Diabetic Cardiomyopathy: What is it and can it be fixed?

Abstract

Diabetic cardiomyopathy was initially described as a human pathophysiological condition in which heart failure occurred in the absence of coronary artery disease, hypertension, and valvular heart disease. Recent studies in diabetic animal models identify decreased cardiomyocyte function as an important mediating mechanism for heart failure. Decreased cardiomyocyte function is in part mediated by abnormal mitochondrial calcium handling and a decreased level of free matrix calcium levels which could be a good target for new therapeutic interventions.

Introduction

Diabetic Cardiomyopathy (DC), is a diabetes mellitus (DM)-induced pathophysiological condition that can result in heart failure (HF). Here we present the point of view that diminished cardiomyocyte contraction is a significant contributor which is currently not included amongst the contributing mechanisms for DC and is largely mediated by changes in the level and/or posttranslational modification of specific cardiomyocyte proteins. Correcting these changes may lead to novel therapeutic approaches.

Heart failure, diabetes, and cardiovascular disease.

DM occurs in 9.3% of the U.S. population. The prevalence of HF is high in patients with DM, ranging from 19–26% (1). HF occurs in both Type 1 Diabetes (T1D) and Type 2 Diabetes (T2D). T2D accounts for 90–95% of DM cases and is frequently linked to obesity.

In diabetic patients cardiovascular disease is the leading cause of death with coronary artery disease (CAD) and ischemic cardiomyopathy as main contributors. In addition to CAD, small vessel disease and diminished cardiac capillary density occurs. Diabetic patients can present initially with impaired diastolic cardiac function (1), but preserved systolic contraction which is termed Heart Failure with Preserved Ejection Fraction (HFpEF) and may account for 50% of all HF.

In 1972 Shirley Rubler identified a new type of cardiomyopathy in diabetic patients termed DC (2). These patients had a history of HF in the absence of CAD, hypertension, or valvular heart disease and it was postulated that the myocardial disease is due to diffuse myocardial fibrosis, cardiac hypertrophy, and diabetic microangiopathy. This definition does not include abnormal cardiomyocyte function and fits the data available at that time. The role of DM in HF was found in the Framingham study (3). Human DC has become a well-documented condition. In T1D and T2D experimental animal models, decreased diastolic and systolic contractile function accompanied by diminished cardiomyocyte contraction and changes in specific cardiomyocyte proteins have been demonstrated (4,5). This leads to the point of view that abnormal contractile function of the diabetic cardiomyocyte could be included in the definition of DC to make it more integrated and conclusive.

In contrast to earlier trials (6), more recent trials in diabetic patients using Sodium/Glucose Exchange Inhibitors and glucagon-like peptide receptor agonists show significant improvements in cardiac contractile function (1). In addition, if myocardial infarct area is adjusted to equal size in non-diabetic and diabetic patients, the incidence of HF is significantly higher in the diabetic patients than in those without the disease. These findings also suggest that ischemic and diabetic cardiomyopathy are frequently interrelated entities amplifying maladaptive contractile effects in DM patients.

Mechanisms contributing to the development of diabetic cardiomyopathy.

Multiple mechanisms contribute to decreased performance of the diabetic heart and have been reviewed (1). They include exposure of the heart to the diabetic milieu of hyperglycemia along with increased fatty acids (FA) and cytokines. Hyperglycemia enhances enzymatic O-GlcNAcylation of cardiomyocyte proteins and is maladaptive. Increased chemical non-enzymatic Advanced Glycation End-product (AGE) formation also occurs with detrimental effects. A diabetic autonomic neuropathy is present and linked to hyperglycemia. Exposure to increased lipid levels including FA and triglycerides causes increased fat droplet accumulation in cardiomyocytes mediating cardiac lipotoxicity. Decreased insulin signaling is a hallmark of T1D and T2D and alterations in other signaling cascades occur, including decreased AMPK signaling and increased PKC and MAPK signaling with maladaptive consequences.

DM-induced changes in specific cardiomyocyte proteins linked to calcium (Ca²⁺) handling.

Exploring DM-induced changes in molecular mechanisms mediating cytosolic and mitochondrial (Mito) Ca²⁺ handling may identify novel therapeutic approaches.

