SYMPOSIUM

Proceedings of the Australian Physiological Society Symposium

Myocardial glycogen dynamics: New perspectives on disease mechanisms

Chanchal Chandramouli,* Upasna Varma,* Ellie M Stevens,† Rui-Ping Xiao,‡ David I Stapleton,*§

Kimberley M Mellor*†1 and Lea MD Delbridge*1

*Department of Physiology, University of Melbourne, Melbourne, Vic., Australia, †Department of Physiology, Universityof Auckland, Auckland, New Zealand, ‡Institute of Molecular Medicine, Peking University, Beijing, China and

§The Florey Institute of Neuroscience, Melbourne, Vic., Australia

SUMMARY

Cardiac glycogen regulation involves a complex interplaybetween multiple signalling pathways, allosteric activation ofenzymes, and sequestration for autophagic degradation. Sig-nalling pathways appear to converge on glycogen regulatoryenzymes via insulin (glycogen synthase kinase 3b, proteinphosphatase 1, allosteric action of glucose-6-phosphate), b–adrenergic (phosphorylase kinase protein phosphatase 1inhibitor), and 50 adenosine monophosphate-activated proteinkinase (allosteric action of glucose-6-phosphate, direct glyco-gen binding, insulin receptor). While cytosolic glycogen syn-thesis and breakdown are relatively well understood, recentfindings relating to phagic glycogen degradation highlight anew area of investigation in the heart. It has been recentlydemonstrated that a specific glycophagy pathway is opera-tional in the myocardium. Proteins involved in recruiting gly-cogen to the forming phagosome have been identified.Starch–binding domain-containing protein 1 is involved inbinding glycogen and mediating membrane anchorage viainteraction with a homologue of the phagosomal protein light-chain 3. Specifically, it has been shown that starch–bindingdomain-containing protein 1 and light-chain 3 have discretephagosomal immunolocalization patterns in cardiomyocytes,indicating that autophagic trafficking of glycogen and proteincargo in cardiomyocytes can occur via distinct pathways.There is strong evidence from glycogen storage diseases thatphagic/lysosomal glycogen breakdown is important for main-taining normal cardiac glycogen levels and does not simplyconstitute a redundant ‘alternative’ breakdown route for gly-cogen. Advancing understanding of glycogen handling in theheart is an important priority with relevance not only togenetic glycogen storage diseases but also to cardiac meta-bolic stress disorders such as diabetes and ischaemia.

Key words: autophagy, cardiac, glucose, glycogen storagediseases, glycophagy, heart, metabolic stress.

PERSPECTIVES ON CARDIAC GLYCOGEN

Unremitting demand for myocardial adenosine triphosphate sup-ply to fuel unrelenting work output confers a particular require-ment for responsive matching of substrate supply and demand inthe heart. Inherent is also a need for ‘defensive’ metabolism toprotect against threatening contingency in a manner that appearsdistinctive to the cardiac muscle operating phenotype. Glycogenstores in the heart have been conventionally understood to pri-marily serve as an energy ‘buffer’ to fine-tune the availability ofglucose (titrating cardiomyocyte glycogen breakdown against glu-cose uptake) to maintain performance. More recent research hasrevealed that glycogen reservoirs have a more sophisticated rolein determining myocardial metabolic health.Early realization that cardiac glycogen regulation is dramati-

cally different to skeletal muscle glycogen regulation generatedmuch interest into comparative tissue investigations. Almost acentury ago it was reported that hypoglycaemia induced glycogendepletion in liver and skeletal muscle with remarkable preserva-tion of cardiac glycogen content.1 Similarly, an early observationwas a reduction in glycogen levels in contracting skeletal musclein vitro but a retention of glycogen levels in the aerobically per-fused isolated working heart.2,3 Additionally, the contrastingeffects of fasting, adrenergic stimulation, hypoxia, insulin, andexercise on skeletal and cardiac muscle glycogen supplies wererecognized in a number of early pioneering studies.1,2 Preserva-tion of cardiac glycogen (relative to non-cardiac tissue glycogen)is apparent in various settings of metabolic stress. These veryearly experimental findings prompted speculation that, in theheart, glucose is not liberated from glycogen during periods ofnormal activity but is reserved for ‘anoxaemic emergencies’.2,4

The fundamental bases for differential glycogen handlingresponses in distinct muscle types are still not understood.Advances in the understanding of the gene/protein origins of

various genetically determined glycogen storage diseases (GSD)have provided considerable new insight into the role of disruptedglycogen handling in the aetiology of cardiac pathology.5,6 Car-diac glycogen macromolecular morphology and localization are

Correspondence: Prof. Lea MD Delbridge, Department of Physiology,University of Melbourne, Parkville, Vic. 3010, Australia. Email:lmd@unimelb.edu.au

1These authors contributed equally to this work.Received 13 October 2014; revision 29 December 2014; accepted 6

January 2015.

Clinical and Experimental Pharmacology and Physiology (2015) 42, 415–425 doi: 10.1111/1440-1681.12370

distinctive and suggest a specific relationship between stored gly-cogen structure and functional access to stores; however, the nat-ure of this relationship has not yet been revealed. New geneticmodels recapitulating various GSD have provided importantinformation about the regulation of glycogen in the heart andmay prove useful in probing the mechanism of glycogen pathol-ogy in non-genetic heart disease.7–9 Finally, there is new evi-dence emerging of a role for lysosomal processing of glycogen, abreakdown route understood to operate in parallel with the well-characterized phosphorylase regulatory degradation pathway.Very recently our investigations of cardiac metabolic stress havedemonstrated the importance of this cardiomyocyte ‘glycophagy’response, and we have identified specific mediator proteinsinvolved.10–12

This review examines the physiology of glycogen handling inthe heart with particular focus on the current controversiesregarding glycogen regulatory signalling pathways (insulin vs b-adrenergic vs 50 adenosine monophosphate-activated proteinkinase (AMPK) signalling). We draw on the information to begleaned from some well-characterized and newly recognizedmuscle GSD for this overview, and we highlight the numerousareas of opportunity where a more complete understanding ofglycogen dynamics in the heart could potentially yield therapeuticbenefits.

GLYCOGEN FLUX PATHWAYS IN THE HEART

Glycogen morphology and subcellular localization incardiomyocytes

Glycogen is a large complex polysaccharide of glucose residueslinked in chains via a–1,4–glycosidic bonds, with branchingoccurring via a-1,6-glycosidic bonds (Fig. 1).13,14 Glycogenexists as either small b particles (less than ~50 nm) or as larger aparticles up to 300 nm in diameter that consist of many b parti-cles aggregating together by an unknown mechanism.15,16 Tissue-specific storage of a- and b-glycogen particles is evident. Liverglycogen is stored predominantly as large a particles with lowsurface-to-volume ratio.15 In skeletal muscle, only high surfacearea b-glycogen particles are evident, which is consistent withrapid glycogenolysis for bursts of muscle activity.15,17,18 Thecharacteristics of particle storage may be reflective of differentialfunctional dynamics in the operation of glycogen stores in differ-ent tissues. Electron micrographs of purified cardiac glycogenreveal that glycogen is stored as both a and b particles in theheart, which may reflect the necessity for a steady glucose supplywith additional capacity for rapid increases in workload.10,18 Thestructure–function relationships pertaining to glycogen storageand usage have yet to be determined.Glycogen levels are higher in foetal cardiomyocytes relative to

adult cardiomyocytes and occupy approximately 30% and 2% ofcell volume, respectively;19 studies with glycogen synthase knock-out mice have demonstrated that glycogen synthesis is critical fornormal heart development.20 In disease settings such as diabetes,up to a four-fold increase in glycogen has been observed in adulthearts, constituting a specific cardiac glycogen pathology.21 Thesubcellular localization of glycogen has been well described inskeletal muscle,22 but less is known with regard to cardiac muscle.Glycogen localized to the cardiac sarcoplasmic reticulum (SR) has

been identified.23,24 It is consistent with subcellular localization ofglycolytic substrates and machinery at the SR and with glycolyticdependence on SR Ca2+ release and SR Ca2+ ATPase activity.25,26

In heart sections, glycogen has also been localized to intermediatefilaments,24 and accumulation of glycogen adjacent to the interca-lated discs has been reported in right heart failure.27 In skeletalmuscle, compartmentalized roles for glycogen have been demon-strated. Intermyofibrillar glycogen localized to the SR was corre-lated with the half-relaxation time of the myocyte twitch,suggesting that this glycogen may provide the substrate for glyco-lytic adenosine triphosphate supply to facilitate Ca2+ reuptake intothe SR and subsequent twitch relaxation.28 Intramyofibrillar gly-cogen was correlated with SR Ca2+ release rate and may beinvolved in coupling glycolytic adenosine triphosphate to plasmamembrane pumps in the t-tubules.29 Two distinct subcellular poolsof cardiac glycogen have been identified by electron microscopy:free cytosolic glycogen and glycogen enclosed in double–mem-brane bound structures likely to be autophagosomes. Free cyto-solic glycogen may play a role in maintaining basal intracellularglucose levels, while membrane-bound glycogen destined for gly-cophagic breakdown of bulk glycogen may play a role in rapidrelease of glucose in response to a large increase in demand. Gly-cogen localized to autophagosomes has been reported to be partic-ularly prominent in neonatal hearts and some forms of GSD.30–32

The manner in which localization of glycogen shifts between sub-sarcolemmal and intra-fibrillar sites in pathophysiologic states andthe conditions that regulate particle flux via lysosomal breakdownhave not yet been determined. Interestingly, distinct separation ofenzymes mediating glycogen-derived glycolytic flux, as opposedto exogenous glucose-derived glycolytic flux, is apparent,33 sug-gesting differential roles for metabolic fuel from these sources. Inthe myocardium these aspects of glycogen dynamics may be par-ticularly important in functional integrity.

