Association of maternal methionine synthase reductase gene polymorphisms with the risk of congenital heart disease in offspring: a hospital-based case-control study

Abstract

Background

Evidence suggests that periconceptional folic acid supplementation may prevent congenital heart disease (CHD). Methionine synthase reductase (MTRR) is one of the key regulatory enzymes in the folate metabolic pathway. This study aimed to comprehensively evaluate the association of single nucleotide polymorphisms (SNPs) in the maternal MTRR gene with CHD risk in offspring.

Methods

A hospital-based case-control study involving 740 mothers of CHD cases and 683 health controls was conducted.

Results

The study showed that maternal MTRR gene polymorphisms at rs1532268 (C/T vs. C/C: aOR = 1.524; T/T vs. C/C: aOR = 3.178), rs1802059 (G/A vs. G/G: aOR = 1.410; A/A vs. G/G: aOR = 3.953), rs2287779 (G/A vs. G/G: aOR = 0.540), rs16879334 (C/G vs. C/C: aOR = 0.454), and rs2303080 (T/A vs. T/T: aOR = 0.546) were associated with the risk of CHD. And seven haplotypes were observed to be associated with the risk of CHD, T-G-A haplotype (OR = 1.298), C-A-C-C (OR = 4.824) and A-G haplotype (OR = 1.751) were associated with increased risk of CHD in offspring; A-A-A (OR = 0.773), T-A-A (OR = 0.557), G-A-C-C (OR = 0.598) and G-C (OR = 0.740) were associated with decreased risk of CHD in offspring.

Conclusions

Maternal MTRR gene polymorphisms were associated with CHD in offspring, and its haplotypes have affected the occurrence of CHD. Furthermore, given the complexity and heterogeneity of CHD, the mechanisms by which these factors influence offspring cardiac development remain unknown, and studies in larger samples in an ethnically diverse population are needed.

Keywords:

• Congenital heart disease
• folic acid
• homocysteine
• methionine synthase reductase
• polymorphisms

Introduction

Congenital heart disease (CHD), one of the most common classes of major congenital malformations, is defined as a structural or functional abnormality of the heart or intrathoracic vessels during the embryonic period [1]. The global reported birth prevalence of CHD is increasing. It is reported that the mean prevalence of CHD worldwide was estimated at 8.22 per thousand live birth [2] and 8.98 per thousand live birth in China [3]. In infants and young children, CHD is responsible for the largest proportion of the total mortality due to birth defects, between 30% and 50%[4]. Over time, the most common cause of death has shifted from communicable diseases to non-communicable diseases, CHD is likely to continue to play an increasingly important role in global infant mortality in the future [5], which will bring a huge burden to society and families, and CHD has become a serious public health problem. Although it is widely believed that the etiology of CHD is multifactorial, where genetic, environmental, and lifestyle factors are thought to interact, little is known about the underlying etiology of CHDs [6]. For this reason, it is necessary to research further the pathogenesis and better control the development of this disease.

Folic acid supplementation may reduce the risk of congenital anomalies, according to some epidemiological evidence [6,7]. As the metabolite in the folic acid pathway, homocysteine is an independent risk factor for CHD [8]. In studies of cardiovascular disease and birth defects, it is widely believed that genetic mutations in enzymes related to folate metabolism cause elevated concentrations of homocysteine in the blood, which in turn affects folate metabolism and leads to the development of birth defects [9]. Therefore, the occurrence of birth defects may be related to the genes of enzymes related to folate metabolism. Methionine synthase reductase (MTRR) maintains sufficient levels of activated cobalamin, which serves as a cofactor for methionine synthase (MS). In the process of MS catalyzed re-methylation of homocysteine to methionine, cobalamin acts as an intermediated methyl carrier between methyltetrahydrofolate and homocysteine. The cobalamin cofactor cycles between the cob(I)alamin and methyl cob(III)alamin, but the cob(I)alamin can be oxidized to the unactivated cob(II)alamin form, and for it to regain activity, cob(II) needs to be converted to the methyl-cob(III)alamin form by obtaining a methyl donor for S-adenosylmethionine catalyzed by MTRR, and this cycle is the reduction system of methionine synthase (MTR)[10–12]. This cycle ensures the activity of MTR, and MTRR acts as a “companion” that plays an important role in keeping MTR in an active state. Therefore, mutations in the MTRR gene may alter the concentration of homocysteine and thus affect normal embryonic development [8]. Thus far, most studies on the association between MTRR gene single nucleotide polymorphisms (SNPs) and the risk of CHD have focused on MTRR polymorphisms A66G [13–16] (i.e. rs1801394), however, these studies have presented conflicting conclusions [17,18]. In addition to the most studied rs1801394, other loci exist in the maternal MTRR gene that have rarely been studied. It is necessary to study these genetic loci in the maternal MTRR gene further to provide an epidemiological basis for the study of genetic susceptibility factors of CHD, and also provide clues for revealing the potential mechanism of folic acid in preventing birth defects.

