Association Between Serine Concentration and Coronary Heart Disease: A Case–Control Study

Introduction

Cardiovascular disease (CVD), as the leading cause of global mortality, accounts for 31% of deaths worldwide. In 2015, CVD accounted for approximately 17.7 million deaths, with coronary heart disease (CHD) being the most common type (7.4 million deaths).^1,2 Although the risk of CHD morbidity and mortality can be reduced by intervention targeted toward traditional risk factors, such as smoking, hypercholesterolemia, hypertension, and diabetes mellitus, residual risks still exist.³ Therefore, the early identification of new risk factors is an urgent problem to be solved.

The key metabolites and associated gene polymorphisms of enzymes involved in one-carbon metabolism, such as homocysteine (Hcy) and methylenetetrahydrofolate reductase (MTHFR), are associated with CHD risk.^4,5 Serine is a non-essential amino acid and an essential metabolite in one-carbon metabolism in the human body. Serine participates in the folic acid cycle and the trans-sulfuration pathway of Hcy. In the folic acid cycle, using pyridoxal phosphate (PLP) as the coenzyme, serine hydroxymethyltransferase (SHMT) catalyzes the interaction between tetrahydrofolate and serine to produce 5,10-methylene tetrahydrofolate (5,10-CH₂-THF) and glycine.⁶ Meanwhile, under the catalysis of PLP-dependent cysteine β-synthase (CBS), Hcy interacts with serine to produce cystathionine, which is transformed into cysteine through trans-sulfuration and finally converted to glutathione.⁷

Serine accelerates the growth of various tumors because its synthesis is linked to tumor cell proliferation.⁸ Serine is also a useful biomarker and potential therapeutic target in senile dementia, as well as schizophrenia, depression, and other mental health conditions.^9–11 In addition, serine synthesis can delay vascular endothelial cell senescence.¹² Serine-related metabolism is also related to atherosclerotic lesions in patients with chronic thromboembolic pulmonary hypertension.¹³ However, few studies have studied the relationship between serine and CHD.

On the basis of the important role of serine in folic acid and Hcy metabolism, as well as its possible regulation of oxidative stress upstream of glutathione production, we speculated that serine may be involved in the formation of coronary atherosclerosis and CHD. The purpose of this case–control study was to preliminarily explore the relationship between serine and coronary atherosclerosis severity. This research provides evidence that serine can be considered as a potential biomarker for CHD.

Materials and Methods

Participants

The flowchart of this case–control study is shown in Figure 1. The participants were hospitalized patients who underwent coronary angiography at the catheterization laboratory of the Department of Cardiology, Peking University First Hospital, from January 1, 2016, to December 31, 2019. Patients whose coronary angiography indicated that the coronary artery stenosis degree (anterior descending branch, circumflex branch, right coronary artery, and any one of the left main arteries) was >70% were enrolled in the case group. Individuals with the degree of coronary artery stenosis less than 30% and no CHD diagnosis upon discharge were included in the control group. The study exclusion criteria were 1) no signed informed consent for biological sample collection; 2) blood samples were not collected or retained; 3) a previous diagnosis of CHD through coronary angiography; 4) prior coronary intervention therapy or coronary artery bypass grafting; and 5) an admission diagnosis of acute myocardial infarction or myocardial injury biomarker elevation greater than the normal upper limit before coronary angiography.

Figure 1 Study flowchart. Abbreviation: CHD, coronary heart disease.

A total of 447 individuals met the criteria for inclusion in the control group and were matched 1:1 with patients with CHD according to the following conditions: 1) identical sex; 2) age range within 2 years; 3) coronary angiography date within 180 days of each other. Finally, 429 case–control pairs were enrolled. Because the serine concentration was too low to be measured in one subject, one case–control pair was further excluded, leading to the inclusion of 428 case–control pairs. This study was approved by the Ethics Committee of Peking University First Hospital, and all of the research processes were conducted in strict accordance with the Declaration of Helsinki.

Data Collection

The following information was collected from the participants’ medical records: general information, including sex, age, body mass index (BMI), and blood pressure at hospital admission; medical history, such as history of hypertension, diabetes mellitus, dyslipidemia, depression, and tumor; lifestyle factors, such as the status of smoking and alcohol consumption; laboratory test results, including routine blood test and biochemical index results; and coronary angiography results.

