A link between excess dietary sugar and cardiac disease is clearly evident and has been largely attributed to systemic metabolic dysregulation. Now a new paradigm is emerging, and a compelling case can be made that fructose-associated heart injury may be attributed to the direct actions of fructose on cardiomyocytes. Plasma and cardiac fructose levels are elevated in patients with diabetes, and evidence suggests that some unique properties of fructose (vs. glucose) have specific cardiomyocyte consequences. Investigations to date have demonstrated that cardiomyocytes have the capacity to transport and utilize fructose and express all of the necessary proteins for fructose metabolism. When dietary fructose intake is elevated and myocardial glucose uptake compromised by insulin resistance, increased cardiomyocyte fructose flux represents a hazard involving unregulated glycolysis and oxidative stress. The high reactivity of fructose supports the contention that fructose accelerates subcellular hexose sugar-related protein modifications, such as O-GlcNAcylation and advanced glycation end product formation. Exciting recent discoveries link heart failure to induction of the specific high-affinity fructose-metabolizing enzyme, fructokinase, in an experimental setting. In this Perspective, we review key recent findings to synthesize a novel view of fructose as a cardiopathogenic agent in diabetes and to identify important knowledge gaps for urgent research focus.
Links between excess fructose consumption, diabetes incidence, and cardiovascular disease risk have been clearly demonstrated (1–4). The systemic effects of high fructose intake have been well described experimentally and include hyperglycemia, dyslipidemia, atherosclerosis, and in some cases hypertension (5). Diabetic cardiomyopathy is recognized as a specific myocardial pathology, the occurrence of which is independent of coronary and hypertensive disease. Diabetic cardiomyopathy is generally characterized by early signs of diastolic dysfunction, which precede progression to systolic failure (6,7).
While the relationship between excess fructose exposure and cardiac disease development has been identified, the underlying mechanisms are as yet only partially understood. Aspects of diabetic cardiomyopathy that may be attributed specifically to high fructose intake or to selective myocardial fructose metabolic dysregulation have not been determined. Whether cardiac vulnerability associated with fructose exposure produces an injury response beyond effects that may be attributed to general overnutrition or to overall excess consumption of refined sugar (either glucose or fructose) in different diabetic settings is also not yet known (8,9).
In this Perspective, these questions relating to fructose, diabetes, and the heart are examined, and the case for a role of fructose in diabetic cardiomyopathy disease etiology is explored. Evidence to support the proposition that fructose is a distinctive cardiopathogenic agent in diabetes and states of metabolic disturbance is considered. Findings from a diverse range of investigative approaches are reviewed to synthesize a novel view of fructose (of exogenous and endogenous origin) as a perpetrator of cardiac damage. New insights into fructose-induced myocardial functional and signaling dysregulation are discussed, and knowledge gaps for priority research focus are identified.
Dietary Fructose and Cardiomyopathy
Dietary Fructose Increases Cardiovascular Risk
The dramatic rise in the prevalence of diabetes has occurred in parallel with an escalation in dietary sugar consumption. In “westernized” cultures, the use of added sweeteners containing fructose (sucrose and high-fructose corn syrup) has increased by approximately 25% over the past three decades (10). Meta-analyses of cohort studies have determined that high intake of fructose-sweetened beverages is associated with a 26% greater risk of cardiometabolic pathology (2). Experimental studies of hepatic fructose metabolism have shown that genetically modified rodents unable to metabolize fructose (fructokinase knockout) are protected from high-carbohydrate–induced metabolic syndrome, supporting the contention that fructose is the toxic component of the sugar complex (11,12). Although increased cardiovascular disease risk may be partially attributed to fructose-induced dyslipidemia, atherosclerosis, hypertension, obesity, or insulin resistance/diabetes/metabolic syndrome (13,14), it is increasingly apparent that fructose-specific cardiac factors are important. Recent studies have demonstrated that dietary fructose is not necessarily associated with changes in blood pressure (15), and the relationship between high sugar intake and increased risk for both type 2 diabetes and cardiovascular disease is independent of BMI (2,16), which indicates that calorie intake and adipose deposition are not the underlying etiology.