Based on the above, we hypothesized that the multiple genetic variants of the folate pathway are associated with susceptibility to CHD. A hospital-based case-control study was undertaken to illustrate the association between maternal MTRR gene polymorphisms and the risk of CHD in offspring.

Methods

Recruitment of study participants and selection criteria

We conducted a hospital-based case-control study to assess the association of maternal MTRR gene polymorphisms with the risk of CHD in offspring. The Hunan Children’s Hospital conducted recruitment from November 2017 to March 2020. All mothers of patients with CHD diagnosed by echocardiography and undergoing cardiac operation were identified as the case group. The control group was defined as the mothers of children who did not have any congenital malformations after physical examination at the same hospital during the same study period. The outcome of interest was non-syndromic CHDs, while patients with syndromic CHDs, such as other organ malformations or known abnormalities, were excluded. To reduce maternal recall bias in exposure information, all participants were recruited into our study when their children were less than one year old. All participants were Han Chinese population taking into account the possible bias of the study due to the differences in ethnicity.

For the case group, the inclusion criteria were: (i) mothers of children with simple CHD diagnosed by echocardiography exploration and surgery; (ii) no familial relationship between individual study participants; and (iii) voluntary enrollment of study subjects. The exclusion criteria were: (i) the child had syndromic CHDs; (ii) singleton or multiple pregnancies; (iii) achieved pregnancy by assisted reproductive technology (i.e. vitro fertilization and intracytoplasmic sperm injection); (iv) no blood relationship between the mother and the child; (v) the mother had a mental disorder that prevented her from completing the survey; (vi) informed consent was not obtained or the questionnaire was not completed or the blood sample was not provided. For the control group, the inclusion criteria were: (i) mothers of children without CHD and/or other congenital disorders diagnosed by medical examination; (ii) no familial relationship with cases and no familial relationship with controls. The exclusion criteria for the control group were the same as those for the case group except for the first criterion.

The study received approval from the ethics committee of Xiangya School of Public Health Central South University (No: XYGW-2018-36) and was conducted based on the Helsinki declaration. The study protocol has been registered with the Chinese Clinical Trial Registry (No: ChiCTR1800018492). We obtained written informed consent from all mothers.

Information collection

A self-designed questionnaire was used to collect relevant information for specially trained investigators through face-to-face interviews, which was established content validity, test-retest reliability (r = 0.867), and internal consistency (α = 0.811). We collected maternal demographic characteristics (i.e. age, body mass index before this pregnancy), abnormal pregnancy history before this pregnancy (i.e. spontaneous abortions, fetal death or stillbirth, preterm birth, hypertension of pregnancy, gestational diabetes mellitus), family history (i.e. consanguineous marriages and congenital malformations), lifestyle and habit before this pregnancy (i.e. active and passive smoking, alcohol consumption), exposure history to the environmentally hazardous substance (i.e. exposure to environmentally harmful substances near the place of residence and history of decorating housing), as well as folic acid consumption before or during pregnancy.

SNP selection and genotyping

After finishing the questionnaires, all mothers donated 3 to 5 milliliters of peripheral venous blood preserved in anticoagulant tubes and then centrifugated into blood and plasma. Blood cells were stored at −80 °C until genotyping analysis. We used the QIAamp DNA Mini Kit (Qiagen, Valencia, CA, USA) to extract genomic DNA. We dissolved it in sterile TBE (Tris-borate-EDTA) buffer following the standard experimental operation protocol supplied by the manufacturer. To ensure that the DNA could be used as a template for the polymerase chain reaction, ultraviolet spectrophotometry was used to detected the DNA solution’s concentration and purity.

