Abstract
Aim: CDKN2A/2B near chromosome 9p21 has been proposed as a potential genetic etiology for both atherosclerosis and arterial calcification. DNA methylation, which can change the expression of CDKN2A/2B, may be an underlying mechanism for this association. This study aimed to evaluate whether CDKN2A/2B methylation is related to aortic arch calcification (AAC) in patients with ischemic stroke.
Methods: DNA methylation levels of CDKN2A/2B was measured using venous blood samples in 322 patients with ischemic stroke. A total of 36 CpG sites around promoter regions of CDKN2A/2B were examined. AAC was quantified with Agatston score based on results of computed tomography angiography.
Results: There were 248 (77.0%) patients with and 74 (23.0%) patients without evident AAC. Compared with patients without AAC, patients with AAC had higher methylation levels of CDKN2B (5.72 vs 4.94, P < 0.001). Using a generalized linear model, positive correlation between methylation levels and log-transformed calcification scores was detected at CDKN2B (β = 0.275 ± 0.116, P = 0.018).
Conclusion: Patients with higher levels of DNA methylation of CDKN2B may bear increased risk for AAC. Further studies to reveal the underlying mechanisms of this association are warranted for establishing a cause–effect relationship.
Keywords: Aortic arch calcification, CDKN2A/2B, DNA methylation, Ischemic stroke
Introduction
Atherosclerotic lesion of the aortic arch is a common etiology for ischemic stroke1, 2). Aortic arch calcification (AAC), a surrogate measure for atherosclerosis, can be readily detected with chest radiography3, 4). AAC was proposed as a satisfactory index for measuring systemic atherosclerosis burden5, 6) and was associated with cardiovascular events2, 7, 8).
Growing evidence indicated that genetic factors may affect the initiation and development of artery calcification9, 10). The human chromosome 9p21 (Chr9p21), for example, has been associated with both atherosclerosis and arterial calcification in several genome-wide association studies11–13). As a chromosome region devoid of protein-coding genes, Chr9p21 only transcribes a long non-coding RNA, namely antisense noncoding RNA in the INK4 locus (ANRIL). The closest protein-coding genes to Chr9p21 locus are two cyclin-dependent kinase inhibitors, CDKN2A and CDKN2B, both of which are involved in cell cycle regulation. Previous studies showed that variations in Chr9p21 may increase the levels of ANRIL transcription, which in turn downregulate CDKN2A/2B expression and enhance cell proliferation, and subsequently promote atherosclerosis14, 15). Studies also confirmed that ANRIL could bind and recruit epigenetic modifiers to CDKN2A/2B and induce DNA methylation16, 17).
Aim
Considering CDKN2A/2B involved in the development of atherosclerotic diseases18, 19), and DNA methylation of CDKN2A/2B has been frequently reported in many conditions20, 21), we hypothesized that CDKN2A/2B methylation may increase the susceptibility of AAC. We tested this hypothesis in a group of Chinese patients with ischemic stroke who were at a high risk of developing atherosclerosis and calcification.
Methods
Study Population
This study was approved by the Ethical Review Board of Jinling Hospital. Informed consent was obtained from all enrolled patients. Consecutive patients with ischemic stroke aged ≥ 18 years were screened from Nanjing Stroke Registry Program (NSRP)22) between July 2012 and September 2013. Patients with malignant neoplasm, severe liver or kidney diseases, autoimmune diseases, parathyroid gland diseases, or calcium–phosphorus metabolism disorders were excluded. As stents may influence the accuracy of calcification assessment, patients with a history of stenting treatment in aortic arch, brachiocephalic trunk, subclavian arteries, and common carotid arteries were also excluded. Finally, 324 patients were enrolled. Demographic characteristics and cardiovascular risk factors, which included age, sex, history of hypertension (HTN) and diabetes mellitus (DM), dyslipidemia, cigarette smoking, and alcohol drinking, were collected.
Artery Calcification Measurement
Each enrolled patient underwent neck computed tomography angiography (CTA) for AAC evaluation. CTA was performed with a dual-source 64 slice CT system (Siemens, Forchheim, Germany). Imaging was acquired by scanning from 4 cm below the aortic arch to the superior border of the orbit in craniocaudal direction. The aortic arch was recognized as a section from the initial segment to the first centimeter of the common carotid, vertebral, and subclavian arteries beyond the origin of the vertebral arteries. Details of scanning parameters have been reported elsewhere23). Calcification scores in the aortic arch were measured with the Syngo Calcium Scoring system (Siemens, Forchheim, Germany). A focus of ≥ 4 contiguous pixels accompanied with a CT density of ≥ 130 Hounsfield units (HU) was defined as calcification according to the method of Agatston score24). For each calcified lesion, the Agatston score was calculated as the product of the area (mm2) and a factor assigned according to the maximum attenuation value of the lesion (HU = 130–199 [1], 200–299 [2], 300–399 [3], > 399 [4]). The total score of the aortic arch was calculated by adding up the scores of all lesions. Finally, patients with Agatston score of 0 or > 0 were dichotomized into groups without or with AAC, respectively. Calcification scores were dual-assessed by two radiologists who were blinded to epi-genotyping results.
DNA Isolation and Genotyping
Venous blood samples were drawn in the morning after an overnight fasting for assaying biochemical parameters and epi-genotyping. Genomic DNA was extracted from whole blood using commercially available kits (TIANGEN Biotech, Beijing, China). DNA was quantified and then diluted to a working concentration of 10 ng/uL. Rs4977574 at Chr9p21, which is significantly associated with calcification in the aorta based on the validation of a previous study25), was selected for genotyping. Single nucleotide polymorphism of rs4977574 (AA, AG, GG) was genotyped via polymerase chain reaction ligase detection reaction with an ABI Prism 377 Sequence Detection System (Applied Biosystems, CA, USA)26). Sequencing primers were CATGCTTTCTGAAACAACACG (forward) and TAATGGAGGTGTGGTCAGCA (reverse). Reproducibility of genotyping was confirmed by randomly selecting 10% of the samples, and the concordance was 100%.
