Increased 2 Subunit–Associated AMPK Activity and PRKAG2 Cardiomyopathy
http://www.100md.com
《循环学杂志》
the Department of Genetics (F.A., M.A., C.E.S., J.G.S.), Harvard Medical School and Howard Hughes Medical Institute, Boston, Mass
Cardiovascular Institute (F.A.), Department of Medicine, and Department of Human Genetics, University of Pittsburgh, Pittsburgh, Pa
Heart Institute (M.A.), Sheba Medical Center and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Research Division (N.M., L.J.G.), Joslin Diabetes Center, Boston, Mass
NMR Laboratory for Physiological Chemistry (H.H., Y.X., R.T., J.S.I.), Cardiovascular Division, Department of Medicine, Brigham and Women’s Hospital, Boston, Mass
Department of Cardiology (C.W., D. Branco, C.I.B.), Children’s Hospital, Boston, Mass
Department of Pathology (A.R.P.-A.), Children’s Hospital, Boston, Mass
Bio21 Molecular Science and Biotechnology Institute (D.S.), University of Melbourne, Melbourne, Australia
Pediatric Medical Genetics (D. Bali), Duke University Medical Center, Research Triangle Park, NC
Cardiovascular Division (C.E.S.), Brigham and Women’s Hospital, Boston, Mass.
Abstract
Background— AMP-activated protein kinase (AMPK) regulatory 2 subunit (PRKAG2) mutations cause a human cardiomyopathy with cardiac hypertrophy, preexcitation, and glycogen deposition. PRKAG2 cardiomyopathy is recapitulated in transgenic mice overexpressing mutant PRKAG2 N488I in the heart (TG2N488I). AMPK is a heterotrimeric kinase consisting of 1 catalytic () and 2 regulatory ( and ) subunits. Two -subunit isoforms, 1 and 2, are expressed in the heart; however, the contribution of AMPK utilization of these subunits to PRKAG2 cardiomyopathy is unknown. Mice overexpressing a dominant-negative 2 subunit of AMPK (TG2DN) provide a tool for selectively inhibiting 2, but not 1, subunit-associated AMPK activity.
Methods and Results— In compound-heterozygous TG2N488I/TG2DN mice, AMPK activity associated with 2 but not 1 was decreased compared with TG2N488I. The TG2DN transgene reduced the disease phenotype of TG2N488I, partially or completely normalizing the ECG, cardiac function, cardiac morphology, and exercise capacity in compound-heterozygous mice. TG2N488I hearts had normal resting levels of high-energy phosphates and could improve cardiac performance during exercise. Cardiac glycogen content decreased in TG2N488I mice after exercise stress, indicating availability of the stored glycogen for metabolic utilization. No differences in glycogen-metabolizing enzymes were observed.
Conclusions— The PRKAG2 N488I mutation causes inappropriate AMPK activation, which leads to glycogen accumulation and conduction system disease. The accumulated glycogen can serve as an energy source, and the animals have contractile reserve during exercise. Because the dominant-negative 2 subunit attenuates the mutant PRKAG2 phenotype, AMPK complexes containing the 2 rather than the 1 subunit are the primary mediators of the effects of PRKAG2 mutations.
Key Words: cardiomyopathy exercise genetics metabolism Wolff-Parkinson-White syndrome
Introduction
Mutations in AMP-activated protein kinase (AMPK) have been discovered to cause a glycogen storage cardiomyopathy that ultimately progresses to heart failure1–3 (for reviews, see Gollob4 and Gollob et al5). AMPK, a heterotrimeric protein composed of , , and subunits, is a metabolite-sensing protein kinase that is activated under conditions of energy depletion manifested by increased cellular AMP levels (reviewed in Hardie6). AMPK is able to phosphorylate >15 known target proteins and is thought to regulate energy metabolism by modulating a variety of metabolic activities, including glucose transport, stimulation of -oxidation of fatty acids, inactivation of cholesterol synthesis, inhibition of creatine kinase, and transcriptional regulation of several genes.6,7 Mutations in the gene PRKAG2, encoding the 2 subunit of AMPK, have been demonstrated to produce a distinct cardiomyopathy in human families characterized by ventricular hypertrophy, ventricular preexcitation, and progressive conduction system disease.1–3,8 Mutations recognized thus far are Exon5:InsLeu, His142Arg, Arg302Gln, Thr400Asn, Tyr487His, Asn488Ile, and Arg531Gly. The cardiomyopathy associated with PRKAG2 mutations is not associated with the myocyte and myofibrillar disarray and cardiac fibrosis characteristic of mutations in genes encoding sarcomeric proteins. Rather, cardiac hypertrophy results from formation within the myocytes of vacuoles filled with glycogen-associated granules.1,9 PRKAG2-mediated cardiomyopathy, unlike other glycogen storage disorders, is characterized by cardiac glycogen accumulation in the absence of other remarkable clinical features, such as skeletal myopathy. The mechanisms by which these PRKAG2 missense mutations mediate cardiomyopathy and the accessibility of the accumulated glycogen as an energy source remain uncertain.
Clinical Perspective p 3148
Both cell and murine models of PRKAG2 cardiomyopathy have been described.1,9–11 Several in vitro studies suggest that PRKAG2 cardiomyopathy is caused by increased AMPK activity.1 Introduction of the Thr400Asn and Asn488Ile mutations into the yeast homolog of the subunit, Snf4, resulted in constitutive activation of the Snf1/Snf4 kinase.1 A similar Arg70Gln mutation in the 1 subunit introduced into COS7 cells and pulmonary fibroblasts caused markedly increased AMPK activity, associated with increased phosphorylation of the subunit Thr172 and increased phosphorylation of 1 of its major substrates, acetyl coenzyme A carboxylase.9 However, introduction of several mutant PRKAG2s into mammalian CCL13 cells caused either a decrease or no change in AMPK activity.10,11 Furthermore, recent studies by Sidhu and colleagues12 have suggested that the PRKAG2 Arg302Gln missense mutation inactivates AMPK and thereby causes glycogen accumulation and cardiac hypertrophy.
TG2N488I mice provide a model of PRKAG2 cardiomyopathy because they express a transgene encoding the Asn488Ile (N488I) mutation, which is expressed only in the heart under control of the -myosin heavy-chain promoter.13 TG2N488I hearts demonstrate elevated AMPK activity, massive glycogen deposition in cardiac myocytes up to 30-fold above normal, dramatic left ventricular (LV) hypertrophy, ventricular preexcitation, and sinus node dysfunction. Disruption of the annulus fibrosus by glycogen-filled myocytes is the anatomic substrate for preexcitation.13,14 TG2N488I extracts have markedly elevated AMPK activity,13 and this increased activity is associated with both the 1 and 2 subunits.
TG2N488I mice provide a unique tool for dissecting the mechanisms leading to PRKAG2 cardiomyopathy. We demonstrate, using a novel murine stress-echocardiography procedure, that although energy metabolism remains unchanged in the resting TG2N488I heart, the accumulated glycogen in the TG2N488I heart can be mobilized and metabolized during exercise. Furthermore, we use a genetic approach to confirm the model that PRKAG2 cardiomyopathy results from hyperactive AMPK, hypothesizing that if the PRKAG2 missense mutations function by inactivating AMPK, then dominant-negative 2-subunit mutations15 would not alter the phenotype of transgenic mice. Because the transgenic allele TG2DN, which expresses a dominant-negative form of the AMPK 2 catalytic subunit, inactivates AMPK containing the 2 but not the 1 subunit,15 we are able to define the relative contributions of AMPK containing these subunits. We conclude that PRKAG2 cardiomyopathy results from overactive AMPK associated primarily with the 2 rather than the 1 subunit.
Methods
Mice
Transgenic mice overexpressing human full-length, wild-type (TG2wt) and N488I (TG2N488I) PRKAG2 have been described.13 In brief, a TA substitution was introduced at nucleotide 1553 in the human PRKAG2 cDNA to alter codon 488 from AsnIle, corresponding to a known human mutation. The transgene is expressed under control of the strong cardiac-specific -myosin heavy-chain promoter. Transgenic mice overexpressing a dominant-negative 2 subunit of AMPK (TG2DN) under control of the same -myosin heavy-chain promoter were generated as previously reported.15 The AspAla substitution at codon 157 in rat PRKAA2 cDNA creates mutant 2 subunits, which bind to 2 subunits, producing enzymatically inactive complexes. The 2DN polypeptide also contains a YPYDVPDYA (HA, a synthetic peptide) tag. All mouse studies were performed in accordance with protocols approved by the institutional animal care and use committee at Harvard Medical School.