We selected the candidate loci of the MTRR gene based on a previous study [19]. Briefly, SNP markers were selected using the SNPBrowser™ program (version 3.0) provided by AppliedBiosystems Inc. This program allowed the selection of SNP markers from the HapMap database. For each target gene, tagging SNPs were selected based on the pairwise r² > 0.8, and SNPs with minor allele frequencies (MAFs) < 10% were excluded. Eventually, we selected rs162036, rs326120, rs1532268, rs162048, rs1802059, rs2287779, rs10380, rs16879334, rs10064631, rs1801394, rs3776455, rs2303080, rs9332 of the MTRR gene as candidate loci for this study. Matrix-assisted laser desorption and ionization time-of-flight mass spectrometry MassARRAY SYSTEM (Agena iPLEX assay, San Diego, CA, USA) was used to assay the maternal MTRR gene polymorphisms. The error rate for genotyping was less than 5%. Experimenters retyped and double-checked all the samples to guarantee the reliability of the experiments. The detail of prime sequences for all SNPs are shown in Supplementary Table 1.

Statistical analysis

Frequencies and percentages were used to describe categorical variables. Differences of unorder categorical variables between two groups were used by Pearson chi-square test or Fisher’s exact probability test. Wilcoxon rank-sum test was used to compare the difference in ordinal categorical variables. Each SNP of the MTRR gene was checked for conformance with the Hardy-Weinberg equilibrium (HWE) in the control group using the goodness-of-fit chi-square test (significance level at p < .10). The odds ratio (OR) and their 95% confidence interval (CI) were used to assess the level of association. Unadjusted OR was calculated by univariable logistic regression, and Adjusted OR was calculated by multivariable logistic regression. We used logistic regression to control the potential confounding factors to evaluate the association of maternal MTRR gene polymorphisms with the risk of CHD in offspring. Further, we applied three genetic models containing the dominant model, recessive model, and additive model to analyze the effect of the MTRR gene polymorphisms comprehensively. The dominant model was defined as homozygous variant + heterozygous variants vs. homozygous wild-type, the recessive model was defined as homozygous variants vs. heterozygous variant + homozygous wild-type, and the additive model was defined as homozygous variant vs. heterozygous variant vs. homozygous wild-types. Linkage disequilibrium test and haplotype analysis were performed to assess the association between haplotype and the risk of CHD in the offspring. In addition, to minimize type I error, the false discovery rate P (FDR_P) was used to obtain a more precise P value when multiple testing was involved. The statistically significant results were those with the false discovery rate P value (FDR_P) < 0.1. All tests were two-sided and performed at a significant level of p < .05, except where otherwise specified. Linkage disequilibrium test and haplotype analysis were performed by Haploview software version 4.2 (http://www.broadinstitute.org/haploview/haploview). Other analyses were performed using R software (version 4.1.1).

Results

Recruitment of study participants and comparison of demographic characteristics

In total, 740 CHD cases and 683 controls were recruited into the study. The demographic characteristics of the two groups are summarized in Table 1. There were statistically significant differences across two groups for maternal age, education level, history of fetal death or stillbirth, history of hypertension of pregnancy, history of gestational diabetes mellitus, family history of consanguineous marriages, family history of congenital malformations, active smoking before pregnancy, passive smoking before pregnancy, alcohol consumption before pregnancy, harmful substances near the place of residence, the situation of house decoration and folic acid consumption before or during pregnancy (all P values < 0.05). These potential confounding factors were processed in multivariate logistic regression to estimate the adjusted odds ratio and 95% confidence intervals for the association between the maternal MTRR gene polymorphisms and the risk of CHD in offspring.

Table 1. Comparison of maternal demographic characteristics in cases and controls.

