A pilot study on DNA methylation changes for non-invasive molecular diagnostics in heart failure

The current therapeutic approach to ischaemic (IsHF) and non-ischaemic (NIsHF) heart failure (HF) mainly overlooks the underlying aetiology owing to a lack knowledge of the differential molecular pathways that contribute to HF with reduced ejection fraction (HFrEF). Alterations in myocardial DNA methylation levels have been identified as potential biomarkers for HF irrespective of its aetiology. Due to the limited availability of cardiac tissues in clinics, our goal is to determine if DNA methylation changes in circulating CD4⁺ T lymphocytes, which are strongly involved in left ventricle remodelling, can help in differentiating IsHF and NIsHF causes among patients with HFrEF and if DNA methylation levels associate with key clinical features.

We performed a post hoc network-oriented analysis of the original PRESMET clinical trial dataset (NCT05475028). Integrating epigenomic data obtained with the high-resolution reduced representation bisulfite sequencing (RRBS) platform and the left-ventricle interactome (protein-protein interaction map) we identified six differentially methylated CpG positions (DMPs), which were able to distinguish IsHF (n = 8) versus NIsHF (n = 4) patients with an area under the curve (AUC) > 0.8. Network-oriented DMPs were significantly hypomethylated in IsHF versus NIsHF and annotated to six genes, namely, cytoskeleton-associated protein 4 (CKAP4), carnitine palmitoyltransferase 1A (CPT1A), eukaryotic translation initiation factor 2 subunit beta (EIF2S2), spectrin beta (SPTB), synaptotagmin 6 (SYT6) and RAB11 family interacting protein 1 (RAB11FIP1) (P < 0.05). We found that the CD4⁺ T cell hypomethylation of EIF2S2 gene significantly correlated with VO₂ max (ρ = 0.73, P = 0.04), hypomethylation of RAB11FIP1 significantly correlated with MECKI score (ρ = -0.85, P = 0.001), whereas SPTB significantly correlated with VO₂ max (ρ = 0.73, P = 0.04), left ventricle mass index (ρ = -0.91, P = 0.005) and left ventricular ejection fraction (ρ = 0.83, P = 0.01) in IsHF patients.

We demonstrate that circulating CD4⁺ T cell-specific methylation levels of network-oriented CKAP4, CPT1A, EIF2S2, SPTB, SYT6 and RAB11FIP1 genes can distinguish between IsHF and NIsHF. In particular, hypomethylation of EIF2S2, SPTB and RAB11FIP1 genes is significantly correlated with key clinical features in isHF patients, highlighting its potential to enhance personalized prognosis for HFrEF patients.