Background

Stroke is one of the leading causes of death and disability [1], with ischemic stroke (cerebral infarction) accounting for 60 to 80% of all strokes [2]. Cerebral infarction is a complex disease that is affected by both environmental risk factors and genetic factors [3, 4]. Common risk factors for cerebral infarction include age, hypertension, diabetes, hyperlipidemia, smoking, and atrial fibrillation, etc. [5]. DNA methylation is an epigenetic modification that links the roles between heredity and the environment [6]. The mechanism depends on the catalysis of DNA methyltransferase (DNMT), which in turn causes the transfer of methyl groups from s-adenosylmethionine (SAM) to specific segments in the DNA sequence, thereby changing the function of the genome [7]. Therefore, the level of SAM is closely related to DNA methylation. Xiao et al. [8] showed that PON1 and PON3 methylation were associated with the risk of male cerebral infarction. There are several DNA methylation studies related to stroke [9]. High homocysteine (Hcy) leads to hypermethylation of the TM gene and further induction of TM gene silencing, which may play an important role in the development and progression of ischemic brain injury [10]. DNA methylation of CDKN2B may play a potential role in arterial calcification in patients with cerebral infarction [11]. People with a lower degree of LINE-1 methylation have a higher risk of stroke [12]. DNA methylation inhibitor Zebularine confers neuroprotective effects on the ischemic rat brain and further supports the hypothesis that DNA methyltransferase promotes delayed ischemic brain injury [13].

In the normal fasting state, the human plasma Hcy concentration is 5–15 μmol. Hcy, as a sulfur-containing amino acid in the human body, can be summarized as: (1) further metabolized to cystathionine and cysteine; (2) re-converted to methionine; (3) as a -Based receptors participate in choline metabolism; (4) participate in the folic acid cycle as an essential substrate. Wald et al. have shown that a 5 μmol/L increase in plasma Hcy leads to a 59% increased risk of cerebral infarction; a decrease in Hcy of 3 μmol/L reduces the risk of cerebral infarction by 24% [14]. In 2011, the American Heart Association and the Association for Cerebral Infarction jointly issued a primary prevention guideline for cerebral infarction, which increases the risk of ischemic cerebral infarction by two to three times as plasma Hcy levels increase [15]. In addition, a meta-analysis of a prospective cohort study found that after a mean follow-up of 7.3 years, serum HHcy was reduced by 3 μmol/L, and stroke risk was reduced by 24% [16]. Ashjazadeh et al. found high serum Hcy associated with stroke in the Iranian population [17]. Hankey et al. found elevated serum Hcy in stroke patients in Australia as an independent risk factor for ischemic stroke [18]. Hyperhomocysteinemia (HHcy) is commonly recognized as an independent risk factor for stroke [19, 20].

AHCY encodes S-adenosine homocysteine hydrolase (SAH), whose function is to catalyze the reversible hydrolysis of SAM to produce adenosine (Ado) and Hcy (Fig. 1). SAH may control the number of methylated donors by regulating the level of intracellular SAM [21]. Plasma Hcy concentrations may increase due to genetic defects in related enzymes [22, 23]. AHCY methylation may affect the activity of the enzyme, causing a metabolic disorder of Hcy, leading to HHcy. However, the association between AHCY methylation and cerebral infarction has not previously been studied.

Fig. 1
figure 1

Hcy metabolic pathway. SAM: S-adenosylmethionine; SAH: S-adenosine homocysteine; Hcy: homocysteine; CBS: cystathionine beta synthase

Full size image

In addition, CBS-encoded cystathionine beta-synthase (CBS) catalyzes the sulfur metabolism of Hcy (Fig. 1). In vivo, Hcy can be irreversibly catalyzed by CBS to produce cysteine (Fig. 1). At present, many studies have explored the relationship between CBS mutations and stroke, of which T833C point mutation is the most common [24,25,26,27]. This mutation replaces threonine with isoleucine, resulting in the inactivation of CBS, blocking Hcy degradation, and ultimately leading to abnormal accumulation of Hcy (HHcy) in the body [28].

The association between AHCY and CBS promoter methylation and stroke studies has not been reported so far. In this study, we examined the levels of AHCY and CBS promoter methylation in peripheral blood from patients with cerebral infarction and healthy controls. Our goal is to assess whether promoter methylation of AHCY and CBS can be used as a diagnostic biomarker for the risk of cerebral infarction.

