ABSTRACT
Globally, ischaemic heart disease is a major contributor to premature morbidity and mortality. A significant number of young Myocardial Infarction (MI) patients (aged <55 y) have subsequent cardiac events within a year of their index event. This study used Next Generation Sequencing (NGS) methylation to understand the pathogenesis in this subset of young MI patients, comparing them to a cohort of patients without recurrent events. Cases and controls were matched for age, gender, ethnicity, and comorbidities. Differential methylation analyses were performed on Reduced Representation Bisulphite Sequencing (RRBS) data. Across the group and within case–control pairs’ variation were analysed. Pairwise comparisons across each matched case–control pair resulted in a list of genes that were consistently significantly differentially methylated between all 16 matched pairs. This gene list was input into pathway analysis databases. Of particular relevance to cardiac pathology the following pathways were identified as over-represented in the patients with recurrent events; cell adhesion, transcription regulation and cardiac electrical conduction, specifically relating to calcium channel activity. This study looked at methylation differences between two populations of young MI patients. There were significantly different methylation profiles between the two groups studied; key pathways were identified as specifically affected in the patients with recurrent cardiac events. Matched pairwise comparisons and detailed interpretations of DNA methylation data may help to elucidate complex pathogeneses within and between clinical subtypes. Further analysis will determine whether these epigenomic differences can be useful as predictive biomarkers of clinical progression.
KEYWORDS: Methylation, next-generation sequencing, RRBS, cardiac, myocardial infarction, young, biomarkers
Introduction
Globally, ischaemic heart disease is a major contributor to premature morbidity and mortality and is associated with significant economic burden [1,2]. A significant proportion of new Myocardial Infarction (MI) patients present as young patients, typically under the age of 55 [3]. A recent study [4] described the clinical characteristics of 1199 patients on the Wellington Acute Coronary Syndrome Registry. In this cohort 12.8% (154/1199) of the patients presented with MI as young patients. Of this group, 36% had none or only one traditional risk factor for MI prior to presenting with their index coronary event and so would have been classified as ‘low cardiovascular risk’. Ten percent of MI patients typically have further events (recurrent myocardial infarction, stroke or death) in the year following the index event [5]. The identification of biomarkers that can assist in the identification and management of young MI patients at increased risk of poor outcomes is therefore needed.
Epigenomic biomarkers are increasingly used in medical diagnostics [6]. Unlike DNA sequences, largely the same in every cell, epigenomic changes can occur because of dietary, behavioural and other environmental exposures, including physical and psychological stressors [7,8]. Abnormalities in DNA methylation are associated with many diseases and typically manifest through inappropriate gene expression [9]. Through comparative analysis of methylome data, we can attain a clearer understanding of the molecular mechanisms that underlie pathology in different population subsets. Epigenomic profiling can provide informative biomarkers with potential for epigenomic reversibility, an attractive target for clinical intervention [10].
Rask-Andersen et al. undertook an epigenome-wide study of MI using an Illumina Infinium HumanMethylation450 BeadChip. They found that individuals with a history of MI have differential methylation in many genomic loci, highlighting potential genes that could be important in contributing to MI pathology [11]. A recent study by Li et al. [12] carried out genome-wide DNA methylation analysis using Reduced Representation Bisulfite Sequencing (RRBS) from DNA extracted from blood leukocytes in a group of patients with cardiac failure. Their study identified three differentially methylated genes in the patient group compared to age and gender-matched controls providing support that an approach investigating DNA extracted from blood leukocytes is a valid method to explore the association between heart failure and DNA methylation. RRBS combines bisulfite conversion and next-generation sequencing (NGS) to give a cost-effective alternative solution to whole genome methylation sequencing [13]. RRBS covers approximately 2.5% of the human genome, but it is highly enriched for promoter-specific CG-enriched areas [14].
Epigenetic mechanisms are likely to be important drivers of cardiac wound healing, with cardiac remodelling crucial in the response and outcome to MI. Understanding the subtle epigenetic differences between MI patients with and without recurrent MI can help to understand the pathogenesis and prognosis of such patients. MI induces myocyte death scarring and inflammation. Understanding how cardiac cells adapt post MI trauma is crucial in understanding the pathogenesis of MI. With NGS-based methylation analysis now readily accessible, this technology can be exploited to aid in understanding subtle variations in pathogenesis contributing to different clinical subtypes. The aim of our study was to identify epigenomic biomarkers that are distinctive to MI patients with recurrent events to understand the pathogenesis in this subset of patients to potentially aid as predictors of disease risk to assist in their management.
Methods
Participants
Sixteen patients with premature MI who experienced recurrent ischaemic events (11 patients went on to have non-fatal MI, 4 had ischemic stroke, and 1 cardiac death) within a year of their initial MI were identified from the Wellington ACS registry. Patients were matched for age, gender, diabetic status and smoking status with 16 controls that had no adverse events within the 1-year period of their index ischaemic event. Demographic details are summarized in Table 1.
Table 1.
