PLOS ONE RESEARCH ARTICLE Dynamic epigenetic age mosaicism in the human atherosclerotic artery Silvio Zaina ID1*, Manel Esteller2,3,4,5, Isabel Gonçalves6, Gertrud Lund7 1 Division of Health Sciences, Department of Medical Sciences, Leon Campus, University of Guanajuato, Leon, Mexico, 2 Josep Carreras Leukemia Research Institute, Badalona, Barcelona, Catalonia, Spain, 3 Centro de Investigación Biomédica en Red Cancer (CIBERONC), Madrid, Spain, 4 Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Catalonia, Spain, 5 Physiological Sciences Department, School of Medicine and Health Sciences, University of Barcelona (UB), Barcelona, Catalonia, Spain, 6 Skåne University Hospital, Clinical Sciences Malmö, Lund University, Malmö, Sweden, 7 Department of Genetic a1111111111 Engineering, CINVESTAV Irapuato Unit, Irapuato, Mexico a1111111111 a1111111111 * szaina@ugto.mx a1111111111 a1111111111 Abstract Accelerated epigenetic ageing, a promising marker of disease risk, has been detected in peripheral blood cells of atherosclerotic patients, but evidence in the vascular wall is lacking. OPEN ACCESS Understanding the trends of epigenetic ageing in the atheroma may provide insights into Citation: Zaina S, Esteller M, Gonçalves I, Lund G mechanisms of atherogenesis or identify targets for molecular therapy. We surveyed DNA (2022) Dynamic epigenetic age mosaicism in the human atherosclerotic artery. PLoS ONE 17(6): methylation age in two human artery samples: a set of donor-matched, paired atheroscle- e0269501. https://doi.org/10.1371/journal. rotic and healthy aortic portions, and a set of carotid artery atheromas. The well-character- pone.0269501 ized pan-tissue Horvath epigenetic clock was used, together with the Weidner whole-blood- Editor: Osman El-Maarri, University of Bonn, specific clock as validation. For the first time, we document dynamic DNA methylation age Institute of Experimental Hematology and mosaicism of the vascular wall that is atherosclerosis-related, switches from acceleration to Transfusion Medicine, GERMANY deceleration with chronological ageing, and is consistent in human aorta and carotid ather- Received: February 22, 2022 oma. At CpG level, the Horvath epigenetic clock showed modest differential methylation Accepted: May 20, 2022 between atherosclerotic and healthy aortic portions, weak association with atheroma histo- Published: June 3, 2022 logical grade and no clear evidence for participation in atherosclerosis-related cellular path- ways. Our data suggest caution when assigning a unidirectional DNA methylation age Peer Review History: PLOS recognizes the benefits of transparency in the peer review change to the atherosclerotic arterial wall. Also, the results support previous conclusions process; therefore, we enable the publication of that epigenetic ageing reflects non-disease-specific cellular alterations. all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0269501 Copyright: © 2022 Zaina et al. This is an open access article distributed under the terms of the Introduction Creative Commons Attribution License, which The concept that ageing is associated with causal changes in DNA, whether genetic or epige- permits unrestricted use, distribution, and reproduction in any medium, provided the original netic, is at least half-a-century old [1]. A consequential hypothesis is that epigenetic and chro- author and source are credited. nological age diverge in disease, thus providing a possible marker of disease risk. Microarray platforms that determine DNA methylation at CpG level with high reproducibility have pro- Data Availability Statement: All relevant data are within the manuscript, its Supporting Information vided opportunities to test that hypothesis in a variety of human conditions. In general, epige- files and the URL indicated in Materials and netic clocks calculate a predicted age as weighted average of a core set of CpG that are