DNA Methylation Analysis
CpG islands adjacent to promoter regions of CDKN2A/2B were selected for measurement according to the following criteria: (1) 200 bp minimum length; (2) ≥ 50% GC content; (3) ≥ 0.60 ratio of observed/expected dinucleotides CpG27). Six CpG regions from CpG islands of CDKN2A and three from those of CDKN2B were sequenced. Bisulfite conversion of 1 ug genomic DNA was performed using the EZ DNA Methylation™-GOLD Kit (ZYMO RESEARCH, CA, USA) according to the manufacturer's protocol. Sodium bisulfite can preferentially deaminate un-methylated cytosine residues to thymines, whereas methyl-cytosines remain unmodified. After PCR amplification (HotStarTaq polymerase kit, TAKARA, Tokyo, Japan) of target CpG regions and library construction, products were sequenced using Illumina MiSeq Benchtop Sequencer (CA, USA) in accordance with the method of BiSulfite Amplicon Sequencing28). Primer sequences used for PCR were shown in Supplemental Table 1. All samples achieved a mean coverage of > 600 X. Methylation levels of 24 CpG sites in CDKN2A and 12 sites in CDKN2B were measured. Each tested CpG site was named as its relative distance (in bp) to transcriptional start site and listed in Supplemental Table 2. The methylation level of each CpG site was calculated as the percentage of the methylated cytosines over total tested cytosines. The average methylation level was calculated using methylation levels of all measured CpG sites within the gene.
Supplemental Table 1. Primer sequences for CDKN2A/2B genes (start and end site were named as its relative distance to transcriptional start site).
| Gene | PCR size (bp) | Start site | End site | Primer | |
|---|---|---|---|---|---|
| CDKN2A | 282 | −1477 | −1197 | forward | GGGATATGGAGGGGGAGAT |
| reverse | CTTCTTCCTCTTTCCTCTTCCC | ||||
| 211 | −1047 | −838 | forward | GGGAAGAGGAAAGAGGAAGAAG | |
| reverse | ATTAAACTAAACCRCTACACRCCTCTAAC | ||||
| 286 | −859 | −574 | forward | AATAAAATAAGGGGAATAGGGGAG | |
| reverse | CCATCTTCCCACCCTCAA | ||||
| 188 | −399 | −212 | forward | GTAGTTAAGGGGGTAGGAGTGG | |
| reverse | ACTACTACCCTAAACRCTAACTCCTCAA | ||||
| 266 | +70 | +335 | forward | TTGAGGAGTTAGYGTTTAGGGTAGTAGT | |
| reverse | TCAATAATACTACRAAAACCACATATCTAAATC | ||||
| 224 | +308 | +531 | forward | GTYGGTTGGTTTTTTATTTTGTTAGAG | |
| reverse | AACCTAAACTCAACTTCATTACCCTC | ||||
| CDKN2B | 255 | −7 | +248 | forward | GAGGGTAATGAAGTTGAGTTTAGGTT |
| reverse | CTATCRCACCTTCTCCACTAATCC | ||||
| 234 | +223 | +455 | forward | GGGGATTAGTGGAGAAGGTG | |
| reverse | TAAAATACACACCTCCRACCAAC | ||||
| 221 | +430 | +650 | forward | TGTTTTTTAAGTTTTTATAGGGTGAGG | |
| reverse | CCAACCTAACCAAAATAATAAAAACC | ||||
Supplemental Table 2. Methylated CpG sites measured in this study.
| Gene | Position | Genomic location* | Relative to TSS, bp |
|---|---|---|---|
| CDKN2A | 1 | Chr9: 21995909 | −1419 |
| 2 | Chr9: 21995896 | −1406 | |
| 3 | Chr9: 21995867 | −1377 | |
| 4 | Chr9: 21995713 | −1223 | |
| 5 | Chr9: 21995470 | −980 | |
| 6 | Chr9: 21995457 | −967 | |
| 7 | Chr9: 21995455 | −965 | |
| 8 | Chr9: 21995354 | −864 | |
| 9 | Chr9: 21995314 | −824 | |
| 10 | Chr9: 21995312 | −822 | |
| 11 | Chr9: 21995305 | −815 | |
| 12 | Chr9: 21995108 | −618 | |
| 13 | Chr9: 21994859 | −369 | |
| 14 | Chr9: 21994782 | −292 | |
| 15 | Chr9: 21994734 | −244 | |
| 16 | Chr9: 21994727 | −237 | |
| 17 | Chr9: 21994286 | +205 | |
| 18 | Chr9: 21994215 | +276 | |
| 19 | Chr9: 21994211 | +280 | |
| 20 | Chr9: 21994208 | +283 | |
| 21 | Chr9: 21994155 | +336 | |
| 22 | Chr9: 21994109 | +382 | |
| 23 | Chr9: 21994076 | +415 | |
| 24 | Chr9: 21993993 | +498 | |
| CDKN2B | 1 | Chr9: 22009259 | +54 |
| 2 | Chr9: 22009179 | +134 | |
| 3 | Chr9: 22009165 | +148 | |
| 4 | Chr9: 22009134 | +179 | |
| 5 | Chr9: 22009000 | +313 | |
| 6 | Chr9: 22008981 | +332 | |
| 7 | Chr9: 22008956 | +357 | |
| 8 | Chr9: 22008890 | +423 | |
| 9 | Chr9: 22008845 | +468 | |
| 10 | Chr9: 22008830 | +483 | |
| 11 | Chr9: 22008815 | +498 | |
| 12 | Chr9: 22008804 | +509 | |
The chromosomal location of each CpG site according to assembly GRCh37/hg19.
Statistical Analysis
Normality of quantitative variables was assessed using Shapiro–Wilk test. As all continuous data in this study did not meet the normality assumption, they were described as median (interquartile range) and compared using Mann–Whitney U test. Categorical variables were compared using Fisher's exact test.
