Serum proteomic signatures of atrial fibrillation in HFpEF/HFmrEF

Background

Atrial fibrillation (AF) frequently complicates the clinical course of patients with heart failure with preserved or mildly reduced ejection fraction (HFpEF/HFmrEF), yet the underlying molecular mechanisms extend far beyond electrical atrial remodelling. High-throughput serum proteomics may provide deeper insights into systemic alterations accompanying AF in this heterogeneous population.

Methods

Serum samples from 86 HFpEF/HFmrEF patients (54 with AF) treated at our center were analysed using the Olink® Explore platform (>5400 proteins). Associations between circulating proteins and AF were assessed using multivariable linear regression. Functional enrichment (GO, WikiPathways), disease association analyses (DisGeNET), and protein-protein interaction (PPI) networks (STRING v12.0) were used to characterize biological signatures. Correlations between AF-associated proteins and clinical parameters were visualized using hierarchical clustering heatmaps. Local ethics approval was obtained.

Results

A total of 126 proteins showed significant associations with AF (p < 0.05). The volcano plot demonstrated clear differential expression, with several proteins exhibiting strong effect sizes and separation between AF and sinus rhythm. The most prominently altered markers included PDE1C, ASS1, PASD1, PM20D1, DBP, WNT7A, CLYBL, TP53INP1, NXPE4, and YU005.

The PPI network revealed seven interconnected protein clusters linked to extracellular matrix remodelling, oxidative stress, lipid metabolism, inflammatory activation, and neurohumoral regulation. Enrichment analyses confirmed pathways related to coagulation, matrisome organisation, superoxide metabolism, lipid catabolism, and JNK/Wnt signalling. Integrated analysis showed that AF in HFpEF/HFmrEF is characterised by a coordinated pattern of structural remodelling, thrombo-inflammatory activity, metabolic and oxidative imbalance, and neurohumoral stress, supporting its nature as a systemic condition rather than a purely cardiac rhythm disorder. Clinical correlation heatmaps further demonstrated strong positive associations of AF-related proteins with markers of atrial and ventricular remodelling, while reduced renal function and impaired right ventricular performance showed inverse associations, reflecting hemodynamic load, inflammation, and metabolic stress.

Conclusions

Comprehensive plasma proteomics reveals a distinct and multidimensional molecular signature of AF in HFpEF/HFmrEF, driven by interconnected pathways of structural remodelling, inflammation, oxidative and metabolic dysregulation, and neurohumoral stress. These systemic alterations may refine our biological understanding of AF in heart failure and highlight several promising biomarker candidates and potential therapeutic targets.