selected Methods. based on their robust correlation with chronological age of healthy blood or tissue sample. PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 1/8 PLOS ONE Epigenetic ageing of the human atheroma Funding: Supported by the following grants to ME: Epigenetic age acceleration or deceleration in a diseased tissue is determined by calculating the European Research Council (ERC) grant EPINORC differential between predicted and chronological age [2]. Notably, epigenetic clock predicted under agreement no. 268626, the MICINN Project: age with higher precision than telomere length in one study [3]. Accelerated epigenetic ageing SAF2011-22803, the Cellex Foundation, the Botin Foundation, the European Community’s Seventh has been detected in peripheral blood cells of atherosclerotic patients, but evidence in the vas- Framework Programme (FP7/2007–2013) from cular wall is limited even outside the atherosclerosis field [4–7]. To our knowledge, epigenetic grant HEALTH-F5-2011–282510–BLUEPRINT and ageing of the vascular wall was addressed by two studies, one that tested for associations with the Health and Science Departments of the haemostatic factors, and another that documented differential DNA methylation age (DNA- Generalitat de Catalunya. SZ was supported by a mAge) between arteries and veins [8, 9]. In the present study, we use a within-donor compari- CONACyT (Mexico) Sabbatical Fellowship no. 166058. ME is an Institució Catalana de Recerca i son design to contrast DNAmAge of atherosclerotic and adjacent normal vascular tissue. Estudis Avançats (ICREA) Research Professor. Strengths of that design include: inter-individual confounders (genetic, environmental, sto- There was no additional external funding received chastic variation) are reduced; it detects any intra-vascular bed mosaicism, a strong indication for this study. that epigenetic ageing is functionally associated with atherosclerosis and not just an epiphe- Competing interests: The authors have declared nomenon; it allows to dissect the role of systemic or local epigenetic ageing in atherogenesis. that no competing interests exist. For those reasons, the description of epigenetic ageing in the atheroma may be relevant, as it potentially provides insights into mechanisms of aberrant gene expression that predispose to or accompany the natural history of atherosclerosis, and may identify potential epigenomic targets for molecular therapy. We surveyed DNAmAge of two sets of human atherosclerotic and healthy arteries for which we previously obtained extensive DNA methylome data, by using the well-characterized, pan-tissue Horvath clock and Weidner whole-blood-specific clock [3, 10, 11]. The Horvath clock uses 353 CpG as core age predictor and has been shown to accurately calculate the age of a variety of healthy human tissues, and to detect DNAmAge deviations from chronological age (i.e. acceleration or deceleration) in a range of diseases. Genes that harbour the 353 Horvath CpG are enriched in cell growth, cell survival and regula- tion of development functions. The Weidner clock has been designed to compute epigenetic age of blood cells. The clock uses three CpG as core predictor, located in genes that encode proteins involved in cell-cell communication (integrin, alpha 2b), aminoacid metabolism (aspartoacylase) and cAMP hydrolysis (phosphodiesterase 4C). The Weidner clock is expected to loosely match the Horvath clock, but we included both in our analysis reasoning that an identical direction of change of DNAmAge in the two clocks would be interpreted as valida- tion of Horvath clock predictions. Materials and methods Donor-matched pairs of atherosclerotic and histologically normal portions of the aorta (A- aorta and N-aorta, respectively; n = 15 pairs) obtained post mortem, carotid atheromas obtained by endarterectomy from symptomatic and asymptomatic patients (A-car; n = 19 each) and