Cytosolic enzymatic synthesis and breakdown of glycogen

Homeostatic glycogen metabolism is primarily mediated byenzymes localized to the glycogen particle. Upon entering thecell, glucose is phosphorylated to glucose-6-phosphate (G6P) byglucokinase. Glucose-6-phosphate can enter the glycolytic path-way via conversion to fructose-6-phosphate or glucose-1-phos-phate and then to uridine diphosphate-glucose, to providesubstrate for glycogen synthesis. Glycogenin catalyses the firststep in glycogen synthesis by covalently attaching glucose toitself. The first few glucose residues of the glycogen strand areassembled by glycogenin and provide a primer for glycogen syn-thase to continue transferring uridine diphosphate–glucose resi-dues onto this polymeric glycogen strand via a-1,4-glycosidicbonds and to release uridine diphosphate.34,35 After 10–14 glu-cose residues (the precise number remains under debate),35 theglycogen-branching enzyme attaches a strand of approximatelyseven glucose residues via an a-1,6-glycosidic bond. Glycogensynthase extends this chain until the glycogen-branching enzymeadds an additional branch point. This process continues resultingin a highly branched glycogen structure. Branching acts toincrease the number of terminal glucose residues available forrelease and increases solubility. The physiologic importance ofthis branching morphology is apparent in genetic conditions inwhich genetic glycogen-branching enzyme deficiency leads to an

416 C Chandramouli et al.

Fig. 1 Glycogen structure and breakdown in the cytosol. Glycogen phosphorylase cleaves a–1,4-glycosidic bonds up to four residues from the branchpoint. The remaining residues are called the limit dextrins. Glycogen-debranching enzyme transfers the terminal 3 residues of the limit dextrins to anadjacent linear strand and then cleaves the a-1,6-glycosidic branch point bond. In the lysosome, acid-a-glucosidase cleaves both a-1,4-glycosidic bondsand a-1,6-glycosidic bonds. Alternative names for acid-a-glucosidase in the literature are lysosomal a-glucosidase, acid maltase, glucoinvertase, glucosi-dosucrase, maltase, glucan-1,4- a-glucosidase, and amyloglucosidase.

Glycogen and glycophagy in the heart 417

accumulation of insoluble glycogen (polyglucosan) with longouter branches and muscle pathology (type IV GSD).36

The breakdown of glycogen is mediated by glycogen phos-phorylase, which cleaves a–1,4–glycosidic bonds to releaseglucose–1–phosphate. The action of glycogen phosphorylase islimited to four residues from the glycogen branch point, whichare referred to as limit dextrins.13 Disassembly of the branchpoints requires the dual-acting glycogen-debranching enzyme(GDE) to first transfer the three terminal residues to anotherlinear strand (via its transferase activity) and then cleave thea-1,6-glycosidic bond (via its glucosidase activity) to liberateone free glucose and release the limit dextrins for phosphory-lase action (Fig. 1).35 Cori’s or Forbes disease (type III GSD)is characterized by GDE deficiency, where short glycogenstrands accumulate in tissues and systemic hypoglycaemia isevident.37

Proteomic analysis of the glycogen particle has identified thatall of these enzymes (with the exception of GDE) are associatedwith glycogen.38 Low levels of phosphate in the glycogen parti-cle have been detected, consistent with phosphorylation of theglycogen molecule.39 The mechanism by which phosphate isincorporated into glycogen is under debate. A role for glycogensynthase was initially proposed but recently disputed by Nitschkeet al.39,40 Laforin has been identified as the phosphatase that dep-hosphorylates glycogen, and its expression is crucial for normalglycogen structure.41 Individuals with Lafora disease exhibitmutations in laforin and present with hyperphosphorylated glyco-gen particles, which are poorly branched and lack solubility.42

Although there have been significant recent advances in theunderstanding of functional branching and debranching dynamicsin relation to glycogen management, major challenges still existwith regard to determining the subtle tissue specificity of theseenzymatic processes and, most particularly, to evaluating howbranching disturbance contributes to myocardial glycogen usagedysfunction.

Lysosomal glycogen breakdown in the heart

A second pathway for glycogenolysis exists via phagic recruit-ment and lysosomal degradation, termed ‘glycophagy’ (i.e. glyco-gen-specific autophagy).10 Electron micrographs identifyglycogen enclosed in double-membrane bound structures (likelyglycophagosomes) in neonatal cardiomyocytes.30 We haverecently demonstrated that a specific glycophagy pathway is oper-ational in the adult heart.10,11 Proteins involved in recruitingglycogen to the forming phagosome have been identified. Starch–binding domain-containing protein 1 is involved in binding gly-cogen and mediating autophagosome anchorage via its auto-phagy-related protein 8 (Atg8)-interacting motif, which appearsto be specific to the Atg8, GABAA receptor-associated protein-like 1, and a homologue of the phagosomal protein light-chain3.12,38,43 We have been able to demonstrate that starch–bindingdomain-containing protein 1 (STBD1) and light-chain 3B havediscrete phagosomal immunolocalization patterns in cardiomyo-cytes, indicating that autophagic trafficking of glycogen andprotein cargo in cardiomyocytes can occur via distinct pathways(Fig. 2).11 STBD1 has recently been shown to interact withglycogen-associated proteins, such as glycogen synthase,debranching enzyme, and laforin.44

In the lysosome, acid a-glucosidase (GAA) cleaves both the a–1,4-glycosidic bonds (without restriction by the limit dextrin) andthe a–1,6–glycosidic bonds (branch points), resulting in completedegradation of the glycogen particle to free a–glucose.45–47

Analysis of purified GAA from the human heart has demon-strated similar substrate specificity and kinetics relative to liverGAA.48 Our recent studies have demonstrated that cardiac glyco-phagy and glycogen content are simultaneously upregulated inresponse to fasting and insulin exposure,10,11 suggesting that pha-gic glycogen breakdown may be triggered by glycogen excess.Interestingly, sex differences in glycophagy were evident underbasal conditions, and women appear particularly susceptible toglycogen and glycophagy increase in response to fasting.10 Fur-ther work is required to delineate the conditions that selectivelypromote glycophagic activity and to understand how phagic andnon-phagic glycogen breakdown processes can operate in com-plementarity.