Emerging evidence suggests that cardiac complications in patients with diabetes (and their attenuation) are not always linked to the degree of blood glucose control. Meta-analyses of large randomized controlled clinical trials report that drugs used to lower blood glucose levels in patients with diabetes may exacerbate heart failure symptoms and increase the risk of heart failure (17,18). While some glucose-lowering agents (metformin, emphafliglozin) have been shown to be cardioprotective in patients with diabetes, others do not improve cardiac dysfunction and can lead to increased heart failure risk (e.g., peroxisome proliferator–activated receptor agonists, dipeptidyl peptidase 4 inhibitors, thiazolidinediones). Thus negative cardiac impacts in diabetes involve mechanisms not necessarily responsive to normalization of circulating glucose levels. As cardiac tissue is both insulin sensitive and glycolysis dependent, increased cardiac vulnerability to fructose in the context of metabolic dysregulation is plausible (5) and supported by an accumulating experimental evidence base. As emphasized in important recent commentaries (4,12), although of caloric equivalence, fructose and glucose are very different sugars. The consequences of these differences in relation to diet-induced dysregulation of cardiomyocyte signaling, metabolism, and energetics are considered below.
Myocardial Metabolic Dysregulation With High Dietary Fructose Intake
Myocardial signaling adaptations in response to high dietary fructose intake are reported. In animal models, tissue insulin resistance is evident, characterized by downregulation of the phosphoinositide 3-kinase (PI3K)/Akt insulin signaling pathway (19). The PI3K/Akt pathway regulates GLUT4 translocation, glucose uptake, and cardiac cell growth and survival, and thus suppression of this major signaling nexus results in metabolic dysregulation and cell death vulnerability. Rodents fed a high-fructose diet for several months display significant decreases in the levels of phosphorylated Akt (Ser473) and the downstream signaling intermediate S6 (Ser235/236) with decreased PI3K activity and phosphorylation of Akt (19,20). Interestingly, insulin growth factor 1 (IGF1) and IGF1 receptor expression are both decreased in this setting, suggesting involvement from the IGF1 signaling pathway in addition to insulin-receptor–mediated effects (20). Fructose feeding also diminishes cardiac glucose uptake (21). Thus despite elevated extracellular glucose levels, intracellular glucose availability is reduced, which is associated with upregulation of cardiac lipid derivatives and transporters (22), indicative of a substrate shift from glucose to fatty acids for energy supply.
Energetic disturbances may manifest as triggers for oxidative stress in response to excess dietary fructose (23). Mitochondrial uncoupling is evident in hearts of fructose-fed rodents (24) and is associated with elevated myocardial production of reactive oxygen species (25). Impaired glucose uptake and utilization has also been linked to an inability to respond to an ischemic challenge. In contrast to controls, fructose-fed animals do not increase cardiomyocyte GLUT4 translocation and glycolytic flux in response to ischemia (26), a response deficit that may underlie chronic ischemic vulnerability in diabetic settings. Interestingly, two studies have reported smaller infarct size in isolated hearts from fructose-fed rats (27,28), suggesting that metabolic adaptation involving modified routes of cardiomyocyte hexose sugar uptake may actually have a role in acute cardioprotection.
Dietary Fructose–Induced Cardiac Cell Loss and Dysfunction
Downregulation of the cardiomyocyte PI3K/Akt “cell survival’’ pathway can promote cell death signaling as inhibition of programmed cell death pathways via the PI3K/Akt axis is relieved. Fructose feeding in rodents is associated with low-level constitutive loss of cardiomyocytes coupled with increased collagen deposition, producing a progressive fibrotic replacement of viable myocardium (19). In feeding interventions where the metabolic disturbance is moderate, apoptotic signaling pathways are not found to be activated but autophagy markers are significantly upregulated (19). Autophagy, a subcellular phagolysosomal degradation process, is essential for physiological turnover of macromolecules and organelles. Sustained, high-level autophagic activity is considered deleterious and is associated with induction of a nonapoptotic form of programmed cell death (29–31). More work is required to establish the nature of the links between dietary fructose, cardiomyocyte loss, and induction of autophagy triggers. In rodent models where the extent of myocardial metabolic disturbance that develops in response to high-fructose feeding is more severe, activation of apoptotic signaling is evident (20). These findings suggest that loss of cardiomyocyte viability with fructose insult may be mediated initially through autophagic pathways, subsequently transitioning to apoptotic demise as cardiopathology progresses.