Spearman correlations were used to evaluate pair-wise correlations of methylation levels between different CpG sites in the same gene. Considering the extremely left-skewed distribution of calcification scores, we added 1 to each calcification score, and the value was then log-transformed as the formula: Ln (calcification score + 1). This transformation may result in a less skewed distribution, as suggested in previous studies13, 25). Generalized linear model was used to explore the association between methylation levels and log-transformed calcification scores, adjusting for potential confounders including age, sex, HTN, DM, dyslipidemia, and smoking. For multiple testing, Bonferroni correction was performed. Kruskal–Wallis test was performed to compare methylation levels across the genotypes of rs4977574 (AA, AG, and GG).
Data were analyzed using IBM SPSS Statistics Version 22.0 (Armonk, NY: IBM Corp.). A two-tailed value of P < 0.05 was considered statistically significant.
Results
Of the 324 enrolled participants, 2 (0.6%) failed in epi-genotyping. Finally, 322 (99.4%) patients were included for data analysis. Baseline characteristics were listed in Table 1. The median age of these 322 patients was 62.0 (55.0–70.0) years, and 229 (71.1%) of them were males. Of these analyzed patients, 250 (77.6%) had a history of HTN and 110 (34.2%) had a history of DM.
Table 1. Baseline characteristics of the study participants.
| Variants | All (n = 322) |
|---|---|
| Age, years | 62.0 (55.0–70.0) |
| Male, n (%) | 229 (71.1) |
| HTN, n (%) | 250 (77.6) |
| DM, n (%) | 110 (34.2) |
| Dyslipidemia, n (%) | 176 (54.7) |
| TC, mmol/L | 4.21 (3.58–5.00) |
| TG, mmol/L | 1.40 (1.09–1.88) |
| HDL-c, mmol/L | 0.98 (0.82–1.15) |
| LDL-c, mmol/L | 2.61 (1.93–3.18) |
| Glucose, mmol/L | 5.3 (4.6–6.6) |
| Smoking, n (%) | 132 (41.0) |
| Drinking, n (%) | 96 (29.8) |
| AAC, n (%) | 248 (77.0%) |
| AAC score | 221.5 (3.8–803.7) |
| Ln (AAC+1) | 5.40 (1.56–6.69) |
Data are presented as number of individuals (%) or median (interquartile range).
AAC, artic arch calcification; HTN, hypertension; DM, diabetes mellitus; TC, total cholesterol; TG, triglyceride; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Based on Agatston scores, there were 248 (77.0%) and 74 (23.0%) patients classified as with and without AAC, respectively. AAC scores presented a highly skewed distribution with a median (interquartile range) of 221.5 (3.8–803.7). Compared with patients without AAC, those with AAC were older (64.0 vs 49.0 years, P < 0.001), had a higher prevalence of HTN (83.1% vs 59.5%, P < 0.001), lower prevalence of dyslipidemia (51.6% vs 64.9%, P = 0.047), and higher HDL-c levels (1.00 vs 0.91, P = 0.048) (Table 2).
Table 2. Comparison of demographic characteristics of patients with and without AAC.
| Variants | AAC |
P value | |
|---|---|---|---|
| With (n = 248) | Without (n = 74) | ||
| Age, years | 64.0 (58.0–72.0) | 49.0 (43.8–58.3) | < 0.001 |
| Male, n (%) | 171 (69.0) | 58 (78.4) | 0.144 |
| HTN, n (%) | 206 (83.1) | 44 (59.5) | < 0.001 |
| DM, n (%) | 86 (34.7) | 24 (32.4) | 0.781 |
| Dyslipidemia, n (%) | 128 (51.6) | 48 (64.9) | 0.047 |
| TC, mmol/L | 4.27 (3.55–5.01) | 4.16 (3.69–4.78) | 0.756 |
| TG, mmol/L | 1.37 (1.06–1.76) | 1.58 (1.15–2.03) | 0.064 |
| HDL-c, mmol/L | 1.00 (0.83–1.16) | 0.91 (0.79–1.06) | 0.048 |
| LDL-c, mmol/L | 2.63 (1.93–3.20) | 2.60 (1.93–3.09) | 0.995 |
| Glucose, mmol/L | 5.3 (4.7–6.7) | 5.3 (4.6–6.6) | 0.670 |
| Smoking, n (%) | 97 (39.1) | 35 (47.3) | 0.227 |
| Drinking, n (%) | 71 (28.6) | 25 (33.8) | 0.390 |
Methylation levels of 36 CpG sites were listed in Supplemental Table 3. Methylation levels of CpG sites measured within CDKN2A were not strongly correlated, whereas those within CDKN2B were well correlated (Supplemental Table 4–5). As shown in Table 3, univariate comparison of these 36 sites and the average methylation levels indicated that methylation levels of CDKN2B were higher in patients with AAC than in those without AAC ( 5.72 vs 4.94, P < 0.001).
Supplemental Table 3. Distribution of methylation levels (%) of 36 CpG sites in CDKN2A/2B genes.