corresponding Infinium HumanMethylation450 BeadChip (Illumina) data were previously described [10, 12]. The characteristics of participating patients are reported in S1 Table. Aortic and carotid artery samples were obtained in two distinct environments (Spain and Sweden, respectively). The observation that 98% of differentially methylated CpG coin- cided between aortic and carotid artery samples suggests that batch or geographic origin effects were minimal [10]. Similarly, any significant effect of post mortem time to collection was ruled out in the original study, strongly supporting previous conclusions that DNA methylation pro- files are stable after death and can be extrapolated to ex vivo samples [10, 13]. DNA methyla- tion array data were extensively validated in the original study [10]. The study conformed to the principles of the Declaration of Helsinki and patients or relatives in the case of post mortem samples gave written consent prior to their participation. The protocol was approved by the Bellvitge Hospital ethical committee (authorization no. PR311/11). DNA methylation microar- ray data are freely available at the Gene Expression Omnibus database (www.ncbi.nlm.nih. PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 2/8 PLOS ONE Epigenetic ageing of the human atheroma gov/geo/) with accession number GSE46401 [10]. To determine DNAmAge, Infinium HumanMethylation450 BeadChip data were formatted according to the calculator require- ments and uploaded to the freely available epigenetic age calculator maintained by Horvath’s laboratory at https://dnamage.genetics.ucla.edu/. The calculator yields predicted ages calcu- lated with Horvath or Weidner clock algorithms and adjusts for sex and cell type. Output files generated by the calculator have been deposited on the University of Guanajuato cloud and are available at the native (non-shortened) URL: https://ugtomx-my.sharepoint.com/:f:/g/ personal/szaina_ugto_mx/Eg-ccQD7c9dHjNekBqjRU1kBkg3zA895OzgJJWF24VZ1NA?e= W1l17j. T-test and one-way ANOVA were used for two group and multiple group compari- sons. Pearson’s r was used for correlation tests. Paired t-test of M-values was used to compare CpG methylation data [14]. Genome-wide significance was set at Bonferroni-corrected p<10−7 [10]. Statistical tests were performed with StatPlus Pro (AnalystSoft Inc.). Functional enrichment analysis of CpG-harbouring genes was performed with the DAVID tool (david. ncifcrf.gov) [15, 16]. Results and discussion Comparison of donor-matched aortas revealed a significant DNAmAge acceleration in A- aorta relative to N-aorta (average +4.3 years, p = 8.0x10-4) (Fig 1A). That differential DNA- mAge was above the performance-assessed Horvath clock accuracy (3.6 years) and ~2 years higher than reported in peripheral blood of recurrent stroke patients [11, 17]. Weidner clock also yielded A-aorta DNAmAge acceleration although ~4.5-fold higher (+16.6 years, p = 9.5x10-5). These discrepancies are in line with previous comparisons of performance among clocks [18]. A-aorta and A-aorta-to-N-aorta differential DNAmAge did not differ between histological grades, except borderline significance in Weidner clock (Horvath: p = 0.147 and p = 0.783, respectively; Weidner: p = 0.056 and p = 0.253, respectively). Unex- pectedly, N-aortas were on average younger relative to chronological age (Horvath: -4.9 years, p = 0.018; Weidner: -39.7 years, p = 1.69x10-9). Conversely, A-aortas DNAmAge matched chronological age (Horvath: -0.4 years, p = 0.804; Weidner: -1.1 years, p = 0.421). Also, the dif- ferential between either A-aorta or N-aorta DNAmAge and chronological age was strongly and inversely correlated with chronological age according to Horvath clock (r = -0.72 and r = -0.89, respectively, p<0.001), while marginally significant in A-aortas by Weidner clock (r = -0.48, p = 0.070, and r = -0.93, p = 10−6). That trend reflected the