Methods

Sample collection

From October 2016 to April 2017, we selected 152 hospitalized patients with cerebral infarction from the 960th Hospital of the People’s Liberation Army. Patients were tested by brain CT or MRI, in strict accordance with the diagnostic criteria of the Fourth National Cerebrovascular Disease Conference [29]. The patients included 112 males and 40 females with an age range of 23 to 88 years and an average age of 60.37 ± 12.02 years. Then, we selected 152 healthy people as controls, including 112 males and 40 females with an age range of 30 to 90 years and an average age of 60.45 ± 12.23 years. All selected individuals excluded serious diseases, including malignant tumors, neurodegenerative diseases, peripheral vascular disease or peripheral vascular thrombosis, blood diseases, severe liver and kidney diseases, endocrine and metabolic diseases, tuberculosis, or connective tissue diseases. We extracted clinical information from participants from the medical records. We used the American Beckman AU5800 automatic biochemical analyzer and its supporting kit to detect the Hcy level of blood. Our research was approved by the hospital’s ethics committee. In addition, all subjects provided informed consent.

DNA extraction and methylation detection

We extracted peripheral blood DNA using a DNA extraction kit (ZhiShan Biotech, Xiamen, China). We then used the EZ DNA Methylation-Gold kit (Zymo Research, USA) for bisulfite modification of peripheral blood DNA [30]. Among them, methylated cytosine remains intact, and all unmethylated cytosines are converted to uracil. Uracil will be expressed as thymine in the subsequent qMSP reaction.

The fragments we detected were located on the CpG island (CGI) of the promoter regions of the two genes, respectively (Fig. 2, AHCY, chr20:34303211–34,303,334; Fig. 3, CBS, chr21:43076986–43,077,078). The relevant qMSP primer sequences are listed in Table 1. The mixture components of the relevant reagents of qMSP are as previously described [31]. qMSP was performed using a 384-well transparent plate in a LightCycler 480 (Roche, Basel, Switzerland). Our qMSP chose ACTB as an internal reference to correct for differences in the loading samples. We used fully methylated human sperm DNA as a positive control and ddH2O as a blank control. Choose a 384-well plate for loading, add the sealing film after adding the sample, centrifuge in a centrifuge to remove air bubbles in the 384-well plate, and then place the 384-well plate in Roche 480 quantitative PCR instrument for measurement. The specific procedures of PCR are as follows: ① Denaturation: 95 °C, 10 min; ② Cycle (45×): 95 °C, 20 s; Annealing temperature, 20 s; 72 °C, 30 s; ③ Melting curve: 95 °C, 15 s; 60 °C, 15 s; 95 °C, continuous; ④ cooling: 40 °C, 10 min; Acquisition 1.5/°C. After the above reaction is completed, the PCR product is placed in a BiOptic Qseq100 automatic nucleic acid analyzer for analysis of the results, and the DNA methylation detection result obtained is expressed by a methylation percentage parameter (PMR). The PMR value of gene was calculated by this equation [PMRgene = 2-△△Ct × 100%, △△Ct = sample DNA (Ct gene– Ct ACTB) - fully methylated DNA (Ctgene – Ct ACTB)]. [PMR CBS = 2-△△Ct × 100%, △△Ct = sample DNA (Ctgene – Ct ACTB) - fully methylated DNA (Ctgene – Ct ACTB)].

Fig. 2
figure 2

The characteristics of AHCY fragment in the qMSP assay. The qMSP primers were underlined and two CpG sites were in grey. F: forward primer; R: reverse primer

Full size image
Fig. 3
figure 3

The characteristics of CBS fragment in the qMSP assay. The qMSP primers were underlined and two CpG sites were in grey. F: forward primer; R: reverse primer

Full size image
Table 1 Primer sequences of the qMSP assays
Full size table

Statistical analysis

We analyzed data using SPSS 17.0 software (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 5.0 software (GraphPad Software, La Jolla, CA). We described data as median (interquartile range) or mean ± standard deviation. We used a nonparametric Mann-Whitney U test or an independent sample t-test to assess differences in DNA methylation levels between groups. We applied chi-square tests or Fisher’s exact tests to analyze the count data, and we calculated the correlation between DNA methylation levels and clinical data by Spearman rank correlation analysis. We used SPSS 17.0 software for receiver operating characteristic (ROC) curve analysis. In the ROC curve, the true positive rate (sensitivity) is plotted on the ordinate, and the false positive rate (1-specificity) is plotted on the abscissa. The best sensitivity and specificity in the ROC curve was found to show the diagnostic value of gene promoter methylation for cerebral infarction. Bilateral P < 0.05 was defined as statistically significant.