| Gene | Position | Min | Q1 | Median | Q3 | Max |
|---|---|---|---|---|---|---|
| CDKN2A | 1 | 0.00 | 2.81 | 4.26 | 5.94 | 17.19 |
| 2 | 0.00 | 5.31 | 6.98 | 8.82 | 19.69 | |
| 3 | 0.00 | 6.53 | 8.08 | 10.17 | 21.40 | |
| 4 | 0.00 | 4.27 | 5.83 | 7.85 | 18.60 | |
| 5 | 0.00 | 4.08 | 4.89 | 5.49 | 9.66 | |
| 6 | 0.00 | 2.29 | 2.71 | 3.29 | 9.23 | |
| 7 | 0.00 | 1.84 | 2.33 | 2.79 | 7.69 | |
| 8 | 1.79 | 3.69 | 4.37 | 5.03 | 10.44 | |
| 9 | 0.00 | 2.44 | 4.42 | 7.90 | 23.53 | |
| 10 | 0.00 | 0.98 | 1.97 | 3.14 | 8.82 | |
| 11 | 0.00 | 2.38 | 3.61 | 4.93 | 13.33 | |
| 12 | 0.00 | 0.57 | 0.93 | 1.33 | 2.85 | |
| 13 | 0.00 | 0.95 | 1.21 | 1.50 | 2.60 | |
| 14 | 0.00 | 0.97 | 1.18 | 1.44 | 2.94 | |
| 15 | 0.00 | 1.64 | 2.05 | 2.43 | 6.86 | |
| 16 | 0.00 | 1.03 | 1.34 | 1.65 | 3.37 | |
| 17 | 0.00 | 2.59 | 3.17 | 3.83 | 25.26 | |
| 18 | 0.49 | 1.72 | 2.18 | 2.58 | 7.49 | |
| 19 | 0.00 | 2.01 | 2.49 | 2.95 | 7.49 | |
| 20 | 0.44 | 2.17 | 2.72 | 3.24 | 8.85 | |
| 21 | 4.86 | 13.74 | 15.51 | 17.20 | 34.15 | |
| 22 | 0.71 | 2.07 | 2.57 | 3.18 | 8.62 | |
| 23 | 0.00 | 3.50 | 4.27 | 5.05 | 10.98 | |
| 24 | 0.00 | 1.26 | 1.70 | 2.46 | 6.17 | |
| Average | 2.39 | 3.61 | 3.94 | 4.25 | 6.15 | |
| CDKN2B | 1 | 1.83 | 4.42 | 5.35 | 6.24 | 12.21 |
| 2 | 0.00 | 3.42 | 4.37 | 5.24 | 10.88 | |
| 3 | 0.00 | 3.10 | 3.89 | 4.82 | 9.32 | |
| 4 | 0.00 | 3.34 | 4.10 | 4.99 | 18.82 | |
| 5 | 4.42 | 6.44 | 7.40 | 8.61 | 16.37 | |
| 6 | 3.44 | 5.54 | 6.66 | 7.78 | 13.05 | |
| 7 | 4.06 | 6.84 | 7.86 | 9.06 | 18.45 | |
| 8 | 0.86 | 2.94 | 3.41 | 3.96 | 12.05 | |
| 9 | 0.00 | 3.23 | 3.74 | 4.36 | 17.09 | |
| 10 | 0.00 | 4.95 | 5.85 | 6.74 | 18.04 | |
| 11 | 3.70 | 6.18 | 7.13 | 8.48 | 27.93 | |
| 12 | 2.43 | 4.80 | 5.48 | 6.45 | 23.90 | |
| Average | 3.45 | 4.83 | 5.54 | 6.15 | 11.08 | |
Q1: 1st quartile (25th percentile), Q3: 3rd quartile (75th percentile).
Supplemental Table 4. Spearman pairwise correlations for CpG sites of CDKN2A.
| Position | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 1.0 | 0.5* | 0.4* | 0.4* | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | −0.1* | 0.1 | −0.1 | 0.0 | −0.1 | −0.1 | −0.1 | 0.0 | −0.1* | 0.0 | 0.1 | 0.1 | 0.2* |
| 2 | 1.0 | 0.5* | 0.4* | 0.0 | 0.0 | 0.0 | 0.1* | 0.0 | 0.0 | 0.0 | 0.0 | −0.1 | −0.1 | 0.1 | −0.1* | 0.0 | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 | 0.1* | 0.3* | |
| 3 | 1.0 | 0.4* | 0.1 | 0.0 | 0.0 | 0.1* | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.1 | −0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1* | 0.2* | 0.3* | ||
| 4 | 1.0 | 0.1 | 0.1 | 0.0 | 0.1 | 0.1 | 0.0 | 0.0 | −0.1 | 0.1 | 0.0 | 0.1 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 | 0.2* | |||
| 5 | 1.0 | 0.2* | 0.3* | 0.2* | 0.0 | 0.0 | 0.0 | 0.1 | 0.2* | 0.1 | 0.1 | 0.2* | 0.0 | 0.1 | 0.1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.1* | ||||
| 6 | 1.0 | 0.3* | 0.3* | −0.1 | 0.1 | 0.1 | 0.1* | 0.1 | 0.2* | 0.2* | 0.2* | 0.1 | 0.1* | 0.1 | 0.0 | 0.0 | 0.1* | 0.1 | 0.2* | |||||
| 7 | 1.0 | 0.2* | 0.0 | 0.1 | 0.0 | 0.1 | 0.0 | 0.1 | 0.2* | 0.2* | 0.0 | 0.1 | 0.1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.1 | ||||||
| 8 | 1.0 | 0.1 | 0.1 | 0.1 | 0.2* | 0.2* | 0.1 | 0.3* | 0.2* | 0.1* | 0.2* | 0.1 | 0.1* | 0.0 | 0.1* | 0.1 | 0.1 | |||||||
| 9 | 1.0 | 0.0 | −0.1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | −0.1 | −0.1 | −0.1* | −0.1* | 0.0 | 0.0 | 0.0 | ||||||||
| 10 | 1.0 | 0.1 | 0.0 | −0.1 | 0.1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | −0.1 | 0.0 | 0.1 | |||||||||
| 11 | 1.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 | 0.1 | 0.1 | 0.0 | 0.1 | 0.2* | 0.0 | 0.0 | ||||||||||
| 12 | 1.0 | 0.0 | 0.0 | 0.1* | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.0 | 0.1 | 0.0 | −0.1 | |||||||||||
| 13 | 1.0 | 0.1 | 0.0 | 0.1* | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | ||||||||||||
| 14 | 1.0 | 0.1* | 0.1 | 0.1 | 0.1 | 0.1 | 0.1* | 0.0 | 0.1* | 0.1* | 0.0 | |||||||||||||
| 15 | 1.0 | 0.1 | 0.0 | 0.1 | 0.0 | 0.1 | 0.0 | 0.1 | 0.1 | 0.2* | ||||||||||||||
| 16 | 1.0 | 0.1 | 0.0 | 0.2* | 0.1 | 0.1 | 0.0 | 0.0 | 0.1 | |||||||||||||||
| 17 | 1.0 | 0.1* | 0.2* | 0.3* | 0.2* | 0.1* | 0.2* | 0.0 | ||||||||||||||||
| 18 | 1.0 | 0.2* | 0.3* | 0.2* | 0.1 | 0.0 | 0.1 | |||||||||||||||||
| 19 | 1.0 | 0.3* | 0.1* | 0.0 | 0.1* | 0.1 | ||||||||||||||||||
| 20 | 1.0 | 0.2* | 0.1 | 0.1* | 0.0 | |||||||||||||||||||
| 21 | 1.0 | 0.1* | 0.1 | 0.1 | ||||||||||||||||||||
| 22 | 1.0 | 0.2* | 0.3* | |||||||||||||||||||||
| 23 | 1.0 | 0.2* | ||||||||||||||||||||||
| 24 | 1.0 |
p < 0.05
Supplemental Table 5. Spearman pairwise correlations for CpG sites of CDKN2B.