AMPK Activity
Resting mice were humanely killed by cervical dislocation without anesthesia, and the heart tissue was immediately frozen in LN2. Freeze-clamped heart samples were homogenized, and lysates were used for AMPK activity assays as described.16 Protein (200 μg) was immunoprecipitated with antibodies recognizing amino acids 339 to 358 of rat AMP 1 or 352 to 366 of 2. The kinase reaction was performed, in the presence of 0.2 mmol/L AMP, with synthetic substrate for AMP-activated protein kinase (SAMS) peptide as a substrate, and AMPK activity was measured as picomoles phosphate incorporated per milligram protein per minute.
Glycogen Enzyme Activity Assays
Hearts from wild-type, TG2wt, and TG2N488I mice were excised. Brancher enzyme or 1,4--glucan branching enzyme [1,4--D-glucan:1,4--D-glucan 6--D-(1,4--D-glucano)transferase; EC 2.4.1.18] and glycogen phosphorylase (1,4--D-glucan:phosphate -D-glucosyltransferase; EC 2.4.1.1) were measured by free phosphate levels with inorganic phosphorus reagent (Roche Diagnostics).17,18 Debrancher enzyme or 4--glucanotransferase (4--glucanotransferase 1,4--D-glucan:1,4--D-glucan 4--D-glycosyltransferase; EC 2.4.1.25) and glycogen phosphorylase kinase (ATP:phosphorylase- phosphotransferase; EC 2.7.1.38) were measured by release of free glucose from Roche glucose reagent according to a standard protocol.17,19,20
Immunoblots
Antibodies recognizing 1, 2, and 2 subunits of AMPK were produced as previously described.15,21,22 A lamin B antibody was used as a loading control (sc-6217, Santa Cruz).
Energetics Assays
Isolated, Langendorff-perfused hearts were freeze-clamped, and myocardial levels of adenine nucleotides and phosphocreatine were determined by a high-performance liquid chromatography method as reported previously.23 31P nuclear magnetic resonance (NMR) spectra were obtained as previously described24 at 161.94 MHz with a GE-400 wide-bore spectrometer. Hearts were placed in a 10-mm glass NMR tube and inserted into a custom-made, 1H/31P double-tuned probe situated in an 89-mm bore, 9.4-T superconducting magnet. The resonance area corresponding to phosphocreatine was fitted to Lorentzian functions and calculated according to a commercially available program (NMR1).
Cardiac Glycogen and Lactate Levels
Glycogen content was determined by perchloric acid extraction and amyloglucosidase digestion, followed by determination of glucose levels with a glucose oxidase kit (Sigma-Aldrich).25 Lactate levels were determined according to Sigma technical bulletin 826-UV with the use of NAD, glycine, and lactate dehydrogenase.
Exercise Stress Testing
To assess exercise capacity, mice were placed on a treadmill with an adjustable belt speed and incline and electric shock bars. Mice were subjected to an endurance protocol as previously described.26 Each mouse was encouraged to run to the limit of its capacity on this protocol, and total exercise duration was recorded.
Stress Echocardiography
The treadmill exercise protocol was modified to bring mice rapidly to maximum exercise capacity. Exercise was initiated at a treadmill incline of 15° and a speed of 10 m/min. The speed was increased by 2.5 m/min every 3 minutes until each mouse reached its maximum capacity and appeared exhausted. Mice were acclimated to the stress protocol by 1 practice exercise session 1 day before stress echocardiography. Echocardiography was performed with an Agilent Sonos 4500 ultrasound machine and a 6- to 15-MHz linear-array transducer as previously described,23 with the exception that no anesthesia was used and that immediate postexercise echocardiography was limited to M-mode views. LV end-diastolic and end-systolic diameters and wall thicknesses were obtained from M-mode tracings from measurements averaged from 3 separate cardiac cycles. LV fractional shortening (in percent) was derived from the equation fractional shortening=[(LV end-diastolic diameter–LV end-systolic diameter)/LV end-diastolic diameter]x100. Echocardiography was completed in unanesthetized animals within 5 minutes before exercise and within 30 seconds after peak exercise.
Static and Continuous Ambulatory (Holter) ECG
ECG was performed as previously described.14,27 Limb-lead ECGs were recorded to detect evidence of ventricular preexcitation, namely, shortened PR intervals and delta waves. In a separate set of experiments, a telemetry device was implanted to allow continuous ambulatory ECG (Holter monitoring) and assessment of the chronotropic response to exercise. Mice were exercised to maximum capacity according to the exercise stress testing protocol described for stress echocardiography and were monitored for another 10 minutes during recovery.
Data Analysis
All data are presented as mean±SD, except as noted in the figure legends. Comparisons between groups were made with Student’s t test for pairwise comparisons and ANOVA for multiple comparisons.
Results
Inhibiting AMPK in TG2N488I Mice With 2 Dominant-Negative Transgene
We have previously presented evidence suggesting that the N488I mutation in the 2 subunit results in increased AMPK activity.1,13 Xing et al15 have described a transgenic mouse, designated TG2DN, that expresses a dominant-negative form of the 2 subunit of AMPK. TG2DN mice have significantly reduced 2 activity but normal 1 activity.15 By mating TG2DN mice with TG2N488I transgenic mice, mice carrying both the TG2N488I transgene and the TG2DN transgene,15 designated TG2N488I/TG2DN, were obtained. Levels of TG2N488I polypeptide were the same in TG2N488I and TG2N488I/TG2DN hearts (Figure 1C). However, the nonfunctional TG2DN peptide, which migrates as a higher-molecular-weight polypeptide owing to its HA tag, completely replaced the lower-molecular-weight, endogenous, wild-type 2 subunit (Figure 1B). The 1-subunit levels in wild-type, TG2N488I, TG2DN, and TG2N488I/TG2DN hearts were not significantly different (Figure 1A), as suggested previously.15
We assessed the amount of AMPK activity associated with the 1 and 2 subunits in 18-day-old transgenic mice (Table 1). TG2N488I/TG2DN mice demonstrated a significant reduction in AMP-stimulated, 2 subunit–associated AMPK activity, to 37% of that seen in TG2N488I mice (Table 1) but no significant reduction in 1-associated AMPK activity compared with TG2N488I mice. Consequently, we proceeded to phenotypic characterization of these mice to define the roles of the 1 and 2 subunit–associated AMPK activities.
Cardiac function and morphology were assessed in TG2N488I, TG2DN, and TG2N488I/TG2DN mice. At age 5 weeks, heart-body mass ratios were indistinguishable between wild-type and TG2DN mice, whereas TG2N488I mice demonstrated massive cardiac hypertrophy (Table 1). In contrast, age-matched TG2N488I/TG2DN mice had near-normal heart-body mass ratios (Table 1). The heart-body mass ratio in TG2N488I/TG2DN mice remained mildly elevated, at a level similar to that found in TG2wt mice (6.9±1.0 versus 7.2±0.3).13 Echocardiographic studies performed in 8-week-old animals demonstrated a similar significant reduction in LV wall thickness in TG2N488I/TG2DN hearts compared with TG2N488I hearts (Table 1; 1.17 versus 1.54 mm; P=0.01). These differences persisted at age 20 weeks (Table 1).
The reduction in cardiac hypertrophy in TG2N488I/TG2DN mice was accompanied by an 8-fold decrease in glycogen content compared with TG2N488I hearts (Table 1). Both light and electron microscopy confirmed the disappearance of glycogen-laden vacuoles, the pathological hallmark of PRKAG2 cardiomyopathy (Figure 2).
TG2N488I mice have ventricular preexcitation, manifested by a short PR interval and delta waves on ECG.13,14 The PR interval was 24% longer in TG2N488I/TG2DN mice relative to TG2N488I mice (P=0.02), representing normal atrioventricular conduction without ventricular preexcitation (Figure 3 and Table 1). As we previously reported, TG2N488I mice develop LV dilatation and systolic dysfunction, phenomena not observed in wild-type and TG2wt mice.13 However, TG2N488I/TG2DN mice demonstrated completely normal LV end-diastolic diameters and fractional shortening at 20 weeks of age (Table 1).
The maximal exercise capacity of wild-type, TG2wt, TG2N488I, and TG2N488I/TG2DN mice was assessed to determine the consequences of cardiac morphological and metabolic changes. Wild-type, TG2wt, TG2DN, and TG2N488I/TG2DN mice demonstrated similar exercise capacities (Figure 4A and data not shown). However, TG2N488I mice demonstrated an exercise capacity approximately half that of wild-type mice (32.4±3.6 versus 58.1±6.0 minutes, P=0.001; Figure 4A). This exercise deficit was completely eliminated in double-transgenic TG2N488I/TG2DN mice.