tendency for accelerated DNAmAge in chronologically younger aortas, followed by a deceleration at older chronologi- cal age (Fig 1A). This coincided with slow DNA methylation ageing across chronological age, as indicated by DNAmAge (y axis)/chronological age (x axis) slope <1 in N-aortas and A-aor- tas (0.46 and 0.57, respectively, in either Horvath or Weidner). Partially echoing our findings, peripheral blood DNAmAge is accelerated in young stroke patients but realigns with chrono- logical age in older patients [4]. We replicated the above analysis in the A-car set. A-car and A-aorta were comparable: con- sidering age-matched data points only (n = 14), A-car DNAmAge did not differ from A-aorta DNAmAge or from chronological age (Horvath: +2.4 years, p = 0.167 and +1.0 years, p = 0.661; Weidner: +2.5 years, p = 0.124 and +1.8 years, p = 0.110). The data confirms the high correlation previously observed between the two DNA methylation array data sets [10]. Conversely, A-car DNAmAge was accelerated relative to N-aorta (Horvath: +7.4 years, p = 2.8x10-4; Weidner: +24.7 years, p = 3.1x10-16). Mirroring the aorta data set, the differential between A-car DNAmAge and chronological age was inversely correlated with chronological age (Horvath: r = -0.75, p<10−6; Weidner: r = -0.56, p = 2.6x10-4). Also, a tendency to DNA- mAge acceleration followed by deceleration with chronological age was observed in A-car, PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 3/8 PLOS ONE Epigenetic ageing of the human atheroma Fig 1. Chronological and Horvath DNA methylation age in aortas and carotid atheromas. Samples are ordered by increasing chronological age (reported on the y-axis) and positioned at regular intervals on the x-dimension. A, donor- matched atherosclerotic and histologically normal portions of the aorta; B, carotid atheromas. Interpolating lines are not regression slopes, rather are intended to convey differences in ageing pace. Notice the tendency for accelerated and decelerated DNA methylation age at early and later ages, respectively. https://doi.org/10.1371/journal.pone.0269501.g001 concomitant with a slower DNA methylation ageing than chronological ageing (DNAmAge (y axis)/chronological age (x axis) slope <1: Horvath: 0.44; Weidner: 0.22) (Fig 1B). Further- more, the A-car data set allowed to explore associations of DNAmAge with asymptomatic/ symptomatic status and with symptom-to-endarterectomy time. DNAmAge was not different between symptomatic and asymptomatic patient atheromas (Horvath: p = 0.116; Weidner: p = 0.151). Likewise, no significant association was detected of DNAmAge with symptom-to- endarterectomy time (Horvath: r = -0.104; Weidner: r = 0.173; p>0.307 in either clock). PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 4/8 PLOS ONE Epigenetic ageing of the human atheroma Table 1. Horvath epigenetic clock CpG (n = 353) that are significantly different between atherosclerotic and healthy portions of the aorta. CpG ID A-N¶ Δβ p§ Chromosome: position‡ Gene symbol Relevance for atherosclerosis -8 cg01353448 0.13 6.23x10 2:74,609,742 HTRA2 Unknown cg01656216 0.16 4.11x10-8 9:85,761,834 C9orf64 Unknown cg14329157 0.05 1.80x10-9 16:23,597,303 PLK1 Overexpressed in atherosclerosis [19] -8 cg22809047 0.08 8.97x10 17:71,487,684 ACOX1 Involved in inflammation resolution [20] ¶ A and N, atherosclerotic and healthy portion of the aorta, respectively. § Comparison of M-values, paired t-test, Bonferroni-corrected (threshold: p = 10−7). ‡ Genome build 36. https://doi.org/10.1371/journal.pone.0269501.t001 In order to gain insights into possible mechanisms underlying the above observations, we assessed the methylation status of the 353 Horvath clock CpG in A-aorta and N-aorta samples [11]. Four CpG (or 1.13%) were differentially methylated between A-aorta and N-aorta at genome-wide significance (p<10−7, Bonferroni correction) (Table 1, S1 Fig). Since differences