Background:
The prevalence of heart failure with preserved ejection fraction (HFpEF) is anticipated to rise in tandem with the prolongation of our lifespan and the increasing burden of comorbidities. Despite the significant public health implications, evidence-based therapeutic interventions for HFpEF are limited, and the underlying pathomechanisms remain elusive, although low-grade systemic inflammation and endothelial dysfunction have been postulated to contribute.
Objective:
Here, we aimed to identify cardiac proteomic and phosphoproteomic profiles that differentiate non-obese HFpEF from heart failure with reduced ejection fraction (HFrEF), using non-failing donor hearts as controls.
Methods & Results:
Left ventricular myocardial tissue from explanted or non-failing donor hearts (Medical University Graz) was classified as Control (Ctrl), HFpEF, or HFrEF (mean EF [%]: 63 [Ctrl], 62 [HFpEF], 24 [HFrEF]; mean age 58.1±9.5; mean BMI 26.6±2.7 kg/m²) and analysed by label-free quantitative mass spectrometry (N=6-7/group). Cardiac proteome analysis revealed 49 cardiac proteins that were differentially abundant (fold change >1.5, p<0.05) in HFpEF and 160 in HFrEF versus Ctrl hearts, out of 1813 detected cardiac proteins. Gene ontology enrichment analysis of the cardiac proteome revealed that proteins associated with “complement activation” (e.g., C6, C4B) and “innate immune response” (e.g., S100A8, STING1) were significantly upregulated in HFpEF versus Ctrl hearts. Both terms also appeared in the enrichment analysis of HFrEF versus Ctrl hearts, indicating low-grade systemic inflammation as a shared feature of both HF subtypes. However, the extent and phenotype differed: complement activation and myeloid-associated inflammation were moderate in HFpEF but more pronounced in HFrEF, which also showed humoral immune activation and thromboinflammatory remodelling. Comparison of significantly altered proteins in HFpEF or HFrEF versus Ctrl hearts identified 19 proteins uniquely regulated in HFpEF. Two of these proteins are of particular interest: upregulated THBS4, indicating activation of the THBS4-ATF6 unfolded protein response axis, and downregulated CDH13, suggesting impaired adiponectin-mediated anti-inflammatory signalling. Phosphoproteome analysis identified 83 differentially phosphorylated phosphosites in cardiac proteins (fold-change >1.5, p<0.05) in HFpEF and 181 in HFrEF versus Ctrl hearts, out of 1045 different phosphosites mapped in 460 cardiac proteins. Phosphoproteomic alterations in both non-obese HFpEF and HFrEF versus Ctrl hearts mainly concerned sarcomeric proteins (e.g., titin, desmin), which exhibited predominant hyper-phosphorylation. In HFpEF versus Ctrl hearts, GO enrichment analysis revealed that proteins associated with the molecular function term “protein kinase binding“ (e.g., protein phosphatase 1 regulatory subunits 12A, 12B and 12C) were hyper-phosphorylated.
Conclusions:
Both HF subtypes showed low-grade systemic inflammation independent of obesity. However, HFpEF showed a milder complement-positive, myeloid-active phenotype compared with HFrEF, which displayed stronger inflammatory activation with additional humoral and thromboinflammatory features. THBS4- and CDH13-related pathways may represent HFpEF-specific targets. In addition, our data suggest a link between systemic inflammation and altered sarcomeric protein phosphorylation.