The myocardial functional consequences of high fructose intake have not been extensively studied. Some insights have been gained from in vivo (hemodynamic) analyses and from in vitro cardiomyocyte contractility and Ca2+ handling studies. In vivo left ventricular dysfunction is evident in rodents administered fructose drinking solution (10%) for only 2 weeks, as characterized by reduced left ventricular end-systolic elastance (a measure of intrinsic left ventricular contractility independent of preload, afterload, and heart rate) (32). In cardiomyocytes isolated from fructose-fed mice, a marked Ca2+ handling disturbance is observed, even with maintenance of twitch contractile performance (33). Contractile myofilament sensitivity to Ca2+ is increased by high fructose intake (33), which may have important implications for impaired relaxation and diastolic dysfunction in vivo. There is also recent evidence that high fructose intake increases cardiomyocyte arrhythmogenic susceptibility. Cardiomyocytes obtained from animals exposed to fructose drinking solution exhibit Ca2+ handling defects, with increased instability of internal Ca2+ stores associated with spontaneous arrhythmogenic events (34).
Collectively these clinical and experimental studies demonstrate the negative impacts of high fructose intake on myocardial structural and functional integrity— effects that may not necessarily be contingent on systemic or cardiac insulin resistance. Evidence suggests that the unique properties of fructose, which differentiate this hexose sugar from glucose, have selective cardiomyocyte metabolic consequences that undermine cellular integrity. These specific aspects of fructose action are considered below.
Specific Myocardial Fructose Actions
Cardiomyocyte Fructose Exposure—A Central Role in Cardiac Pathology?
Although systemic fructose levels are maintained within the micromolar range by hepatic clearance of absorbed fructose, plasma fructose concentration is elevated in patients with diabetes (35). Thus it is feasible that elevated plasma fructose levels exert cardiomyocyte influence. Work with isolated cardiomyocytes has demonstrated that these cells have the capacity to transport and utilize exogenously supplied fructose. A key study has shown that the high-affinity fructose-specific transporter, GLUT5, is expressed in adult rat cardiomyocytes (36). Importantly GLUT5-mediated glucose uptake is negligible, and thus the operation of this transporter is not influenced by fructose-glucose competition. In isolated cardiomyocytes, the contractile deficit induced by inhibition of glucose oxidation is abrogated by fructose supplementation, providing direct evidence that cardiomyocyte fructose uptake and utilization is operational (36). Production of fructose 1-phosphate (F1P) from exogenous fructose has been demonstrated in cultured neonatal cardiomyocytes using 13C-radiolabeled fructose (37). These findings not only establish that fructose has a role in acute modulation of cardiomyocyte fuel usage but also confirm that fructose has direct cardiomyocyte “assault” access (36). Although not yet well characterized in cardiac tissues, increased fructose metabolic flux has the potential to cause damage as a consequence of increased conversion of fructose to F1P. In noncardiac tissues, this reaction step results in the ultimate production of uric acid and is linked with cell nucleotide depletion (4,38). Further investigation is required to fully characterize the extent and regulation of fructose fuel usage in the heart in both physiological and pathophysiological circumstances.