REGULATED ENZYMATIC MANAGEMENT OFCARDIAC GLYCOGEN

Regulation of cardiac glycogen synthesis by insulin

Activation of the insulin signalling pathway is linked to a varietyof downstream effects, one of which is glycogen regulation.49 Inisolated, perfused rat hearts, insulin increases the rate of glycogensynthesis and decreases the rate of glycogenolysis.50 Insulin acti-vates the phosphatidylinositiol 30-kinase (PI3K)/Akt signallingpathway by binding to the sarcolemmal tyrosine kinase insulinreceptor.51 Both in vitro and in vivo studies have demonstratedthat, in cardiomyocytes, insulin-induced activation of Akt leadsto glycogen synthesis by relieving the glycogen synthase kinase3 (GSK3)-mediated inhibition (phosphorylation) of glycogen syn-thase.52–54 Two isoforms of GSK3, a and b, are expressed in theheart; the levels of the GSK3b isoform are four-fold those of a.52

Akt inhibits both GSK3a and b by phosphorylating Ser21 andSer9 residues, respectively.52 In mice expressing constitutivelyactive GSK3a and/or b (via alanine substitution for Ser21 or

Fig. 2 Distinct cardiomyocyte immunohistologic localization of glyco-gen autophagic trafficking cargo in vitro. Starch–binding domain-contain-ing protein 1 (green) and light-chain 3B (red) do not co-localize inneonatal ventricular rat myocytes. Confocal image, 963 magnification.40,6-Diamidino-2-phenylindole (blue) shows nuclei. Reprinted with per-mission from Mellor et al.11

418 C Chandramouli et al.

Ser9), insulin activation of glycogen synthase in the heart wasreduced.52 The effect was more marked in the constitutivelyactive GSK3b mouse hearts, suggesting that this isoform plays agreater role in insulin regulation of glycogen storage.52 Addition-ally, insulin promotes activation of glycogen synthase (bydephosphorylation) by stimulating protein phosphatase-1 (PP1).55

Protein targeting to glycogen (PTG) is a scaffold protein that tar-gets PP1 to glycogen; it is crucial for regulation of glycogen con-tent. Global deletion of PTG has been observed to lowerglycogen content in the heart.56 Protein phosphatase-1 also sup-presses glycogen phosphorylase-mediated glycogen breakdown(by phosphorylase dephosphorylation).57 Furthermore, insulin-induced glucose uptake facilitated through glucose transporter 4translocation provides another stimulus for glycogen synthesis viaallosteric activation of glycogen synthase by G6P.58 Thus, insulinsignalling promotes glycogen storage via multiple targets(Fig. 3).

b-Adrenergic signalling regulation of cardiac glycogenbreakdown

The b-adrenergic signalling pathway promotes glycogen break-down by simultaneously inhibiting glycogen synthase and activat-ing glycogen phosphorylase. All three b–adrenergic sarcolemmalG-protein coupled receptors (b1–3) are expressed in the heart,with relative expression levels (b1 > b2 > b3) under normal con-ditions.59,60 When heart failure occurs, there is an increase in the

ratio of b2 : b1 receptors in the sarcolemma.61 Activation ofeither b1 or b2 receptors has been shown to produce similar ele-vation in cyclic adenosine monophosphate, but only the b1-recep-tor agonist has been observed to result in activation of proteinkinase A (PKA) and glycogen phosphorylase in vivo.62 Similarly,in the ex vivo setting, glycogen degradation was stimulated byboth b1- and b2-receptor agonists, but to a lesser extent by b2.63

Direct activation (phosphorylation) of glycogen phosphorylasekinase by PKA has been shown to lead to phosphorylation ofglycogen phosphorylase and to promote glycogen breakdown. Inhearts perfused with isoproterenol, a non-selective b–receptoragonist, PKA directly phosphorylated PP1 inhibitor.64 Whenphosphorylated, PP1 inhibitor suppresses PP1-mediated dephos-phorylation of both glycogen synthase and phosphorylase; thus,these two enzymes remain in a phosphorylated state that corre-sponds to the deactivation of synthase and activation of phos-phorylase. In isolated cardiomyocytes, PP1 activity was reducedby isoproterenol and restored by the beta-blocker propranolol.65

Thus, PKA promotes glycogen breakdown by simultaneouslyphosphorylating glycogen phosphorylase via the phosphorylasekinase and inhibiting its dephosphorylation via the PP1 inhibitor(Fig. 3).

AMPK of glycogen in the heart

AMPK is a highly conserved intracellular energy sensor complexcomprising a catalytic subunit (a) and two regulatory subunits (band c).66 The link between AMPK and glycogen content in theheart has primarily been demonstrated by activation mutations inthe c2 regulatory subunit of AMPK, which induce cardiac GSDand cardiomyopathy in humans.67,68 Patients with pre-excitationconduction abnormalities classed as Wolff–Parkinson–White syn-drome commonly exhibit mutations in the c2 regulatory subunitof AMPK.69 It has been demonstrated in mouse models of thesemutations that the electrophysiological abnormalities can bedirectly linked to disruption of the annulus fibrosis by glycogen-filled myocytes.70 Early in vitro studies using purified proteinsreported that AMPK can directly phosphorylate glycogen syn-thase and phosphorylase kinase, suggesting inhibition of glyco-gen synthesis and stimulation of glycogen degradation.71,72

However, a later study demonstrated that AMPK does not phos-phorylate phosphorylase kinase.73 More recently, in skeletal mus-cle of mice with glycogen synthase knock–in mutations, it wasdemonstrated that AMPK increased glycogen storage throughincreased glucose uptake and the subsequent G6P–mediated allo-steric activation of glycogen synthase.74 Cardiomyocyte-specificoverexpression of the c2 subunit of AMPK induced a five toseven-fold increase in sodium-dependent glucose transporter 1messenger RNA, suggesting that AMPK-mediated glucose uptakemay play a similar role in promoting G6P allosteric activation ofglycogen synthase in the heart.75 The b subunit of AMPK con-tains a glycogen–binding domain,76,77 and cytoplasmic AMPKco–localization with glycogen was abrogated by mutation of thissite.78 There is some evidence that high glycogen levels caninhibit AMPK activity,72,79 thus providing a negative feedbacksystem, but at physiological glycogen levels, glycogen binding toAMPK did not alter AMPK activation.76

In mice with c2 AMPK mutations (gene name: PRKAG2),increased cardiac glycogen content could be explained by a

Fig. 3 Regulated enzymatic management of glycogen flux: insulin andb-adrenergic modulation. Stimulation of the b-adrenergic receptor leads toprotein kinase A (PKA) activation. Protein kinase A promotes glycogenbreakdown by stimulating glycogen phosphorylase kinase-mediated phos-phorylation and inhibiting protein phosphatase 1 (PP1)-mediated dephos-phorylation of glycogen phosphorylase. Stimulation of the insulinreceptor leads to insulin receptor substrate (IRS) 1/2-mediated Akt activa-tion. Akt promotes glycogen synthesis by relieving glycogen synthasekinase 3b (GSK3b) inhibition of glycogen synthase and stimulating PP1-mediated dephosphorylation of glycogen phosphorylase. cAMP, cyclicadenosine monophosphate; G6P, glucose-6-phosphate; GP, glycogenphosphorylase.

Glycogen and glycophagy in the heart 419

compensatory overactivation of the a2 AMPK subunit.80 Upregu-lation of AMPK c2 regulatory subunit activity (by either muta-tion or overexpression) induces glycogen accumulation in theheart.81,82 In response to fasting and dexamethasone treat-ment,10,83 glycogen accumulation in the heart is observed coinci-dent with AMPK activation. Conversely, perfusion of isolated rathearts with aminoimidazole carboxamide riboside (AICAR), anAMPK agonist, induced a slight decrease in glycogen contentassociated with a dramatic increase in glycogen degradation.Interestingly, no changes in phosphorylation states of glycogensynthase or glycogen phosphorylase were observed, suggestinginvolvement from allosteric regulation. In this study, AICARinduced a four-fold increase in 5–aminoimidazole–4–carboxamide1-b-D-ribofuranotide (ZMP), a metabolite of AICAR that mimicsthe allosteric effects of adenosine monophosphate on glycogenphosphorylase activity and may explain the observed increase inglycogen degradation.84 There are conflicting reports on theeffect of AICAR on glycogen content, and in both cardiac andskeletal muscle tissue, AICAR has been reported to increase,85,86

decrease,84 or not change glycogen content.86–88 Treatment withAICAR may induce differential effects on glycogen content dueto extent of ZMP allosteric activation of glycogen phosphorylaserelative to AMPK-mediated G6P allosteric activation of glycogensynthase, and the net effect may depend on duration of AICARexposure. When AICAR was administered to rats in vivo, skeletalmuscle glycogen was not altered after 60 min, but it increasedtwo-fold after 5 days of AICAR maintained treatment.86 Also,AMPK–mediated phosphorylation of the insulin receptor was evi-dent in glucose-deprived cardiomyocytes, which may provide aroute for AMPK-driven glycogen storage via stimulation of theinsulin signalling pathway.89 Thus, there is some evidence,mainly from AMPK mutation studies, suggesting that AMPK hasa pro-glycogen storage effect on the heart, but the effect of phar-macological activation of AMPK on cardiac glycogen requiresfurther investigation. Understanding the role of AMPK in glyco-gen regulation is particularly important in a diabetic setting ofcardiac glycogen excess, as AMPK agonists commonly used aspharmacotherapy (e.g. metformin, AICAR) have the potential toexacerbate the glycogen pathology.A specific understanding of how insulin, adrenergic, and

AMPK-mediated responses are affected in a tissue-specific man-ner is still developing. In particular, the role of disturbed myocar-dial energy utilization signalling in cardiomyocyte glycogenpathology metabolic disease states has not been determined.