Energy Reserve in TG2N488I Mice
To determine the physiological basis underlying impaired exercise capacity in TG2N488I mice, we measured resting cardiac levels of phosphocreatine by NMR and of ATP, ADP, and AMP by high-performance liquid chromatography and assessed chronotropic and contractile responses to exercise. Because defects in glycogen metabolism, as well as cardiomyopathies secondary to mutations in genes encoding sarcomeric proteins, have been found to exhibit altered energetics,24,28,29 we hypothesized that TG2N488I mice might have abnormal energy stores. However, we detected no significant differences in phosphocreatine, ATP, ADP, and AMP levels among wild-type, TG2wt, and TG2N488I in isolated, perfused hearts from 5-week-old mice (Table 2).
On stress echocardiography, TG2N488I mice had lower baseline cardiac contractility (Table 3). However, TG2N488I mice had an enhanced inotropic response to exercise, increased fractional shortening, and decreased LV end-systolic dimension, thereby suggesting adequate contractile reserve. Contrasted with wild-type mice, TG2N488I mice demonstrated significantly reduced heart rates at baseline, during exercise, and during recovery (Figure 4B; P=0.004). Heart rates of wild-type, TG2wt, TG2DN, and TG2N488I/TG2DN mice were not significantly different (data not shown). Therefore, decreased exercise performance in TG2N488I mice appears to be attributable to the impairment in cardiac contractility and lower heart rates at baseline and during exercise, but not to defective energetics.
To determine whether the increased glycogen in the cardiac myocytes of TG2N488I mice could be metabolized, we measured cardiac tissue glycogen and lactate levels in hearts from resting and exercised mice (Table 4). Exercised mice were humanely killed within 10 seconds after completion of maximal exercise. Wild-type and TG2wt mice demonstrated significant (at least 40%) decreases in glycogen levels with exercise. Glycogen levels decreased by 25% in TG2N488I hearts, which represented a very large absolute quantity of glycogen. Concomitantly, more lactate was produced, consistent with the model that stored glycogen was utilized as an energy source during exercise.
Glycogen Metabolism in TG2N488I Hearts
In an attempt to define the mechanism leading to glycogen accumulation in TG2N488I hearts, we assayed the activity of 4 enzymes involved in either glycogen synthesis or glycogenolysis, which are known to cause human glycogen storage cardiomyopathies when defective. No significant differences were detected in the activities of brancher enzyme, debrancher enzyme, glycogen phosphorylase, and glycogen phosphorylase kinase in 5-week-old TG2N488I and wild-type hearts (Table 2).
Discussion
TG2N488I mice provide a useful tool for gaining further insights into the mechanisms by which mutations in the 2 subunit of AMPK cause cardiac glycogen accumulation, preexcitation, cardiomyopathy, and heart failure. Here we provide evidence that (1) PRKAG2 cardiomyopathy results from activation of AMPK, which is mediated primarily via 2 subunit–associated AMPK, and (2) there is no cardiac energy depletion in PRKAG2 cardiomyopathy at rest, and the presence of contractile reserve in response to exercise challenge is observed on stress exercise echocardiography. Despite the pivotal role of AMPK in maintaining energy metabolism, inappropriate activation of AMPK has minimal effects on resting cardiac high-energy phosphate levels.
Murine models of hypertrophic cardiomyopathy resulting from mutations in genes encoding sarcomeric proteins have demonstrated impaired cellular energetics.24,28 In contrast, TG2N488I mice exhibit normal resting cardiac cellular energetics (Table 2). Thus, although AMPK is a central regulator of cellular energy levels, the N488I mutation does not perturb resting cellular energy levels in the heart. Presumably, homeostatic mechanisms preserve normal energetics, at least in the absence of stress. The functional impairment observed in these mice is likely the result of glycogen deposition in the heart rather than energy depletion. The normal resting cellular energetics observed in TG2N488I mice also underscores the fact that the mechanism of cardiac hypertrophy and cardiomyopathy resulting from this mutation is distinct from the hypertrophic cardiomyopathy caused by sarcomere protein gene mutations.
Stress echocardiography allows direct assessment of cardiac function in response to exercise challenge and is a standard technique for evaluating myocardial ischemia in humans with coronary heart disease. Although this test has been found to have prognostic value in patients with hypertrophic or dilated cardiomyopathy, with impaired contractile response to stress connoting a poorer prognosis,30 few if any patients with PRKAG2 missense mutations have undergone this test. To evaluate further cardiac energetics during exercise in TG2N488I mice, we developed a stress echocardiography protocol for assessing cardiac function in mice. Stressed TG2N488I mice demonstrated marked improvement in cardiac function, as fractional shortening increased from 31% to 46% (P<0.0001; Table 3). The improved cardiac function of exercised TG2N488I mice may have resulted from any 1 of several mechanisms, including increased glycogen utilization. In contrast, fractional shortening was maximal at rest in wild-type and TG2wt mice and did not improve further with exercise.
Hypothesizing that improved cardiac function could result from increased glycogen utilization, we investigated whether the excess stored glycogen in TG2N488I mice could be used during exercise. TG2N488I mice demonstrated a decrease in cardiac glycogen levels with exercise, indicating that the excess glycogen in their hearts could be mobilized and metabolized (Table 4). TG2N488I mice also demonstrated an increase in cardiac lactate, presumably from increased glycogen metabolism. The large decrease in glycogen levels was even more remarkable, considering that TG2N488I mice have a markedly reduced exercise tolerance. These results raise the possibility that mobilization of cardiac glycogen allows improved cardiac function.
We and others have previously suggested that PRKAG2 cardiomyopathy results from inappropriate activation of AMPK.1,9,13 However, other investigators have proposed other models, including a biphasic response manifested by increased resting activity but impaired responsiveness to metabolic stress.10–12,31 We have used a genetic approach to determine whether the principal effect of the N488I mutation is to increase or decrease AMPK activity. If the phenotypic effects of this mutation result from an increase in AMPK 2-subunit activity, then they should be attenuated or abolished in the presence of a dominant-negative AMPK 2 subunit. The features of double-transgenic mice carrying both mutations (TG2N488I/TG2DN) are consistent with the model that PRKAG2 with the N488I mutation inappropriately activates the AMPK 2 subunit. In TG2N488I/TG2DN mice, normal catalytic 2 subunits were replaced by an inactive form of the 2 subunit (Figure 1). As a result, TG2N488I/TG2DN mice had a significant reduction in 2-, but not 1-, subunit activity, and exhibited normal or near-normal cardiac function, cardiac morphology, and conduction system function (Table 1 and Figures 2 through 4 ). Hence, reducing AMPK-associated 2-subunit activity restores cardiac physiology to nearly normal, consistent with the model that in the presence of the N488I 2 subunit, the AMPK 2 subunit is inappropriately active. Moreover, despite the 2-fold increase in 1 subunit–associated AMPK activity observed in TG2N488I mice,13 cardiac hypertrophy was almost completely reversed in the presence of a dominant-negative 2 subunit, indicating that hypertrophy is mediated largely through the effects of mutations in 2-subunit activity. Although the mechanism is uncertain, these data suggest that in vivo 2-associated AMPK and 1-associated AMPK have different biological targets. This target specificity may be determined directly by the subunit or by a preference of the subunit for a particular or isoform, which may be responsible for target specificity.
Although inappropriately active 2 subunit–associated AMPK activity is responsible for the phenotype of this cardiomyopathy, the mechanism of glycogen accumulation is still unclear. The activities of branching enzyme, debranching enzyme, glycogen phosphorylase, and glycogen phosphorylase kinase are unchanged in adult TG2N488I hearts (Table 2). These enzymes are involved in the synthesis or degradation of glycogen, and deficiencies in each of these enzymes cause glycogen storage disease. However, regulation of 1 or more glycogen metabolic enzyme activities must be altered in PRKAG2 cardiomyopathy.
PRKAG2 cardiomyopathy in TG2N488I mice provides a unique opportunity to investigate the role of AMPK and glycogen in regulating cardiac energy metabolism. Eventually, cardiac AMPK targets and the mechanisms leading from AMPK activation to glycogen storage will be defined. Identification of these targets, coupled with an understanding of the physiological pathways by which glycogen accumulates in cardiac myocytes, will provide novel therapeutic targets for a variety of conditions associated with defective cardiac energetics.
Acknowledgments
This work was supported by the Howard Hughes Medical Institute (F.A., M.A., C.E.S., J.G.S.), the Canadian Institutes of Health Research (F.A.), the National Institutes of Health (HL 67970 to R.T. and HL 52320 and HL 63985 to J.S.I.), the American Heart Association (R.T.), and the American Diabetes Association (N.M., L.G.).