Endogenous fructose production is also potentially a major factor in determining local exposure to this hexose sugar. In hepatic and renal tissues, there is clear evidence of significant endogenous fructose production via augmented polyol pathway throughput (conversion of glucose to fructose via sorbitol) in disease states (11,39). Conditions of enhanced polyol activity in myocardial tissues are less well described, but an important observation is that myocardial fructose content is measured to be 60-fold higher in diabetic rats (40). This observation supports the proposition that stimulated intracellular production of fructose via the polyol pathway in the diabetic heart may also be significant and pathological. A recent pivotal report has demonstrated that cardiac fructose uptake, transporter expression, and content are elevated in three mouse models of heart failure (1-kidney-1-clip [1K1C], transverse aortic constriction [TAC], and chronic isoproterenol perfusion) and in cardiac biopsies from humans with aortic stenosis and hypertrophic cardiomyopathy (37). Crucially, the finding that cardiac-specific knockdown of Sf3b1, a positive regulator of fructokinase (C isoform, also known as ketohexokinase-C), attenuates 1K1C- and TAC-induced elevated cardiac fructose levels and cardiac dysfunction and hypertrophy (37) demonstrates a central role for fructose metabolism in cardiac pathology. This study also determined that TAC-induced cardiac dysfunction was prevented in global fructokinase knockout mice (37), a model previously reported to exhibit elevated plasma fructose levels (11). Thus cardioprotection was achieved despite cardiomyocyte exposure to elevated plasma fructose. These findings suggest that direct negative effects of fructose exposure on cardiomyocytes may be dependent on the pathological setting and involvement of hepatic fructose dysregulation (e.g., extent of concomitant insulin resistance/diabetes, increased lactic acid, increased uric acid). Clearly more work is needed to elucidate the hepatic dependent and independent aspects of myocardial fructose vulnerability.
Increased intracellular fructose availability has the potential to significantly modify cardiomyocyte metabolic processes. In contrast to the tightly regulated process of glucose metabolism, fructose can bypass the glycolytic rate-limiting enzyme, phosphofructokinase, and proceed through glycolysis to pyruvate and lactate end products in a relatively unregulated manner (41). Accumulation of lactate in particular has been shown to have adverse effects in cardiomyocytes (42). Fructose can also enter the hexosamine biosynthesis pathway (HBP) to generate uridine diphosphate (UDP)-GlcNAc, the precursor for O-GlcNAcylation (see below). The importance of the HBP in myocardial metabolic control and disease vulnerability is increasingly recognized (43,44). Regulatory factors that determine the balance between glycolytic and HBP fructose flux have not yet been identified. Qualitative and quantitative shifts in cardiomyocyte fructose shunting likely play an important role in determining the impacts of altered intracellular fructose on cardiomyocytes. New metabolic studies in this area will be especially informative in relation to understanding adverse actions of cardiomyocyte fructose metabolism. In particular the application of metabolic tracing methodologies to track fructose uptake and/or production and disposal in pathophysiological conditions will yield valuable insights (45).
Fructose and Cardiomyocyte O-GlcNAcylation in the Heart
O-GlcNAcylation is a major posttranslational modification involved in normal physiological signaling regulation and has also been implicated in a number of pathological processes (43,46). O-GlcNAcylation is a dynamic/reversible covalent attachment of an O-GlcNAc moiety onto serine or threonine residues of target proteins, catalyzed by O-GlcNAc transferase, and its removal is catalyzed by O-GlcNAcase. Attachment of the O-GlcNAc substrate can influence transcription, translation, nuclear transport, and cell signaling, and there is evidence of interaction and/or competition for the target amino acid residues between phosphorylation and O-GlcNAcylation (47,48). O-GlcNAcylation has been identified as an important mediator of diabetic heart pathology. Promotion of O-GlcNAc removal by overexpression of O-GlcNAcase restores contractile function and Ca2+ handling in cardiomyocytes of diabetic rodents (47). More recently, it has been demonstrated that specific O-GlcNAcylation of Ca2+-dependent calmodulin kinase II, a key regulator of myocyte Ca2+ cycling and contractility, mediates cardiac dysfunction and arrhythmias in diabetic hearts (49). Thus O-GlcNAcylation may prove to be an effective therapeutic target in diabetic cardiomyopathy.