TERMINAL CARDIAC GLYCOGENDEGRADATION FUNCTION AND DYSFUNCTION

The importance of physiologic management of terminal glycogenbreakdown in the heart is most clearly evidenced by the nature ofcardiopathologies, which occur with mutations in those proteinsthat mediate the final stages of glycogen degradation. In mostcases, the cardiac effects of these mutations are more severe thanin non-cardiac tissues,90 despite the evidence that, under normalconditions, glycogen appears to have a more vital role in facilitat-ing muscle contraction in skeletal muscle than in myocardium.As failure of distal glycogen breakdown processes are typicallyassociated with increased cardiomyocyte glycogen content, thedescriptor of ‘glycogen storage disease’ is used, although this

‘storage’ terminology is not necessarily apt. Clinically, GSD(occurrence rate 1 in 20 000–43 000) are categorized (type 0–XIII) based on the protein mutations/deficiencies and tissueinvolvement (Table 1).5 Given the emerging understanding of thegenetic bases of GSD, it has become possible to gain newinsights into cardiac-specific glycogen-handling/mishandlingmechanisms.

Lysosomal GAA deficiency

Lysosomal GAA is the enzyme that catalyses the completebreakdown of glycogen to free a-glucose via lysosome degrada-tion, thus constituting the final stage of the recently recognizedprocess of glycogen autophagy. Deficiency of GAA results inlysosomal glycogen accumulation and formation of excessivevesicular structures (presumably arrested glycogen phagosomes)in skeletal, cardiac, and smooth muscles.95 This deficiency, whichis an autosomal recessive disorder classed as glycogen storagedisease type II, or Pompe’s disease, has a global incidence of1 in 40 000.96 The pathophysiology of GAA deficiency in theheart is not completely understood but is clinically observed ashypertrophic or dilated cardiomyopathy.32 When GAA deficiencyis severe, cardiac abnormalities include short cardiac PRinterval,97 ventricular wall thickening followed by progression todilation,98 reduced ejection fraction, and diastolic dysfunc-tion.32,95,97,99 In some patients, conduction abnormalities consis-tent with Wolff–Parkinson–White syndrome are evident andassociated with increased risk of sudden cardiac death.100,101

Experimentally, in hearts from mouse models of GAA defi-ciency, electron microscopy has revealed marked glycogendeposits in intracellular vacuoles, consistent with an accumulationof lysosomal glycogen.102 Enzyme replacement therapy (ERT)with recombinant human GAA has been shown to attenuate gly-cogen accumulation, cardiac hypertrophy, and conductanceabnormality in the hearts of GAA-deficient mice.103–105 Clini-cally, ERT with recombinant human GAA has also been reportedto diminish cardiac hypertrophy by reducing left ventricular massindex, decreasing interventricular septal and left ventricular wallthickness, and improving systolic and diastolic function.98 Theclinical efficacy of ERT may be limited due to the instability ofGAA and problems related to tissue-specific targeting andimmune response.106,107 Adjunct therapies have proven beneficialin increasing the efficacy of GAA ERT through mammalian tar-get or rapamycin (mTOR) inhibition and b2-agonist treat-ment,106,108 and pharmacological chaperones have facilitatedenzyme delivery.107 In late-onset cardiopathology, beta–blockers,angiotensin-converting-enzyme inhibitors, and diuretics are usedas supportive therapy. The phenotype of GAA deficiency demon-strates that lysosomal GAA-mediated glycogen breakdown isessential to maintain normal glycogen content in the heart. How-ever, further work is required to elucidate the interactionsbetween the two glycogen degradation pathways, cytosolic phos-phorylase and lysosomal GAA action.

Lysosome-associated membrane protein-2 (LAMP2)deficiency

Mutations in LAMP2 generate cardiac glycogen pathology simi-lar to that caused by GAA deficiency. This rare X-linked

420 C Chandramouli et al.

dominant lysosomal GSD was first clinically recognized as a con-dition mimicking features of Pompe’s disease in the absence ofGAA deficiency and was identified as Danon disease.109 Themutation was later identified to be on LAMP2,110 and this muta-tion has been reported in 1% of cases of hypertrophic cardiomy-opathy.111 Danon disease phenotype involves hypertrophiccardiomyopathy and skeletal myopathy, with Wolff–Parkinson–White syndrome’s electrocardiogram pre–excitation pattern.112–114 Danon disease has also been classed as a subtype of autopha-gic vacuolar myopathies because of the presence of glycogen-filled intracellular autophagic vacuoles in human skeletal and car-diac myocytes.110,114,115 The symptoms of Danon disease present~15 years later in women than in men and with less severe car-diac complications, which may be attributed to a gene dosageeffect because LAMP2 is located on the X chromosome.116

In an experimental mouse model of LAMP2-deficiency,reduced fractional shortening, ejection fraction, and cardiac out-put were observed with in vivo echocardiography.117 Lowertwitch force development and blunted increase to inotropic stimu-lation with Ca2+ and isoproterenol suggested impaired contractilefunction in isolated cardiac trabeculae, but the expression ofCa2+-handling proteins, specifically sarcoendoplasmic reticulumcalcium transport ATPase, phospholamban, calsequestrin, and tro-ponin I, in the ventricles were unaltered.117 The accumulation ofautophagic vacuoles was evident in skeletal and heart musclesand was correlated with increased mortality.118 In cultured he-patocytes, LAMP2 deficiency impaired recycling of the mannose-6-phosphate receptor, which is responsible for sorting lysosomalenzymes in the trans-Golgi network, leading to an accumulationof autophagic vacuoles.119 Thus, the cardiac phenotype related toLAMP2 deficiency further validates an important role for

lysosomal breakdown of glycogen in maintaining glycogen con-tent at a sub-pathological level in the heart.

Glycogen-debranching enzyme deficiency

Mutation in the GDE gene (gene name: AGL) results in an accumu-lation of an intermediate form of glycogen with short branches offour glucose residues, the limit dextrins, in the heart.120 Glycogen-debranching enzyme deficiency is an autosomal recessive disorderclassed as GSD type III and has a reported prevalence of1 in 100 000.6 The clinical manifestations of GDE deficiencyinclude cardiac fibrosis, severe vacuolation of cardiac myocytes,and glycogen accumulation in the atrioventricular node and smoothmuscle cells of intramyocardiac arteries.120 Echocardiographyanalysis has reported left ventricular hypertrophy and diastolic dys-function with preserved systolic function.121 In human primaryskeletal muscle cells of GDE-deficient individuals, inhibition ofmTOR by rapamycin reduced glycogen content, glucose uptake,and the expression of glycogen synthase, thus promoting mTORinhibition as a potential therapy.122 Recombinant human GAA hasalso effectively reduced glycogen levels in skeletal muscles ofGSD III individuals, suggesting that upregulation of the lysosomalglycogen breakdown pathway may bypass the GDE obstruction inthe phosphorylase-glycogen breakdown route.123

Experimentally, a canine model of GDE deficiency caused bynaturally occurring GDE mutations has been used to gain mecha-nistic insight into the disease phenotype.124 Atrial fibrillation wasobserved in some cases, but other conduction system characteris-tics appeared normal and overt signs of cardiac failure were notevident in this model.124 Thus, the cardiac phenotype of thecanine model of GDE deficiency appears milder than the clini-

Table 1 Overview of glycogen storage diseases (GSD)

GSD type Other names Subtype Enzyme deficiency Tissue involvementidentified to date

References

Type 0 Aglycogenosis – Glycogen synthase Hepatic 6,91Type I Von Gierke disease,

hepato-renal glycogenosisType Ia Catalytic unit of

glucose-6-phosphataseHepatic, renal, interstinal mucosa 6,91,92

Type Ib Glucose-6-phosphate translocase Hepatic, renal, interstinal mucosa 6,91,92Type Ic Variation of type Ib/non-Ia Hepatic, renal, interstinal mucosa 92,93Type Id Variation of type Ib/non-Ia Hepatic, renal, interstinal mucosa 92,93

Type II Pompe’s Type II Acid a-glucosidase (GAA) Cardiac, skeletal muscle, hepatic 95,99,102Danon Type IIb LAMP2 mutation with normal GAA Cardiac, skeletal muscle, hepatic 110,113,114

Type III Cori’s, Forbes,limit dextrinosis

Type IIIa Glycogen-debranching enzyme (GDE) Cardiac, skeletal muscle, hepatic 6,94Type IIIb GDE Hepatic 6,94Type IIIc Glucosidase activity of GDE Cardiac, skeletal muscle, hepatic 6,94Type IIId Transferase activity of GDE Cardiac, skeletal muscle, hepatic 6,94