of methylation β were below 16% in all cases, those CpG did not pass differential methylation criteria in the original analysis of A-aorta and N-aorta [10]. Those four CpG map to proximal (cg01656216) or distal (cg01353448, cg14329157, cg22809047) promoters. Comparatively little or no published evidence links the four corresponding genes to atherosclerosis. HtrA serine peptidase 2 (HTRA2) and chromosome 9 open reading frame 64 (C9orf64) have not been implicated in atherogenesis. Limited evidence indicates that polo like kinase 1 (PLK1), encod- ing a stimulator of cell proliferation, and acyl-CoA oxidase 1 (ACOX1), encoding a player in fatty acid metabolism, are upregulated in peripheral blood cells of atherosclerotic patients and linked to resolution of inflammation, respectively [19, 20]. Furthermore, less than 10% of Hor- vath clock CpG were correlated with chronological age or histological grade (30 and 14, respectively) at nominal significance level (overall lowest significance p = 4.7x10-4), but none reached genome-wide significance (p<10−7;) (S2 Table, S1 Fig). None of those 44 CpG was differentially methylated between A-aorta and N-aorta. No significant functional enrichment (Gene_Ontology, Pathways, Protein_Interactions, Tissue_Expression) was observed for the corresponding genes in either case. The majority (~78.6%) of histological grade-associated Horvath clock CpG were also correlated with chronological age, reflecting the fact that chro- nological age tended to be lower in grade III atheromas than in grade VII counterparts (p = 0.055; ANOVA, Fisher LSD post hoc). Taken together, the analysis of Horvath clock CpG confirmed previous conclusions that epigenetic ageing reflects alterations in fundamental, rather than disease-specific, cellular epigenetic maintenance systems (the concept of EMS dis- cussed in [11]). A few studies support that notion. Peripheral blood cell profiling of six cohorts spanning a wide range of chronological age and pathological conditions, revealed that a minority of Horvath clock CpG correlate with expression of genes in cis that are not signifi- cantly enriched in any Gene Ontology term [21]. Epigenetic ageing was weakly associated with classical cardiovascular risk factors in the large MESA and HRS cohorts [22]. Furthermore, no association of epigenetic age with development of cardiovascular risk factors was observed in a cohort of 6–10 years old children [23]. The results of the latter two studies agree with our observation that epigenetic ageing was not associated with asymptomatic/symptomatic status in the carotid atheroma. As caveat, Horvath clock CpG are associated with gene expression in trans in loci that regulate T cell function and may therefore be relevant for atherosclerosis [21, 24, 25]. As the atheroma discontinuously expands by bursts in macrophage infiltration and cellular proliferation, it is possible that the epigenetic clock is a readout of mitotic age in ath- erosclerosis [26]. Yet, that hypothesis runs counter to the fact that the Horvath epigenetic PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 5/8 PLOS ONE Epigenetic ageing of the human atheroma clock was previously found to be independent of mitotic age and was not associated with histo- logical grade—a proxy of the proliferative history of the atheroma—in this study. In the light of the mentioned evidence, epigenetic ageing may be the consequence of poorly understood cellular mechanisms [27], or may simply be a useful epiphenomenon for marker identification purposes. In conclusion, we document for the first time dynamic epigenetic age mosaicism of the vas- cular wall that is atherosclerosis-related, switches from acceleration to deceleration with chro- nological ageing, and is consistent in human aorta and carotid atheromas. Our data suggest caution when assigning unidirectional DNAmAge change to the atherosclerotic artery. Addi- tionally, Horvath