Integrating Clinical Phenotype With Multiomics Analyses of Human Cardiac Tissue Unveils Divergent Metabolic Remodeling in Genotype-Positive and Genotype-Negative Patients With Hypertrophic Cardiomyopathy

Abstract

Background: Hypertrophic cardiomyopathy (HCM) is caused by sarcomere gene mutations (genotype-positive HCM) in ≈50% of patients and occurs in the absence of mutations (genotype-negative HCM) in the other half of patients. We explored how alterations in the metabolomic and lipidomic landscape are involved in cardiac remodeling in both patient groups.

Methods: We performed proteomics, metabolomics, and lipidomics on myectomy samples (genotype-positive N=19; genotype-negative N=22; and genotype unknown N=6) from clinically well-phenotyped patients with HCM and on cardiac tissue samples from sex- and age-matched and body mass index-matched nonfailing donors (N=20). These data sets were integrated to comprehensively map changes in lipid-handling and energy metabolism pathways. By linking metabolomic and lipidomic data to variability in clinical data, we explored patient group-specific associations between cardiac and metabolic remodeling.

Results: HCM myectomy samples exhibited (1) increased glucose and glycogen metabolism, (2) downregulation of fatty acid oxidation, and (3) reduced ceramide formation and lipid storage. In genotype-negative patients, septal hypertrophy and diastolic dysfunction correlated with lowering of acylcarnitines, redox metabolites, amino acids, pentose phosphate pathway intermediates, purines, and pyrimidines. In contrast, redox metabolites, amino acids, pentose phosphate pathway intermediates, purines, and pyrimidines were positively associated with septal hypertrophy and diastolic impairment in genotype-positive patients.

Conclusions: We provide novel insights into both general and genotype-specific metabolic changes in HCM. Distinct metabolic alterations underlie cardiac disease progression in genotype-negative and genotype-positive patients with HCM.

Keywords: cardiomyopathies; genotype; hypertrophy; lipidomics; metabolism; metabolomics.

Figure 1.

Overview of metabolomic differences between hypertrophic cardiomyopathy (HCM) myectomy tissue and cardiac tissue from nonfailing (NF) donors. A, Principal component analysis (PCA) showing distinct clustering of patients with HCM and NF donors. B, Heatmap showing z scores of all significantly different metabolites in HCM vs NF donor hearts (Q<0.05). C, Top 10 enriched metabolite sets in HCM vs NF donors analyzed via MetaboAnalyst 5.0. dAMP indicates deoxyadenosine monophosphate; Gluconate-6P, gluconate-6-phosphate; GMP, guanosine monophosphate; Hexose-P, hexose-phosphate; IMP, inosine monophosphate; NADH, reduced nicotinamide adenine dinucleotide; NADPH, reduced nicotinamide adenine dinucleotide phosphate; PRPP, phosphoribosyl pyrophosphate; and TCA, tricarboxylic acid.

Figure 2.

Overview of lipidomic differences between hypertrophic cardiomyopathy (HCM) myectomy tissue and cardiac tissue from nonfailing (NF) donors. A, Principal component analysis (PCA) showing distinct clustering of patients with HCM and NF donors. B through E, log2-transformed data for total amounts of lipids within a class, grouped per theme. Abbreviations of lipid classes are detailed in Table S3. F through J, Major alterations in individual lipid species abundance within a lipid class as dot plots, in which each dot represents a significantly altered lipid species. Vertical axes indicate double bond number, and horizontal axes indicate the carbon chain length of lipid species. The color and size of dots indicate fold change (FC) between groups and the significance of these changes, respectively.

Figure 3.

Summary of proteomic differences between hypertrophic cardiomyopathy (HCM) and nonfailing (NF) donors. A, Principal component analysis (PCA) showing distinct clustering of patients with HCM and NF donors. Top 5 downregulated clusters of interacting proteins (B) and the top 5 upregulated clusters of interacting proteins (C) according to cluster size. The color and size of protein nodes indicate fold change between groups and the significance of these changes, respectively.

Figure 4.

Comprehensive overview of alterations in energy metabolism and lipid-handling pathways in patients with hypertrophic cardiomyopathy (HCM) vs nonfailing (NF) donors integrating metabolomics, lipidomics, and proteomics data. Metabolomic, lipidomic, and proteomic data were integrated to provide an overview of changes in major energy metabolism and lipid-handling pathways. Metabolites and lipids are indicated by objects with a dark fill and white text; proteins are indicated by objects with a light fill and black text. Blue, orange, and gray indicate lower, higher, and unchanged abundances in HCM vs NF donors, respectively. Created with BioRender.com.

Figure 5.

Correlations between cardiac phenotype and metabolites and lipids in patients with hypertrophic cardiomyopathy (HCM). Correlations were tested via linear regression and are described by the Pearson R coefficient. A, All significant correlations between cardiac parameters and metabolites in genotype-negative (G_negative; G−) and genotype-positive (G_positive; G+) patients. B, All significant correlations between cardiac parameters and lipids classes in G_negative and G_positive patients. Blue and orange lines indicate negative and positive correlations, respectively. Line thickness corresponds with the significance level of the correlation. Exact R coefficients and P values, as well as group comparison statistics between nonfailing donors and G− or G+ patients of metabolites and lipids in these correlations, are available as Supplemental Material (Correlation statistics). IVSi indicates septal thickness indexed to body surface area; and LADi, left atrial diameter indexed to body surface area.

Figure 6.

Contrasting metabolic alterations underlie cardiac remodeling in genotype-negative (G−) and genotype-positive (G+) patients with hypertrophic cardiomyopathy (HCM). Matrix summarizing correlation patterns between metabolite classes and (functional) cardiac parameters in G− and G+ patients with HCM. Negative and positive correlation patterns are colored blue and orange, respectively. The color intensity reflects the strength of correlation patterns. IVSi indicates interventricular septum thickness indexed to body surface area; LADi, left atrial diameter indexed to body surface area; and PPP, pentose phosphate pathway;

Figure 7.

Summary of the main findings of this study. Genotype-independent changes in hypertrophic cardiomyopathy (HCM) included downregulation of fatty acid oxidation and increased glucose metabolism and glycogen turnover; decreased lipid storage and ceramide formation and increased sphingomyelin abundance; and proteasome upregulation and downregulation of mitochondrial translation proteins. G_negative patients with HCM displayed negative associations between cardiac remodeling and fatty acid oxidation, mitochondrial function, and levels of biosynthetic metabolites (ie, pentose phosphate pathway [PPP] intermediates, purines, pyrimidines, and amino acids). Conversely, G_positive patients with HCM displayed positive associations between biosynthetic metabolites and cardiac remodeling. Thus, opposite metabolic changes are linked to cardiac disease in G_positive and G_negative patients. Created with BioRender.com.

Figure 8.

Limited agreement among studies into metabolomic and lipidomic changes in tissue of patients with hypertrophic cardiomyopathy (HCM) compared with nonfailing (NF) donor tissue. A, Pie chart visualizing agreement in metabolite findings in HCM vs NF donor cardiac tissue samples; 114 metabolites were detected in at least 2 studies. Of these metabolites, 92 displayed inconsistent outcomes in HCM vs NF donor tissue. B, Table displaying consistency in lipidomic findings in HCM vs NF donor tissue. Blue and orange indicate consistently lowered and elevated levels, respectively, in HCM vs NF donors. *Derived from both metabolomics and lipidomics.