Type IV Amylopectinosis, Anderson’s – Glycogen-branching enzyme Cardiac, hepatic, spleen 91Type V McArdle – Glycogen phosphorylase Skeletal muscle 91Type VI Hers – Glycogen phosphorylase Hepatic 94Type VII Tarui – Phosphofructokinase Skeletal muscle, erythrocytes 91Type IX – Type IXa a subunit of phosphorylase kinase Hepatic, erythrocytes, leukocytes 94

Type IXb b subunit of phosphorylase kinase Hepatic, skeletal muscle,erythrocytes, leukocytes

94

Type IXc c subunit of phosphorylase kinase Hepatic 94Type XI Fanconi-Bickel – Glucose transporter 2 Hepatic, renal 6,91Not classified Lafora progressive

myoclonic epilepsy– Laforin phosphatase/malin

ubiquitin ligaseCardiac, skeletal muscle,hepatic, brain, renal

126,128–130

Glycogen and glycophagy in the heart 421

cally observed condition involving similar mutations. Onlyrecently has an experimental mouse model of GDE deficiencybeen generated and validated as recapitulating human diseasewith severe accumulation of glycogen in the heart.7,8 In vitrostudies have demonstrated that mutations causing inactivation ofeither a–1,6-glucosidase or transferase GDE activity are sufficientto cause glycogen mishandling.125 Mutation in the carbohydrate-binding domain inhibits GDE binding to glycogen and inactivatesboth enzyme activities of GDE, leading to a more severe pheno-type and to instability of the GDE protein with rapid ubiquitina-tion by the E3 ligase, malin.125,126 Intact GDE activity appearscrucial for normal glycogen breakdown in the heart, and the pres-ence of cardiomyocyte vacuoles in tissues of GDE–deficient indi-viduals suggests that when the glycogen breakdown route viaphosphorylase is impaired, the phagic–lysosomal route of glyco-gen degradation is upregulated.

Laforin and malin mutations (Lafora disease)

Laforin is a dual-specificity phosphatase that localizes protein toglycogen and correlates with the amount of glycogen in cells.127

The specific role of laforin in glycogen metabolism remainsunclear. In vitro experiments have revealed that malin, an E3ubiquitin ligase, targets laforin, glycogen synthase, PTG, andGDE for degradation.126,128–130 Mutations in the genes epilepsy,progressive myoclonus type 2a and epilepsy, progressive myoclo-nus type 2b, which respectively encode laforin phosphatase andmalin, have been documented as ‘rare’ in the USA and Canadabut are relatively frequent in Southern Europe, North Africa, andAsia.131 These mutations result in an accumulation of polygluco-san inclusions (Lafora bodies) containing hyperphosphorylated,poorly branched, insoluble glycogen polymers in neurons, skele-tal muscle, heart, and liver.9 Lafora bodies are similar in appear-ance to the amylopectin-like structures that accumulate withdeficiency of glycogen-branching enzyme.6 Clinical manifesta-tions primarily include symptoms of epilepsy such as seizure,ataxia, myoclonus, and dementia, with an onset in late childhoodor adolescence.126

Experimental mouse models of laforin and malin knockoutexhibit abundant Lafora bodies in the brain, heart, and skeletalmuscle.132 Contrary to in vitro studies, malin knockout miceshow no changes in glycogen synthase, GDE, or PTG, and therole of malin in brain or skeletal muscle has not been recapitu-lated.132 Studies looking at molecular mechanisms have reportedcontradictory findings in insulin sensitivity and altered glucosemetabolism. Enhanced insulin response through overactivation ofthe Akt pathway and subsequent increased glycogen andpolyglucosan inclusions was evident in laforin knockout mousehearts.133 In contrast, another report found no changes in glucosedisposal, insulin sensitivity, or the downstream markers of insulinin muscle, liver, or heart.134 Echocardiography function alsoremained unchanged in both laforin and malin knockout mice.134

A recent study has reported comparable reductions in the levelsof the autophagy marker, light-chain 3, and increases inphosphorylation of S6 in both models; however, reduced GABAA

receptor-associated protein-like 1 and LAMP1 were observedonly in mice lacking malin, not laforin.9 This suggestsreduced mTOR-dependent autophagy in laforin-deficient condi-

tions and an independent role for malin in lysosomal glycogendegradation.

CONCLUSIONS

Cardiac glycogen regulation involves a complex interplaybetween multiple signalling pathways, allosteric activation ofenzymes, and sequestration for phagic degradation. Signalingpathways appear to converge on glycogen regulatory enzymesvia insulin (GSK3b, PP1, allosteric action of G6P), b–adrenergic(phosphorylase kinase, PP1 inhibitor), and AMPK (allostericaction of G6P, direct glycogen binding, insulin receptor). Whilecytosolic glycogen synthesis and breakdown are relatively wellunderstood, recent findings relating to phagic glycogen degrada-tion highlight a new area of investigation in the heart. There isstrong evidence from GSD that phagic/lysosomal glycogen break-down is important for maintaining normal cardiac glycogen levelsand does not simply constitute a redundant alternative breakdownroute for glycogen. Advancing understanding of glycogen han-dling in the heart is an important priority with relevance not onlyto genetic GSD but also to cardiac metabolic stress disorderssuch as diabetes and ischaemia.

REFERENCES

1. Dudley HW, Marrian GF. The effect of insulin on the glycogen inthe tissues of normal animals. Biochem. J. 1923; 17: 435–8.

2. Evans G. The glycogen content of the rat heart. J. Physiol. 1934;82: 468–80.

3. Visscher MB, Mulder AG. The carbohydrate metabolism of theheart. Am. J. Physiol. 1930; 94: 630–40.

4. Rabinowitz M. Control of metabolism and synthesis of macromol-ecules in normal and ischemic heart. J. Mol. Cell. Cardiol. 1971;2: 277–92.

5. Shin YS. Glycogen storage disease: Clinical, biochemical, andmolecular heterogeneity. Semin. Pediatr. Neurol. 2006; 13: 115–20.

6. Hicks J, Wartchow E, Mierau G. Glycogen storage diseases: A briefreview and update on clinical features, genetic abnormalities, patho-logic features, and treatment. Ultrastruct. Pathol. 2011; 35: 183–96.

7. Liu KM, Wu JY, Chen YT. Mouse model of glycogen storage dis-ease type III. Mol. Genet. Metab. 2014; 111: 467–76.

8. Pagliarani S, Lucchiari S, Ulzi G et al. Glycogen storage diseasetype III: A novel Agl knockout mouse model. Biochim. Biophys.Acta 2014; 1842: 2318–28.

9. Garyali P, Segvich DM, DePaoli-Roach AA, Roach PJ. Proteindegradation and quality control in cells from laforin and malinknockout mice. J. Biol. Chem. 2014; 289: 20606–14.

10. Reichelt ME, Mellor KM, Curl CL, Stapleton D, Delbridge LM.Myocardial glycophagy – A specific glycogen handling responseto metabolic stress is accentuated in the female heart. J. Mol. Cell.Cardiol. 2013; 65: 67–75.

11. Mellor KM, Varma U, Stapleton D, Delbridge LMD. Cardiomyo-cyte glycophagy is regulated by insulin and exposure to highextracellular glucose. Am. J. Physiol. Heart Circ. Physiol. 2014;306: H1240–5.

12. Jiang S, Wells CD, Roach PJ. Starch-binding domain-containingprotein 1 (Stbd1) and glycogen metabolism: Identification of theAtg8 family interacting motif (AIM) in Stbd1 required for interac-tion with GABARAPL1. Biochem. Biophys. Res. Commun. 2011;413: 420–5.

13. Larner J, Illingworth B, Cori GT, Cori CF. Structure of glycogensand amylopectins. II. Analysis by stepwise enzymatic degradation.J. Biol. Chem. 1952; 199: 641–51.

422 C Chandramouli et al.

14. Meyer KH, Bernfeld P, Boissonnas RA, Gurtler P, Noelting G.Starch solutions and pastes and their molecular interpretation. J.Phys. Colloid Chem. 1949; 53: 319.

15. Ryu JH, Drain J, Kim JH et al. Comparative structural analyses ofpurified glycogen particles from rat liver, human skeletal muscleand commercial preparations. Int. J. Biol. Macromol. 2009; 45:478–82.

16. Drochmans P. Morphologie du glycog�ene: Etude au microscope�electronique de colorations n�egatives du glycog�ene particulaire. J.Ultrastruct. Res. 1962; 6: 141–63.

17. Marchand I, Chorneyko K, Tarnopolsky M et al. Quantification ofsubcellular glycogen in resting human muscle: Granule size, num-ber, and location. J. Appl. Physiol. (1985) 2002; 93: 1598–607.