| Position | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 1.0 | 0.5 | 0.5 | 0.4 | 0.6 | 0.6 | 0.6 | 0.3 | 0.3 | 0.4 | 0.5 | 0.4 |
| 2 | 1.0 | 0.5 | 0.5 | 0.5 | 0.5 | 0.5 | 0.3 | 0.3 | 0.4 | 0.3 | 0.3 | |
| 3 | 1.0 | 0.4 | 0.5 | 0.5 | 0.5 | 0.3 | 0.3 | 0.3 | 0.3 | 0.3 | ||
| 4 | 1.0 | 0.5 | 0.5 | 0.5 | 0.3 | 0.3 | 0.3 | 0.4 | 0.4 | |||
| 5 | 1.0 | 0.8 | 0.7 | 0.4 | 0.4 | 0.6 | 0.5 | 0.5 | ||||
| 6 | 1.0 | 0.8 | 0.4 | 0.4 | 0.6 | 0.5 | 0.6 | |||||
| 7 | 1.0 | 0.3 | 0.4 | 0.5 | 0.5 | 0.5 | ||||||
| 8 | 1.0 | 0.3 | 0.3 | 0.3 | 0.3 | |||||||
| 9 | 1.0 | 0.6 | 0.6 | 0.5 | ||||||||
| 10 | 1.0 | 0.8 | 0.7 | |||||||||
| 11 | 1.0 | 0.7 | ||||||||||
| 12 | 1.0 |
All p < 0.001
Table 3. Differences of methylation levels between patients with and without AAC.
| Gene | Position | AAC |
P value | |
|---|---|---|---|---|
| With | Without | |||
| CDKN2A | 1 | 4.20 (2.82–5.88) | 4.33 (2.74–6.19) | 0.571 |
| 2 | 6.89 (5.30–8.90) | 7.13 (5.51–8.55) | 0.844 | |
| 3 | 8.09 (6.62–10.4) | 8.02 (6.25–10.0) | 0.379 | |
| 4 | 5.86 (4.25–7.81) | 5.80 (4.38–7.87) | 0.864 | |
| 5 | 4.80 (4.04–5.49) | 5.08 (4.33–5.59) | 0.132 | |
| 6 | 2.79 (2.34–3.36) | 2.56 (2.09–3.07) | 0.038 | |
| 7 | 2.33 (1.87–2.83) | 2.27 (1.77–2.71) | 0.174 | |
| 8 | 4.43 (3.78–5.08) | 4.01 (3.56–4.67) | 0.037 | |
| 9 | 4.30 (2.37–7.33) | 5.26 (3.27–8.83) | 0.130 | |
| 10 | 2.04 (0.96–3.23) | 1.75 (1.00–2.56) | 0.233 | |
| 11 | 3.57 (2.38–4.89) | 3.70 (2.22–5.13) | 0.885 | |
| 12 | 0.96 (0.57–1.34) | 0.88 (0.56–1.25) | 0.363 | |
| 13 | 1.20 (0.96–1.49) | 1.34 (0.94–1.52) | 0.441 | |
| 14 | 1.21 (1.03–1.46) | 1.11 (0.88–1.41) | 0.039 | |
| 15 | 2.05 (1.66–2.41) | 1.93 (1.54–2.48) | 0.630 | |
| 16 | 1.33 (1.02–1.66) | 1.35 (1.08–1.62) | 0.906 | |
| 17 | 3.23 (2.64–3.97) | 3.01 (2.48–3.46) | 0.005 | |
| 18 | 2.19 (1.71–2.61) | 2.14 (1.74–2.53) | 0.487 | |
| 19 | 2.49 (2.01–2.97) | 2.48 (1.96–2.87) | 0.827 | |
| 20 | 2.75 (2.18–3.27) | 2.60 (2.10–3.07) | 0.259 | |
| 21 | 15.5 (13.5–17.2) | 15.6 (14.4–17.1) | 0.313 | |
| 22 | 2.52 (2.08–3.15) | 2.62 (2.07–3.20) | 0.762 | |
| 23 | 4.27 (3.50–5.09) | 4.32 (3.51–4.84) | 0.853 | |
| 24 | 1.69 (1.26–2.46) | 1.77 (1.31–2.48) | 0.631 | |
| Average | 3.95 (3.58–4.27) | 3.92 (3.67–4.23) | 0.909 | |
| CDKN2B | 1 | 5.43 (4.51–6.44) | 4.97 (4.04–5.81) | 0.005 |
| 2 | 4.45 (3.48–5.34) | 3.99 (3.06–5.06) | 0.014 | |
| 3 | 3.89 (3.10–4.87) | 3.86 (3.10–4.64) | 0.700 | |
| 4 | 4.20 (3.53–5.11) | 3.89 (3.14–4.41) | 0.007 | |
| 5 | 7.66 (6.65–8.80) | 6.65 (5.80–7.68) | < 0.001 | |
| 6 | 6.95 (5.99–8.12) | 6.00 (5.04–6.80) | < 0.001 | |
| 7 | 8.12 (7.04–9.49) | 7.10 (6.41–7.92) | < 0.001 | |
| 8 | 3.52 (3.03–4.11) | 3.07 (2.76–3.49) | < 0.001 | |
| 9 | 3.94 (3.34–4.47) | 3.41 (2.76–3.82) | < 0.001 | |
| 10 | 6.08 (5.21–7.03) | 5.11 (4.20–5.92) | < 0.001 | |
| 11 | 7.53 (6.32–8.72) | 6.56 (5.47–7.05) | < 0.001 | |
| 12 | 5.74 (5.02–6.70) | 5.01 (4.37–5.60) | < 0.001 | |
| Average | 5.72 (5.03–6.34) | 4.94 (4.48–5.47) | < 0.001 | |
As shown in Table 4, generalized liner model detected a positive correlation between average methylation levels of CDKN2B and log-transformed calcification scores (β = 0.275 ± 0.116, P = 0.018) after adjusting for age, sex, HTN, DM, dyslipidemia, and smoking. The association still remained after further correction for multiple comparison (corrected P = 0.036).
