https://doi.org/10.1007/s00392-025-02625-4
1Universitätsmedizin der Johannes Gutenberg-Universität Mainz Präventive Kardiologie und Medizinische Prävention Mainz, Deutschland; 2Universitätsmedizin der Johannes Gutenberg-Universität Mainz Zentrum für Kardiologie Mainz, Deutschland; 3Universitätsmedizin der Johannes Gutenberg-Universität Mainz Kardiologie 1, Zentrum für Kardiologie Mainz, Deutschland; 4University Medical Center of the Johannes Gutenberg University Mainz, Germany; Institute of Clinical Chemistry and Laboratory Medicine, Mainz, Deutschland
Background Monocytes and fibroblasts contribute to heart failure (HF) through inflammation and fibrotic remodeling of the left ventricle (LV). This study aims to identify proteomic markers of monocyte and fibroblasts involvement in HF and their association with cardiac structure, function, and HF outcome.
Methods Data from the MyoVasc study (N=3,289; NCT04064450), a HF cohort with deep clinical and molecular phenotyping, were analyzed. Plasma levels of 538 proteins were measured using a targeted immuno-qPCR-based proteomics assay (Olink Proteomics, Sweden). The monocyte protein signature was derived with elastic net linear regression models by predicting monocyte quantification in three orthogonal cell count estimation methods in blood: (1) differential blood counts, (2) RNA sequencing-based cell deconvolution, and (3) DNA methylation-based cell deconvolution. Proteins consistently associated with monocyte concentration across the methods were included in a monocyte protein score derived by principal component analysis (PCA). A fibroblast protein score was created based on a validated fibroblast gene set. Both protein scores were evaluated against measures of cardiac function and structure, as well as clinical outcome over 4 years, stratifying for HF with reduced EF (HFrEF, EF≤40%) and preserved EF (HFpEF, EF ≥50%). Pathway enrichment analysis was performed with GWAS catalogue, Cytoscape, PanglaoDB, and STRINGdb.
Results Monocyte counts and protein data were available for N=3,189 individuals, with RNAseq-based monocyte data in N=1,644 and DNA methylation-based monocyte data in N=1,004. Across the three cell count estimation methods, 33 proteins (top 3: IFN-γ, TNFSF13B, LILRB1) were consistently associated with monocytes. A total of 19 proteins were identified as part of the fibroblast gene set and confirmed to be fibroblast-specific through enrichment analysis. Pathways were involved in immune response, cytokine activity and fibrosis. In HFrEF, the monocyte protein score was associated with systolic (EF per standard deviation (SD) -0.08 [-0.14; -0.02], p=0.013), and diastolic dysfunction (E/E’SD 0.15 [0.01; 0.29], p=0.036), and LV hypertrophy (Relative Wall Thickness (RWTSD) 0.17 [0.06; 0.27], p=0.0024). In HFpEF, it was related to diastolic dysfunction (E/E’SD 0.10 [0.04; 0.16], p=0.0014) and LV hypertrophy (RWTSD 0.11 [0.05; 0.18], p=0.0009). The monocyte protein score predicted worsening of HF in HFpEF (Hazard Ratio (HR) 1.35 [1.13; 1.63], p=0.0012), but not in HFrEF (HRSD 0.99 [0.79; 1.25], p=0.96; interaction p<0.001) independently of the clinical profile and NT-proBNP. The fibroblast protein score showed strong associations with systolic (HFrEF: EFSD -0.12 [-0.19; -0.06], p=0.0001) and diastolic dysfunction (HFrEF: E/E’SD 0.24 [0.10; 0.38], p=0.00089; HFpEF: E/E’SD 0.13[0.07;0.19] <0.0001), and LV remodeling (HFrEF: (RWTSD 0.18 [0.07; 0.28] p=0.0017); HFpEF: RWTSD 0.15 [0.08; 0.21] <0.0001). The fibroblast protein score was associated with worsening of HF (HR 1.43 [1.26; 1.63], p<0.0001) independently of the monocyte protein score. In contrast, the monocyte score was not associated with worsening of HF when adjusted for the fibroblast score (HR 1.02 [0.89; 1.15], p=0.81).
Conclusion Monocyte and fibroblast protein signatures were prognostically relevant for worsening of HF, particularly in HFpEF. The fibroblast signature predicts HF outcome, emphasizing fibroblasts as mediators of monocyte-driven remodeling and dysfunction in HF.