These results show that the major downstream

HEPATOLOGY, February 2013 Fig. 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 ¼ 8 for each group. (B) Glucose AUC (area under curve) calculatedbased on data in (A). (C) Plasma insulin levels in HFD (N ¼ 3) and Chow-diet fed mice (N ¼ 10) in IPGTT. (D) Glucose-induced insulin secre-tion calculated by integrating the AUC after baseline subtraction based on data in (C). (E) ITT revealed that HFD mice (N ¼ 4) exhibited lowerinsulin sensitivity than Chow-diet fed control (N ¼ 10). (F) Insulin-induced glucose level in plasma calculated by integrating AAC (area abovecurve) based on data in (E). Data are presented as mean 6 SEM. *P < 0.05; **P < 0.01.
pathways of pyruvate can be monitored with hyperpolar- [1-13C]pyruvate and [4-13C]OAA, kpyr->oaa, showed a more than 3-fold increase in the HFD group, relative to We next examined the metabolic changes in gluco- Chow-fed mice. The corresponding time courses over neogensis in HFD mice. When compared to control 60 seconds illustrated the relatively faster production of these four-carbon metabolites in the steatotic liver (Sup- [4-13C]OAA/tCarbon, [1-13C]aspartate/tCarbon, and porting Fig. S3). Together, the flux measurements show- [1-13C]alanine/tCarbon were significantly larger in fatty ing increased PC activity in HFD-fed mouse livers sug- livers of HFD-fed mice (Fig. 2C), whereas no significant gest that PC may be a central player in enhanced change in [1-13C]lactate/tCarbon and [1-13C]bicar/ gluconeogenesis in the prediabetic stage.
tCarbon ratios was observed, suggesting higher PC and Increased Malate, Aspartate, and Alanine 13C ALT activities, but not LDH or PDH activities. Consis- Metabolite Signals in Fatty Liver. With the observa- tently, the rate of 13C label exchange between [1-13C]py- tion that [1-13C]malate and [1-13C]aspartate signals ruvate and [1-13C]alanine, kpyr->ala, was increased by were significantly increased in fatty liver, we next more than 70% in HFD-fed mice, whereas no signifi- sought to understand the mechanism underlying the cant change was detected in its exchange with changes. Because each pathway involves two mediating [1-13C]lactate, kpyr->lac (Fig. 2D). The exchange rate enzymes, PC/MDH and PC/AST, respectively, it is between [1-13C]pyruvate and [1-13C]malate, kpyr->mal, essential to distinguish each enzyme's contribution to which is dependent on both PC and MDH enzyme ca- the 13C metabolite signal. Ex vivo enzyme-activity talysis, increased significantly in fatty liver. The exchange assays of liver extracts obtained from both HFD- and between [1-13C]pyruvate and [1-13C]aspartate, kpyr->asp, Chow-fed mice revealed a significant up-regulation of mediated by both PC and AST enzymes, was also ele- PC activity in fatty liver (Fig. 3A). However, there was vated in the steatotic liver. Exchange rate between no apparent increase in AST activity (Fig. 3B),

HEPATOLOGY, Vol. 57, No. 2, 2013 Fig. 2. Elevated pyruvate ana- plerosis in mice with hepatic stea-tosis is detected by hyperpolarized13C MRS. (A) Representation of thein vivo measured hyperpolarized13C spectra in fatty liver. Metabo-lites that were detected include[1-13C]lactate ppm), [1-13C]alanine (176.4 ppm),[4-13C]OAA [1-13C]pyruvate (170.8 ppm), and[1-13C]bicarbonate (160.8 ppm).