Table 4. Association of methylation levels and log-transformed calcification scores detected by generalized liner model.
| β | SE | P value | |
|---|---|---|---|
| Model 1 | |||
| CDKN2A | −0.011 | 0.220 | 0.961 |
| Age | 0.159 | 0.011 | < 0.001 |
| Sex | −0.065 | 0.297 | 0.827 |
| HTN | 0.600 | 0.299 | 0.044 |
| DM | 0.039 | 0.257 | 0.880 |
| Dyslipidemia | −0.137 | 0.245 | 0.576 |
| Smoking | −0.016 | 0.272 | 0.954 |
| Model 2 | |||
| CDKN2B | 0.275 | 0.116 | 0.018 |
| Age | 0.148 | 0.012 | < 0.001 |
| Sex | −0.057 | 0.295 | 0.847 |
| HTN | 0.653 | 0.296 | 0.027 |
| DM | 0.042 | 0.253 | 0.869 |
| Dyslipidemia | −0.077 | 0.244 | 0.751 |
| Smoking | −0.049 | 0.269 | 0.854 |
Generalized liner model was adjusted for age, sex, HTN, DM, dyslipidemia and smoking.
Further, we assessed the association between rs4977574 and methylation of CDKN2B. After adjusting for potential risk factors, rs4977574 (G as coded allele) was associated with AAC in the study population (β = 0.414 ± 0.171, P = 0.015). There were no differences in average methylation levels of CDKN2B among three genotypes of rs4977574 (AA vs AG vs GG: 5.56 vs 5.58 vs 5.46, P = 0.626). When rs4977574 was further added into the generalized linear model, the average methylation levels of CDKN2B still correlated with log-transformed calcification scores (β = 0.292 ± 0.115, P = 0.011) (Table 5).
Table 5. Association of CDKN2B methylation levels and AAC after adjustment of rs4977574 and other confounders.
| Variants | β | SE | P value |
|---|---|---|---|
| CDKN2B | 0.292 | 0.115 | 0.011 |
| Age | 0.148 | 0.012 | < 0.001 |
| Sex | 0.01 | 0.293 | 0.972 |
| HTN | 0.645 | 0.293 | 0.028 |
| DM | 0.112 | 0.252 | 0.656 |
| Dyslipidemia | −0.072 | 0.242 | 0.766 |
| Smoking | −0.125 | 0.268 | 0.641 |
| Rs4977574 | 0.439 | 0.170 | 0.010 |
Discussion
This study observed that methylation levels of CDKN2B were relatively higher in patients with AAC than those in patients without AAC. A positive correlation between CDKN2B methylation and AAC load was detected. These results verified our hypothesis that DNA methylation in CDKN2B may increase the susceptibility of artery calcification.
CDKN2B is a well-characterized tumor suppressor gene which is involved in cell cycle regulation via retinoblastoma (Rb) pathway29). The p15INK4b protein encoded by CDKN2B can specifically bind to CDKN4/6 and result in G1 phase arrest and cell proliferation interruption12). Methylation in CpG islands around promoter regions can generally reduce gene expression30). Evidence that CDKN2B methylation represses expression and leads to unlimited cell proliferation has been confirmed in a spectrum of cancers21, 31).
Both inflammatory responses and migration of proliferating vascular smooth muscle cells (VSMCs) are considered essential for the development of atherosclerosis32). Chronic vascular inflammation arising from atherosclerosis also contributes to arterial calcification10). Under certain circumstances, a subpopulation of VSMCs may be predisposed to differentiate into osteoblastic and proliferative phenotypes. They can acquire osteoblast-like characteristics and become calcifying vascular cells, participating in spontaneous mineral deposition33–35). As the expression of CDKN2B is repressed, Rb proteins may lose control and result in increased proliferation of macrophages and VSMCs15, 19).
The association of CDKN2B methylation and coronary artery disease (CAD) has been previously observed20). Zhuang and colleagues found that the methylation levels of CDKN2B were significantly higher in CAD patients than in controls. Based on quantitative assessment of calcification, our study observed similar results in the aortic arch. Therefore, the higher the methylation level, the more serious the artery calcification might be.
Methylation levels of CDKN2B were not directly linked to genotypes of rs4977574, which was associated with AAC in previous and in our studies. It was possible that genetic variants directly contribute to ANRIL expression rather than to CDKN2B methylation according to evidence from previous studies12, 36). CDKN2B methylation was likely to be modulated by ANRIL or other epigenetic changes.
There are several limitations to our study. First, the nature of the cross-sectional study limited us to reach a causal inference. We cannot determine if the observed associations is attributed to methylation effects on AAC or vice versa. Second, CDKN2A/2B expression was not tested in this study due to lack of fresh leukocytes, which prevented us from evaluating the interactions between methylation variation and CDKN2A/2B gene expression. Therefore, future studies need to be conducted to provide more functional evidence. Third, considering varied predisposition of DNA methylation in different tissues, methylation measured from white blood cells may not represent that of vessel walls, although the role of white blood cells in atherogenesis is well-defined32). Because of the difficulty in obtaining vascular tissues from human body via invasive therapy, methylation tests from peripheral blood is still a convenient and rational method for investigation. In addition, a larger sample size is favorable for confirmation and more reliable results. The study was conducted in subjects with ischemic stroke, which may lead to selection bias as the prevalence of AAC was higher than that in the general population. Therefore, further exploration in the population with health controls is more convincible.