Variables	Control (n = 740)	Case (n = 683)	Univariate analysis
Maternal age (years)
＜25	61 (8.2%)	138 (20.2%)	Z = −4.761, p＜.001^a
25–29	266 (35.9%)	231 (33.8%)
30–34	242 (32.7%)	180 (26.4%)
≥35	171 (23.1%)	134 (19.6%)
Body mass index before this pregnancy (kg/m^2)			Z= −1.510, p = .131^a
＜18.5	192 (26.4%)	138 (20.4%)
18.5–23.9	401 (55.2%)	421 (62.3%)
24.0–27.9	92 (12.7%)	83 (12.3%)
≥28	41 (5.6%)	34 (5.0%)
Education
Primary school and lower	9 (1.2%)	98 (14.3%)	Z= −15.471, p＜.001^a
Junior high school	144 (19.5%)	282 (41.3%)
Senior middle school	246 (33.2%)	196 (28.7%)
College and above	341 (46.1%)	107 (15.7%)
History of spontaneous abortions (yes)	68 (9.2%)	79 (11.6%)	χ^2=2.167, p = .141
History of fetal death or stillbirth (yes)	3 (0.4%)	23 (3.4%)	χ^2=17.373, p＜.001
History of preterm birth (yes)	6 (0.8%)	10 (1.5%)	χ^2=1.364, p = .243
History of hypertension of pregnancy (yes)	17 (2.3%)	46 (6.7%)	χ^2=16.531, p＜.001
History of gestational diabetes mellitus (yes)	23 (3.1%)	67 (9.8%)	χ^2=26.924, p＜.001
Family history of consanguineous marriages (yes)	3 (0.4%)	25 (3.7%)	χ^2=19.508, p＜.001
Family history of congenital malformations (yes)	7 (0.9%)	44 (6.4%)	χ^2=31.050, p＜.001
Active smoking before pregnancy (yes)	16 (2.2%)	50 (7.3%)	χ^2=21.369, p＜.001
Passive smoking before pregnancy (yes)	265 (35.8%)	357 (52.3%)	χ^2=39.104, p＜.001
Alcohol consumption before pregnancy (yes)	54 (7.3%)	88 (12.9%)	χ^2=12.342, p＜.001
Harmful substances near place of residence (yes)	53 (7.2%)	135 (19.8%)	χ^2=49.206, p＜.001
House decoration (yes)	65 (8.8%)	37 (5.4%)	χ^2=6.049, p = .014
Folic acid consumption before or during pregnancy (yes)	687 (92.8%)	577 (84.5%)	χ^2=24.996, p＜.001

^aDifferences between controls and cases were tested by Wilcoxon rank-sum test.

Maternal MTRR gene polymorphisms and the risk of CHD in offspring

Genotype frequencies for each SNPs of the maternal MTRR gene and P values of the Hardy-Weinberg equilibrium test in the control group are summarized in Supplementary Table 2. The genotype distributions in the control group were consistent with the HWE (P values >0.10) except for rs10064631 because no variant genotype was observed. Therefore, this locus was not included in the subsequent analysis.

Table 2 summarized the association between the maternal MTRR genetic variant and the risk of CHD in offspring using the logistic regression model. After adjustment for baseline characteristics with significant differences among the two groups, multivariable logistic regression showed that maternal MTRR gene polymorphisms at rs1532268, rs1802059, rs2287779, rs16879334 and rs2303080 were significantly associated with the risk of CHD. For rs1532268, mothers with C/T genotype (aOR = 1.524, 95%CI = 1.123 to 2.070) and T/T genotype (aOR = 3.178, 95%CI = 1.448 to 6.976) compared with those with C/C genotype, had a significantly increased risk of CHD in offspring (FDR_ p < .1); the dominant model (aOR = 1.664, 95%CI = 1.244 to 2.226), recessive model (aOR = 2.904, 95%CI = 1.327 to 6.357) and additive model (aOR = 1.617, 95%CI = 1.260 to 2.073) were significantly associated with risk of CHD (FDR_ p < .1). For rs1802059, mothers with G/A genotype (aOR = 1.410, 95%CI = 1.045 to 1.903) and A/A genotype (aOR = 3.953, 95%CI = 1.789 to 8.737) compared with those with G/G genotype, had a significantly increased risk of CHD in offspring (FDR_ p < .1); the dominant model (aOR = 1.581, 95%CI = 1.188 to 2.105), recessive model (aOR = 3.636, 95%CI = 1.650 to 8.010) and additive model (aOR = 1.599, 95%CI = 1.251 to 2.043) were significantly associated with risk of CHD (FDR_ p < .1). For rs2287779, mothers with G/A genotype (aOR = 0.540, 95%CI = 0.406 to 0.718) compared with those with G/G genotype, had a significantly increased risk of CHD in offspring (FDR_ p < .1); the dominant model (aOR = 0.535, 95%CI = 0.407 to 0.704) and additive model (aOR = 0.597, 95%CI = 0.472 to 0.756) were significantly associated with risk of CHD (FDR_ p < .1). For rs16879334, mothers with C/G genotype (aOR = 0.454, 95%CI = 0.339 to 0.608) compared with those with C/C genotype, had a significantly increased risk of CHD in offspring (FDR_ p < .1); the dominant model (aOR = 0.468, 95%CI = 0.354 to 0.617) and additive model (aOR = 0.557, 95%CI = 0.439 to 0.706) were significantly associated with risk of CHD (FDR_ p < .1). For rs2303080, mothers with T/A genotype (aOR = 0.546, 95%CI = 0.393 to 0.757) compared with those with T/T genotype, had a significantly increased risk of CHD in offspring (FDR_ p < .1); the dominant model (aOR = 0.536, 95%CI = 0.388 to 0.740) and additive model (aOR = 0.554, 95%CI = 0.409 to 0.750) were significantly associated with risk of CHD (FDR_ p < .1).