Hypertension, diabetes mellitus, dyslipidemia, depression, and tumors were all determined from the discharge diagnosis. Smoking was defined as continuous smoking for more than half a year with smoking ≥1 cigarette per day. Quitting smoking was defined as being a former smoker who had not smoked for more than 6 months. Alcohol consumption was defined as drinking at least once per week for more than half a year. Abstinence was defined as abstaining from alcohol for more than 6 months. Drug history was based on admission records. Routine blood and biochemical index results were obtained from the first laboratory test after admission before coronary angiography. Coronary artery stenosis was calculated as the diameter stenosis rather than the area stenosis.

Serine Examination

With the informed consent of each patient, 10 mL blood was collected from the median cubital vein before coronary angiography and loaded into ethylenediaminetetraacetic acid anticoagulant blood tubes (Solebo Technology Co., Ltd., Beijing, China). All samples were immediately sent to the cardiology laboratory and centrifuged at 3000 rpm for 5 minutes (Xiangyi Laboratory Instrument Development Co., Ltd., Hunan, China). After centrifugation, the upper layer of clear plasma was absorbed, placed into an EP tube (1.5 mL) and stored at −80°C for later use.

Serine was identified by liquid chromatography–tandem mass spectrometry (LC–MS/MS) at Shenzhen Tailored Medical Laboratory. The specific method of LC–MS/MS was as follows. The plasma protein precipitation was pre-treated, and serine-d3 was used as the internal standard (standard solution was from Inorganic Ventures, VA, US). The Waters ACQUITY UPLC^® BEH HILIC (2.1 × 100 mm, 1.7 µm) chromatographic column (US Waters Company) was used, with 0.5% acetic acid water (containing 10 mmoL ammonium acetate) and 95% acetonitrile water (containing 0.5% acetic acid, 10 mmoL ammonium acetate) used as the mobile phase for gradient elution (for each 10 µL injection, the elution flow rate was set to 0.4 mL/min). An electrospray ion source, positive ion ionization mode, and the multi-reactive ion monitoring mode were adopted for mass spectrometry detection. The corresponding internal standard ion pair m/z of serine was 106.1→ 60.1.

Statistical Analysis

The data are expressed as the mean ± standard deviation for normally distributed continuous variables, or as the median (interquartile range) for non-normally distributed continuous variables. The Student’s t-test or Mann–Whitney U-test was used to identify differences between the cases and controls. Categorical data are expressed as number (percentage), and differences between the two groups were identified using the Chi-square test.

First, a smooth curve (restricted cubic spline) was used to explore the dose–effect relationship between serine and CHD risk. Then, conditional logistic regression was applied to analyze the association of serine (as a continuous variable, as quartiles, and as a dichotomic variable divided by cut-off point) with CHD. Adjusted variables included BMI, systolic blood pressure (SBP), fasting plasma glucose (FPG), low-density lipoprotein cholesterol (LDL-C), plasma creatinine, smoking status, drinking status, hypertension, diabetes mellitus, hyperlipidemia, anti-hypertensive treatment, hypoglycemic treatment, and lipid-lowering treatment. Finally, because of the different characteristics of the above covariates, further subgroup and interaction analyses were performed to explore the differences in the relationship between serine and CHD in the different subgroups. Statistical analyses were two-sided, and P < 0.05 was considered statistically significant. All analyses were performed using EmpowerStats software (Version 2.0, http://www.empowerstats.com) and R software (Version 4.3.1, http://www.R-project.org).

Results

Baseline Characteristics of All Participants

In this study, 856 subjects were enrolled, including 428 cases and 428 controls (Table 1). The mean age of the cases and controls was 63.89 ± 10.49 years and 63.06 ± 10.27 years, respectively. Participants in the case group had a lower LDL-C; higher FPG and creatinine levels; and a higher prevalence of current smoking, hypertension, diabetes mellitus, hyperlipidemia, hypoglycemic, and lipid-lowering medications. For the other baseline characteristics, there was no significant difference between cases and controls. Table S1 shows that no significant differences were found among the participants in the four serine concentration quartiles, except for SBP and drinking status.

Table 1 Characteristics of the Cases and Controls

Relationships Between Serine and CHD

The restricted cubic spline demonstrated a linear negative relationship between serine and CHD risk (Figure 2). In the univariable conditional logistic regression analysis, CHD risk showed a downward trend with an increase in the serine concentration. After adjusting for other confounding covariates, for every 1 μg/mL increase in serine concentration, the risk of CHD significantly decreased by 6% (95% CI 0.90–0.99; P = 0.010) (Table 2). Furthermore, the risk of CHD in the third and fourth serine concentration quartiles was reduced by 45% and 46% (P_trend = 0.008) compared with the first quartile. Compared with subjects with a serine concentration of <13.41 μg/mL, those with a serine concentration of ≥13.41 μg/mL had a 43% lower risk of CHD (OR 0.57, 95% CI 0.39–0.84; P = 0.004).