In conclusion, CDKN2B methylation is independently associated with AAC. Patients with higher methylation levels in CDKN2B may have increased risk for AAC. Further studies on the underlying mechanisms of this association are warranted to establishing a cause–effect relationship.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (NSFC, # 81571143 to GX, # 81530038 to XL and # 81220108008 to XL). We thank all patients for participating in this study. We are also grateful to Center for Genetics & Genomic Analysis, Genesky Biotechnologies Inc. (Shanghai, 201203) for their technical support in sequencing.
Conflict of Interest
None.
References
- 1). Atherosclerotic disease of the aortic arch as a risk factor for recurrent ischemic stroke The French Study of Aortic Plaques in Stroke Group, N Engl J Med, 1996; 334: 1216-1221 [DOI] [PubMed] [Google Scholar]
- 2). Iribarren C, Sidney S, Sternfeld B, Browner WS: Calcification of the aortic arch: risk factors and association with coronary heart disease, stroke, and peripheral vascular disease, Jama, 2000; 283: 2810-2815 [DOI] [PubMed] [Google Scholar]
- 3). Hyman JB, Epstein FH: A study of the correlation between roentgenographic and post-mortem calcification of the aorta, Am Heart J, 1954; 48: 540-543 [DOI] [PubMed] [Google Scholar]
- 4). Hashimoto H, Iijima K, Hashimoto M, Son BK, Ota H, Ogawa S, Eto M, Akishita M, Ouchi Y: Validity and usefulness of aortic arch calcification in chest X-ray, J Atheroscler Thromb, 2009; 16: 256-264 [DOI] [PubMed] [Google Scholar]
- 5). Bos D, Leening MJ, Kavousi M, Hofman A, Franco OH, van der Lugt A, Vernooij MW, Ikram MA: Comparison of Atherosclerotic Calcification in Major Vessel Beds on the Risk of All-Cause and Cause-Specific Mortality: The Rotterdam Study, Circ Cardiovasc Imaging, 2015; 8 [DOI] [PubMed] [Google Scholar]
- 6). Allison MA, Criqui MH, Wright CM: Patterns and risk factors for systemic calcified atherosclerosis, Arterioscler Thromb Vasc Biol, 2004; 24: 331-336 [DOI] [PubMed] [Google Scholar]
- 7). Li J, Galvin HK, Johnson SC, Langston CS, Sclamberg J, Preston CA: Aortic calcification on plain chest radiography increases risk for coronary artery disease, Chest, 2002; 121: 1468-1471 [DOI] [PubMed] [Google Scholar]
- 8). Iijima K, Hashimoto H, Hashimoto M, Son BK, Ota H, Ogawa S, Eto M, Akishita M, Ouchi Y: Aortic arch calcification detectable on chest X-ray is a strong independent predictor of cardiovascular events beyond traditional risk factors, Atherosclerosis, 2010; 210: 137-144 [DOI] [PubMed] [Google Scholar]
- 9). Rutsch F, Nitschke Y, Terkeltaub R: Genetics in arterial calcification: pieces of a puzzle and cogs in a wheel, Circ Res, 2011; 109: 578-592 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10). Demer LL, Tintut Y: Inflammatory, Metabolic, and Genetic Mechanisms of Vascular Calcification, Arterioscler Thromb Vasc Biol, 2014; 34: 715-723 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11). Ye S, Willeit J, Kronenberg F, Xu Q, Kiechl S: Association of genetic variation on chromosome 9p21 with susceptibility and progression of atherosclerosis: a populationbased, prospective study, J Am Coll Cardiol, 2008; 52: 378-384 [DOI] [PubMed] [Google Scholar]
- 12). Holdt LM, Teupser D: Recent studies of the human chromosome 9p21 locus, which is associated with atherosclerosis in human populations, Arterioscler Thromb Vasc Biol, 2012; 32: 196-206 [DOI] [PubMed] [Google Scholar]
- 13). O'Donnell CJ, Kavousi M, Smith AV, Kardia SL, Feitosa MF, Hwang SJ, Sun YV, Province MA, Aspelund T, Dehghan A, Hoffmann U, Bielak LF, Zhang Q, Eiriksdottir G, van Duijn CM, Fox CS, de Andrade M, Kraja AT, Sigurdsson S, Elias-Smale SE, Murabito JM, Launer LJ, van der Lugt A, Kathiresan S, Krestin GP, Herrington DM, Howard TD, Liu Y, Post W, Mitchell BD, O'Connell JR, Shen H, Shuldiner AR, Altshuler D, Elosua R, Salomaa V, Schwartz SM, Siscovick DS, Voight BF, Bis JC, Glazer NL, Psaty BM, Boerwinkle E, Heiss G, Blankenberg S, Zeller T, Wild PS, Schnabel RB, Schillert A, Ziegler A, Munzel TF, White CC, Rotter JI, Nalls M, Oudkerk M, Johnson AD, Newman AB, Uitterlinden AG, Massaro JM, Cunningham J, Harris TB, Hofman A, Peyser PA, Borecki IB, Cupples LA, Gudnason V, Witteman JC: Genome-wide association study for coronary artery calcification with follow-up in myocardial infarction, Circulation, 2011; 124: 2855-2864 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14). Visel A, Zhu Y, May D, Afzal V, Gong E, Attanasio C, Blow MJ, Cohen JC, Rubin EM, Pennacchio LA: Targeted deletion of the 9p21 non-coding coronary artery disease risk interval in mice, Nature, 2010; 464: 409-412 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15). Motterle A, Pu X, Wood H, Xiao Q, Gor S, Ng FL, Chan K, Cross F, Shohreh B, Poston RN, Tucker AT, Caulfield MJ, Ye S: Functional analyses of coronary artery disease associated variation on chromosome 9p21 in vascular smooth muscle cells, Hum Mol Genet, 2012; 21: 4021-4029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16). Zaina S: Unraveling the DNA methylome of atherosclerosis, Curr Opin Lipidol, 2014; 25: 148-153 [DOI] [PubMed] [Google Scholar]