References
Arad M, Benson DW, Perez-Atayde AR, McKenna WJ, Sparks EA, Kanter RJ, McGarry K, Seidman JG, Seidman CE. Constitutively active AMP kinase mutations cause glycogen storage disease mimicking hypertrophic cardiomyopathy. J Clin Invest. 2002; 109: 357–362.
Blair E, Redwood C, Ashrafian H, Oliveira M, Broxholme J, Kerr B, Salmon A, Ostman-Smith I, Watkins H. Mutations in the (2) subunit of AMP-activated protein kinase cause familial hypertrophic cardiomyopathy: evidence for the central role of energy compromise in disease pathogenesis. Hum Mol Genet. 2001; 10: 1215–1220.
Gollob MH, Green MS, Tang AS, Gollob T, Karibe A, Ali Hassan AS, Ahmad F, Lozado R, Shah G, Fananapazir L, Bachinski LL, Roberts R, Hassan AS. Identification of a gene responsible for familial Wolff-Parkinson-White syndrome. N Engl J Med. 2001; 344: 1823–1831.
Gollob MH. Glycogen storage disease as a unifying mechanism of disease in the PRKAG2 cardiac syndrome. Biochem Soc Trans. 2003; 31: 228–231.
Gollob MH, Green MS, Tang AS, Roberts R. PRKAG2 cardiac syndrome: familial ventricular preexcitation, conduction system disease, and cardiac hypertrophy. Curr Opin Cardiol. 2002; 17: 229–234.
Hardie DG. Minireview: the AMP-activated protein kinase cascade: the key sensor of cellular energy status. Endocrinology. 2003; 144: 5179–5183.
Kemp BE, Mitchelhill KI, Stapleton D, Michell BJ, Chen ZP, Witters LA. Dealing with energy demand: the AMP-activated protein kinase. Trends Biochem Sci. 1999; 24: 22–25.
Gollob MH, Seger JJ, Gollob TN, Tapscott T, Gonzales O, Bachinski L, Roberts R. Novel PRKAG2 mutation responsible for the genetic syndrome of ventricular preexcitation and conduction system disease with childhood onset and absence of cardiac hypertrophy. Circulation. 2001; 104: 3030–3033.
Hamilton SR, Stapleton D, O’Donnell JB Jr, Kung JT, Dalal SR, Kemp BE, Witters LA. An activating mutation in the 1 subunit of the AMP-activated protein kinase. FEBS Lett. 2001; 500: 163–168.
Scott JW, Hawley SA, Green KA, Anis M, Stewart G, Scullion GA, Norman DG, Hardie DG. CBS domains form energy-sensing modules whose binding of adenosine ligands is disrupted by disease mutations. J Clin Invest. 2004; 113: 274–284.
Daniel T, Carling D. Functional analysis of mutations in the 2 subunit of AMP-activated protein kinase associated with cardiac hypertrophy and Wolff-Parkinson-White syndrome. J Biol Chem. 2002; 277: 51017–51024.
Sidhu JS, Rajawat YS, Rami TG, Gollob MH, Wang Z, Yuan R, Marian AJ, DeMayo FJ, Weilbacher D, Taffet GE, Davies JK, Carling D, Khoury DS, Roberts R. Transgenic mouse model of ventricular preexcitation and atrioventricular reentrant tachycardia induced by an AMP-activated protein kinase loss-of-function mutation responsible for Wolff-Parkinson-White syndrome. Circulation. 2005; 111: 21–29.
Arad M, Moskowitz IP, Patel VV, Ahmad F, Perez-Atayde AR, Sawyer DB, Walter M, Li GH, Burgon PG, Maguire CT, Stapleton D, Schmitt JP, Guo XX, Pizard A, Kupershmidt S, Roden DM, Berul CI, Seidman CE, Seidman JG. Transgenic mice overexpressing mutant PRKAG2 define the cause of Wolff-Parkinson-White syndrome in glycogen storage cardiomyopathy. Circulation. 2003; 107: 2850–2856.
Patel VV, Arad M, Moskowitz IP, Maguire CT, Branco D, Seidman JG, Seidman CE, Berul CI. Electrophysiologic characterization and postnatal development of ventricular pre-excitation in a mouse model of cardiac hypertrophy and Wolff-Parkinson-White syndrome. J Am Coll Cardiol. 2003; 42: 942–951.
Xing Y, Musi N, Fujii N, Zou L, Luptak I, Hirshman MF, Goodyear LJ, Tian R. Glucose metabolism and energy homeostasis in mouse hearts overexpressing dominant negative 2 subunit of AMP-activated protein kinase. J Biol Chem. 2003; 278: 28372–29377.
Musi N, Fujii N, Hirshman MF, Ekberg I, Froberg S, Ljungqvist O, Thorell A, Goodyear LJ. AMP-activated protein kinase (AMPK) is activated in muscle of subjects with type 2 diabetes during exercise. Diabetes. 2001; 50: 921–927.
Brown BI. Diagnosis of glycogen storage disease. In: Wapnir PA, ed. Congenital Metabolic Diseases. New York: Dekker; 1985: 227–250.
Brown BI, Brown DH. Branching enzyme activity of cultured amniocytes and chorionic villi: prenatal testing for type IV glycogen storage disease. Am J Hum Genet. 1989; 44: 378–381.
Hays AP, Hallett M, Delfs J, Morris J, Sotrel A, Shevchuk MM, DiMauro S. Muscle phosphofructokinase deficiency: abnormal polysaccharide in a case of late-onset myopathy. Neurology. 1981; 31: 1077–1086.
Tuchman M, Brown BI, Burke BA, Ulstrom RA. Clinical and laboratory observations in a child with hepatic phosphorylase kinase deficiency. Metabolism. 1986; 35: 627–633.
Musi N, Hayashi T, Fujii N, Hirshman MF, Witters LA, Goodyear LJ. AMP-activated protein kinase activity and glucose uptake in rat skeletal muscle. Am J Physiol Endocrinol Metab. 2001; 280: E677–E684.
Stapleton D, Mitchelhill KI, Gao G, Widmer J, Michell BJ, Teh T, House CM, Fernandez CS, Cox T, Witters LA, Kemp BE. Mammalian AMP-activated protein kinase subfamily. J Biol Chem. 1996; 271: 611–614.
Bak MI, Ingwall JS. Acidosis during ischemia promotes adenosine triphosphate resynthesis in postischemic rat heart: in vivo regulation of 5'-nucleotidase. J Clin Invest. 1994; 93: 40–49.
Spindler M, Saupe KW, Christe ME, Sweeney HL, Seidman CE, Seidman JG, Ingwall JS. Diastolic dysfunction and altered energetics in the MHC403/+ mouse model of familial hypertrophic cardiomyopathy. J Clin Invest. 1998; 101: 1775–1783.
Raben N, Danon M, Lu N, Lee E, Shliselfeld L, Skurat AV, Roach PJ, Lawrence JC Jr, Musumeci O, Shanske S, DiMauro S, Plotz P. Surprises of genetic engineering: a possible model of polyglucosan body disease. Neurology. 2001; 56: 1739–1745.
Fewell JG, Osinska H, Klevitsky R, Ng W, Sfyris G, Bahrehmand F, Robbins J. A treadmill exercise regimen for identifying cardiovascular phenotypes in transgenic mice. Am J Physiol. 1997; 273: H1595–H1605.
Berul CI, McConnell BK, Wakimoto H, Moskowitz IP, Maguire CT, Semsarian C, Vargas MM, Gehrmann J, Seidman CE, Seidman JG. Ventricular arrhythmia vulnerability in cardiomyopathic mice with homozygous mutant myosin-binding protein C gene. Circulation. 2001; 104: 2734–2739.
Javadpour MM, Tardiff JC, Pinz I, Ingwall JS. Decreased energetics in murine hearts bearing the R92Q mutation in cardiac troponin T. J Clin Invest. 2003; 112: 768–775.
Vissing J, Haller RG. The effect of oral sucrose on exercise tolerance in patients with McArdle’s disease. N Engl J Med. 2003; 349: 2503–2509.
Wu WC, Bhavsar JH, Aziz GF, Sadaniantz A. An overview of stress echocardiography in the study of patients with dilated or hypertrophic cardiomyopathy. Echocardiography. 2004; 21: 467–475.