18. Besford QA, Sullivan MA, Zheng L, Gilbert RG, Stapleton D,Gray-Weale A. The structure of cardiac glycogen in healthy mice.Int. J. Biol. Macromol. 2012; 51: 887–91.

19. Depre C, Vanoverschelde JL, Taegtmeyer H. Glucose for theheart. Circulation 1999; 99: 578–88.

20. Pederson BA, Chen H, Schroeder JM, Shou W, DePaoli-RoachAA, Roach PJ. Abnormal cardiac development in the absence ofheart glycogen. Mol. Cell. Biol. 2004; 24: 7179–87.

21. Alfarano C, Suffredini S, Fantappie O et al. The effect of losartantreatment on the response of diabetic cardiomyocytes to ATPdepletion. Pharmacol. Res. 2011; 63: 225–32.

22. Graham TE, Yuan Z, Hill AK, Wilson RJ. The regulation of mus-cle glycogen: The granule and its proteins. Acta Physiol. (Oxf.)2010; 199: 489–98.

23. Entman ML, Bornet EP, Van Winkle WB, Goldstein MA, Sch-wartz A. Association of glycogenolysis with cardiac sarcoplasmicreticulum: II. Effect of glycogen depletion, deoxycholate solubili-zation and cardiac ischemia: Evidence for a phorphorylase kinasemembrane complex. J. Mol. Cell. Cardiol. 1977; 9: 515–28.

24. Rybicka K. Binding of glycosomes to endoplasmic reticulum andto intermediate filaments in cardiac conduction fibers. J. Histo-chem. Cytochem. 1981; 29: 553–60.

25. Pierce GN, Philipson KD. Binding of glycolytic enzymes to car-diac sarcolemmal and sarcoplasmic reticular membranes. J. Biol.Chem. 1985; 260: 6862–70.

26. Kockskamper J, Zima AV, Blatter LA. Modulation of sarcoplas-mic reticulum Ca2+ release by glycolysis in cat atrial myocytes. J.Physiol. 2005; 564: 697–714.

27. Laks MM, Van de Velde S, Swan HJ. The presence and locationof glycogen in the hypertrophied right ventricle of the canineheart. Cardiovasc. Res. 1969; 3: 272–6.

28. Nielsen J, Schroder HD, Rix CG, Ortenblad N. Distinct effects ofsubcellular glycogen localization on tetanic relaxation time andendurance in mechanically skinned rat skeletal muscle fibres. J.Physiol. 2009; 587: 3679–90.

29. Ortenblad N, Nielsen J, Saltin B, Holmberg HC. Role of glycogenavailability in sarcoplasmic reticulum Ca2+ kinetics in human skel-etal muscle. J. Physiol. 2011; 589: 711–25.

30. Kondomerkos DJ, Kalamidas SA, Kotoulas OB, Hann AC. Glyco-gen autophagy in the liver and heart of newborn rats. The effectsof glucagon, adrenalin or rapamycin. Histol. Histopathol. 2005;20: 689–96.

31. Kondomerkos DJ, Kalamidas SA, Kotoulas OB. An electronmicroscopic and biochemical study of the effects of glucagon onglycogen autophagy in the liver and heart of newborn rats. Mi-crosc. Res. Tech. 2004; 63: 87–93.

32. Fayssoil A. Cardiomyopathy in Pompe’s disease. Eur. J. Intern.Med. 2008; 19: 57–9.

33. Anousis N, Carvalho RA, Zhao P, Malloy CR, Sherry AD. Com-partmentation of glycolysis and glycogenolysis in the perfused ratheart. NMR Biomed. 2004; 17: 51–9.

34. Roach PJ, Skurat AV. Self-glucosylating initiator proteins andtheir role in glycogen biosynthesis. Prog. Nucleic Acid Res. Mol.Biol. 1997; 57: 289–316.

35. Roach PJ, Depaoli-Roach AA, Hurley TD, Tagliabracci VS. Gly-cogen and its metabolism: Some new developments and oldthemes. Biochem. J. 2012; 441: 763–87.

36. Nase S, Kunze KP, Sigmund M, Schroeder JM, Shin Y, HanrathP. A new variant of type IV glycogenosis with primary cardiacmanifestation and complete branching enzyme deficiency. In vivodetection by heart muscle biopsy. Eur. Heart J. 1995; 16: 1698–704.

37. Shen JJ, Chen YT. Molecular characterization of glycogen storagedisease type III. Curr. Mol. Med. 2002; 2: 167–75.

38. Stapleton D, Nelson C, Parsawar K et al. Analysis of hepatic gly-cogen-associated proteins. Proteomics 2010; 10: 2320–9.

39. Tagliabracci VS, Heiss C, Karthik C et al. Phosphate incorpora-tion during glycogen synthesis and Lafora disease. Cell Metab.2011; 13: 274–82.

40. Nitschke F, Wang P, Schmieder P et al. Hyperphosphorylation ofglucosyl C6 carbons and altered structure of glycogen in the neu-rodegenerative epilepsy Lafora disease. Cell Metab. 2013; 17:756–67.

41. Tagliabracci VS, Turnbull J, Wang W et al. Laforin is a glycogenphosphatase, deficiency of which leads to elevated phosphoryla-tion of glycogen in vivo. Proc. Natl Acad. Sci. USA 2007; 104:19262–6.

42. Tagliabracci VS, Girard JM, Segvich D et al. Abnormal metabo-lism of glycogen phosphate as a cause for Lafora disease. J. Biol.Chem. 2008; 283: 33 816–25.

43. Jiang S, Heller B, Tagliabracci VS et al. Starch binding domain-containing protein 1/genethonin 1 is a novel participant in glyco-gen metabolism. J. Biol. Chem. 2010; 285: 34 960–71.

44. Zhu Y, Zhang M, Kelly AR, Cheng A. The carbohydrate-bindingdomain of overexpressed STBD1 is important for its stability andprotein–protein interactions. Biosci. Rep. 2014; 34: 311–320.

45. Onodera S, Matsui H, Chiba S. Substrate specificity and subsiteaffinities of rabbit liver acid alpha-glucosidase. J. Biochem. 1994;116: 7–11.

46. Palmer TN. The substrate specificity of acid-glucosidase from rab-bit muscle. Biochem. J. 1971; 124: 701–11.

47. Tashiro K, Iwamasa T, Kato H, Ogata S, Anai M. Purification andcharacterization of two components of acid alpha-glucosidase frompig liver. J. Biochem. 1986; 99: 693–701.

48. Chambers JP, Williams JC. Acid alpha-glucosidase from humanheart. Enzyme 1983; 29: 109–19.

49. Kandel ES, Hay N. The regulation and activities of the multifunc-tional serine/threonine kinase Akt/PKB. Exp. Cell Res. 1999; 253:210–29.

50. Goodwin GW, Arteaga JR, Taegtmeyer H. Glycogen turnover inthe isolated working rat heart. J. Biol. Chem. 1995; 270: 9234–40.

51. Cohen P. The Croonian Lecture 1998. Identification of a proteinkinase cascade of major importance in insulin signal transduc-tion. Philos. Trans. R. Soc. Lond. B Biol. Sci. 1999; 354: 485–95.

52. Mora A, Sakamoto K, McManus EJ, Alessi DR. Role of thePDK1-PKB-GSK3 pathway in regulating glycogen synthase andglucose uptake in the heart. FEBS Lett. 2005; 579: 3632–8.

53. Wang L, Wang X, Proud CG. Activation of mRNA translation inrat cardiac myocytes by insulin involves multiple rapamycin-sensi-tive steps. Am. J. Physiol. Heart Circ. Physiol. 2000; 278:H1056–68.

54. Wolleben CD, Jaspers SR, Miller TB Jr. Use of adult rat cardio-myocytes to study cardiac glycogen metabolism. Am. J. Physiol.1987; 252: E673–8.

55. De Luca JP, Garnache AK, Rulfs J, Miller TB Jr. Wortmannininhibits insulin-stimulated activation of protein phosphatase 1 inrat cardiomyocytes. Am. J. Physiol. 1999; 276: H1520–6.

56. Crosson SM, Khan A, Printen J, Pessin JE, Saltiel AR. PTG genedeletion causes impaired glycogen synthesis and developmentalinsulin resistance. J. Clin. Invest. 2003; 111: 1423–32.

Glycogen and glycophagy in the heart 423

57. Brady MJ, Printen JA, Mastick CC, Saltiel AR. Role of proteintargeting to glycogen (PTG) in the regulation of protein phospha-tase-1 activity. J. Biol. Chem. 1997; 272: 20198–204.