Results

As shown in Table 2, our case group and control group were gender- and age-matched (P > 0.05). In addition, hyperlipidemia and GLU in the case group were significantly higher than those in the control group (P < 0.05). The levels of TC, HDL-C, and LDL-C in the case group were lower than those in the control group (P < 0.05). There was no significant difference in HHcy, Hcy, TG between the case group and the control group (P > 0.05).

Table 2 Comparison of clinical data in the case group and the control group
Full size table

In this study, we examined the relative methylation levels of the AHCY and CBS genes in 152 patients with cerebral infarction and 152 healthy controls. We found that the PMR of the AHCY promoter region was not statistically different between the case group and the control group (P = 0.115, Fig. 4a). However, the CBS methylation level in the case group was significantly lower than that in the control group (median PMR: 70.20% vs. 104.10%, P = 3.71E-10, Fig. 5a).

Fig. 4
figure 4

Comparison of AHCY methylation levels between control and case groups. a: Methylation level of AHCY gene in control group and case group, scatter point indicates the methylation level of experimental subjects. b: Methylation level of AHCY gene in males between control group and case group, scatter point indicates the methylation level of experimental subjects. c: Methylation level of AHCY gene in females between control group and case group, scatter point indicates the methylation level of experimental subjects. d: Methylation level of AHCY gene in the young group between control group and case group, scatter point indicates the methylation level of experimental subjects. e: Methylation level of AHCY gene in the middle-aged group between control group and case group, scatter point indicates the methylation level of experimental subjects. f: Methylation level of AHCY gene in the elderly group between control group and case group, scatter point indicates the methylation level of experimental subjects

Full size image
Fig. 5
figure 5

Comparison of CBS methylation levels between control and case groups. a: Methylation level of CBS gene in control group and case group, scatter point indicates the methylation level of experimental subjects. b: Methylation level of CBS gene in males between control group and case group, scatter point indicates the methylation level of experimental subjects. c: Methylation level of CBS gene in females between control group and case group, scatter point indicates the methylation level of experimental subjects. d: Methylation level of CBS gene in the young group between control group and case group, scatter point indicates the methylation level of experimental subjects. e: Methylation level of CBS gene in the middle-aged group between control group and case group, scatter point indicates the methylation level of experimental subjects. f: Methylation level of CBS gene in the elderly group between control group and case group, scatter point indicates the methylation level of experimental subjects

Full size image

Subsequently, we also conducted a subgroup analysis of gender and age stratification. In our male cases, the PMR of AHCY was lower than that of the control group (0.14 vs. 0.26, P = 0.045, Fig. 4b, Table 3), while the PMR of AHCY in women was not significantly different between the two groups (P > 0.05, Fig. 4c Table 3). Interestingly, the PMR of AHCY was higher in the middle-aged case group than in the middle-aged control group (0.14 vs. 0.03, P = 0.016, Fig. 4e, Table 3), while the PMR in AHCY was lower in the elderly case group than in the elderly control group (0.18 vs. 0.37, P = 4.05E-04, Fig. 4f, Table 3).

Table 3 Subgroup analysis of AHCY and CBS methylation levels by gender or age
Full size table

As shown in Table 4, we performed a breakdown analysis of AHCY methylation and stroke by both age and gender. We found that AHCY hypomethylation in the elderly group (age > 59) was associated with stroke risk in both men and women, whereas in middle-aged (aged between 45 and 59) women, AHCY hypermethylation was associated with stroke risk. This opposite result suggests that the risk of AHCY methylation and stroke in women needs further analysis in the future.

Table 4 Subgroup analysis of AHCY methylation levels with stoke by both gender and age
Full size table

At the same time, we found that the PMR levels of the CBS gene were significantly lower in the stratified analysis of various ages and genders (P < 0.01, Fig. 5, Table 3).

Therefore, we continue to conduct a ROC curve analysis of CBS methylation. We found that the area under the ROC curve (AUC) was 0.713 (95% CI: 0.652–0.773), the sensitivity was 67.4%, and the specificity was 74.0% (Fig. 6). This indicated that CBS hypomethylation might be used as a potential biomarker for the diagnosis of cerebral infarction.