Zou L, Shen M, Arad M, He H, Lofgren B, Ingwall JS, Seidman CE, Seidman JG, Tian R. N488I mutation of the 2-subunit results in bidirectional changes in AMP-activated protein kinase activity. Circ Res. 2005; 97: 323–328.(Ferhaan Ahmad, MD, PhD; M)
Cardiovascular Institute (F.A.), Department of Medicine, and Department of Human Genetics, University of Pittsburgh, Pittsburgh, Pa
Heart Institute (M.A.), Sheba Medical Center and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Research Division (N.M., L.J.G.), Joslin Diabetes Center, Boston, Mass
NMR Laboratory for Physiological Chemistry (H.H., Y.X., R.T., J.S.I.), Cardiovascular Division, Department of Medicine, Brigham and Women’s Hospital, Boston, Mass
Department of Cardiology (C.W., D. Branco, C.I.B.), Children’s Hospital, Boston, Mass
Department of Pathology (A.R.P.-A.), Children’s Hospital, Boston, Mass
Bio21 Molecular Science and Biotechnology Institute (D.S.), University of Melbourne, Melbourne, Australia
Pediatric Medical Genetics (D. Bali), Duke University Medical Center, Research Triangle Park, NC
Cardiovascular Division (C.E.S.), Brigham and Women’s Hospital, Boston, Mass.
Abstract
Background— AMP-activated protein kinase (AMPK) regulatory 2 subunit (PRKAG2) mutations cause a human cardiomyopathy with cardiac hypertrophy, preexcitation, and glycogen deposition. PRKAG2 cardiomyopathy is recapitulated in transgenic mice overexpressing mutant PRKAG2 N488I in the heart (TG2N488I). AMPK is a heterotrimeric kinase consisting of 1 catalytic () and 2 regulatory ( and ) subunits. Two -subunit isoforms, 1 and 2, are expressed in the heart; however, the contribution of AMPK utilization of these subunits to PRKAG2 cardiomyopathy is unknown. Mice overexpressing a dominant-negative 2 subunit of AMPK (TG2DN) provide a tool for selectively inhibiting 2, but not 1, subunit-associated AMPK activity.
Methods and Results— In compound-heterozygous TG2N488I/TG2DN mice, AMPK activity associated with 2 but not 1 was decreased compared with TG2N488I. The TG2DN transgene reduced the disease phenotype of TG2N488I, partially or completely normalizing the ECG, cardiac function, cardiac morphology, and exercise capacity in compound-heterozygous mice. TG2N488I hearts had normal resting levels of high-energy phosphates and could improve cardiac performance during exercise. Cardiac glycogen content decreased in TG2N488I mice after exercise stress, indicating availability of the stored glycogen for metabolic utilization. No differences in glycogen-metabolizing enzymes were observed.
Conclusions— The PRKAG2 N488I mutation causes inappropriate AMPK activation, which leads to glycogen accumulation and conduction system disease. The accumulated glycogen can serve as an energy source, and the animals have contractile reserve during exercise. Because the dominant-negative 2 subunit attenuates the mutant PRKAG2 phenotype, AMPK complexes containing the 2 rather than the 1 subunit are the primary mediators of the effects of PRKAG2 mutations.
Key Words: cardiomyopathy exercise genetics metabolism Wolff-Parkinson-White syndrome
Introduction
Mutations in AMP-activated protein kinase (AMPK) have been discovered to cause a glycogen storage cardiomyopathy that ultimately progresses to heart failure1–3 (for reviews, see Gollob4 and Gollob et al5). AMPK, a heterotrimeric protein composed of , , and subunits, is a metabolite-sensing protein kinase that is activated under conditions of energy depletion manifested by increased cellular AMP levels (reviewed in Hardie6). AMPK is able to phosphorylate >15 known target proteins and is thought to regulate energy metabolism by modulating a variety of metabolic activities, including glucose transport, stimulation of -oxidation of fatty acids, inactivation of cholesterol synthesis, inhibition of creatine kinase, and transcriptional regulation of several genes.6,7 Mutations in the gene PRKAG2, encoding the 2 subunit of AMPK, have been demonstrated to produce a distinct cardiomyopathy in human families characterized by ventricular hypertrophy, ventricular preexcitation, and progressive conduction system disease.1–3,8 Mutations recognized thus far are Exon5:InsLeu, His142Arg, Arg302Gln, Thr400Asn, Tyr487His, Asn488Ile, and Arg531Gly. The cardiomyopathy associated with PRKAG2 mutations is not associated with the myocyte and myofibrillar disarray and cardiac fibrosis characteristic of mutations in genes encoding sarcomeric proteins. Rather, cardiac hypertrophy results from formation within the myocytes of vacuoles filled with glycogen-associated granules.1,9 PRKAG2-mediated cardiomyopathy, unlike other glycogen storage disorders, is characterized by cardiac glycogen accumulation in the absence of other remarkable clinical features, such as skeletal myopathy. The mechanisms by which these PRKAG2 missense mutations mediate cardiomyopathy and the accessibility of the accumulated glycogen as an energy source remain uncertain.
Clinical Perspective p 3148
Both cell and murine models of PRKAG2 cardiomyopathy have been described.1,9–11 Several in vitro studies suggest that PRKAG2 cardiomyopathy is caused by increased AMPK activity.1 Introduction of the Thr400Asn and Asn488Ile mutations into the yeast homolog of the subunit, Snf4, resulted in constitutive activation of the Snf1/Snf4 kinase.1 A similar Arg70Gln mutation in the 1 subunit introduced into COS7 cells and pulmonary fibroblasts caused markedly increased AMPK activity, associated with increased phosphorylation of the subunit Thr172 and increased phosphorylation of 1 of its major substrates, acetyl coenzyme A carboxylase.9 However, introduction of several mutant PRKAG2s into mammalian CCL13 cells caused either a decrease or no change in AMPK activity.10,11 Furthermore, recent studies by Sidhu and colleagues12 have suggested that the PRKAG2 Arg302Gln missense mutation inactivates AMPK and thereby causes glycogen accumulation and cardiac hypertrophy.
TG2N488I mice provide a model of PRKAG2 cardiomyopathy because they express a transgene encoding the Asn488Ile (N488I) mutation, which is expressed only in the heart under control of the -myosin heavy-chain promoter.13 TG2N488I hearts demonstrate elevated AMPK activity, massive glycogen deposition in cardiac myocytes up to 30-fold above normal, dramatic left ventricular (LV) hypertrophy, ventricular preexcitation, and sinus node dysfunction. Disruption of the annulus fibrosus by glycogen-filled myocytes is the anatomic substrate for preexcitation.13,14 TG2N488I extracts have markedly elevated AMPK activity,13 and this increased activity is associated with both the 1 and 2 subunits.
TG2N488I mice provide a unique tool for dissecting the mechanisms leading to PRKAG2 cardiomyopathy. We demonstrate, using a novel murine stress-echocardiography procedure, that although energy metabolism remains unchanged in the resting TG2N488I heart, the accumulated glycogen in the TG2N488I heart can be mobilized and metabolized during exercise. Furthermore, we use a genetic approach to confirm the model that PRKAG2 cardiomyopathy results from hyperactive AMPK, hypothesizing that if the PRKAG2 missense mutations function by inactivating AMPK, then dominant-negative 2-subunit mutations15 would not alter the phenotype of transgenic mice. Because the transgenic allele TG2DN, which expresses a dominant-negative form of the AMPK 2 catalytic subunit, inactivates AMPK containing the 2 but not the 1 subunit,15 we are able to define the relative contributions of AMPK containing these subunits. We conclude that PRKAG2 cardiomyopathy results from overactive AMPK associated primarily with the 2 rather than the 1 subunit.
Methods
Mice
Transgenic mice overexpressing human full-length, wild-type (TG2wt) and N488I (TG2N488I) PRKAG2 have been described.13 In brief, a TA substitution was introduced at nucleotide 1553 in the human PRKAG2 cDNA to alter codon 488 from AsnIle, corresponding to a known human mutation. The transgene is expressed under control of the strong cardiac-specific -myosin heavy-chain promoter. Transgenic mice overexpressing a dominant-negative 2 subunit of AMPK (TG2DN) under control of the same -myosin heavy-chain promoter were generated as previously reported.15 The AspAla substitution at codon 157 in rat PRKAA2 cDNA creates mutant 2 subunits, which bind to 2 subunits, producing enzymatically inactive complexes. The 2DN polypeptide also contains a YPYDVPDYA (HA, a synthetic peptide) tag. All mouse studies were performed in accordance with protocols approved by the institutional animal care and use committee at Harvard Medical School.
