Defects in the Proteome and Metabolome in Human Hypertrophic Cardiomyopathy

Abstract

Background: Defects in energetics are thought to be central to the pathophysiology of hypertrophic cardiomyopathy (HCM); yet, the determinants of ATP availability are not known. The purpose of this study is to ascertain the nature and extent of metabolic reprogramming in human HCM, and its potential impact on contractile function.

Methods: We conducted proteomic and targeted, quantitative metabolomic analyses on heart tissue from patients with HCM and from nonfailing control human hearts.

Results: In the proteomic analysis, the greatest differences observed in HCM samples compared with controls were increased abundances of extracellular matrix and intermediate filament proteins and decreased abundances of muscle creatine kinase and mitochondrial proteins involved in fatty acid oxidation. These differences in protein abundance were coupled with marked reductions in acyl carnitines, byproducts of fatty acid oxidation, in HCM samples. Conversely, the ketone body 3-hydroxybutyrate, branched chain amino acids, and their breakdown products, were all significantly increased in HCM hearts. ATP content, phosphocreatine, nicotinamide adenine dinucleotide and its phosphate derivatives, NADP and NADPH, and acetyl CoA were also severely reduced in HCM compared with control hearts. Functional assays performed on human skinned myocardial fibers demonstrated that the magnitude of observed reduction in ATP content in the HCM samples would be expected to decrease the rate of cross-bridge detachment. Moreover, left atrial size, an indicator of diastolic compliance, was inversely correlated with ATP content in hearts from patients with HCM.

Conclusions: HCM hearts display profound deficits in nucleotide availability with markedly reduced capacity for fatty acid oxidation and increases in ketone bodies and branched chain amino acids. These results have important therapeutic implications for the future design of metabolic modulators to treat HCM.

Keywords: cardiomyopathy, hypertrophic; mitochondrial proteins; phosphate; phosphocreatine; proteomics.

Figure 1.. Volcano plots summarizing findings from proteomics analysis of human HCM and control hearts.

Proteins with increased (+ log) or decreased abundance (- log) are shown as a function of fold change over control hearts for HCM samples with (A) no sarcomere variants, (B) MYBPC3 variants, and (C) MYH7 variants. Proteins that were significantly different from control, after Bonferroni correction for multiple comparisons (P<0.0001), are shown in red and a subset of the individual proteins are labeled. N=6 controls, 18 HCM-MYBPC3, 9 HCM-MYH7, and 12 HCM - No sarcomere variant.

Figure 2.. Metabolomic analysis of HCM and control hearts.

(A) PCA plot showing the overall distinction between the control hearts and the 3 HCM groups. The 3 HCM groups overlap with each other with more variability in the HCM – no sarcomere variant group. (B) Heat map showing the individual metabolites that were significantly different in the HCM compared to the control heart samples after adjustment by false discovery rate. N=8 for controls, HCM- MYH7, HCM-no sarcomere variant and N=12 for HCM-MYBCP3.

Figure 3.. Metabolomic analysis showing representative abundances of long-chain acyl-carnitines and ketone bodies in HCM and control hearts.

Abundances of C10, C12, C14, C16 and C18 were markedly and uniformly depleted in all HCM groups, independent of genotype, compared to control hearts. Ketone bodies were increased in abundance across all 3 HCM groups. ANOVA with Tukey post-hoc comparison were used to compare groups with P<0.0001 for all comparisons of HCM groups to control. N=8 for controls, HCM- MYH7, HCM-no sarcomere variant and N=12 for HCM-MYBCP3.

Figure 4.. Metabolomic analysis showing increased abundance of branched chain amino acids in the hearts of patients with HCM compared to controls.

N=8 for controls, HCM- MYH7, HCM-no sarcomere variant and N=12 for HCM-MYBCP3. P<0.0001 for all comparisons of HCM to control hearts.

Figure 5.. Metabolomic analysis showing reductions of acetyl CoA, trinucleotides, NADH and NADPH with increased abundance of NMN, the precursor of NAD in HCM samples compared to controls.

N=8 for controls, HCM- MYH7, HCM-no sarcomere variant and N=12 for HCM-MYBCP3. P<0.0001 for all comparisons of HCM to control samples.

Figure 6.. Rate of myosin cross-bridge detachment slows at lower MgATP concentrations.

A. A quick stretch of 1% muscle length (ML) was applied to demembranated epicardial muscle that had been biopsied from patients undergoing coronary bypass grafting. B. The recorded tension response (solid lines) at maximum Ca^2+-activation was fit to the double-exponential function (dotted lines). The rate of myosin crossbridge detachment is reflected in the rate of tension release (krel) immediately following the stretch. C. Lower concentrations of MgATP resulted in slower krel and would be expected to reduce relaxation function in the cardiac cycle during transition from late systole to early diastole. Mean ± SEM, n=32 samples from 14 patients. D. Inverse correlation between left atrial diameter (mm) and the content of ATP in the corresponding heart tissue for each patient. Correlation analysis was performed using the Spearman rank sum non-parametric test, yielding a r of −0.47, P=0.01. N=40 individual values.