Demographics of cohort studied.
| Cases (recurrent ischaemic events) n = 16 | Controls (single ischaemic event) n = 16 | |
|---|---|---|
| Age | 49.88 ± 3.94 | 50.06 ± 4.00 |
| Male | 14 (88%) | 14 (88%) |
| Female | 2 (12%) | 2 (12%) |
| Ethnicity | ||
| European | 9 | 9 |
| Maori | 6 | 2 |
| Other | 1 | 5 |
| Diabetes | 2 | 2 |
| Current smoker | 7 |
RRBS analysis
Peripheral blood samples were obtained from all patients upon admission at their index MI. The study was reviewed and approved by the Central Regional Ethics Committee (URA/11/05/2016). Patient consent was voluntary and included the use of samples for the identification of biomarkers. DNA was extracted from the blood samples, assessed for quantity and purity and 20 µL aliquots were used for RRBS library preparation and sequencing. Custom Science Sequencing Service was contracted to prepare and sequence the 32 RRBS bisulphite treated libraries. An Illumina HiSeq sequencer was used to generate 150 bp paired-end reads. From the paired-end reads only the R1 reads were used to minimize overlap that would bias CpG mapping. Sequence files were quality (phred 33) and adaptor trimmed to 100 bp to maximize the number of reads that map to the genome. After trimming, all sequence files were checked with fastqc to ensure adaptors had been removed and reads were of sufficient quality. For the diffmeth DMAP step, the criteria set were 2 CpGs in each fragment needed to qualify for the criteria of 10 or more hits on all fragments. Additionally valid fragments were further restricted to needing to appear in the majority of samples, 9/11 was the proportion used, this was the parameter used within the software.
Results
Each of the 32 data files comprised on average approximately 45 million trimmed sequence reads, with each file approximately 7 GB in size. Each file was then mapped to the human genome (GRCh38) using bismark to first build bisulfite genome libraries and then to map the sequence files in order to produce bam files for subsequent analyses with DMAP [15]. Mapping efficiency averaged at 64% for the 32 samples, which is in line with other research using bismark to map bisulfite sequence data. Numbers of cytosines analysed, as well as the percentage of which were analysed in the CpG context, were also produced for each data file. For each of the 32 samples, on average 661 million cytosines were analysed, with approximately 58% of these methylated in the CpG context, 1% in the CHG context, and 0.6% in the CHH context, these data are summarized in Table 2.
Table 2.
Genome-wide methylation analysis from DNA extracted from peripheral blood.
| Number of reads | C’s analysed | %meth CpG | %meth CHG | %meth CHH | Bismark mapping efficiency | |
|---|---|---|---|---|---|---|
| Mean for 32 samples analysed | 45,346,487.25 | 661,929,241.3 | 57.66875 | 1.08125 | 0.6 | 64.19375 |
| Cases | 44,187,373 | 6.6E+08 | 58.08125 | 1.0625 | 0.60625 | 64.21875 |
| Control | 46,505,601 | 6.64E+08 | 57.25625 | 1.1 | 0.59375 | 64.16875 |
Differential methylation analysis of case v control groups
In total, 403,045 loci were identified through ANOVA as differentially methylated between the two groups of patients, a plot of the distribution of p-values show that the data conform to a uniform distribution. (Figure 1), Regions were ranked on p-value then a Benjamin Hochberg correction for FDR (q) was applied to the results with an FDR of 10%. This resulted in a list of 260 loci that were identified as significantly different, with an FDR corrected value (q) of ≤ 0.01 and with at least 14/16 cases contributing to the ANOVA statistic (Supplementary Table 1). There were 122 loci had greater methylation in the control group and 138 in the case group. The top 20 (highest q value) are displayed in Table 3. Of the 20 loci, nine had significantly increased methylation in the group with recurrent events (case) and 11 loci in the control group. Within these, there are candidate gene regions identified, that are of potential interest to cardiac pathology. DGAT1 encodes a protein that functions as a key metabolic enzyme and is potentially associated with obesity and other metabolic diseases. OSBPL5 encodes a protein believed to play a key role in the maintenance of cholesterol balance in the body. UBP1 has previously identified as an important contributing factor to blood pressure [16]. Repression of stress response gene Hsph1 was identified in a murine study investigating sustained ligand-activated preconditioning in pre- and post-ischaemic myocardium [17]. HDAC11 is a histone deacetylase and is a regulatory molecule in Th2 response and plays a critical role in the restriction of the biased IL-13 expression in CD4 + T cells of the heart in patients with myocarditis [18]. EIFSL has recently been identified as one of 21 newly identified genes to be significantly associated with early onset coronary artery disease in a large exome wide association study on a Japanese population of 7256 individuals [19]. In murine studies OTOP1 has been shown to be protective for obesity-induced metabolic dysfunction, another study has identified the protein crucial to forming proton-selective ion channel and important in pH regulation [20]. MYLBL2 (along with FOXO3) has been identified as one of two transcription factors involved in the regulations of key differentially expressed genes in patients with acute myocardial infarction [21]. Sox18 has been identified as important in cardiomyocyte differentiation in murine studies investigating cardiomyogenesis [22]
Figure 1.
Plot of p-value distribution.
Table 3.