(B) depicting the simultaneous produc-tion between mice fed on Chow andHFD. (D) Corresponding 13C labelexchange rates (N ¼ 8 for eachgroup).
mean 6 SEM. *P < 0.05; **P <0.01. Values are displayed in Sup-porting Table 1.
indicating that the larger [1-13C]aspartate signal was between measured in vivo hyperpolarized 13C exchange primarily the result of increased PC activity. Hepatic and actual hepatic enzyme activity. First, the exchange MDH activity, on the other hand, was up-regulated in rate between hyperpolarized [1-13C]pyruvate and diabetic mice (Fig. 3C). Therefore, the higher [1-13C]alanine, kpyr->ala, correlated with ex vivo hepatic [1-13C]malate signal could be attributed to a combina- ALT activity (Fig. 4A). Because ALT is a key regulatory tion of increased PC and MDH activities. This com- enzyme in pyruvate recycling and urea production, bined effect probably led to increased 13C label in vivo kpyr->ala could be used as an important bio- exchange between OAA, malate, and fumarate, thus marker of liver dysfunction. Second, faster 13C label contributing to an elevated [4-13C]OAA signal (Fig.
exchange between [1-13C]pyruvate and [1-13C]aspar- 2C). These results further support the critical role of tate, kpyr->asp, correlated well with higher PC activity PC in gluconeogensis in the prediabetic stage.
in hepatocytes (Fig. 4B). Therefore, kpyr->asp could be Another key enzyme in gluconeogenesis, PEPCK, a potential biomarker to reflect in vivo gluconeogenic was concomitantly up-regulated in the insulin-resistant flux in the liver. Together, these results demonstrate liver (Fig. 3D). This further corroborated the observa- that hyperpolarized 13C metabolic signals may be used tion that elevated pyruvate anaplerosis was required to as relevant diagnostic biomarkers of liver dysfunction, support the increased hepatic glucose production in such as in diabetes.
diabetic mice. The higher exchange rate between Detecting Changes in Liver Metabolism Induced [1-13C]pyruvate and [1-13C]alanine indicated faster by Glucagon and Metformin With Hyperpolarized transamination, which was confirmed in the biochemi- 13C MRS. To assess the detection sensitivity of hyper- cal ALT activity assay (Fig. 3E).
polarized 13C MRS on changes in liver metabolism, we Hyperpolarized 13C Metabolic Fluxes as Potential first examined glucagon-induced glucose production in Biomarkers of Liver Function. We next determined Chow-fed animals. Higher aspartate, bicarbonate, and the potential of hyperpolarized 13C metabolic signals OAA signals were recorded 10 minutes after IV gluca- as relevant diagnostic biomarkers of liver dysfunction gon injection (Fig. 5A). The corresponding 13C-label in the diabetic state by examining the relationship exchanges rates (kpyr->asp, kpyr->bic, and kpyr->oaa) were

HEPATOLOGY, February 2013 Fig. 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 glu-coneogenesis in the insulin-resistant liver. (E) Abnormal liver function in diabetes was evident in the increased ALT flux (N ¼ 8 for each group).
Data are presented as mean 6 SEM. *P < 0.05; **P < 0.01.
also significantly increased (Fig. 5B). Elevated kpyr->asp and kpyr->oaa are signatures of enhanced hepatic gluco-neogenesis (see above), whereas higher k Novelty in Measuring Liver Metabolism In pyr->bic indicates up-regulated pyruvate dehydrogenase (PDH) activity.13 Vivo. Although it is recognized that glucose homeostasis Conversely, metformin treatment successfully reduced maintained by the tissue trio (muscle, liver, and fat) is hepatic gluconeogenesis, as evidenced by the signifi- disturbed in diabetes,14 it has not been possible to detect cantly lower malate and aspartate signals, and the abate- and measure the underlying hepatic metabolic aberra- ment of their corresponding exchange rates, k tions noninvasively in real time. In this study, we dem- onstrate, for the first time, the novel use of hyperpolar- pyr->asp (Fig. 5C,D). Blood glucose level was decreased by 24% as well (Supporting Table 3). These ized 13C MRS to quantify and assess enzyme fluxes results show that hyperpolarized 13C MRS appears to specific to the liver in a type 2 diabetes mouse model in be sufficiently sensitive for measurement of induced vivo. By measuring gluconeogenic fluxes, we identify PC metabolic changes in the liver.
and that its downstream MDH activities are up- Fig. 4. Correlation between PC and ALT enzyme activities and the corresponding in vivohyperpolarized 13C metabolic flux biomarkers.
Faster [1-13C]pyruvate to (A) [1-13C]alanine,and (B) [1-13C]aspartate exchanges correlateto higher intracellular ALT and PC activities,respectively.