58. Bouskila M, Hunter RW, Ibrahim AF et al. Allosteric regulationof glycogen synthase controls glycogen synthesis in muscle. CellMetab. 2010; 12: 456–66.

59. McNeel RL, Mersmann HJ. Distribution and quantification ofbeta1-, beta2-, and beta3-adrenergic receptor subtype transcripts inporcine tissues. J. Anim. Sci. 1999; 77: 611–21.

60. Woo AY, Xiao RP. beta-Adrenergic receptor subtype signaling inheart: From bench to bedside. Acta Pharmacol. Sin. 2012; 33:335–41.

61. Xiao RP, Zhang SJ, Chakir K et al. Enhanced G(i) signalingselectively negates beta2-adrenergic receptor (AR)–but not beta1-AR-mediated positive inotropic effect in myocytes from failing rathearts. Circulation 2003; 108: 1633–9.

62. Kuschel M, Zhou YY, Spurgeon HA et al. beta2-adrenergiccAMP signaling is uncoupled from phosphorylation of cytoplas-mic proteins in canine heart. Circulation 1999; 99: 2458–65.

63. McConville P, Lakatta EG, Spencer RG. Greater glycogen utiliza-tion during 1- than 2-adrenergic receptor stimulation in the iso-lated perfused rat heart. Am. J. Physiol. Endocrinol. Metab. 2007;293: E1828–35.

64. Neumann J, Gupta RC, Schmitz W, Scholz H, Nairn AC, Watana-be AM. Evidence for isoproterenol-induced phosphorylation ofphosphatase inhibitor-1 in the intact heart. Circ. Res. 1991; 69:1450–7.

65. Gupta RC, Neumann J, Watanabe AM, Sabbah HN. Inhibition oftype 1 protein phosphatase activity by activation of beta-adreno-ceptors in ventricular myocardium. Biochem. Pharmacol. 2002;63: 1069–76.

66. Carling D. The AMP-activated protein kinase cascade–a unifyingsystem for energy control. Trends Biochem. Sci. 2004; 29: 18–24.

67. Gollob MH. Glycogen storage disease as a unifying mechanism ofdisease in the PRKAG2 cardiac syndrome. Biochem. Soc. Trans.2003; 31: 228–31.

68. Dyck JR, Lopaschuk GD. AMPK alterations in cardiac physiol-ogy and pathology: Enemy or ally? J. Physiol. 2006; 574: 95–112.

69. Murphy RT, Mogensen J, McGarry K et al. Adenosine mono-phosphate-activated protein kinase disease mimicks hypertrophiccardiomyopathy and Wolff-Parkinson-White syndrome: Naturalhistory. J. Am. Coll. Cardiol. 2005; 45: 922–30.

70. Arad M, Moskowitz IP, Patel VV et al. Transgenic mice overex-pressing mutant PRKAG2 define the cause of Wolff-Parkinson-White syndrome in glycogen storage cardiomyopathy. Circulation2003; 107: 2850–6.

71. Carling D, Hardie DG. The substrate and sequence specificity ofthe AMP-activated protein kinase. Phosphorylation of glycogensynthase and phosphorylase kinase. Biochim. Biophys. Acta 1989;1012: 81–6.

72. Wojtaszewski JF, Jorgensen SB, Hellsten Y, Hardie DG, RichterEA. Glycogen-dependent effects of 5-aminoimidazole-4-carboxa-mide (AICA)-riboside on AMP-activated protein kinase and glyco-gen synthase activities in rat skeletal muscle. Diabetes 2002; 51:284–92.

73. Beyer A, Kitzerow A, Crute B, Kemp BE, Witters LA, HeilmeyerLM Jr. Muscle phosphorylase kinase is not a substrate of AMP-activated protein kinase. Biol. Chem. 2000; 381: 457–61.

74. Hunter RW, Treebak JT, Wojtaszewski JF, Sakamoto K. Molecu-lar mechanism by which AMP-activated protein kinase activationpromotes glycogen accumulation in muscle. Diabetes 2011; 60:766–74.

75. Banerjee SK, Wang DW, Alzamora R et al. SGLT1, a novel car-diac glucose transporter, mediates increased glucose uptake inPRKAG2 cardiomyopathy. J. Mol. Cell. Cardiol. 2010; 49: 683–92.

76. Polekhina G, Gupta A, Michell BJ et al. AMPK beta subunit tar-gets metabolic stress sensing to glycogen. Curr. Biol. 2003; 13:867–71.

77. Polekhina G, Gupta A, van Denderen BJ et al. Structural basis forglycogen recognition by AMP-activated protein kinase. Structure2005; 13: 1453–62.

78. Hudson ER, Pan DA, James J et al. A novel domain in AMP-acti-vated protein kinase causes glycogen storage bodies similar tothose seen in hereditary cardiac arrhythmias. Curr. Biol. 2003; 13:861–6.

79. Banerjee SK, Ramani R, Saba S et al. A PRKAG2 mutationcauses biphasic changes in myocardial AMPK activity and doesnot protect against ischemia. Biochem. Biophys. Res. Commun.2007; 360: 381–7.

80. Ahmad F, Arad M, Musi N et al. Increased alpha2 subunit-associ-ated AMPK activity and PRKAG2 cardiomyopathy. Circulation2005; 112: 3140–8.

81. Ofir M, Arad M, Porat E et al. Increased glycogen stores due togamma-AMPK overexpression protects against ischemia and rep-erfusion damage. Biochem. Pharmacol. 2008; 75: 1482–91.

82. Luptak I, Shen M, He H et al. Aberrant activation of AMP-acti-vated protein kinase remodels metabolic network in favor of car-diac glycogen storage. J. Clin. Invest. 2007; 117: 1432–9.

83. Puthanveetil P, Wang F, Kewalramani G et al. Cardiac glycogenaccumulation after dexamethasone is regulated by AMPK. Am. J.Physiol. Heart Circ. Physiol. 2008; 295: H1753–62.

84. Longnus SL, Wambolt RB, Parsons HL, Brownsey RW, AllardMF. 5-Aminoimidazole-4-carboxamide 1-beta-D-ribofuranoside(AICAR) stimulates myocardial glycogenolysis by allosteric mech-anisms. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2003; 284:R936–44.

85. Aschenbach WG, Hirshman MF, Fujii N, Sakamoto K, HowlettKF, Goodyear LJ. Effect of AICAR treatment on glycogen metab-olism in skeletal muscle. Diabetes 2002; 51: 567–73.

86. Holmes BF, Kurth-Kraczek EJ, Winder WW. Chronic activationof 50-AMP-activated protein kinase increases GLUT-4, hexoki-nase, and glycogen in muscle. J. Appl. Physiol. (1985) 1999; 87:1990–5.

87. Russell RR 3rd, Bergeron R, Shulman GI, Young LH. Transloca-tion of myocardial GLUT-4 and increased glucose uptake throughactivation of AMPK by AICAR. Am. J. Physiol. 1999; 277:H643–9.

88. Kristiansen SB, Solskov L, Jessen N et al. 5-Aminoimidazole-4-carboxamide-1-beta-D-ribofuranoside increases myocardial glucoseuptake during reperfusion and induces late pre-conditioning:Potential role of AMP-activated protein kinase. Basic Clin. Phar-macol. Toxicol. 2009; 105: 10–6.

89. Chopra I, Li HF, Wang H, Webster KA. Phosphorylationof the insulin receptor by AMP-activated protein kinase(AMPK) promotes ligand-independent activation of the insulinsignalling pathway in rodent muscle. Diabetologia 2012; 55:783–94.

90. Di Sant’Agnese PA, Andersen DH, Mason HH. Glycogen storagedisease of the heart. II. Critical review of the literature. Pediatrics1950; 6: 607–24.

91. Chen M. Glycogen storage diseases. In: Monga SPS (ed.). Molec-ular Pathology of Liver Diseases. Springer, Boston. 2011; Ch. 45.

92. Moses SW. Historical highlights and unsolved problems in glyco-gen storage disease type 1. Eur. J. Pediatr. 2002; 161(Suppl. 1):S2–9.

93. Nordlie RC, Sukalski KA, Munoz JM, Baldwin JJ. Type Ic, anovel glycogenosis. Underlying mechanism. J. Biol. Chem. 1983;258: 9739–44.

94. Wolfsdorf JI, Weinstein DA. Glycogen storage diseases. Rev. En-docr. Metab. Disord. 2003; 4: 95–102.

95. Lim JA, Li L, Raben N. Pompe disease: From pathophysiology totherapy and back again. Front. Aging Neurosci. 2014; 6: 177.