Top 20 differentially methylated genomic regions identified between case and control groups.
| Chr. position | Gene ID | Gene name/function | > Meth | p-Value | Q value | Fold diff | Genic region | CpG island region |
|---|---|---|---|---|---|---|---|---|
| Chr3. 33,439,984 | UBP1 | Upstream Binding Protein 1/transcriptional activator, angiogenesis, blood pressure | Case | 9.46E-06 | 0.0002 | 2.45 | on_exon | InterCpG |
| Chr13. 31,183,198 | HSPH1 | Heat Shock Protein Family H (Hsp110) Member 1/inhibits aggregation of misfolded proteins, stress response | Case | 3.52E-05 | 0.0004 | 1.05 | 3’UTR | InterCpG |
| Chr3. 13,290,659 | HDAC11 | Histone Deacetylase 11 | Case | 3.7E-05 | 0.0004 | 1.03 | 5ʹUTR | InterCpG |
| Chr19. 55,359,429 | FAM71E2 | Family with Sequence Similarity 71 Member E2 | Case | 4.02E-05 | 0.0004 | 1.08 | on_exon | InterCpG |
| Chr3. 100,417,592 | LNP1 | Leukaemia NUP98 Fusion Partner 1 | Case | 0.0001 | 0.0006 | 1.03 | on_intron | InterCpG |
| Chr14. 32,115,522 | ARHGAP5 | Rho GTPase Activating Protein 5/endothelial cytoskeleton and permeability | Case | 0.0001 | 0.0008 | 1.04 | on_intron | InterCpG |
| Chr3. 87,375,514 | POU1F1 | POU Class 1 Homeobox 1/transcription factor, growth and development | Case | 0.0001 | 0.0008 | 1.04 | 3’UTR | InterCpG |
| Chr9. 80,344,759 | SPATA31D4 | SPATA31 Subfamily D Member 4/may play a role in spermatogenesis. | Case | 0.0002 | 0.00010 | 1.06 | 5’UTR | InterCpG |
| Chr7. 90,497,286 | CLDN12 | Claudin 12/tight junction protein | Case | 0.0002 | 0.00010 | 1.02 | on_intron | InterCpG |
| Chr11. 3,204,666 | OSBPL5 | Oxysterol Binding Protein Like 5 | Control | 8.58E-06 | 0.0001 | 1.19 | 3ʹUTR | CpGI_shore |
| Chr8. 144,327,288 | DGAT1 | Diacylglycerol O-Acyltransferase 1 | Control | 9.96E-06 | 0.0002 | 1.65 | 3ʹUTR | CpGI_core |
| Chr22. 37,828,716 | EIF3L | Eukaryotic Translation Initiation Factor 3 Subunit L | Control | 1.13E-05 | 0.0002 | 1.07 | 5’UTR | InterCpG |
| Chr14. 99,831,302 | EML1 | Echinoderm Microtubule Associated Protein Like 1/recognition of histone marks | Control | 1.17E-05 | 0.0002 | 1.05 | on_intron | InterCpG |
| Chr4. 4,227,293 | OTOP1 | Otopetrin 1 | Control | 1.24E-05 | 0.0003 | 1.28 | 3’UTR | CpGI_core |
| Chr3. 51,713,043 | GRM2 | Glutamate Metabotropic Receptor 2 | Control | 1.92E-05 | 0.0003 | 1.04 | on_exon | InterCpG |
| Chr20. 43,655,560 | MYBL2 | MYB Proto-Oncogene Like 2 | Control | 0.0001 | 0.0004 | 1.03 | 5ʹUTR | InterCpG |
| Chr20. 64,057,589 | SOX18 | SRY-Box 18/cardiomyogenesis, endothelial angiogenic function | Control | 0.0001 | 0.0008 | 1.58 | 3’UTR | InterCpG |
| Chr11. 61,752,535 | MYRF | Myelin Regulatory Factor | Control | 0.0001 | 0.0009 | 1.48 | 5ʹUTR | CpGI_core |
| Chr20. 62,134,895 | LSM14B | LSM Family Member 14B | Control | 0.0001 | 0.0009 | 1.11 | on_intron | InterCpG |
| Chr9. 72,060,140 | C9orf57 | Chromosome 9 Open Reading Frame 57 | Control | 0.0002 | 0.0009 | 1.02 | on_intron | InterCpG |
Differential methylation analysis of matched pairwise comparisons
The above results consider comparative variance between the two groups. As the group size was small and heterogeneous we also studied the data through pairwise comparison of each carefully matched patient pair using Fishers Exact test. Patients were matched with respect to age, gender, comorbidity with diabetes and current smoking status. Pairwise comparison using FE test comprised 16 pairwise comparisons, the mean number of comparisons that resulted in differential methylation per pair was 184,184.5 (sd 12,831.6), across the whole genome (all chromosomes) this included multiple hits within the same gene location. The data resulting from each pairwise comparison were filtered to only include the lowest probability < 0.0000003. The data were then further analysed to identify genes significantly differentially methylated in common for the entire set of 16 pairwise comparisons. This resulted in a list of 764 gene ids; these were run in Reactome [23] and GOrilla [24] gene ontology enrichment analyses to see if particular pathways were over-represented. Cell adhesion, calcium channel and ion binding as well as transcriptional regulation appeared to be critical and over-represented in both pathway analyses, the output from GOrilla is provided in Table 4.
Table 4.
