DNA methylation, the addition of methyl groups to cytosine bases within CpG dinucleotides, is one of the principal epigenetic mechanisms regulating gene expression. Methylation of promoter-associated CpG islands generally silences transcription, whereas hypomethylation can permit gene activation. MicroRNAs, small single-stranded non-coding RNAs typically sixteen to twenty-seven nucleotides in length, regulate gene expression post-transcriptionally and are known to dichotomously control virtually every stage of atherosclerosis, from endothelial senescence and inflammation through smooth muscle cell proliferation and migration to the final rupture of an unstable plaque. Yet while methylation-dependent regulation of microRNA genes has been studied extensively in cancer, its contribution to vascular disease has remained poorly understood.

The Tomsk investigators, led by Iuliia A. Koroleva of the Research Institute of Medical Genetics, profiled DNA methylation of regulatory elements and gene bodies of five microRNA genes, MIR10B, MIR21, MIR100, MIR127, and MIR143, using targeted bisulfite amplicon sequencing on the Illumina MiSeq platform. The cohort comprised ninety-two patients with advanced carotid atherosclerosis, seventy percent men with a median age of sixty-four years, all with severe stenosis exceeding sixty-five percent, an indication for endarterectomy. Thirty-two subclinical atherosclerosis controls, with stenosis of roughly twenty-eight percent and intima-media thickness of about 0.8 millimeters, provided peripheral blood leukocytes for comparison. During surgery, the team collected atherosclerotic plaque samples from all ninety-two patients, macroscopically intact adjacent carotid artery tissue from thirty-four, and great saphenous vein samples from twenty-three.

Each plaque sample was divided into three fragments with similar macroscopic characteristics. One fragment underwent histological verification according to the American Heart Association classification system, allowing assessment of plaque stability and pathomorphological features such as intraplaque hemorrhage, foam cell content, fibrous cap rupture, calcification, and surface thrombus. A second fragment was used for DNA extraction, sodium bisulfite treatment, and methylation analysis. The third was preserved in RNAlater for subsequent microRNA expression profiling by small RNA sequencing on a HiSeq 1500 instrument.

The team analyzed eight genomic regions related to the five microRNA genes, including the CpG island overlaying the MIR10B gene body with twenty-three CpG sites, the literature-defined putative promoter of MIR21 with nine CpG sites, the MIR21 CpG island with sixteen sites, the MIR21 gene body with five sites, the E-box region containing the transcription factor ZEB1 binding site representing the putative regulatory region of MIR100 with seven CpG sites, the CpG island overlapping the MIR127 gene body with forty sites, the MIR143HG promoter with ten sites, and the MIR143 gene body with five sites. Across all eight regions, median DNA methylation patterns proved profoundly tissue-specific. Methylation profiles were highly similar between peripheral blood leukocytes from patients and controls but differed markedly between blood and vascular tissues.

Compared with intact adjacent carotid artery, atherosclerotic plaques demonstrated significantly lower median DNA methylation in the MIR10B region by 8.5 percent, the MIR21 gene body by 11.7 percent, the MIR100 E-box region by 9.7 percent, and the MIR143 gene body by 6.9 percent. Median methylation in these regions was also lower in plaques than in great saphenous vein, and the same pattern was observed for the MIR127 region. By contrast, the MIR143HG promoter, represented by the single cg01558401 site, exhibited higher methylation in plaques than in unaffected vessels, ranging from 34.2 to 54.4 percent versus 16.7 to 32.6 percent. To validate these findings, the team compared their sequencing results with a lab-internal epigenome-wide association study using the Illumina Infinium MethylationEPIC BeadChip and with five publicly available datasets from the Gene Expression Omnibus repository. Methylation patterns for MIR10B, MIR21, and MIR143 were highly consistent across independent studies, analytical platforms, and vascular tissues.

Cellular heterogeneity represents a critical potential confounder in methylation studies of vascular tissue, because blood vessels contain variable proportions of smooth muscle cells, vascular endothelial cells, fibroblasts, and immune cell populations, each characterized by distinct methylation profiles. To address this, the team estimated cell-type proportions using the HEpiDISH algorithm with original DNA methylation reference matrices across 102 vascular samples derived from combined external and lab-internal datasets. The deconvolution analysis revealed a striking mirror-image dichotomy. Higher methylation in the MIR10B and MIR127 loci and the MIR143 gene body served as an epigenetic signature of smooth muscle cells, whereas methylation of the MIR21 and MIR143HG promoters reflected immune cell infiltration. The strongest positive association was observed between methylation of cg01558401 in the MIR143HG promoter and the estimated immune cell proportion, with a Spearman correlation coefficient of 0.96. Two CpG sites in the MIR143 gene body showed similarly strong positive correlations with the estimated smooth muscle cell proportion.