424 C Chandramouli et al.

96. Kishnani PS, Hwu WL, Mandel H et al. A retrospective, multina-tional, multicenter study on the natural history of infantile-onsetPompe disease. J. Pediatr. 2006; 148: 671–6.

97. Fernandez C, Legido A, Jethva R, Marks HG. Correction of ashort cardiac PR interval in a 12-year-old girl with late-onsetPompe disease following enzyme replacement therapy. Genet.Med. 2012; 14: 757–8.

98. Van den Hout JM, Kamphoven JH, Winkel LP et al. Long-termintravenous treatment of Pompe disease with recombinant humanalpha-glucosidase from milk. Pediatrics 2004; 113: e448–57.

99. Fayssoil A, Nardi O, Annane D, Orlikowski D. Right ventricularfunction in late-onset Pompe disease. J. Clin. Monit. Comput.2014; 28: 419–21.

100. Bulkley BH, Hutchins GM. Pompe’s disease presenting as hyper-trophic myocardiopathy with Wolff-Parkinson-White syndrome.Am. Heart J. 1978; 96: 246–52.

101. Francesconi M, Auff E. Cardiac arrhythmias and the adult form oftype II glycogenosis. N. Engl. J. Med. 1982; 306: 937–8.

102. Raben N, Nagaraju K, Lee E et al. Targeted disruption of the acidalpha-glucosidase gene in mice causes an illness with critical fea-tures of both infantile and adult human glycogen storage diseasetype II. J. Biol. Chem. 1998; 273: 19 086–92.

103. Mah C, Pacak CA, Cresawn KO et al. Physiological correction ofPompe disease by systemic delivery of adeno-associated virusserotype 1 vectors. Mol. Ther. 2007; 15: 501–7.

104. Xu F, Ding E, Migone F et al. Glycogen storage in multiple mus-cles of old GSD-II mice can be rapidly cleared after a single intra-venous injection with a modified adenoviral vector expressinghGAA. J. Gene Med. 2005; 7: 171–8.

105. Zhu Y, Jiang JL, Gumlaw NK et al. Glycoengineered acid alpha-glucosidase with improved efficacy at correcting the metabolicaberrations and motor function deficits in a mouse model ofPompe disease. Mol. Ther. 2009; 17: 954–63.

106. Farah BL, Madden L, Li S et al. Adjunctive beta2-agonist treat-ment reduces glycogen independently of receptor-mediated acidalpha-glucosidase uptake in the limb muscles of mice with Pompedisease. FASEB J. 2014; 28: 2272–80.

107. Khanna R, Flanagan JJ, Feng J et al. The pharmacological chaper-one AT2220 increases recombinant human acid alpha-glucosidaseuptake and glycogen reduction in a mouse model of Pompe dis-ease. PLoS ONE 2012; 7: e40776.

108. Ashe KM, Taylor KM, Chu Q et al. Inhibition of glycogen bio-synthesis via mTORC1 suppression as an adjunct therapy forPompe disease. Mol. Genet. Metab. 2010; 100: 309–15.

109. Danon MJ, Oh SJ, DiMauro S et al. Lysosomal glycogen storagedisease with normal acid maltase. Neurology 1981; 31: 51–7.

110. Nishino I, Fu J, Tanji K et al. Primary LAMP-2 deficiency causesX-linked vacuolar cardiomyopathy and myopathy (Danon disease).Nature 2000; 406: 906–10.

111. Charron P, Villard E, Sebillon P et al. Danon’s disease as a causeof hypertrophic cardiomyopathy: A systematic survey. Heart2004; 90: 842–6.

112. Bui YK, Renella P, Martinez-Agosto JA, Verity A, Madikians A,Alejos JC. Danon disease with typical early-onset cardiomyopathyin a male: Focus on a novel LAMP-2 mutation. Pediatr. Trans-plant. 2008; 12: 246–50.

113. Boucek D, Jirikowic J, Taylor M. Natural history of Danon dis-ease. Genet. Med. 2011; 13: 563–8.

114. Yang Z, Vatta M. Danon disease as a cause of autophagic vacuo-lar myopathy. Congenit. Heart. Dis. 2007; 2: 404–9.

115. Nishino I. Autophagic vacuolar myopathies. Curr. Neurol. Neuro-sci. Rep. 2003; 3: 64–9.

116. Arad M, Maron BJ, Gorham JM et al. Glycogen storage diseasespresenting as hypertrophic cardiomyopathy. N. Engl. J. Med.2005; 352: 362–72.

117. Stypmann J, Janssen PM, Prestle J et al. LAMP-2 deficient miceshow depressed cardiac contractile function without significantchanges in calcium handling. Basic Res. Cardiol. 2006; 101: 281–91.

118. Tanaka Y, Guhde G, Suter A et al. Accumulation of autophagicvacuoles and cardiomyopathy in LAMP-2-deficient mice. Nature2000; 406: 902–6.

119. Eskelinen EL, Illert AL, Tanaka Y et al. Role of LAMP-2 in lyso-some biogenesis and autophagy.Mol. Biol. Cell 2002; 13: 3355–68.

120. Austin SL, Proia AD, Spencer-Manzon MJ, Butany J, WechslerSB, Kishnani PS. Cardiac pathology in glycogen storage diseasetype III. JIMD Rep. 2012; 6: 65–72.

121. Olson LJ, Reeder GS, Noller KL, Edwards WD, Howell RR,Michels VV. Cardiac involvement in glycogen storage disease III:Morphologic and biochemical characterization with endomyocardi-al biopsy. Am. J. Cardiol. 1984; 53: 980–1.

122. Yi H, Brooks ED, Thurberg BL, Fyfe JC, Kishnani PS, Sun B.Correction of glycogen storage disease type III with rapamycin ina canine model. J. Mol. Med. (Berl.) 2014; 92: 641–50.

123. Sun B, Fredrickson K, Austin S et al. Alglucosidase alfa enzymereplacement therapy as a therapeutic approach for glycogen stor-age disease type III. Mol. Genet. Metab. 2013; 108: 145–7.

124. Yi H, Thurberg BL, Curtis S et al. Characterization of a caninemodel of glycogen storage disease type IIIa. Dis. Model Mech.2012; 5: 804–11.

125. Cheng A, Zhang M, Okubo M, Omichi K, Saltiel AR. Distinctmutations in the glycogen debranching enzyme found in glycogenstorage disease type III lead to impairment in diverse cellularfunctions. Hum. Mol. Genet. 2009; 18: 2045–52.

126. Cheng A, Zhang M, Gentry MS, Worby CA, Dixon JE, Saltiel AR.A role for AGL ubiquitination in the glycogen storage disorders ofLafora and Cori’s disease. Genes Dev. 2007; 21: 2399–409.

127. Ganesh S, Agarwala KL, Ueda K et al. Laforin, defective in theprogressive myoclonus epilepsy of Lafora type, is a dual-specific-ity phosphatase associated with polyribosomes. Hum. Mol. Genet.2000; 9: 2251–61.

128. Gentry MS, Worby CA, Dixon JE. Insights into Lafora disease:Malin is an E3 ubiquitin ligase that ubiquitinates and promotes thedegradation of laforin. Proc. Natl Acad. Sci. USA 2005; 102:8501–6.

129. Lohi H, Ianzano L, Zhao XC et al. Novel glycogen synthasekinase 3 and ubiquitination pathways in progressive myoclonusepilepsy. Hum. Mol. Genet. 2005; 14: 2727–36.

130. Worby CA, Gentry MS, Dixon JE. Malin decreases glycogenaccumulation by promoting the degradation of protein targeting toglycogen (PTG). J. Biol. Chem. 2008; 283: 4069–76.

131. Acharya JN, Satishchandra P, Asha T, Shankar SK. Lafora’s dis-ease in south India: A clinical, electrophysiologic, and pathologicstudy. Epilepsia 1993; 34: 476–87.

132. DePaoli-Roach AA, Tagliabracci VS, Segvich DM, Meyer CM, Ir-imia JM, Roach PJ. Genetic depletion of the malin E3 ubiquitinligase in mice leads to lafora bodies and the accumulation ofinsoluble laforin. J. Biol. Chem. 2010; 285: 25 372–81.

133. Vernia S, Heredia M, Criado O et al. Laforin, a dual specificityphosphatase involved in Lafora disease, regulates insulin responseand whole-body energy balance in mice. Hum. Mol. Genet. 2011;20: 2571–84.

134. DePaoli-Roach AA, Segvich DM, Meyer CM et al. Laforin andmalin knockout mice have normal glucose disposal and insulinsensitivity. Hum. Mol. Genet. 2012; 21: 1604–10.

Glycogen and glycophagy in the heart 425