GOrilla output of overrepresented pathways involving the 764 commonly differentially methylated gene regions in all 16 pairwise comparisons.
| GO term | Description | |
|---|---|---|
| Component | GO:0016342 | Catenin complex |
| Process | GO:0007156 | Homophilic cell adhesion via plasma membrane adhesion molecules |
| GO:0048856 | Anatomical structure development | |
| GO:0098742 | Cell-cell adhesion via plasma-membrane adhesion molecules | |
| GO:0032502 | Developmental process | |
| GO:0098609 | Cell-cell adhesion | |
| GO:0001708 | Cell fate specification | |
| GO:0044057 | Regulation of system process | |
| GO:0048731 | System development | |
| GO:0016339 | Calcium-dependent cell-cell adhesion via plasma membrane cell adhesion molecules | |
| GO:0051148 | Negative regulation of muscle cell differentiation | |
| GO:0042462 | Eye photoreceptor cell development | |
| GO:0048468 | Cell development | |
| GO:0032501 | Multicellular organismal process | |
| GO:0048869 | Cellular developmental process | |
| GO:0023052 | Signaling | |
| GO:0042461 | Photoreceptor cell development | |
| GO:0048666 | Neuron development | |
| GO:0014807 | Regulation of somitogenesis | |
| GO:0021527 | Spinal cord association neuron differentiation | |
| GO:0030534 | Adult behavior | |
| GO:0045944 | Positive regulation of transcription by RNA polymerase II | |
| GO:0022610 | Biological adhesion | |
| GO:0007405 | Neuroblast proliferation | |
| GO:0035137 | Hindlimb morphogenesis | |
| GO:0007155 | Cell adhesion | |
| Function | GO:0000981 | DNA-binding transcription factor activity, RNA polymerase II-specific |
| GO:0003700 | DNA-binding transcription factor activity | |
| GO:0043565 | Sequence-specific DNA binding | |
| GO:0140110 | ranscription regulator activity | |
| GO:0001228 | DNA-binding transcription activator activity, RNA polymerase II-specific | |
| GO:1,990,837 | Sequence-specific double-stranded DNA binding | |
| GO:0044212 | Transcription regulatory region DNA binding | |
| GO:0001067 | Regulatory region nucleic acid binding | |
| GO:0000976 | Transcription regulatory region sequence-specific DNA binding | |
| GO:0022843 | Voltage-gated cation channel activity | |
| GO:0000977 | RNA polymerase II regulatory region sequence-specific DNA binding | |
| GO:0001012 | RNA polymerase II regulatory region DNA binding | |
| GO:0003690 | Double-stranded DNA binding | |
| GO:0001221 | Transcription cofactor binding | |
| GO:0000977 | RNA polymerase II regulatory region sequence-specific DNA binding | |
| GO:0001012 | RNA polymerase II regulatory region DNA binding | |
| GO:0003690 | Double-stranded DNA binding | |
| GO:0001221 | Transcription cofactor binding | |
| GO:0099604 | Ligand-gated calcium channel activity | |
| GO:0015278 | Calcium-release channel activity | |
| GO:0005217 | Intracellular ligand-gated ion channel activity | |
| GO:0005251 | Delayed rectifier potassium channel activity | |
| GO:0005509 | Calcium ion binding | |
| GO:0005262 | Calcium channel activity |
For each pairwise comparison, the significantly differentially methylated regions were sorted by significance and by greatest fold difference, excluding the X chromosome, to produce 10 top differentially methylated hits for each pairwise comparison (Table 5). From the 16 pairwise comparisons, genes that were identified as being greater methylated only in the cases and in more than one pairwise comparisons included SEPT9, EIF2AK3, MADCAM1, PKP3, RGPD6, SPAG1, and ZNF664. Genes that were identified as being greater methylated in the controls, in more than one pairwise comparison, and not in the cases, were ALG1L, ARL4D, DLGAP2, RYR1, TLE1, and ZNF518A.
Table 5.
Top 10 hits for each pairwise comparison based on fold difference of methylation.