While glucose has been established as the main contributor toward O-GlcNAcylation, recent findings suggest that fructose may also play a significant role. Fructose can be converted to fructose 6-phosphate (F6P), the substrate for the HBP, providing for consequent production of UDP-O-GlcNAc. This conversion is either via direct phosphorylation of fructose by hexokinase to F6P (low affinity) (50,51) or via phosphorylation of fructose to F1P by fructokinase, followed by generation of F6P involving intermediate steps via dihydroxyacetone phosphate or glyceraldehyde, glyceraldehyde-3-phosphate, and fructose 1,6-biphosphate (51), as detailed in Fig. 1. Literature reporting the direct effects of fructose on the HBP and O-GlcNAcylation is limited, and measurement of fructose metabolic flux into the HBP is warranted. Left ventricular tissues from fructose-fed rodents exhibit significant elevation in the level of O-GlcNAcylation (52). Direct comparison of cardiomyocyte glucose- and fructose-induced O-GlcNAcylation activity in vitro has yet to be reported. In human hepatocarcinoma HepG2 lineage cells, incubation with either glucose or fructose produces a similar increase in UDP-O-GlcNAc levels in a 24-h period (51). Although involving noncardiac cell types, these findings are consistent with the contention that fructose can act as a substrate for the HBP and subsequent O-GlcNAcylation. These data suggest that fructose can induce O-GlcNAcylation to a similar extent as glucose, but further work with cardiac cell types and mapping fructose-induced O-GlcNAc to downstream cardiac consequences using in vivo experimental models is required.
Fructose-Induced Glycation of Cardiac Proteins
In contrast to the dynamic, regulated nature of O-GlcNAcylation reactions, advanced glycation end products (AGEs) are adducts produced from the irreversible nonenzymatic glycation and oxidation of proteins and lipids. Clinical and experimental studies have demonstrated an association between AGE formation and dysfunction in the diabetic heart (53,54). In general, literature has focused on extracellular AGEs and in particular the cross-linking of collagen resulting in myocardial stiffness and diastolic dysfunction of extracellular matrix origin. There is some evidence that intracellular AGEs may also elicit significant impact on cardiomyocyte function (55,56).
AGEs are formed from the attachment of a single hexose molecule by its aldehyde group to the NH2-terminal of a basic amino acid residue (usually lysine or arginine) to form a Schiff base (57). Schiff bases are then rearranged into Amadori products, which can develop into reactive intermediates such as 3-deoxyglucosone and glyoxal (58). A series of oxidation-based Maillard reactions involving chemical cleavage, cross-linking, and conformational change transform the reactive intermediates into irreversible AGEs (57), as detailed in Fig. 2. Not only do the reducing properties of the bound AGE alter protein structure and function, but AGEs can also bind to cell membranes via receptors for AGEs (RAGEs) to activate signaling cascades involving oxidative stress, pathological growth, and induction of cell death processes (53,59,60). The combination of hyperglycemia and reactive oxygen species in the diabetic heart provides a conducive environment for extracellular AGE production.
AGE formation within the extracellular matrix of the diabetic heart has been mostly attributed to hyperglycemia (53,61). But with evidence of intracellular AGEs in the context of impaired glucose uptake in diabetic cardiomyocytes, a recognition of the importance of alternative substrates for AGE production, including fructose, is emerging. In vitro studies with purified proteins have shown that fructose-related AGEs (Fru-AGEs) are more reactive than their Glu-AGE equivalents (62,63). Fructose can undergo nonenzymatic condensation with protein amino groups to form Schiff bases in a similar manner to glucose (58,62,63), which subsequently undergo Heyns rearrangement to form Heyns products (a fructose homolog to the glucose-derived Amadori product) (58). Fructose naturally exists in its open-chain conformation more often than glucose (63), which promotes faster glycation kinetics (58,62,64). Hence, the conversion of Heyns products into Fru-AGEs occurs more rapidly than the conversion of the glucose equivalent, which has significant implications for greater Fru-AGE production (65). Glu-AGEs are thought to be developed over periods ranging from weeks to months, whereas Fru-AGE formation is believed to be much more rapid (Fig. 2) (62,64) and subsequently may have more severe protein damage outcomes. New investigative initiatives are required to understand the pathological importance of Fru-AGEs and Glu-AGEs in the myocardium and the cardiomyocyte. In diabetic cardiomyocytes, Glu-AGEs are detected even on short-lived proteins involved in electromechanical transduction and Ca2+ handling (55,56), with reported half-life times of 3–8 days (66,67). Given the established timelines of AGE formation, this suggests that AGEs may impair protein turnover—a positive feedback scenario that would facilitate even more extensive AGE formation. In a cellular environment where fructose-driven AGE formation is particularly promoted, the accumulation of Fru-AGE adducts conferring functional and structural deformation of affected proteins could be much accelerated.