AMPK Activity
Resting mice were humanely killed by cervical dislocation without anesthesia, and the heart tissue was immediately frozen in LN2. Freeze-clamped heart samples were homogenized, and lysates were used for AMPK activity assays as described.16 Protein (200 μg) was immunoprecipitated with antibodies recognizing amino acids 339 to 358 of rat AMP 1 or 352 to 366 of 2. The kinase reaction was performed, in the presence of 0.2 mmol/L AMP, with synthetic substrate for AMP-activated protein kinase (SAMS) peptide as a substrate, and AMPK activity was measured as picomoles phosphate incorporated per milligram protein per minute.
Glycogen Enzyme Activity Assays
Hearts from wild-type, TG2wt, and TG2N488I mice were excised. Brancher enzyme or 1,4--glucan branching enzyme [1,4--D-glucan:1,4--D-glucan 6--D-(1,4--D-glucano)transferase; EC 2.4.1.18] and glycogen phosphorylase (1,4--D-glucan:phosphate -D-glucosyltransferase; EC 2.4.1.1) were measured by free phosphate levels with inorganic phosphorus reagent (Roche Diagnostics).17,18 Debrancher enzyme or 4--glucanotransferase (4--glucanotransferase 1,4--D-glucan:1,4--D-glucan 4--D-glycosyltransferase; EC 2.4.1.25) and glycogen phosphorylase kinase (ATP:phosphorylase- phosphotransferase; EC 2.7.1.38) were measured by release of free glucose from Roche glucose reagent according to a standard protocol.17,19,20
Immunoblots
Antibodies recognizing 1, 2, and 2 subunits of AMPK were produced as previously described.15,21,22 A lamin B antibody was used as a loading control (sc-6217, Santa Cruz).
Energetics Assays
Isolated, Langendorff-perfused hearts were freeze-clamped, and myocardial levels of adenine nucleotides and phosphocreatine were determined by a high-performance liquid chromatography method as reported previously.23 31P nuclear magnetic resonance (NMR) spectra were obtained as previously described24 at 161.94 MHz with a GE-400 wide-bore spectrometer. Hearts were placed in a 10-mm glass NMR tube and inserted into a custom-made, 1H/31P double-tuned probe situated in an 89-mm bore, 9.4-T superconducting magnet. The resonance area corresponding to phosphocreatine was fitted to Lorentzian functions and calculated according to a commercially available program (NMR1).
Cardiac Glycogen and Lactate Levels
Glycogen content was determined by perchloric acid extraction and amyloglucosidase digestion, followed by determination of glucose levels with a glucose oxidase kit (Sigma-Aldrich).25 Lactate levels were determined according to Sigma technical bulletin 826-UV with the use of NAD, glycine, and lactate dehydrogenase.
Exercise Stress Testing
To assess exercise capacity, mice were placed on a treadmill with an adjustable belt speed and incline and electric shock bars. Mice were subjected to an endurance protocol as previously described.26 Each mouse was encouraged to run to the limit of its capacity on this protocol, and total exercise duration was recorded.
Stress Echocardiography
The treadmill exercise protocol was modified to bring mice rapidly to maximum exercise capacity. Exercise was initiated at a treadmill incline of 15° and a speed of 10 m/min. The speed was increased by 2.5 m/min every 3 minutes until each mouse reached its maximum capacity and appeared exhausted. Mice were acclimated to the stress protocol by 1 practice exercise session 1 day before stress echocardiography. Echocardiography was performed with an Agilent Sonos 4500 ultrasound machine and a 6- to 15-MHz linear-array transducer as previously described,23 with the exception that no anesthesia was used and that immediate postexercise echocardiography was limited to M-mode views. LV end-diastolic and end-systolic diameters and wall thicknesses were obtained from M-mode tracings from measurements averaged from 3 separate cardiac cycles. LV fractional shortening (in percent) was derived from the equation fractional shortening=[(LV end-diastolic diameter–LV end-systolic diameter)/LV end-diastolic diameter]x100. Echocardiography was completed in unanesthetized animals within 5 minutes before exercise and within 30 seconds after peak exercise.
Static and Continuous Ambulatory (Holter) ECG
ECG was performed as previously described.14,27 Limb-lead ECGs were recorded to detect evidence of ventricular preexcitation, namely, shortened PR intervals and delta waves. In a separate set of experiments, a telemetry device was implanted to allow continuous ambulatory ECG (Holter monitoring) and assessment of the chronotropic response to exercise. Mice were exercised to maximum capacity according to the exercise stress testing protocol described for stress echocardiography and were monitored for another 10 minutes during recovery.
Data Analysis
All data are presented as mean±SD, except as noted in the figure legends. Comparisons between groups were made with Student’s t test for pairwise comparisons and ANOVA for multiple comparisons.
Results
Inhibiting AMPK in TG2N488I Mice With 2 Dominant-Negative Transgene
We have previously presented evidence suggesting that the N488I mutation in the 2 subunit results in increased AMPK activity.1,13 Xing et al15 have described a transgenic mouse, designated TG2DN, that expresses a dominant-negative form of the 2 subunit of AMPK. TG2DN mice have significantly reduced 2 activity but normal 1 activity.15 By mating TG2DN mice with TG2N488I transgenic mice, mice carrying both the TG2N488I transgene and the TG2DN transgene,15 designated TG2N488I/TG2DN, were obtained. Levels of TG2N488I polypeptide were the same in TG2N488I and TG2N488I/TG2DN hearts (Figure 1C). However, the nonfunctional TG2DN peptide, which migrates as a higher-molecular-weight polypeptide owing to its HA tag, completely replaced the lower-molecular-weight, endogenous, wild-type 2 subunit (Figure 1B). The 1-subunit levels in wild-type, TG2N488I, TG2DN, and TG2N488I/TG2DN hearts were not significantly different (Figure 1A), as suggested previously.15
We assessed the amount of AMPK activity associated with the 1 and 2 subunits in 18-day-old transgenic mice (Table 1). TG2N488I/TG2DN mice demonstrated a significant reduction in AMP-stimulated, 2 subunit–associated AMPK activity, to 37% of that seen in TG2N488I mice (Table 1) but no significant reduction in 1-associated AMPK activity compared with TG2N488I mice. Consequently, we proceeded to phenotypic characterization of these mice to define the roles of the 1 and 2 subunit–associated AMPK activities.
Cardiac function and morphology were assessed in TG2N488I, TG2DN, and TG2N488I/TG2DN mice. At age 5 weeks, heart-body mass ratios were indistinguishable between wild-type and TG2DN mice, whereas TG2N488I mice demonstrated massive cardiac hypertrophy (Table 1). In contrast, age-matched TG2N488I/TG2DN mice had near-normal heart-body mass ratios (Table 1). The heart-body mass ratio in TG2N488I/TG2DN mice remained mildly elevated, at a level similar to that found in TG2wt mice (6.9±1.0 versus 7.2±0.3).13 Echocardiographic studies performed in 8-week-old animals demonstrated a similar significant reduction in LV wall thickness in TG2N488I/TG2DN hearts compared with TG2N488I hearts (Table 1; 1.17 versus 1.54 mm; P=0.01). These differences persisted at age 20 weeks (Table 1).
The reduction in cardiac hypertrophy in TG2N488I/TG2DN mice was accompanied by an 8-fold decrease in glycogen content compared with TG2N488I hearts (Table 1). Both light and electron microscopy confirmed the disappearance of glycogen-laden vacuoles, the pathological hallmark of PRKAG2 cardiomyopathy (Figure 2).
TG2N488I mice have ventricular preexcitation, manifested by a short PR interval and delta waves on ECG.13,14 The PR interval was 24% longer in TG2N488I/TG2DN mice relative to TG2N488I mice (P=0.02), representing normal atrioventricular conduction without ventricular preexcitation (Figure 3 and Table 1). As we previously reported, TG2N488I mice develop LV dilatation and systolic dysfunction, phenomena not observed in wild-type and TG2wt mice.13 However, TG2N488I/TG2DN mice demonstrated completely normal LV end-diastolic diameters and fractional shortening at 20 weeks of age (Table 1).
The maximal exercise capacity of wild-type, TG2wt, TG2N488I, and TG2N488I/TG2DN mice was assessed to determine the consequences of cardiac morphological and metabolic changes. Wild-type, TG2wt, TG2DN, and TG2N488I/TG2DN mice demonstrated similar exercise capacities (Figure 4A and data not shown). However, TG2N488I mice demonstrated an exercise capacity approximately half that of wild-type mice (32.4±3.6 versus 58.1±6.0 minutes, P=0.001; Figure 4A). This exercise deficit was completely eliminated in double-transgenic TG2N488I/TG2DN mice.