clock CpG methylation profiles reinforce the notion that epigenetic ageing reflects poorly understood, non-disease-specific cellular events. Supporting information S1 Fig. Horvath clock CpG that were differentially methylated between atherosclerotic and healthy aortic portions, or associated with chronological age or atheroma histological grade. Differential methylation was significant at genome-wide level (p<10−7, Bonferroni cor- rection; see Table 1 for details). Other associations were significant at nominal significance level (p<0.05; see S1 Table for details). (PPTX) S1 Table. Patient information. (DOCX) S2 Table. CpG associated with chronological age or histological grade. (XLSX) Author Contributions Conceptualization: Silvio Zaina, Gertrud Lund. Funding acquisition: Manel Esteller. Writing – original draft: Silvio Zaina. Writing – review & editing: Manel Esteller, Isabel Gonçalves, Gertrud Lund. References 1. Hayflick L. Current theories of biological aging. Fed. Proc. 1975, 34, 9–13. PMID: 1088947 2. Bell C.G.; Lowe R.; Adams P.D.; Baccarelli A.A.; Beck S.; Bell J.T.; et al. DNA methylation aging clocks: challenges and recommendations. Genome Biol. 2019, 20, 249, https://doi.org/10.1186/s13059-019- 1824-y PMID: 31767039 3. Weidner C.I.; Lin Q.; Koch C.M.; Eisele L.; Beier F.; Ziegler P.; et al. Aging of blood can be tracked by DNA methylation changes at just three CpG sites. Genome Biol. 2014, 15, R24, https://doi.org/10. 1186/gb-2014-15-2-r24 PMID: 24490752 4. Soriano-Tárraga C.; Giralt-Steinhauer E.; Mola-Caminal M.; Vivanco-Hidalgo R.M.; Ois A.; Rodrı́guez- Campello A.; et al. Ischemic stroke patients are biologically older than their chronological age. Aging (Albany. NY). 2016, 8, 2655–2666, https://doi.org/10.18632/aging.101028 PMID: 27922817 5. Lind L.; Ingelsson E.; Sundström J.; Siegbahn A.; Lampa E. Methylation-based estimated biological age and cardiovascular disease. Eur. J. Clin. Invest. 2017, 48, e12872, https://doi.org/10.1111/eci. 12872 PMID: 29231988 6. Roetker N.S.; Pankow J.S.; Bressler J.; Morrison A.C.; Boerwinkle E. Prospective Study of Epigenetic Age Acceleration and Incidence of Cardiovascular Disease Outcomes in the ARIC Study (Atherosclero- sis Risk in Communities). Circ. Genomic Precis. Med. 2018, 11, e001937, https://doi.org/10.1161/ CIRCGEN.117.001937 PMID: 29555670 PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 6/8 PLOS ONE Epigenetic ageing of the human atheroma 7. Joyce B.; Gao T.; Zheng Y.; Ma J.; Hwang S.-J.; Liu L.; et al. Epigenetic Age Acceleration Reflects Long-Term Cardiovascular Health. Circ. Res. 2021, https://doi.org/10.1161/CIRCRESAHA.121.318965 PMID: 34428927 8. Ward-Caviness C.K.; Huffman J.E.; Everett K.; Germain M.; van Dongen J.; Hill W.D.; et al. DNA meth- ylation age is associated with an altered hemostatic profile in a multiethnic meta-analysis. Blood 2018, 132, 1842–1850, https://doi.org/10.1182/blood-2018-02-831347 PMID: 30042098 9. Nazarenko M.S.; Markov A. V; Lebedev I.N.; Freidin M.B.; Sleptcov A.A.; Koroleva I.A.; et al. A compar- ison of genome-wide DNA methylation patterns between different vascular tissues from patients with coronary heart disease. PLoS One 2015, 10, e0122601, https://doi.org/10.1371/journal.pone.0122601 PMID: 25856389 10. Zaina S.; Heyn H.; Carmona F.J.; Varol N.; Sayols S.; Condom E.; et al. DNA Methylation Map of Human Atherosclerosis. Circ. Cardiovasc. Genet. 2014, 7, 692–700, https://doi.org/10.1161/ CIRCGENETICS.113.000441 PMID: 25091541 11. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013, 14, R115, https://doi.org/10.1186/gb-2013-14-10-r115 PMID: 24138928 12. Valencia-Morales M. del P.; Zaina S.; Heyn H.; Carmona F.J.; Varol N.; Sayols S.; et al. The DNA meth- ylation drift of the atherosclerotic aorta increases with lesion progression. BMC Med. Genomics 2015, 8, 7, https://doi.org/10.1186/s12920-015-0085-1 PMID: 25881171 13. Barrachina M.; Ferrer I. DNA methylation of Alzheimer disease and tauopathy-related genes in post- mortem brain. J. Neuropathol. Exp. Neurol. 