HEPATOLOGY, Vol. 57, No. 2, 2013 Fig. 5. Hyperpolarized 13C MRS detects changes in hepatic gluconeogenesis uponpharmacological intervention. (A) Significantlyhigher [1-13C]aspartate and [4-13C]OAA sig-nals and (B) exchange rates were detected 10minutes after glucagon stimulation. (C) Signifi-cantly lower [1-13C]malate and [1-13C]aspar-tate signals and (D) exchange rates weredetected in metformin-treated mice (N ¼ 5 foreach group). Data are presented as mean 6SEM. *P < 0.05; **P < 0.01. Values are dis-played in Supporting Tables 2 and 3.
regulated and suggest the PC pathway as a critical com- pression of key gluconeogenic enzymes PEPCK and glu- ponent in the development of hyperglycemia and addition, previous studies utilizing radioisotopic analysis Heightened Gluconeogenesis in Diabetes Drives also showed that carboxylation of pyruvate into OAA is Pyruvate Anaplerosis. Through validation with spec- up-regulated in the diabetic rat liver, concomitant with trophotometric assays of liver tissue extracts, we dem- dramatic increases in PC,16 PEPCK, and G-6-Pase15 expression. These studies corroborate our finding that [1-13C]aspartate metabolite signal may be attributed to both PC and PEPCK enzyme activities are increased in a higher PC flux, whereas the increased [1-13C]malate the fatty liver, leading to larger 13C-malate, -aspartate, signal is a combined effect of increased PC and MDH and -OAA signals as well as higher rates of chemical exchange with pyruvate. Indeed, higher hepatic PC ac- [1-13C]malate are derived from [1-13C]oxaloacetate tivity correlated with increased PEPCK activity (r2 ¼ through anaplerosis from [1-13C]pyruvate, and a larger 0.82; P < 0.0001) (Supporting Fig. 4), further support- [4-13C]oxaloacetate pool is indeed detected simultane- ing the hypothesis that both PC and PEPCK are impor- ously, we hypothesize that the increase in these down- tant regulators in gluconeogenesis.7 stream metabolite pools is a manifestation of elevated Increased glucagon effect. In diabetes, pathological demand for the intermediate OAA, in response to alteration of the precise balance between insulin and higher anabolic rates of gluconeogenesis and fatty acid glucagon action results in excessive hepatic gluconeo- synthesis (FAS) (Fig. 6). A few factors may contribute genesis and glycogenolysis, both of which induce to this phenomenon in fatty liver, as described below.
hyperglycemia. Moreover, inadequate suppression of Insulin resistance. Insulin insensitivity in the fatty postprandial glucagon secretion by insulin in the dia- liver is detrimental to the hormone's inhibitory role in betic state causes hyperglucagonemia and evokes ele- gluconeogenesis, primarily through the inactivation of vated HGP, as observed in HFD mice. We previously the phosphatidylinositol 3-kinase/serine/threonine ki- reported that combined defects in insulin secretion nase–signaling pathway,15 thereby enfeebling the sup-

HEPATOLOGY, February 2013 Fig. 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 molecularevents triggered by (1) insulin resistance, (2) dominant glucagon effect, and (3) obesity-induced de novo FAS (see text). Metabolites in redboxes can be detected by hyperpolarized 13C MRS.
hyperglycemia in the absence of dysregulated glucagon Elevated FAS. De novo FAS in the liver is regulated secretion in a mouse model with deletion of calcium- by three known transcription factors: steroid regulatory sensing protein synaptotagmin-7.17 Indeed, glucagon element-binding protein (SREBP-1c)22; carbohydrate plays a major role in promoting gluconeogenesis in response element-binding protein (ChREBP)23; and enhancing G-6-Pase activity and PEPCK transcription peroxisome proliferator-activated receptor gamma.24 in the liver, likely through the protein kinase A–signal- Hyperinsulinemia induces hepatic SREBP-1c expres- ing cascade mechanism.18 Thereafter, up-regulated glu- sion while hyperglycemia stimulates ChREBP activity.