Energy Reserve in TG2N488I Mice
To determine the physiological basis underlying impaired exercise capacity in TG2N488I mice, we measured resting cardiac levels of phosphocreatine by NMR and of ATP, ADP, and AMP by high-performance liquid chromatography and assessed chronotropic and contractile responses to exercise. Because defects in glycogen metabolism, as well as cardiomyopathies secondary to mutations in genes encoding sarcomeric proteins, have been found to exhibit altered energetics,24,28,29 we hypothesized that TG2N488I mice might have abnormal energy stores. However, we detected no significant differences in phosphocreatine, ATP, ADP, and AMP levels among wild-type, TG2wt, and TG2N488I in isolated, perfused hearts from 5-week-old mice (Table 2).
On stress echocardiography, TG2N488I mice had lower baseline cardiac contractility (Table 3). However, TG2N488I mice had an enhanced inotropic response to exercise, increased fractional shortening, and decreased LV end-systolic dimension, thereby suggesting adequate contractile reserve. Contrasted with wild-type mice, TG2N488I mice demonstrated significantly reduced heart rates at baseline, during exercise, and during recovery (Figure 4B; P=0.004). Heart rates of wild-type, TG2wt, TG2DN, and TG2N488I/TG2DN mice were not significantly different (data not shown). Therefore, decreased exercise performance in TG2N488I mice appears to be attributable to the impairment in cardiac contractility and lower heart rates at baseline and during exercise, but not to defective energetics.
To determine whether the increased glycogen in the cardiac myocytes of TG2N488I mice could be metabolized, we measured cardiac tissue glycogen and lactate levels in hearts from resting and exercised mice (Table 4). Exercised mice were humanely killed within 10 seconds after completion of maximal exercise. Wild-type and TG2wt mice demonstrated significant (at least 40%) decreases in glycogen levels with exercise. Glycogen levels decreased by 25% in TG2N488I hearts, which represented a very large absolute quantity of glycogen. Concomitantly, more lactate was produced, consistent with the model that stored glycogen was utilized as an energy source during exercise.
Glycogen Metabolism in TG2N488I Hearts
In an attempt to define the mechanism leading to glycogen accumulation in TG2N488I hearts, we assayed the activity of 4 enzymes involved in either glycogen synthesis or glycogenolysis, which are known to cause human glycogen storage cardiomyopathies when defective. No significant differences were detected in the activities of brancher enzyme, debrancher enzyme, glycogen phosphorylase, and glycogen phosphorylase kinase in 5-week-old TG2N488I and wild-type hearts (Table 2).
Discussion
TG2N488I mice provide a useful tool for gaining further insights into the mechanisms by which mutations in the 2 subunit of AMPK cause cardiac glycogen accumulation, preexcitation, cardiomyopathy, and heart failure. Here we provide evidence that (1) PRKAG2 cardiomyopathy results from activation of AMPK, which is mediated primarily via 2 subunit–associated AMPK, and (2) there is no cardiac energy depletion in PRKAG2 cardiomyopathy at rest, and the presence of contractile reserve in response to exercise challenge is observed on stress exercise echocardiography. Despite the pivotal role of AMPK in maintaining energy metabolism, inappropriate activation of AMPK has minimal effects on resting cardiac high-energy phosphate levels.
Murine models of hypertrophic cardiomyopathy resulting from mutations in genes encoding sarcomeric proteins have demonstrated impaired cellular energetics.24,28 In contrast, TG2N488I mice exhibit normal resting cardiac cellular energetics (Table 2). Thus, although AMPK is a central regulator of cellular energy levels, the N488I mutation does not perturb resting cellular energy levels in the heart. Presumably, homeostatic mechanisms preserve normal energetics, at least in the absence of stress. The functional impairment observed in these mice is likely the result of glycogen deposition in the heart rather than energy depletion. The normal resting cellular energetics observed in TG2N488I mice also underscores the fact that the mechanism of cardiac hypertrophy and cardiomyopathy resulting from this mutation is distinct from the hypertrophic cardiomyopathy caused by sarcomere protein gene mutations.
Stress echocardiography allows direct assessment of cardiac function in response to exercise challenge and is a standard technique for evaluating myocardial ischemia in humans with coronary heart disease. Although this test has been found to have prognostic value in patients with hypertrophic or dilated cardiomyopathy, with impaired contractile response to stress connoting a poorer prognosis,30 few if any patients with PRKAG2 missense mutations have undergone this test. To evaluate further cardiac energetics during exercise in TG2N488I mice, we developed a stress echocardiography protocol for assessing cardiac function in mice. Stressed TG2N488I mice demonstrated marked improvement in cardiac function, as fractional shortening increased from 31% to 46% (P<0.0001; Table 3). The improved cardiac function of exercised TG2N488I mice may have resulted from any 1 of several mechanisms, including increased glycogen utilization. In contrast, fractional shortening was maximal at rest in wild-type and TG2wt mice and did not improve further with exercise.
Hypothesizing that improved cardiac function could result from increased glycogen utilization, we investigated whether the excess stored glycogen in TG2N488I mice could be used during exercise. TG2N488I mice demonstrated a decrease in cardiac glycogen levels with exercise, indicating that the excess glycogen in their hearts could be mobilized and metabolized (Table 4). TG2N488I mice also demonstrated an increase in cardiac lactate, presumably from increased glycogen metabolism. The large decrease in glycogen levels was even more remarkable, considering that TG2N488I mice have a markedly reduced exercise tolerance. These results raise the possibility that mobilization of cardiac glycogen allows improved cardiac function.
We and others have previously suggested that PRKAG2 cardiomyopathy results from inappropriate activation of AMPK.1,9,13 However, other investigators have proposed other models, including a biphasic response manifested by increased resting activity but impaired responsiveness to metabolic stress.10–12,31 We have used a genetic approach to determine whether the principal effect of the N488I mutation is to increase or decrease AMPK activity. If the phenotypic effects of this mutation result from an increase in AMPK 2-subunit activity, then they should be attenuated or abolished in the presence of a dominant-negative AMPK 2 subunit. The features of double-transgenic mice carrying both mutations (TG2N488I/TG2DN) are consistent with the model that PRKAG2 with the N488I mutation inappropriately activates the AMPK 2 subunit. In TG2N488I/TG2DN mice, normal catalytic 2 subunits were replaced by an inactive form of the 2 subunit (Figure 1). As a result, TG2N488I/TG2DN mice had a significant reduction in 2-, but not 1-, subunit activity, and exhibited normal or near-normal cardiac function, cardiac morphology, and conduction system function (Table 1 and Figures 2 through 4 ). Hence, reducing AMPK-associated 2-subunit activity restores cardiac physiology to nearly normal, consistent with the model that in the presence of the N488I 2 subunit, the AMPK 2 subunit is inappropriately active. Moreover, despite the 2-fold increase in 1 subunit–associated AMPK activity observed in TG2N488I mice,13 cardiac hypertrophy was almost completely reversed in the presence of a dominant-negative 2 subunit, indicating that hypertrophy is mediated largely through the effects of mutations in 2-subunit activity. Although the mechanism is uncertain, these data suggest that in vivo 2-associated AMPK and 1-associated AMPK have different biological targets. This target specificity may be determined directly by the subunit or by a preference of the subunit for a particular or isoform, which may be responsible for target specificity.
Although inappropriately active 2 subunit–associated AMPK activity is responsible for the phenotype of this cardiomyopathy, the mechanism of glycogen accumulation is still unclear. The activities of branching enzyme, debranching enzyme, glycogen phosphorylase, and glycogen phosphorylase kinase are unchanged in adult TG2N488I hearts (Table 2). These enzymes are involved in the synthesis or degradation of glycogen, and deficiencies in each of these enzymes cause glycogen storage disease. However, regulation of 1 or more glycogen metabolic enzyme activities must be altered in PRKAG2 cardiomyopathy.
PRKAG2 cardiomyopathy in TG2N488I mice provides a unique opportunity to investigate the role of AMPK and glycogen in regulating cardiac energy metabolism. Eventually, cardiac AMPK targets and the mechanisms leading from AMPK activation to glycogen storage will be defined. Identification of these targets, coupled with an understanding of the physiological pathways by which glycogen accumulates in cardiac myocytes, will provide novel therapeutic targets for a variety of conditions associated with defective cardiac energetics.
Acknowledgments
This work was supported by the Howard Hughes Medical Institute (F.A., M.A., C.E.S., J.G.S.), the Canadian Institutes of Health Research (F.A.), the National Institutes of Health (HL 67970 to R.T. and HL 52320 and HL 63985 to J.S.I.), the American Heart Association (R.T.), and the American Diabetes Association (N.M., L.G.).