Cytosolic Ca²⁺ handling in diabetic cardiomyocytes.

In a T1D rat model depressed sarcoplasmic reticulum (SR) function was reported by James Scheuer, et al. in 1981 (4). Subsequent studies in a T1D mouse model found a significant decrease in SERCa2a protein. Decreased contractile function occurred and was rescued by expressing a SERCa2a transgene (tge) (5). These findings support that cardiomyocyte dysfunction can be included in the definition of DC and suggest that enhancing SERCa2a function can be a therapeutic target.

Decreased contractile function also occurs in the T2D db/db mouse model. An abnormal cytosolic Ca²⁺ transient and increased SR Ca²⁺ leak occur, accompanied by decreases in SERCa2a and RyR2 levels (not statistically significant). Increased phospholamban (Pln) and decreased Pln phosphorylation mediate increased SERCa2a inhibition. In addition, an abnormal cardiomyocyte Ca²⁺ transient and a marked decrease in RyR2 protein levels were identified (8). T2D db/db mice are leptin receptor deficient. Leptin has adaptive cardiovascular effects making it difficult to distinguish if decreased cardiac function is only due to DM or if absent leptin signaling contributes.

The T2D model of Otsuka Long Evans Tokushima Fatty (OLETF) rats have reduced SERCa2a protein levels and exhibit impaired diastolic function. Treatment with adenoviral vector (Adv)-based SERCa2a tge expression improves contractile function. Increased cardiomyocyte size in T2D hearts is restored to normal, but no influence on collagen production occurs. Although the exact mechanism(s) for these effects are uncertain, increased SERCa2a expression in isolated cardiomyocytes has been shown to increase the expression of genes linked to insulin signaling (9).

For human DC, only limited results for myocardial contractile function or the level of cardiomyocyte proteins are available. Acto-myosin cross-bridge kinetics and work output at varying calcium concentrations were determined in male and female patients with DM. The findings suggest that changes in diabetic cardiac muscle function contribute to the incidence and mortality of DM-induced HF (10).

Conclusion for DM and cytosolic Ca²⁺ handling.

The data from T1D and T2D animals justify the inclusion of abnormal cytosolic Ca²⁺ handling in cardiomyocytes as an important contributor to DC. Studies in specific T1D and T2D animal models imply that different proteins linked to Ca²⁺ handling are potential targets. Data from T1D mice and T2D rats point to SERCa2a, whereas results from T2D db/db mice also identify proteins linked to SR Ca²⁺ release by the ryanodine receptor and its regulatory proteins.

Although quantitation of proteins linked to cytosolic Ca²⁺ handling in human diabetic hearts is not available, patients with HF due to other causes exhibit decreased SERCa2a expression and diminished contractile function. Based on findings in animals, restoring cytosolic Ca²⁺ handling in human diabetic cardiomyocytes may be a promising strategy. Discussing the Calcium Upregulation by Percutaneous Administration of Gene Therapy In Patients with Cardiac Disease (CUPID) trials, using AAV1 SERCa2a tge expression in human HF patients, is relevant because nearly half of the patients had DM. CUPID1 enrolled 39 HF patients and was a phase1/2 trial using percutaneous intracoronary infusion of AAV1/SERCa2a (11). Improvements in clinical and cardiac measurements occurred in the high dose AAV1/SERCa2a group. A subsequent larger phase2b CUPID2 trial, however, failed to demonstrate a reduction in recurrent HF hospitalization. A specific cause for the different results of CUPID1 versus CUPID2 is difficult to identify, but the patient numbers in CUPID1 were small and the improvements may have been the play of chance despite positive statistics. Alternatively, a low level of cardiac AAV1/SERCa2a expression occurred in CUPID2 and is therefore a potential contributor for the failure to demonstrate benefit. The preparation of AAV1/SERCa2a for CUPID1 versus CUPID2 was different and may have contributed to the discrepant results (11). In contrast to humans, successful peripheral intravenous administration of AAV9.45 tges occurs in rodents, with tge expression in a high percent of cardiomyocytes. For humans, the AAV tge approach needs further development, but has been curtailed by limited funding (12). HF is a multifactorial condition and focusing only on SERCa2a rectification may be insufficient.

Mitochondrial Ca²⁺ handling in diabetic cardiomyocytes.