| Pair 1 Age 45,46 M,M Cau, Cau |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 9 Age 54, 54 M,M Mao, Cau |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 3 | 13 | Control | 30.89 | CNTN3 | 6 | 12 | Control | 75.88 | FOXO3 | ||
| 21 | 5 | Control | 16.76 | SLC19A1 | 19 | 10 | Case | 67.04 | HOOK2 | ||
| 17 | 8 | Case | 16.75 | NLRP1 | 6 | 14 | Control | 50.55 | DUSP22 | ||
| 21 | 23 | Case | 9.56 | KCNE1B | 6 | 14 | Control | 45.42 | DUSP22 | ||
| 8 | 9 | Control | 8.52 | DLGAP2 | 19 | 9 | Case | 25.99 | MADCAM1 | ||
| 3 | 5 | Control | 8 | IQCG | 18 | 19 | Case | 17.43 | SIGLEC15 | ||
| Y | 7 | Control | 7.87 | VCY | 16 | 14 | Case | 16.77 | ZFPM1 | ||
| 20 | 6 | Control | 7.7 | PROKR2 | 6 | 14 | Control | 12.86 | DUSP22 | ||
| 17 | 5 | Control | 7.63 | MFSD6L | 16 | 10 | Case | 12.19 | ZFPM1 | ||
| 2 | 14 | Case | 7.59 | CD8B | 16 | 11 | Case | 11.41 | ZFPM1 | ||
| 4 | 8 | Case | 7.24 | TAPT1 | 16 | 13 | Case | 10.14 | ZFPM1 | ||
| Pair 2 Age 48, 48 M,M Cau, PI |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 10 Age 53,53 M,M Cau, Chi |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 8 | 17 | Control | - | LAPTM4B | 7 | 11 | Case | 24.45 | COX19 | ||
| 2 | 27 | Control | 288.26 | RABL2A | 6 | 12 | Case | 18.53 | FOXO3 | ||
| 10 | 19 | Case | 80.49 | BMPR1A | 20 | 10 | Control | 15.59 | PPP1R3D | ||
| 6 | 14 | Control | 61.49 | DUSP22 | 6 | 7 | Control | 11.88 | SLC35B3 | ||
| 2 | 12 | Case | 51.1 | RGPD8 | 8 | 8 | Case | 11.78 | C8orf22 | ||
| 6 | 14 | Control | 47.64 | DUSP22 | 16 | 10 | Control | 11.25 | ZFPM1 | ||
| 7 | 10 | Control | 47.18 | ARMC10 | 2 | 14 | Case | 10.99 | CD8B | ||
| 2 | 11 | Case | 44.3 | RGPD8 | 5 | 7 | Control | 10.62 | IRGM | ||
| 6 | 12 | Case | 40.47 | FOXO3 | 2 | 7 | Case | 10.5 | SNTG2 | ||
| 2 | 13 | Case | 37.72 | RGPD8 | 8 | 9 | Control | 9.49 | AC138696.1 | ||
| 6 | 14 | Control | 24.35 | DUSP22 | 8 | 8 | Control | 9.1 | ERICH1 | ||
| Pair 3 Age 53, 53 M,M Cau, PI Sm, Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 11 Age 51, 51 F,F Mao, Mao Sm,Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 2 | 16 | Control | 71.39 | CD8B | 8 | 15 | Case | 42.13 | SPAG1 | ||
| 2 | 14 | Control | 61.35 | CD8B | 8 | 13 | Case | 36.07 | SPAG1 | ||
| 12 | 8 | Case | 30.4 | ZNF664 | |||||||
| 8 | 15 | Case | 31.54 | SPAG1 | |||||||
| 5 | 7 | Control | 12.3 | IRGM | 6 | 12 | Control | 22.96 | FOXO3 | ||
| 2 | 27 | Control | 9.12 | RABL2A | 8 | 16 | Control | 11.63 | NAPRT | ||
| 11 | 8 | Case | 8.65 | PKP3 | 2 | 7 | Case | 7.04 | EIF2AK3 | ||
| 2 | 7 | Control | 7.66 | ANO7 | 9 | 10 | Control | 7.02 | TLE1 | ||
| 10 | 8 | Case | 7.17 | ADGRA1 | 2 | 15 | Case | 6.9 | EIF2AK3 | ||
| 18 | 8 | Control | 7.07 | ENOSF1 | 10 | 12 | Case | 6.62 | MAP3K8 | ||
| 2 | 12 | Control | 6.02 | RABL2A | 5 | 9 | Control | 6.31 | RUFY1 | ||
| 11 | 13 | Case | 5.79 | PKP3 | 3 | 13 | Control | 5.75 | CNTN3 | ||
| Pair 4 Age 49, 49 M,M Cau, Cau |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 12 Age 47,47 M,M Mao, Cau Sm,Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 6 | 14 | Control | 173.41 | DUSP22 | 6 | 12 | Case | 39.53 | FOXO3 | ||
| 2 | 15 | Case | 44.61 | RGPD6 | 8 | 9 | Control | 36.71 | ERICH1 | ||
| 17 | 13 | Control | 27.69 | LRRC37A | 8 | 6 | Control | 34.08 | THEM6 | ||
| 12 | 8 | Case | 15.41 | ZNF664 | 19 | 10 | Control | 33.25 | HOOK2 | ||
| 7 | 11 | Control | 14.68 | COX19 | 8 | 8 | Control | 33.1 | ERICH1 | ||
| 17 | 11 | Control | 13.32 | PLEKHH3 | 17 | 13 | Control | 31.05 | ARL4D | ||
| 17 | 10 | Case | 10.09 | SLC47A2 | 17 | 11 | Case | 24.71 | PLEKHM1 | ||
| 17 | 8 | Case | 9.66 | Sep-09 | 10 | 10 | Control | 19.12 | ZNF518A | ||
| 17 | 10 | Control | 9.61 | WNT3 | 5 | 7 | Case | 16.91 | IRGM | ||
| 10 | 16 | Case | 8.59 | CPXM2 | 19 | 15 | Control | 15.94 | RYR1 | ||
| 2 | 12 | Case | 8.2 | AL845331.2 | 20 | 7 | Control | 14.56 | ACTL10 | ||
| Pair 5 Age 54, 54 M,M Cau,Cau |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 13 Age 51, 51 M,M Mao, Cau Sm,Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 3 | 10 | Control | 152.79 | FNDC3B | 19 | 12 | Case | 128.37 | MADCAM1 | ||