Fig. 6
figure 6

ROC curve analysis of CBS methylation in the cerebral infarction patients. a: ROC curve of methylation of the CBS gene. AUC is 0.713. b: ROC curve of methylation of the CBS gene in males between the control group and the case group. AUC is 0.709. c: ROC curve of methylation of the CBS gene in females between the control group and the case group. AUC is 0.722. d: ROC curve of methylation of the CBS gene in the young group between the control group and the case group. AUC is 0.814. e: ROC curve of methylation of the CBS gene in the middle-aged group between the control group and the case group. AUC is 0.736. f: ROC curve of methylation of the CBS gene in the elderly group between the control group and the case group. AUC is 0.681

Full size image

Discussion

In this study, we first analyzed the general data between the case group and the control group and found that the two groups matched in gender and age. The prevalence of Hhcy and the level of Hcy were not different between the cerebral infarction group and the healthy control group, and it was inconsistent with the results of previous studies [32, 33]. Analysis of the reasons for no significant difference in Hhcy prevalence and Hcy level between the two groups may be related to the use of folic acid, vitamin B12, and vitamin B6 in the case group for drug intervention, insufficient sample size, and lack of sample information. The hyperlipidemia and GLU in the case group were higher than those in the control group, which was consistent with previous studies. The TC, HDL-C, and LDL-C levels in the case group were lower than those in the control group, which was considered to be related to the patients’ oral statins.

In this study, we analyzed the methylation levels of the AHCY and CBS promoters between the case and control groups. Our results indicated that hypomethylation in the CBS promoter region was significantly associated with cerebral infarction and might serve as a potential biomarker for the diagnosis of cerebral infarction. In addition, our results showed that differences in the methylation of the AHCY gene between the case group and the control group were related to gender and age. Specifically, the AHCY methylation level in male patients was lower than that in the control group. To analyze the correlation between AHCY promoter methylation and male patients with cerebral infarction. The AHCY methylation level in middle-aged patients was significantly higher than that in the control group, while the AHCY methylation level in elderly patients was lower than that in the control group. Fuso A et al. [34,35,36] found that the degree of bias in DNA methylation is related to many factors. When the degree of DNA methylation gradually increases, the degree of DNA methylation bias also changes, and this is mainly related to non-CpG methylation in the genome. In the current study, the correlation between the AHCY gene methylation level and ischemic stroke is inconsistent in the middle-aged and elderly groups. The bias between the middle-aged and elderly groups needs to be further explored in the future.

CBS catalyzes the irreversible synthesis of cystathionine and water from Hcy, which then produces cysteine and alpha-ketoacetate catalyzed by gamma-cystatin. We hypothesized that abnormal DNA methylation may affect enzyme activity or gene expression, which in turn leads to disorder of Hcy metabolism, but due to conditions and other limitations, it could not be further verified. To the best of our knowledge, no one has assessed the relationship between CBS methylation and cerebral infarction. Our results showed that the CBS methylation level in the case group was significantly lower than that in the control group (Fig. 5). Usually, insufficient methylation will lead to overexpression of genes, but our research shows that the CBS gene in the case group is in a state of insufficient methylation, there is no functional mechanism linking the CBS gene hypomethylation to cerebral infarction. Further research is needed to prove the effect of general methylation regulation on CBS enzyme activity and/or DNA methylation regulation on CBS gene expression impact.

Our ROC curve analysis showed CBS hypomethylation has an AUC of 0.713, a sensitivity of 67.4%, and a specificity of 74.0% (Fig. 6). Specifically, we found that 73.0% (111/152) of the 152 patients with cerebral infarction had hypomethylation of the CBS gene promoter region in contrast to only 30.0% (46/152) of the 152 healthy controls. Therefore, methylation of the CBS promoter region can be used as a potential biomarker for diagnosing cerebral infarction.

There are some limitations in this study. First, all subjects in this study were from the same region, and the sample size was not large. Therefore, the application of our results should be validated in other populations. Second, due to the limitations of the sample, we were unable to further detect CBS protein and mRNA expression. Finally, we only studied a limited area on the CGI of the two genes, and other areas of the genes for methylation need further study.

Conclusions

In summary, our study found that hypomethylation of the CBS promoter region is a risk factor for cerebral infarction and it can serve as a potential biomarker for the diagnosis of cerebral infarction. There is an age and gender difference in the association between methylation of AHCY and cerebral infarction, and further research is needed in the future.