Using novel Fru-AGE antibodies, it has been demonstrated that patients with type 1 diabetes exhibit fourfold higher serum Fru-AGE levels than patients without nondiabetes (65). No studies to date have directly explored the presence of intracellular Fru-AGEs in cardiomyocytes, but some literature from in vitro studies working with purified proteins and using noncardiomyocyte cell culture experiments is available. In experiments involving incubation of purified proteins with fructose, it has been shown that Fru-AGE formation is increased in parallel with significant modification of the protein function. In these experiments, Fru-AGE formation is found to be markedly more rapid and/or more extensive than Glu-AGE formation and renders the protein more resistant to biological enzymatic breakdown (62,64,65,68). In a noncardiac cell culture system, when gene transfer methods are used to promote intracellular fructose synthesis by stimulation of the polyol pathway, a coincident increase in the level of intracellular Fru-AGEs is observed (65). Together these various experimental approaches demonstrate that intracellular formation of Fru-AGEs may be an important pathophysiological event with specific cellular structural and functional adverse outcomes. Exploration of the role of cardiomyocyte Fru-AGE formation as a substrate of heart damage, particularly in the context of high cardiac fructose in diabetes, is required. The development of selective and sensitive molecular tools for identifying fructose-specific adduct types will allow new research, mapping the evolution of fructose-dependent AGE pathology in the myocardium.
Overview and New Directions
Consideration of the literature from epidemiological, clinical, and experimental perspectives provides an abundance of evidence attesting to fructose participation in the etiology of diabetic cardiomyopathy—including involvement in cardiac metabolic, structural, and electromechanical pathologies. The findings indicate that when dietary provocation is a factor in diabetes induction, fructose (more than glucose) constitutes a particular “toxic” sugar challenge. Moreover, in the insulin-resistant/deficient diabetic cardiac milieu, abnormalities in fructose metabolism have the potential to contribute directly to myocardial disease evolution. Bringing together key clinical and experimental observations, the evidence suggests that dysregulated tissue fructose metabolism, and not specifically systemic glycemic exposure, is associated with the ultimate progression of diabetic cardiomyopathy to cardiac failure state.
Fructose is increasingly recognized as a critical cellular energy intermediate and signaling agent in many cell types. The available evidence suggests that cardiomyocyte fructose vulnerability could arise from exposure to elevated extracellular fructose (both direct and indirect consequences of dietary conditions) and to augmented intracellular fructose production (with polyol synthetic pathway involvement). In particular fructose-driven extracellular and intracellular posttranslational modifications, which exert both dynamic and permanent influence on myocardial structure and function (Fig. 3), have been identified as potential pathological provocateurs.
As urgent investigation priorities, fructose-driven AGE formation and O-GlcNAcylation processes, as well as the involvement of these events in inflicting cardiac damage, are highlighted. New studies that track diabetic disease induction and parameters of cardiac function mapped against shifts in systemic and myocardial fructose handling are required. Defining the pathophysiological attributes of the fructose-damaged heart can potentially provide impetus in establishing a case for dietary fructose intake limitation as a cardioprotective measure. Characterizing the role of cardiomyocyte fructose dysregulation in the development of diabetic cardiomyopathy will provide a substrate for identifying targeted interventions to achieve damage remediation.
Acknowledgments. We acknowledge Brendan Ma from the University of Melbourne and Andrew Lim from the University of Auckland for their assistance in the early stages of literature compilation for manuscript development. We acknowledge funding support from the Diabetes Australia Research Trust.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.