2009, 68, 880–891, https://doi.org/10.1097/NEN. 0b013e3181af2e46 PMID: 19606065 14. Du P.; Zhang X.; Huang C.-C.; Jafari N.; Kibbe W.A.; Hou L.; et al. Comparison of Beta-value and M- value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics 2010, 11, 587, https://doi.org/10.1186/1471-2105-11-587 PMID: 21118553 15. Huang D.W.; Sherman B.T.; Lempicki R.A. Bioinformatics enrichment tools: paths toward the compre- hensive functional analysis of large gene lists. Nucleic Acids Res. 2009, 37, 1–13, https://doi.org/10. 1093/nar/gkn923 PMID: 19033363 16. Huang D.W.; Sherman B.T.; Lempicki R.A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 2009, 4, 44–57, https://doi.org/10.1038/nprot.2008.211 PMID: 19131956 17. Soriano-Tárraga C.; Lazcano U.; Jiménez-Conde J.; Ois A.; Cuadrado-Godia E.; Giralt-Steinhauer E.; et al. Biological age is a novel biomarker to predict stroke recurrence. J. Neurol. 2021, 268, 285–292, https://doi.org/10.1007/s00415-020-10148-3 PMID: 32789606 18. Daunay A.; Baudrin L.G.; Deleuze J.-F.; How-Kit A. Evaluation of six blood-based age prediction mod- els using DNA methylation analysis by pyrosequencing. Sci. Rep. 2019, 9, 8862, https://doi.org/10. 1038/s41598-019-45197-w PMID: 31222117 19. Du N.; Li M.; Yang D. Hsa_circRNA_102541 regulates the development of atherosclerosis by targeting miR-296-5p/PLK1 pathway. Ir. J. Med. Sci. 2021, https://doi.org/10.1007/s11845-021-02708-x PMID: 34251586 20. Vamecq J.; Andreoletti P.; El Kebbaj R.; Saih F.E.; Latruffe N.; El Kebbaj M.H.S.; et al. Peroxisomal Acyl-CoA Oxidase Type 1: Anti-Inflammatory and Anti-Aging Properties with a Special Emphasis on Studies with LPS and Argan Oil as a Model Transposable to Aging. Oxid. Med. Cell. Longev. 2018, 2018, https://doi.org/10.1155/2018/6986984 PMID: 29765501 21. Jonkman T.H.; Dekkers K.F.; Slieker R.C.; Grant C.D.; Ikram M.A.; van Greevenbroek M.M.J.; et al. Functional genomics analysis identifies T and NK cell activation as a driver of epigenetic clock progres- sion. Genome Biol. 2022, 23, 24, https://doi.org/10.1186/s13059-021-02585-8 PMID: 35031073 22. Schmitz L.; Zhao W.; Ratliff S.; Goodwin J.; Miao J.; Lu Q.; et al. The Socioeconomic Gradient in Epige- netic Ageing Clocks: Evidence from the Multi-Ethnic Study of Atherosclerosis and the Health and Retire- ment Study. Epigenetics 2021, 1–23, https://doi.org/10.1080/15592294.2021.1939479 PMID: 34227900 23. Monasso G.S.; Jaddoe V.W. V; Küpers L.K.; Felix J.F. Epigenetic age acceleration and cardiovascular outcomes in school-age children: The Generation R Study. Clin. Epigenetics 2021, 13, 205, https://doi. org/10.1186/s13148-021-01193-4 PMID: 34784966 24. Jonasson L.; Holm J.; Skalli O.; Bondjers G.; Hansson G.K. Regional accumulations of T cells, macro- phages, and smooth muscle cells in the human atherosclerotic plaque. Arteriosclerosis 1986, 6, 131– 138, https://doi.org/10.1161/01.atv.6.2.131 PMID: 2937395 25. Depuydt M.A.; Prange K.H.; Slenders L.; Örd T.; Elbersen D.; Boltjes A.; et al. Microanatomy of the Human Atherosclerotic Plaque by Single-Cell Transcriptomics. Circ. Res. 2020, CIRCRE- SAHA.120.316770, https://doi.org/10.1161/CIRCRESAHA.120.316770 PMID: 32981416 PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 7/8 PLOS ONE Epigenetic ageing of the human atheroma 26. Libby P. Inflammation during the life cycle of the atherosclerotic plaque. Cardiovasc. Res. 2021, 117, 2525–2536, https://doi.org/10.1093/cvr/cvab303 PMID: 34550337 27. Gibson J.; Russ T.C.; Clarke T.-K.; Howard D.M.; Hillary R.F.; Evans K.L.; et al. A meta-analysis of genome-wide association studies of epigenetic age acceleration. PLOS Genet. 2019, 15, e1008104, https://doi.org/10.1371/journal.pgen.1008104 PMID: 31738745 PLOS ONE | https://doi.org/10.1371/journal.pone.0269501 June 3, 2022 8/8