As part of the ERC UCARE project, we identified a plaque-specific epigenetic signal in cell-free DNA (cfDNA) from the plasma of women with CAD. This signal, consisting of hypomethylated regions specific to plaques, can be detected in plasma cfDNA. Our methods thus far have utilized whole genome sequencing, but to study the predictive value of these regions in large cohorts and explore commercial potential, we need to develop a more specific methodology. The current U-BiomarCARE project aims to determine whether the methylation status of plaque-specific regions can be detected in plasma using a droplet digital PCR (ddPCR)-based method. This method targets the most plaque-specific CpG candidate regions, using predesigned methylated and unmethylated probes, allowing us to enrich, amplify, and quantify the methylated and unmethylated signals at these genomic regions.
In previous analyses, we demonstrated that 18-53% of plaque DNA is unmethylated at this CpG site, while other tissues and cell types show less than 5% unmethylation. We confirmed the plaque-specific nature of this signal by applying the ddPCR assay to genomic DNA from 16 plaques, where unmethylated fragments were found in all samples with an average of 46%. These fragments were absent in heart and peripheral blood mononuclear cell (PBMC) DNA, further emphasizing their plaque-specificity. Additionally, we observed that smooth muscle cell (SMC)-derived myofibroblasts were the primary source of this plaque-specific methylation pattern, while endothelial cells showed no detectable unmethylated signal.
To support the role of this CpG in plaque biology, we integrated bulk and single-cell RNA data from the same atherosclerotic plaques. The top CpG is located upstream of MYH10, a well-known vascular SMC gene, and we showed that MYH10 expression in plaques inversely correlates with methylation status. High MYH10 expression was particularly prominent in plaques with a fibrous phenotype. This is further supported by single-cell RNA sequencing (scRNAseq), which revealed that genes negatively associated with this CpG were predominantly expressed in SMCs.
To explore the potential for detecting plaque-specific DNA in circulation, we applied the ddPCR assay to plasma cfDNA from 52 patients with suspected CAD. Among CAD patients, 40% tested positive for an unmethylated signal, although this was not significantly different from the 20% positivity rate observed in patients without significant atherosclerosis. While there was a trend toward higher proportions of unmethylated fragments in CAD patients, the difference did not reach statistical significance.
By demonstrating that methylation-based circulating plaque markers can be measured using this sequencing-free ddPCR method, we are advancing towards the ability to detect plaque-specific DNA in circulation. Once proven reproducible and comparable to whole genome sequencing, this methodology can expand to a capture-based sequencing method that will enable large biomarker studies in both women and men suspected of heart disease. This would significantly improve early and accurate detection of CAD, potentially revolutionizing the healthcare sector by allowing for specific, timely identification of dangerous atherosclerotic plaques in both women and men. For this, we have applied for a new ERC PoC that will develop a larget set of CpGs.