coneogenesis increases the demand for OAA. In this These events lead to transcriptional activation of all li- work, we demonstrated up-regulated PC activity in pogenic genes including adenosine triphosphate citrate glucagon-stimulated HGP in Chow-fed animals, as lyase, acetyl-CoA carboxylase, and FAS,25 effectively detected in vivo with hyperpolarized 13C MRS, increasing FAS flux. Because citrate formed in the through the biomarker kpyr->asp. Concomitantly, gluca- TCA cycle and shuttled to the cytosol is the primary gon increases PDH activity.19 This technology appears metabolite required in the production of fatty acids, to possess sufficient sensitivity to detect this phenom- there is inevitably an increase in demand for this inter- enon as well, as evident from the higher kpyr->bic mediate. Therefore, it is unsurprising that a recent 13C exchange rate. Treatment with a glucagon-receptor an- isotopomer study found a 2-fold increase of hepatic tagonist appears to alleviate HGP in the diabetic TCA cycle flux in patients with nonalcoholic fatty liver liver,20 and reducing glucagon signaling is being disease.26 Because only pyruvate that enters the TCA explored as a potential therapy for diabetes.21 It will cycle through PC produces a net increase in cycle be interesting to measure corresponding changes in he- intermediates, whereas pyruvate entering through patic metabolism upon therapeutic intervention with a PDH is restricted to energy production only,27 the ele- glucagon-receptor antagonist in diabetic animals, and vated PC flux and OAA pool observed in diabetic that forms the next phase of our research.
mice must have also catered to the increased FAS HEPATOLOGY, Vol. 57, No. 2, 2013 demand. This agrees with our recent observation in than fasted, in part because of the abundance of other the hypertrophied heart, in which a larger 13C-citrate alternative metabolic substrates in the circulation (e.g., signal (from increased pyruvate anaplerosis) was unlabeled pyruvate), which limits the uptake of the recorded.28 Moreover, detection of citrate pool with infused hyperpolarized 13C-labeled pyruvate. This ob- hyperpolarized [2-13C]pyruvate substrate has recently servation is observed in the lower metabolite peaks in been demonstrated to be feasible in the study of myo- the spectrum of fed mice (Supporting Fig. 3A).
cardial TCA flux29; therefore, similar measurements in In summary, we demonstrate the application of the insulin-resistant liver will undoubtedly aid in vali- hyperpolarized 13C MRS in probing metabolic events dating the hypothesis that an enlarged citrate pool sup- in the liver and its correlation with enzyme activities.
ports FAS. Metformin is used clinically to counter ele- We identify an important role of the PC pathway in vated FAS and gluconeogenesis in diabetes, primarily the development of hyperglycemia and diabetes. We through its activation of adenosine-monophosphate– also demonstrated the capability of this technique to activated protein kinase.30 In this study, we demon- probe changes in hepatic metabolism upon therapeutic strated that metformin treatment leads to reduced intervention, paving the way for longitudinal assess- HGP by, at least in part, decreasing PC activity, as well as production of malate and aspartate from exchange rates and hepatic enzyme activities illustrates the potential of these indices as biomarkers of liver The advent of hyperpolarized 13C MRS has enabled function in diabetes and a wide range of other visualization of real-time metabolism in the in vivo mouse liver, in particular, the anaplerosis of pyruvate The authors thank Dr. Hongyu into the TCA cycle. The distinct patterns in down- Li for technical assistance in the glucose and insulin stream metabolite progression suggest that hyperpolar- tolerance tests.
ized 13C MRS is sensitive to subtle differences in met-abolic conversions. It is worth noting that LDH-,ALT-, MDH-, and AST-mediated conversions are re- versible. Therefore, the appearance of lactate, alanine, 1. Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to malate, aspartate, and OAA peaks resulted from the insulin resistance and type 2 diabetes. Nature 2006;444:840-846.
enzyme-mediated exchange of the hyperpolarized 13C 2. Cohen SM. Simultaneous 13C and 31P NMR studies of perfused rat label, in which equilibrium is dependent on the con- liver. Effects of insulin and glucagon and a 13C NMR assay of freeMg2þ. J Biol Chem 1983;258:14294-14308.
centrations of both substrate and product, as well as 3. Ardenkjaer-Larsen JH, Fridlund B, Gram A, Hansson G, Hansson L, the redox potential. Hence, these metabolite signals Lerche MH, et al. Increase in signal-to-noise ratio of >10,000 times in reflect the concentration of each metabolite that al- liquid-state NMR. Proc Natl Acad Sci U S A 2003;100:10158-10163.