Mito Calcium Uniporter Complex (MCUC)-based cardiomyocyte Mito Ca²⁺ import, as well as Mito Ca²⁺ export by the Mito Sodium Calcium Lithium exchanger, and the influence which the free Mito matrix calcium level ([Ca²⁺]_m) has on Mito energetic function have been reviewed (13). The MCUC is a highly selective channel that moves Ca²⁺ ions across the Mito inner membrane driven by the Mito membrane potential (Δψ_m)_. Four MCU homodimers form the Ca²⁺ channel pore. EMRE is a 10 kDa protein which is essential for MCUC Ca²⁺ conductance. The Ca²⁺-sensing proteins MICU1 and MICU2 form dimers and interact electrostatically with EMRE and directly with MCU. Mito Ca²⁺ import and export determine [Ca²⁺]_m which stimulates oxidative phosphorylation by enhancing the activity of Complex I, III, IV and the V_max of Complex V, leading to increased ATP formation. The pyruvate dehydrogenase complex (PDC), mediating glucose oxidation (GOX), is also activated by [Ca²⁺]_m. With increased GOX the oxidation of fatty acids (FAOX) decreases, diminishing oxygen consumption for contractile work, making it more energetically efficient.

In T1D hearts MCU and EMRE are significantly decreased. Decreased transcriptional expression of MCU is mediated by high glucose exposure. Maladaptive consequences resulting from decreased [Ca²⁺]_m include decreased PDC acitivity with diminished GOX, increased FAOX, decreased Δψ_m, increased oxidative stress, and increased apoptotic cardiomyocyte death. Adv-based MCU tge expression in cardiomyocytes eliminates all of the high glucose-induced maladaptive effects. Similar results are obtained by in vivo studies in T1D mice with AAV9.45 MCU tge expression restoring [Ca²⁺]_m (14). The beneficial energetic, metabolic, and contractile effects occur and MI size is significantly smaller in T1D+AAV-MCU mice versus T1D mice without MCU restoration. Restoration of [Ca²⁺]_m in T2D hearts resulted in similar adaptive effects including restoration of cardiac contractile function (unpublished results). No maladaptive consequences could be identified from MCU restoration.

Conclusion for DM and Mito Ca²⁺ handling.

Restoring [Ca²⁺]_m in T1D and T2D hearts improves several functions including enhanced Mito energetic and metabolic activity, as well as improved metabolic fuel flux with increased GOX and decreased FAOX with less oxygen consumption for increased cardiac work. In addition, SERCa2a gene expression and protein levels increase in DM cardiomyocytes improving cytosolic Ca²⁺ flux with enhanced diastolic and systolic cardiac function. In contrast, restoring SERCa2a protein levels in DM cardiomyocytes results in a more limited response primarily enhancing cytosolic Ca²⁺ handling and contractile function. It is therefore our point of view that restoring Mito Ca²⁺ handling is the preferred approach over improving only cytosolic Ca²⁺ flux. One potential mechanism for increased SERCa2a gene expression with improved Mito Ca²⁺ handling may be retrograde Mito-Nuclear signaling, but the precise nature of this signal is currently unclear.

A potential problem with the restoration of MCUC in diabetic cardiomyocytes could be Mito Ca²⁺ overload triggering apoptotic cardiomyocyte death. We have not observed this detrimental consequence and the approach for the restoration of MCUC function may make a difference. For example, restoring the decreased levels of the MCU-bracketing protein EMRE, instead of MCU, may make it less likely that MCUC levels exceed the normal range and excessive [Ca²⁺]_m results with detrimental effects.

Summary:

HF due to DC was described over 40 years ago and in the last three decades new knowledge, largely derived from diabetic animal models, has accumulated. These results have shown that decreased function of diabetic cardiomyocytes is a significant contributor for the development of HF. Decreased cardiomyocyte function of the diabetic heart was, however, not included amongst the contributing factors in the original description (2). Currently only a very small number of studies from diabetic patients exploring myocardial function and the molecular mechanisms mediating it have been reported. If one accepts guidance from T1D and T2D animal models, improving cytosolic Ca²⁺ and especially Mito Ca²⁺ handling of the human diabetic heart are appropriate targets for future therapeutic interventions.