- 17). Kotake Y, Nakagawa T, Kitagawa K, Suzuki S, Liu N, Kitagawa M, Xiong Y: Long non-coding RNA ANRIL is required for the PRC2 recruitment to and silencing of p15(INK4B) tumor suppressor gene, Oncogene, 2011; 30: 1956-1962 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18). Jarinova O, Stewart AF, Roberts R, Wells G, Lau P, Naing T, Buerki C, McLean BW, Cook RC, Parker JS, McPherson R: Functional analysis of the chromosome 9p21. 3 coronary artery disease risk locus, Arterioscler Thromb Vasc Biol, 2009; 29: 1671-1677 [DOI] [PubMed] [Google Scholar]
- 19). Hannou SA, Wouters K, Paumelle R, Staels B: Functional genomics of the CDKN2A/B locus in cardiovascular and metabolic disease: what have we learned from GWASs?, Trends Endocrinol Metab, 2015; 26: 176-184 [DOI] [PubMed] [Google Scholar]
- 20). Zhuang J, Peng W, Li H, Wang W, Wei Y, Li W, Xu Y: Methylation of p15INK4b and expression of ANRIL on chromosome 9p21 are associated with coronary artery disease, PloS one, 2012; 7: e47193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21). Esteller M: CpG island hypermethylation and tumor suppressor genes: a booming present, a brighter future, Oncogene, 2002; 21: 5427-5440 [DOI] [PubMed] [Google Scholar]
- 22). Liu X, Xu G, Wu W, Zhang R, Yin Q, Zhu W: Subtypes and one-year survival of first-ever stroke in Chinese patients: The Nanjing Stroke Registry, Cerebrovasc Dis, 2006; 22: 130-136 [DOI] [PubMed] [Google Scholar]
- 23). Lu L, Zhang LJ, Poon CS, Wu SY, Zhou CS, Luo S, Wang M, Lu GM: Digital subtraction CT angiography for detection of intracranial aneurysms: comparison with three-dimensional digital subtraction angiography, Radiology, 2012; 262: 605-612 [DOI] [PubMed] [Google Scholar]
- 24). Pham PH, Rao DS, Vasunilashorn F, Fishbein MC, Goldin JG: Computed tomography calcium quantification as a measure of atherosclerotic plaque morphology and stability, Invest Radiol, 2006; 41: 674-680 [DOI] [PubMed] [Google Scholar]
- 25). van Setten J, Isgum I, Smolonska J, Ripke S, de Jong PA, Oudkerk M, de Koning H, Lammers JW, Zanen P, Groen HJ, Boezen HM, Postma DS, Wijmenga C, Viergever MA, Mali WP, de Bakker PI: Genome-wide association study of coronary and aortic calcification implicates risk loci for coronary artery disease and myocardial infarction, Atherosclerosis, 2013; 228: 400-405 [DOI] [PubMed] [Google Scholar]
- 26). Xiao Z, Xiao J, Jiang Y, Zhang S, Yu M, Zhao J, Wei D, Cao H: A novel method based on ligase detection reaction for low abundant YIDD mutants detection in hepatitis B virus, Hepatol Res, 2006; 34: 150-155 [DOI] [PubMed] [Google Scholar]
- 27). Gardiner-Garden M, Frommer M: CpG islands in vertebrate genomes, J Mol Biol, 1987; 196: 261-282 [DOI] [PubMed] [Google Scholar]
- 28). Masser DR, Berg AS, Freeman WM: Focused, high accuracy 5-methylcytosine quantitation with base resolution by benchtop next-generation sequencing, Epigenetics Chromatin, 2013; 6: 33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29). Gil J, Peters G: Regulation of the INK4b-ARF-INK4a tumour suppressor locus: all for one or one for all, Nat Rev Mol Cell Biol, 2006; 7: 667-677 [DOI] [PubMed] [Google Scholar]
- 30). Portela A, Esteller M: Epigenetic modifications and human disease, Nat Biotechnol, 2010; 28: 1057-1068 [DOI] [PubMed] [Google Scholar]
- 31). Herman JG, Baylin SB: Gene silencing in cancer in association with promoter hypermethylation, N Engl J Med, 2003; 349: 2042-2054 [DOI] [PubMed] [Google Scholar]
- 32). Wierda RJ, Geutskens SB, Jukema JW, Quax Paul HA, van den Elsen PJ: Epigenetics in atherosclerosis and inflammation, J Cell Mol Med, 2010; 14: 1225-1240 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33). Doherty TM, Asotra K, Fitzpatrick LA, Qiao JH, Wilkin DJ, Detrano RC, Dunstan CR, Shah PK, Rajavashisth TB: Calcification in atherosclerosis: bone biology and chronic inflammation at the arterial crossroads, Proc Natl Acad Sci U S A, 2003; 100: 11201-11206 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34). Johnson RC, Leopold JA, Loscalzo J: Vascular calcification: pathobiological mechanisms and clinical implications, Circ Res, 2006; 99: 1044-1059 [DOI] [PubMed] [Google Scholar]
- 35). Zhu D, Mackenzie NC, Farquharson C, Macrae VE: Mechanisms and clinical consequences of vascular calcification, Front Endocrinol (Lausanne), 2012; 3: 95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36). Holdt LM, Beutner F, Scholz M, Gielen S, Gäbel G, Bergert H, Schuler G, Thiery J, Teupser D: ANRIL expression is associated with atherosclerosis risk at chromosome 9p21, Arterioscler Thromb Vasc Biol, 2010; 30: 620-627 [DOI] [PubMed] [Google Scholar]