| 2 | 13 | Control | 53.62 | RGPD8 | 3 | 6 | Control | 26.13 | ALG1L | ||
| 10 | 16 | Case | 33.47 | CPXM2 | 9 | 14 | Case | 25.41 | CACNA1B | ||
| 12 | 8 | Case | 25.78 | ZNF664 | 3 | 9 | Control | 23.79 | ALG1L | ||
| 10 | 10 | Control | 25.18 | ZNF518A | 6 | 10 | Case | 19.45 | DUSP22 | ||
| 19 | 10 | Control | 24.24 | HOOK2 | 8 | 8 | Control | 16.72 | ERICH1 | ||
| 10 | 15 | Case | 17.47 | CPXM2 | 20 | 12 | Case | 11.37 | NPBWR2 | ||
| 7 | 10 | Control | 17.16 | ARMC10 | 2 | 27 | Control | 9.76 | RABL2A | ||
| 8 | 16 | Case | 16.75 | NAPRT | 19 | 6 | Control | 9.69 | ZNF714 | ||
| 19 | 7 | Control | 15.55 | ZNF714 | 17 | 13 | Control | 8.25 | ARL4D | ||
| 19 | 6 | Control | 11.99 | ZNF714 | 13 | 10 | Control | 7.78 | SKA3 | ||
| Pair 6 Age 46, 47 M,M Cau,Mao Sm,Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 14 Age 48, 48 F,F Mao, Cau Dia, Dia Sm,Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 17 | 11 | Control | 66.14 | PLEKHM1 | 2 | 27 | Case | 57.59 | RABL2A | ||
| 19 | 14 | Control | 41.4 | RETN | 7 | 10 | Case | 56.66 | ARMC10 | ||
| 7 | 10 | Control | 26.07 | ARMC10 | 6 | 12 | Control | 36.08 | FOXO3 | ||
| 6 | 10 | Control | 25.56 | DUSP22 | 19 | 15 | Control | 21.41 | RYR1 | ||
| 2 | 7 | Control | 19.9 | MFF | 19 | 6 | Control | 20.1 | ZNF714 | ||
| 20 | 10 | Case | 18.82 | PPP1R3D | 8 | 8 | Case | 20.05 | ERICH1 | ||
| 12 | 8 | Case | 17.02 | ZNF664 | 20 | 7 | Control | 18.66 | ACTL10 | ||
| 9 | 14 | Control | 15.02 | CACNA1B | 2 | 9 | Case | 18.49 | RABL2A | ||
| 11 | 8 | Case | 12.15 | PKP3 | 9 | 10 | Control | 7.76 | TLE1 | ||
| 11 | 9 | Case | 11.36 | PKP3 | 20 | 17 | Case | 10.12 | PPP1R3D | ||
| 20 | 8 | Case | 10.78 | PPP1R3D | 9 | 6 | Control | 7.38 | LHX6 | ||
| Pair 7 Age 53, 53 M,M Ind, Asi Diab, Diab |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 15 Age 40, 40 M,M Cau, Cau |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 2 | 8 | Control | - | RABL2A | 7 | 10 | Control | 272.22 | ARMC10 | ||
| 16 | 10 | Case | 161.06 | PRR25 | 8 | 15 | Case | 148.52 | ERICH1 | ||
| 19 | 10 | Control | 50.06 | HOOK2 | 6 | 14 | Control | 84.23 | DUSP22 | ||
| 2 | 12 | Control | 46.02 | RABL2A | 8 | 8 | Case | 44.09 | ERICH1 | ||
| 11 | 12 | Control | 31.92 | DGKZ | 19 | 10 | Control | 29.61 | HOOK2 | ||
| 7 | 10 | Control | 30.77 | ARMC10 | 20 | 17 | Control | 25.05 | GNAS | ||
| 5 | 14 | Case | 18.49 | LPCAT1 | 6 | 14 | Control | 24.22 | DUSP22 | ||
| 2 | 9 | Control | 18.2 | RABL2A | 20 | 10 | Control | 21.26 | PCMTD2 | ||
| 5 | 7 | Control | 11.76 | IRGM | 2 | 27 | Control | 19.44 | RABL2A | ||
| 10 | 14 | Control | 9.24 | BMS1 | 18 | 11 | Case | 18.78 | ZNF519 | ||
| 16 | 17 | Control | 7.38 | PDZD9 | 20 | 15 | Control | 12.94 | SOX18 | ||
| Pair 8 Age 54, 55 M,M Mao, PI Sm,Sm |
#Chr | CpGs | >Meth | Fold_diff | GeneID | Pair 16 Age 52,52 M,M Cau,Cau |
#Chr | CpGs | >Meth | Fold_diff | GeneID |
| 6 | 12 | Control | 123.85 | FOXO3 | 19 | 7 | Case | 42.9 | ZNF714 | ||
| 17 | 11 | Case | 43.13 | PLEKHM1 | 4 | 8 | Control | 40.32 | MYL5 | ||
| 8 | 16 | Control | 36.49 | NAPRT | 2 | 13 | Control | 38.72 | RGPD8 | ||
| 6 | 10 | Case | 24.55 | DUSP22 | 10 | 19 | Control | 32.67 | BMPR1A | ||
| 19 | 6 | Case | 16.24 | ZNF714 | 3 | 6 | Control | 23.18 | ALG1L | ||
| 13 | 7 | Control | 14.03 | MRPL57 | 19 | 6 | Case | 16.45 | ZNF714 | ||
| 16 | 9 | Control | 12.68 | PRR25 | 10 | 16 | Control | 16.34 | CPXM2 | ||
| 17 | 8 | Case | 8.07 | Sep-09 | 16 | 17 | Control | 7.42 | ZNF469 | ||
| 11 | 8 | Case | 7.66 | PKP3 | 17 | 12 | Control | 6.35 | SLC47A2 | ||
| 7 | 7 | Case | 6.89 | AC115220.1 | 13 | 12 | Control | 4.48 | PABPC3 | ||
| 11 | 13 | Case | 6.42 | PKP3 | 17 | 10 | Control | 4.42 | SLC47A2 |