References
Arad M, Benson DW, Perez-Atayde AR, McKenna WJ, Sparks EA, Kanter RJ, McGarry K, Seidman JG, Seidman CE. Constitutively active AMP kinase mutations cause glycogen storage disease mimicking hypertrophic cardiomyopathy. J Clin Invest. 2002; 109: 357–362.
Blair E, Redwood C, Ashrafian H, Oliveira M, Broxholme J, Kerr B, Salmon A, Ostman-Smith I, Watkins H. Mutations in the (2) subunit of AMP-activated protein kinase cause familial hypertrophic cardiomyopathy: evidence for the central role of energy compromise in disease pathogenesis. Hum Mol Genet. 2001; 10: 1215–1220.
Gollob MH, Green MS, Tang AS, Gollob T, Karibe A, Ali Hassan AS, Ahmad F, Lozado R, Shah G, Fananapazir L, Bachinski LL, Roberts R, Hassan AS. Identification of a gene responsible for familial Wolff-Parkinson-White syndrome. N Engl J Med. 2001; 344: 1823–1831.
Gollob MH. Glycogen storage disease as a unifying mechanism of disease in the PRKAG2 cardiac syndrome. Biochem Soc Trans. 2003; 31: 228–231.
Gollob MH, Green MS, Tang AS, Roberts R. PRKAG2 cardiac syndrome: familial ventricular preexcitation, conduction system disease, and cardiac hypertrophy. Curr Opin Cardiol. 2002; 17: 229–234.
Hardie DG. Minireview: the AMP-activated protein kinase cascade: the key sensor of cellular energy status. Endocrinology. 2003; 144: 5179–5183.
Kemp BE, Mitchelhill KI, Stapleton D, Michell BJ, Chen ZP, Witters LA. Dealing with energy demand: the AMP-activated protein kinase. Trends Biochem Sci. 1999; 24: 22–25.
Gollob MH, Seger JJ, Gollob TN, Tapscott T, Gonzales O, Bachinski L, Roberts R. Novel PRKAG2 mutation responsible for the genetic syndrome of ventricular preexcitation and conduction system disease with childhood onset and absence of cardiac hypertrophy. Circulation. 2001; 104: 3030–3033.
Hamilton SR, Stapleton D, O’Donnell JB Jr, Kung JT, Dalal SR, Kemp BE, Witters LA. An activating mutation in the 1 subunit of the AMP-activated protein kinase. FEBS Lett. 2001; 500: 163–168.
Scott JW, Hawley SA, Green KA, Anis M, Stewart G, Scullion GA, Norman DG, Hardie DG. CBS domains form energy-sensing modules whose binding of adenosine ligands is disrupted by disease mutations. J Clin Invest. 2004; 113: 274–284.
Daniel T, Carling D. Functional analysis of mutations in the 2 subunit of AMP-activated protein kinase associated with cardiac hypertrophy and Wolff-Parkinson-White syndrome. J Biol Chem. 2002; 277: 51017–51024.
Sidhu JS, Rajawat YS, Rami TG, Gollob MH, Wang Z, Yuan R, Marian AJ, DeMayo FJ, Weilbacher D, Taffet GE, Davies JK, Carling D, Khoury DS, Roberts R. Transgenic mouse model of ventricular preexcitation and atrioventricular reentrant tachycardia induced by an AMP-activated protein kinase loss-of-function mutation responsible for Wolff-Parkinson-White syndrome. Circulation. 2005; 111: 21–29.
Arad M, Moskowitz IP, Patel VV, Ahmad F, Perez-Atayde AR, Sawyer DB, Walter M, Li GH, Burgon PG, Maguire CT, Stapleton D, Schmitt JP, Guo XX, Pizard A, Kupershmidt S, Roden DM, Berul CI, Seidman CE, Seidman JG. Transgenic mice overexpressing mutant PRKAG2 define the cause of Wolff-Parkinson-White syndrome in glycogen storage cardiomyopathy. Circulation. 2003; 107: 2850–2856.
Patel VV, Arad M, Moskowitz IP, Maguire CT, Branco D, Seidman JG, Seidman CE, Berul CI. Electrophysiologic characterization and postnatal development of ventricular pre-excitation in a mouse model of cardiac hypertrophy and Wolff-Parkinson-White syndrome. J Am Coll Cardiol. 2003; 42: 942–951.
Xing Y, Musi N, Fujii N, Zou L, Luptak I, Hirshman MF, Goodyear LJ, Tian R. Glucose metabolism and energy homeostasis in mouse hearts overexpressing dominant negative 2 subunit of AMP-activated protein kinase. J Biol Chem. 2003; 278: 28372–29377.
Musi N, Fujii N, Hirshman MF, Ekberg I, Froberg S, Ljungqvist O, Thorell A, Goodyear LJ. AMP-activated protein kinase (AMPK) is activated in muscle of subjects with type 2 diabetes during exercise. Diabetes. 2001; 50: 921–927.
Brown BI. Diagnosis of glycogen storage disease. In: Wapnir PA, ed. Congenital Metabolic Diseases. New York: Dekker; 1985: 227–250.
Brown BI, Brown DH. Branching enzyme activity of cultured amniocytes and chorionic villi: prenatal testing for type IV glycogen storage disease. Am J Hum Genet. 1989; 44: 378–381.
Hays AP, Hallett M, Delfs J, Morris J, Sotrel A, Shevchuk MM, DiMauro S. Muscle phosphofructokinase deficiency: abnormal polysaccharide in a case of late-onset myopathy. Neurology. 1981; 31: 1077–1086.
Tuchman M, Brown BI, Burke BA, Ulstrom RA. Clinical and laboratory observations in a child with hepatic phosphorylase kinase deficiency. Metabolism. 1986; 35: 627–633.
Musi N, Hayashi T, Fujii N, Hirshman MF, Witters LA, Goodyear LJ. AMP-activated protein kinase activity and glucose uptake in rat skeletal muscle. Am J Physiol Endocrinol Metab. 2001; 280: E677–E684.
Stapleton D, Mitchelhill KI, Gao G, Widmer J, Michell BJ, Teh T, House CM, Fernandez CS, Cox T, Witters LA, Kemp BE. Mammalian AMP-activated protein kinase subfamily. J Biol Chem. 1996; 271: 611–614.
Bak MI, Ingwall JS. Acidosis during ischemia promotes adenosine triphosphate resynthesis in postischemic rat heart: in vivo regulation of 5'-nucleotidase. J Clin Invest. 1994; 93: 40–49.
Spindler M, Saupe KW, Christe ME, Sweeney HL, Seidman CE, Seidman JG, Ingwall JS. Diastolic dysfunction and altered energetics in the MHC403/+ mouse model of familial hypertrophic cardiomyopathy. J Clin Invest. 1998; 101: 1775–1783.
Raben N, Danon M, Lu N, Lee E, Shliselfeld L, Skurat AV, Roach PJ, Lawrence JC Jr, Musumeci O, Shanske S, DiMauro S, Plotz P. Surprises of genetic engineering: a possible model of polyglucosan body disease. Neurology. 2001; 56: 1739–1745.
Fewell JG, Osinska H, Klevitsky R, Ng W, Sfyris G, Bahrehmand F, Robbins J. A treadmill exercise regimen for identifying cardiovascular phenotypes in transgenic mice. Am J Physiol. 1997; 273: H1595–H1605.
Berul CI, McConnell BK, Wakimoto H, Moskowitz IP, Maguire CT, Semsarian C, Vargas MM, Gehrmann J, Seidman CE, Seidman JG. Ventricular arrhythmia vulnerability in cardiomyopathic mice with homozygous mutant myosin-binding protein C gene. Circulation. 2001; 104: 2734–2739.
Javadpour MM, Tardiff JC, Pinz I, Ingwall JS. Decreased energetics in murine hearts bearing the R92Q mutation in cardiac troponin T. J Clin Invest. 2003; 112: 768–775.
Vissing J, Haller RG. The effect of oral sucrose on exercise tolerance in patients with McArdle’s disease. N Engl J Med. 2003; 349: 2503–2509.
Wu WC, Bhavsar JH, Aziz GF, Sadaniantz A. An overview of stress echocardiography in the study of patients with dilated or hypertrophic cardiomyopathy. Echocardiography. 2004; 21: 467–475.
Zou L, Shen M, Arad M, He H, Lofgren B, Ingwall JS, Seidman CE, Seidman JG, Tian R. N488I mutation of the 2-subunit results in bidirectional changes in AMP-activated protein kinase activity. Circ Res. 2005; 97: 323–328.(Ferhaan Ahmad, MD, PhD; M)