Crucially, after adjustment for estimated vascular cell proportions in regression models, methylation at four CpG sites, two in the MIR100 E-box region and two in the MIR127 region, remained significantly associated with carotid artery state. This finding suggests that methylation differences at these sites are not fully accounted for by vascular cell composition and may therefore reflect atherosclerosis-associated functional changes in the carotid artery tissue itself. The team also quantified sequencing-depth-associated uncertainty using a Monte Carlo simulation, generating ten thousand methylation values per CpG site with a binomial model based on observed methylation proportion and coverage. The ninety-five percent Monte Carlo confidence intervals for regional methylation were generally below one percentage point, confirming that the larger observed differences between tissues were unlikely to be explained solely by coverage-derived measurement variability.

The associations with clinical characteristics proved equally compelling. Increased methylation at a single CpG site, chr11:122025143 on the GRCh37/hg19 reference genome, located in the MIR100 E-box region was associated with metabolic syndrome, with patients exhibiting the syndrome showing 92.2 percent methylation versus 89.1 percent in those without. Because this E-box is bound by the transcription factor ZEB1, which has been reported to enhance miR-100 expression, altered methylation at this site could potentially affect regulatory activity. Reduced miR-100 expression has previously been reported in macrophage-rich and lipid-rich atherosclerotic plaques and in metabolic conditions associated with obesity and diabetes, and experimental studies have suggested a protective role of miR-100 against high-fat diet-induced metabolic syndrome in mice. Within plaques, methylation levels of MIR10B, MIR21, and MIR127 were linked with indicators of histological instability and a history of acute cerebrovascular events. Hypomethylation of individual CpG sites within MIR10B and the MIR21 putative promoter was associated with thrombus formation on the plaque surface, intraplaque hemorrhage, and the presence of foam cells. Two CpG sites in the MIR127 region exhibited increased methylation in patients with a history of acute cerebrovascular events and were also inversely associated with fibrous cap thickness and necrotic core size.

Integration with paired microRNA expression data from the same plaques revealed the most highly expressed microRNAs as miR-143-3p, miR-21-5p, miR-100-5p, miR-145-5p, miR-145-3p, and miR-10b-5p. Eight significant inverse correlations were identified between methylation of seven CpG sites and expression of four microRNAs. Methylation of the promoter-associated cg27160395 site in MIR10B was inversely correlated with both miR-10b-5p and miR-10b-3p expression, consistent with previous reports linking hypermethylation at this site to reduced miR-10b expression. Within the MIR143HG locus, which overlaps the sequences encoding miR-145 and miR-143, CpG sites near the transcription start site were inversely associated with expression of miR-143-3p and miR-145-3p, whereas sites within the gene body showed positive associations. This spatially coordinated dual pattern represents a novel finding of the study. In peripheral blood, moderate hypomethylation was observed predominantly at individual CpG sites within the regulatory regions of MIR10B, MIR21, and the MIR100 E-box region in patients compared with controls. Blood methylation of MIR10B and MIR143 correlated with lipid metabolism parameters and the degree of carotid stenosis, while the MIR21 CpG island showed strong correspondence between blood and plaque methylation, with correlation coefficients exceeding 0.7 for four consecutive CpG sites. These findings suggest that peripheral blood leukocytes may serve as a surrogate material for indirectly assessing plaque methylation, and they highlight promising candidate loci and biomarkers for predicting the complicated course of atherosclerosis, warranting further validation in larger independent cohorts.

Subject of Research: Epigenetic regulation of microRNA genes in advanced carotid atherosclerosis

Article Title: Tissue-specific moderate DNA methylation alterations featuring hypomethylation in microRNA regulatory regions associated with advanced carotid atherosclerosis

Article References: Koroleva, I. A., Zarubin, A. A., Markov, A. V., Sleptcov, A. A., Kuznetsov, M. S., Kozlov, B. N., Muslimova, E. F., Afanasiev, S. A., Babushkina, N. P., Bragina, E. Y., Goncharova, I. A., Golubenko, M. V., Kucher, A. N., & Nazarenko, M. S. (2026). Tissue-specific moderate DNA methylation alterations featuring hypomethylation in microRNA regulatory regions associated with advanced carotid atherosclerosis. Frontiers in Cell and Developmental Biology, 14, Article 1948449. https://doi.org/10.3389/fcell.2026.1948449

Image Credits: AI Generated

DOI: 10.3389/fcell.2026.1948449

Keywords: DNA methylation, carotid atherosclerosis, microRNA, epigenetics, bisulfite sequencing, plaque instability, MIR10B, MIR21, MIR100, MIR127, MIR143, biomarkers