In the cases group SEPT9, MADCAM1, and PKP3 were found to show greater methylation in multiple pairwise comparisons, so potentially these genes have a suppressed function in the individuals with recurrent events. SEPT9 is involved in cytokinesis, cell cycle control and cadherin binding, important in cell-cell adhesion. MADCAM1 is also involved in cell adhesion, specifically integrin binding in cell-matrix adhesion. PKP3 has a role in alpha-catenin binding, cadherin binding and cell adhesion binding with a possible role in junctional plaques. It could be that the normal function of these genes is to promote wound healing post cardiac trauma, and their possible suppression could indicate a maladaptive responsive in these individuals and thus an increased susceptibility to secondary cardiac events. However, with the limitations of this study, this is conjecture and further studies including modelling in cardiomyocytes are required to see if this is the case. Oxidative stress has been proposed as a key factor contributing to remodelling of the extracellular matrix with excessive oxidative stress leading to cardiovascular injury [25]. It is possible that this occurs through extended exposure to oxidative stress, which in turn causes a change to the methylation status of key genes involved in the remodelling of the cardiac matrix, thus suppressing their function and impeding cardiac repair. In a recent article by Paiva et al., the communication between cardiomyocytes and macrophages was cited as a crucial factor disrupted by ischaemic injury likely to impact on cardiac remodelling post MI, with the adhesion of macrophages to fibronectin higher in cellular conditions replicating cardiomyocytes under ischaemia [26]. In the control group, RYR1 was found to be significantly more methylated than in the individuals with recurrent events, suggesting suppressed function of the ryanodine receptor in these individuals compared with the individuals with recurrent events. The ryanodine receptors play an important role in releasing calcium from the sarcoplasmic reticulum to activate cardiac muscle contraction [27]. Perhaps the increased methylation of RYR1 in the control group conveys stability to this subset of patients and relative increased activity of RYR1 could contribute to increased susceptibility to recurrent events.
The data generated from this study are interesting, and further studies are now warranted to investigate a change in methylation status as a pathogenic mechanism contributing to recurrent infarction/arterio-thrombotic events in young MI patients. The role of methylation on maladaptive changes to the cardiomyocytes needs to be explored, specifically in the context of whether methylation changes contribute to the pathological accumulation of extracellular matrix proteins, tissue fibrosis, and increased susceptibility to cardiac failure.
Limitations
A limitation of our study is the small sample size but despite this, by adjusting for statistical significance and carrying out detailed pairwise comparisons we have identified crucial areas differentially methylated between the two groups that can aid in understanding the different pathogenesis between the two groups. Another limitation is the use of DNA extracted from peripheral blood as opposed to cardiac tissue; however, previous studies have demonstrated a good correlation between methylation patterns identified in peripheral blood and in cardiac tissue extracted at biopsy [28].
Conclusion
In this study, we identified differential methylation between two groups of young cardiac MI patients, those with and without recurrent events. We first compared the two groups and then looked at each matched pairwise comparison. Both analyses identified genomic regions with relevance to cardiac function, repair, and response to injury. There is significant overlap in the molecular pathogeneses identified by the whole group and individual pairwise comparisons in that cell adhesion, and cardiac electrical conduction, specifically relating to calcium channel activity appear to be significantly differentially methylated with respect to individuals who do and do not have recurrent MI events. Thus, points to these both as potentially key pathological processes that contribute to multiple cardiac events in young MI patients.
Supplementary Material
Funding Statement
This work was supported by the Wellington Medical Research Foundation [2017/289].
Disclosure statement
No potential conflict of interest was reported by the authors.