Table 2 Association of the Serine Concentration with CHD Risk

Figure 2 Smooth curve of the association between the serine concentration and CHD risk. The restricted cubic spline was adjusted for sex, age, BMI, SBP, FPG, LDL-C, Crea, smoking status, drinking status, hypertension, diabetes mellitus, hyperlipidemia, anti-hypertensive treatment, hypoglycemic treatment, and lipid-lowering treatment. Abbreviations: BMI, body mass index; CHD, coronary heart disease; Crea, plasma creatinine; FPG, fasting plasma glucose; LDL-C, low-density lipoprotein cholesterol; OR, odds ratio; SBP, systolic blood pressure.

Subgroup and Interaction Analyses

The subgroup analyses based on different sexes, age groups, BMI values, creatinine concentrations, smoking and drinking statuses, hypertension, diabetes mellitus, and hyperlipidemia are shown in Table 3. Sex modified the relationship between serine concentration and CHD, which was more significant (OR 0.93, 95% CI 0.87–0.98; P = 0.013) in males, while no correlation was seen in females (P_interaction = 0.039). The interaction tests of the other covariates were not significant. In terms of the characteristics of males and females, Table S2 shows that compared with females, males were younger; had lower SBP, LDL-C, and FPG values; and had a lower prevalence of hypertension, diabetes mellitus, and all three types of medication. Moreover, males more frequently smoked and consumed alcohol, and they had a higher creatinine concentration.

Table 3 Subgroup and Interaction Analyses for the Association Between the Serine Concentration and CHD

Although the subgroup and interaction analyses based on smoking status (never, ever, and current) were not statistically significant, a non-significant reverse association between the serine concentration and CHD was observed in ever and current smokers, but not in never smokers. Further subgroup analyses (Table S3) identified similar results, especially when combining ever and current smokers as one group and dividing the serine concentration into quartiles, where the P_interaction value changed to 0.023, suggesting that the protection effect of serine may mainly exist in ever/current smokers.

Sensitivity Analyses

A positive correlation between serine and glycine (r = 0.70) was found with regard to the bidirectional transformation between serine and glycine. We performed sensitivity analyses adjusted for the glycine concentration (Table S4), and the results showed that the relationship between serine and CHD remained.

We included Hcy; other related nutrients, including methionine, vitamin B6, and B12; and white blood cell count into the multivariable-adjusted model and found that the main results did not change substantially (Table S5).

A history of depression (n = 10) and tumor (n = 96) was further adjusted in the multivariable conditional logistic regression analyses (Table S6). The main results did not change substantially. Further multivariable logistic regression analyses excluding participants with a history of depression and tumor produced similar results (not shown).

Discussion

To our knowledge, this is the first case–control study to observe a negative correlation between the serine concentration and CHD in Chinese individuals. Individuals with a serine concentration of ≥13.41 μg/mL had a significantly reduced risk of CHD (43%) compared to those with a serine concentration of <13.41 μg/mL, suggesting that serine may be one of the protective factors for coronary artery atherosclerosis.

Recently, a cohort study investigated the relationship between CVD onset and plasma-free amino acid profiles in the general Japanese population, observing that elevated serine concentration was significantly associated with a decreased CVD risk.¹⁴ Another trans-ancestry Mendelian randomization (MR) analysis investigated the association between 20 types of circulating amino acid and CVD.¹⁵ Consistent with our study, per 1 unit increase in serine, the risk of CHD decreased by 12.6% (OR 0.874, 95% CI 0.836–0.914; P_IVW-MRE = 3.37 × 10⁻⁹) in the East Asian population, which indicated the potential of serine as a biomarker or therapeutic target for CHD in clinical scenarios. However, the specific mechanism underpinning the possible protective effect of serine against CHD remains unclear. We speculate that the following mechanisms may explain this phenomenon.