4. Merritt ME, Harrison C, Sherry AD, Malloy CR, Burgess SC. Flux ready exists within the cellular environment and in the through hepatic pyruvate carboxylase and phosphoenolpyruvate carbox- plasma, rather than net metabolite production.31 ykinase detected by hyperpolarized 13C magnetic resonance. Proc Natl Nonetheless, the validity of hyperpolarized Acad Sci U S A 2011;108:19084-19089.
exchange rates to investigate metabolic changes in pa- 5. Hu S, Chen AP, Zierhut ML, Bok R, Yen YF, Schroeder MA, et al.
In vivo carbon-13 dynamic MRS and MRSI of normal and fasted rat thology has been demonstrated in cardiac diseases28,32 liver with hyperpolarized 13C-pyruvate. Mol Imaging Biol 2009;11: 6. Burgess SC, He T, Yan Z, Lindner J, Sherry AD, Malloy CR, et al. Cy- tosolic phosphoenolpyruvate carboxykinase does not solely control the kpyr->mal. Although both malate and aspartate signals rate of hepatic gluconeogenesis in the intact mouse liver. Cell Metab were increased in the fatty liver, kpyr->asp, rather than 7. Groen AK, van Roermund CW, Vervoorn RC, Tager JM. Control of pyr->mal (data not shown), correlated linearly to PC activity. In addition, k gluconeogenesis in rat liver cells. Flux control coefficients of the pyr->asp appeared to be more sen- enzymes in the gluconeogenic pathway in the absence and presence of sitive in detecting glucagon-induced up-regulation in glucagon. Biochem J 1986;237:379-389.
gluconeogenesis because it increased significantly upon 8. Gustavsson N, Lao Y, Maximov A, Chuang JC, Kostromina E, Repa glucagon injection. However, with metformin treat- JJ, et al. Impaired insulin secretion and glucose intolerance in synapto- ment, the changes in gluconeogenesis were large tagmin-7 null mutant mice. Proc Natl Acad Sci U S A 2008;105:3992-3997.
enough to manifest in both kpyr->asp and kpyr->mal. This 9. Longo R, Pollesello P, Ricci C, Masutti F, Kvam BJ, Bercich L, et al.
confounding phenomenon warrants further investiga- Proton MR spectroscopy in quantitative in vivo determination of fat tion, in particular, the influence of MDH activity. In content in human liver steatosis. J Magn Reson Imaging 1995;5:281-285.
addition, from our experience, the metabolite peaks 10. Schroeder MA, Swietach P, Atherton HJ, Gallagher FA, Lee P, Radda measured in vivo are generally lower in the fed state GK, et al. Measuring intracellular pH in the heart using hyperpolarized HEPATOLOGY, February 2013 carbon dioxide and bicarbonate: a 13C and 31P magnetic resonance 23. Dentin R, Benhamed F, Hainault I, Fauveau V, Foufelle F, Dyck JR, spectroscopy study. Cardiovasc Res 2010;86:82-91.
et al. Liver-specific inhibition of ChREBP improves hepatic steatosis 11. Clee SM, Attie AD. The genetic landscape of type 2 diabetes in mice.
and insulin resistance in ob/ob mice. Diabetes 2006;55:2159-2170.
Endocr Rev 2007;28:48-83.
24. Reddy JK, Rao MS. Lipid metabolism and liver inflammation. II. Fatty 12. Buldain G, de los Santos C, Frydman B. Carbon-13 nuclear magnetic liver disease and fatty acid oxidation. Am J Physiol Gastrointest Liver resonance spectra of the hydrate, keto and enol forms of oxalacetic acid. Magn Reson Chem 1985;23:478-481.
25. Browning JD, Horton JD. Molecular mediators of hepatic steatosis and 13. Schroeder MA, Cochlin LE, Heather LC, Clarke K, Radda GK, Tyler liver injury. J Clin Invest 2004;114:147-152.