References
- [1].Abraham G, Havulinna AS, Bhalala OG, et al. Genomic prediction of coronary heart disease. Eur Heart J. 2016;37(43):3267–3278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Khera AV, Kathiresan S.. Genetics of coronary artery disease: discovery, biology and clinical translation. Nat Rev Genet. 2017;18(6):331–344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Shah N, Kelly AM, Cox N, et al. Myocardial infarction in the “young”: risk factors, presentation, management and prognosis. Heart Lung Circ. 2016;25(10):955–960. [DOI] [PubMed] [Google Scholar]
- [4].Matsis K, Holley A, Al-Sinan A, et al. Differing clinical characteristics between young and older patients presenting with myocardial infarction. Heart Lung Circ. 2017;26(6):566–571. [DOI] [PubMed] [Google Scholar]
- [5].Larsen PD, Holley AS, Sasse A, et al. Comparison of multiplate and verifyNow platelet function tests in predicting clinical outcome in patients with acute coronary syndromes. Thromb Res. 2017;152:14–19. [DOI] [PubMed] [Google Scholar]
- [6].Ogino S, Nishihara R, VanderWeele TJ, et al. Review article: the role of molecular pathological epidemiology in the study of neoplastic and non-neoplastic diseases in the era of precision medicine. Epidemiology. 2016;27(4):602–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Bodden C, van Den Hove D, Lesch KP, et al. Impact of varying social experiences during life history on behaviour, gene expression, and vasopressin receptor gene methylation in mice. Sci Rep. 2017;7(1):8719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Vel Szic KS, Declerck K, Vidaković M, et al. From inflammaging to healthy aging by dietary lifestyle choices: is epigenetics the key to personalized nutrition? Clin Epigenetics. 2015;7(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Cao DJ. Epigenetic regulation and heart failure. Expert Rev Cardiovasc Ther. 2014;12(9):1087–1098. [DOI] [PubMed] [Google Scholar]
- [10].Mathur S, Sutton J. Personalized medicine could transform healthcare. Biomed Rep. 2017;7(1):3–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Rask-Andersen M, Martinsson D, Ahsan M, et al. Epigenome-wide association study reveals differential DNA methylation in individuals with a history of myocardial infarction. Hum Mol Genet. 2016;25(21):4739–4748. [DOI] [PubMed] [Google Scholar]
- [12].Li B, Feng ZH, Sun H, et al. The blood genome-wide DNA methylation analysis reveals novel epigenetic changes in human heart failure. Eur Rev Med Pharmacol Sci. 2017;21(8):1828–1836. [PubMed] [Google Scholar]
- [13].Walker DL, Bhagwate AV, Baheti S, et al. DNA methylation profiling: comparison of genome-wide sequencing methods and the infinium human methylation 450 bead chip. Epigenomics. 2015;7(8):1287–1302. [DOI] [PubMed] [Google Scholar]
- [14].Chatterjee A, Rodger EJ, Stockwell PA, et al. Technical considerations for reduced representation bisulfite sequencing with multiplexed libraries. J Biomed Biotechnol. 2012;2012:741542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Stockwell PA, Chatterjee A, Rodger EJ, et al. DMAP: differential methylation analysis package for RRBS and WGBS data. Bioinformatics. 2014;30(13):1814–1822. [DOI] [PubMed] [Google Scholar]
- [16].Koutnikova H, Laakso M, Lu L, et al. Identification of the UBP1 locus as a critical blood pressure determinant using a combination of mouse and human genetics. PLoS Genet. 2009;5(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Ashton KJ, Tupicoff A, Williams-Pritchard G, et al. Unique transcriptional profile of sustained ligand-activated preconditioning in pre- and post-ischemic myocardium. PLoS One. 2013;8(8):e72278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Yuan L, Chen X, Cheng L, et al. HDAC11 regulates interleukin-13 expression in CD4+ T cells in the heart. J Mol Cell Cardiol. 2018;122:1–10. [DOI] [PubMed] [Google Scholar]
- [19].Yamada Y, Yasukochi Y, Kato K, et al. Identification of 26 novel loci that confer susceptibility to early-onset coronary artery disease in a Japanese population. Biomed Rep. 2018;9(5):383–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Tu YH, Cooper AJ, Teng B, et al. An evolutionarily conserved gene family encodes proton-selective ion channels. Science. 2018;359(6379):1047–1050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Cheng M, An S, Li J. Identifying key genes associated with acute myocardial infarction. Medicine (Baltimore). 2017;96(42):e7741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Afouda BA, Lynch AT, Alves ED, et al. Genome-wide transcriptomics analysis identifies sox7 and sox18 as specifically regulated by gata4 in cardiomyogenesis. Dev Biol. 2018;434(1):108–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Fabregat A, Jupe S, Matthews L, et al. The reactome pathway knowledgebase. Nucleic Acids Res. 2018;46(D1):D649–D655. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Eden E, Navon R, Steinfeld I, et al. GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics. 2009;10:48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Jacob-Ferreira AL, Schulz R. Activation of intracellular matrix metalloproteinase-2 by reactive oxygen-nitrogen species: consequences and therapeutic strategies in the heart. Arch Biochem Biophys. 2013;540(1–2):82–93. [DOI] [PubMed] [Google Scholar]
- [26].Almeida Paiva R, Martins‐Marques T, Jesus K, et al. Ischaemia alters the effects of cardiomyocyte‐derived extracellular vesicles on macrophage activation. J Cell Mol Med. 2019;23(2):1137–1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Dulhunty AF, Beard NA, Casarotto MG. Recent advances in understanding the ryanodine receptor calcium release channels and their role in calcium signalling. F1000Res. 2018;7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Ma BS, Wilker EH, Willis-Owen SAG, et al. Predicting DNA methylation level across human tissues. Nucleic Acids Res. 2014;42(6):3515–3528. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.