First, one way by which serine metabolism promotes cancer cell growth is by controlling the antioxidant and methylation abilities of cells,¹⁶ which is speculated to be involved in the different effects of serine on tumors and CHD. Oxidative stress plays an important role in atherosclerosis, vascular damage, and CHD.^17–20 The metabolites of serine metabolism participate in various reactions that regulate oxidative stress. For example, tetrahydrobiopterin, participating in nitric oxide generation, is involved in serine metabolism.²¹ Nitric oxide has vasodilatory, antiplatelet, anti-proliferative, and anti-inflammatory effects.²² Furthermore, serine and Hcy can be transformed into glutathione, a crucial antioxidant cofactor.²³ Insufficient serine may inhibit the trans-sulfuration pathway, reduce glutathione production, and consequently decrease antioxidant power.²⁴ Therefore, we hypothesized that the antioxidant effect of serine metabolism might protect against CHD due to its role in one-carbon metabolism and glutathione generation.

In addition, some genetic mutations in serine metabolism have also demonstrated associations with CHD. The gene mutation SHMT C1420T (rs1979277) decreases SHMT activity, resulting in reduced serine metabolism.²⁵ Previous studies have found that CHD cases compared with control cases had a lower proportion of individuals with SHMT C1420T mutations and higher levels of oxidative stress markers.²⁶ Epigenomic studies have also shown that mutations such as SHMT C1420T are associated with the expression of genes related to oxidative stress, thereby influencing CHD susceptibility.²⁷ Interestingly, the SHMT 1420TT genotype was found to protect against CHD risk in a risk prediction model.²⁸ However, the frequency of SHMT C1420T gene polymorphism was not available in this study, and thus its relationship with serine and CHD could not be analyzed.

Serine and glycine can be transformed into each other,²⁹ and glycine may have a protective effect on the cardiovascular system. A cohort study in northern Europe found that glycine was inversely associated with myocardial infarction risk in angina patients.³⁰ A genome-wide association study demonstrated that genetically determined serum glycine protected against CHD in the Singaporean Chinese population,³¹ consistent with findings from two other MR studies.^15,32 In addition, glycine supplementation could mitigate atherosclerosis development, reduce tissue injury, and treat metabolic disorders in multiple animal models.^33–35 Our data also showed a positive correlation between serine and glycine (r = 0.70, 95% CI 0.66–0.73; P < 0.001). However, the supplementary analysis (Table S3) showed that the serine-CHD relationship remained after further adjustment for the glycine concentration, suggesting that the protective effect of serine on CHD was not entirely dependent on glycine.

We also found that sex and smoking modified the relationship between serine and CHD, and their inverse relationship was mainly observed in males and ever/current smokers. However, the specific mechanism is unknown and needs to be further explored. Oxidative stress is speculated to play a critical role in sex differences, smoking, and serine concentration. The prevalence of CHD was higher in males than in females across all age groups.³⁶ Possible reasons for this include higher oxidative stress and lower antioxidant potential in males, making them more susceptible to oxidative stress.³⁷ Moreover, smoking has been proven to increase inflammation, endothelial dysfunction, and oxidative stress to initiate cardiovascular dysfunction.¹⁸ Conversely, serine may attenuate oxidative stress and play a protective role in coronary atherosclerosis. Thus, serine may counteract the harmful effects of risk factors, such as male sex and smoking, on atherosclerosis.

This study has several limitations. First, this study’s case–control design did not allow causality determination. Cohort and randomized trials should be conducted to provide further evidence. Second, a family history of CVD; inflammation markers, such as C-reactive protein; and other potential confounding factors that may influence serine concentrations, were not available from the participants’ medical records and were thus not included in the analyses. Finally, we chose patients with coronary artery stenosis of <30% rather than those with normal coronary arteries (no stenosis) as control subjects. Nevertheless, none of the controls were diagnosed with CHD at discharge, and most had nearly normal coronary angiography. Meanwhile, we chose those with coronary artery stenosis of >70% as cases, as all of these individuals were considered as having definite CHD. Patients whose coronary angiography indicated that the coronary artery stenosis degree was 30%–70%, which is an intermediate status, were not enrolled in this study. Therefore, whether the conclusions can be extrapolated to this population should be further evaluated in the future.

Conclusions

This case–control study of Chinese individuals is the first to report a negative correlation between the serine concentration and CHD, suggesting that serine may play a protective role in coronary atherosclerosis. This relationship was more robust in males, suggesting the existence of an interaction by sex. However, these findings need to be further verified in more extensive cohort studies, and the specific pathophysiological mechanisms need to be explored through animal and cell experiments.