DJ. In vivo assessment of pyruvate dehydrogenase flux in the heart 26. Sunny NE, Parks EJ, Browning JD, Burgess SC. Excessive hepatic mi- using hyperpolarized carbon-13 magnetic resonance. Proc Natl Acad tochondrial TCA cycle and gluconeogenesis in humans with nonalco- Sci U S A 2008;105:12051-12056.
holic fatty liver disease. Cell Metab 2011;14:804-810.
14. Barthel A, Schmoll D. Novel concepts in insulin regulation of hepatic 27. Cohen SM. 13C NMR study of effects of fasting and diabetes on the metabolism of pyruvate in the tricarboxylic acid cycle and the utiliza- tion of pyruvate and ethanol in lipogenesis in perfused rat liver. Bio-chemistry 1987;26:581-589.
15. Saltiel AR, Kahn CR. Insulin signalling and the regulation of glucose 28. Atherton HJ, Dodd MS, Heather LC, Schroeder MA, Griffin JL, and lipid metabolism. Nature 2001;414:799-806.
Radda GK, et al. Role of pyruvate dehydrogenase inhibition in the de- 16. Weinberg MB, Utter MF. Effect of streptozotocin-induced diabetes velopment of hypertrophy in the hyperthyroid rat heart: a combined mellitus on the turnover of rat liver pyruvate carboxylase and pyruvate magnetic resonance imaging and hyperpolarized magnetic resonance dehydrogenase. Biochem J 1980;188:601-608.
spectroscopy study. Circulation 2011;123:2552-2561.
17. Gustavsson N, Seah T, Lao Y, Radda GK, Sudhof TC, Han W.
29. Schroeder MA, Atherton HJ, Ball DR, Cole MA, Heather LC, Griffin Delayed onset of hyperglycaemia in a mouse model with impaired glu- JL, et al. Real-time assessment of Krebs cycle metabolism using hyper- cagon secretion demonstrates that dysregulated glucagon secretion pro- polarized 13C magnetic resonance spectroscopy. FASEB J 2009;23: motes hyperglycaemia and type 2 diabetes. Diabetologia 2011;54: 30. Zhou G, Myers R, Li Y, Chen Y, Shen X, Fenyk-Melody J, et al. Role 18. Jiang G, Zhang BB. Glucagon and regulation of glucose metabolism.
of AMP-activated protein kinase in mechanism of metformin action.
Am J Physiol Endocrinol Metab 2003;284:E671-E678.
J Clin Invest 2001;108:1167-1174.
19. Oviasu OA, Whitton PD. Hormonal control of pyruvate dehydrogen- 31. Tyler DJ, Schroeder MA, Cochlin LE, Clarke K, Radda GK. Applica- ase activity in rat liver. Biochem J 1984;224:181-186.
tion of hyperpolarized magnetic resonance in the study of cardiac me- 20. Qureshi SA, Rios Candelore M, Xie D, Yang X, Tota LM, Ding VD, tabolism. Appl Magn Reson 2008;34:523-531.
et al. A novel glucagon receptor antagonist inhibits glucagon-mediated 32. Schroeder MA, Atherton HJ, Dodd MS, Lee P, Cochlin LE, Radda biological effects. Diabetes 2004;53:3267-3273.
GK, et al. The Cycling of acetyl-CoA through acetylcarnitine buffers 21. Moller DE. New drug targets for type 2 diabetes and the metabolic cardiac substrate supply: a hyperpolarised 13C magnetic resonance syndrome. Nature 2001;414:821-827.
study. Circ Cardiovasc Imaging 2012;5:201-209.
22. Horton JD, Goldstein JL, Brown MS. SREBPs: activators of the com- 33. Zierhut ML, Yen YF, Chen AP, Bok R, Albers MJ, Zhang V, et al. Ki- plete program of cholesterol and fatty acid synthesis in the liver. J Clin netic modeling of hyperpolarized 13C1-pyruvate metabolism in normal rats and TRAMP mice. J Magn Reson 2010;202:85-92.

Source: http://oar.a-star.edu.sg/jspui/bitstream/123456789/531/1/In%20Vivo%20hyperpolarized%20carbon-13%20magnetic%20resonance%20spectroscopy%20reveals%20increased%20pyruvate%20carboxylase%20flux%20in%20an%20insulin-resistant%20mouse%20model.pdf