Table 2. Association of maternal MTRR gene polymorphisms with risk of CHD.

SNPs of MTRR gene	Univariate logistic regression		Multivariable logistic regression
	Unadjusted OR (95%CI)	p Value	Adjusted OR (95%CI)^a	p Value	FDR_P
rs162036
A/A	1	–	1	–	–
G/A	0.940 (0.740 to 1.194)	.614	0.895 (0.674 to 1.188)	.442	0.580
G/G	1.179 (0.635 to 2.187)	.602	1.347 (0.659 to 2.756)	.414	0.580
Dominant model	0.962 (0.765 to 1.211)	.744	0.935 (0.713 to 1.227)	.629	0.633
Recessive model	1.198 (0.648 to 2.216)	.564	1.389 (0.680 to 2.825)	.369	0.492
Additive model	0.990 (0.812 to 1.207)	.745	0.985 (0.780 to 1.244)	.902	0.902
rs326120
A/A	1	–	1	–	–
A/G	0.893 (0.716 to 1.114)	.316	0.968 (0.739 to 1.267)	.810	0.845
G/G	1.190 (0.842 to 1.681)	.324	1.467 (0.983 to 2.191)	.061	0.146
Dominant model	0.947 (0.769 to 1.168)	.613	1.064 (0.825 to 1.371)	.633	0.633
Recessive model	1.259 (0.907 to 1.747)	.168	1.491 (1.022 to 2.176)	.038	0.114
Additive model	1.02 (0.875 to 1.194)	.745	1.136 (0.945 to 1.365)	.175	0.263
rs1532268
C/C	1	–	1	–	–
C/T	1.527 (1.182 to 1.972)	.001	1.524 (1.123 to 2.070)	.007	0.028
T/T	3.708 (1.908 to 7.205)	<.001	3.178 (1.448 to 6.976)	.004	0.020
Dominant model	1.704 (1.335 to 2.173)	＜.001	1.664 (1.244 to 2.226)	.001	0.003
Recessive model	3.376 (1.741 to 6.543)	＜.001	2.904 (1.327 to 6.357)	.008	0.048
Additive model	1.667 (1.356 to 2.051)	＜.001	1.617 (1.260 to 2.073)	＜.001	＜0.001
rs162048
A/A	1	–	1	–	–
G/A	0.950 (0.761 to 1.186)	.651	1.108 (0.845 to 1.454)	.459	0.580
G/G	1.220 (0.872 to 1.709)	.246	1.321 (0.891 to 1.957)	.166	0.307
Dominant model	1.005 (0.816 to 1.238)	.964	1.156 (0.897 to 1.490)	.262	0.477
Recessive model	1.251 (0.909 to 1.721)	.169	1.257 (0.867 to 1.823)	.227	0.289
Additive model	1.055 (0.906 to 1.229)	.351	1.138 (0.949 to 1.364)	.163	0.263
rs1802059
G/G	1	–	1	–	–
G/A	1.472 (1.145 to 1.892)	.003	1.410 (1.045 to 1.903)	.024	0.082
A/A	3.479 (1.782 to 6.793)	<.001	3.953 (1.789 to 8.737)	.001	0.006
Dominant model	1.624 (1.278 to 2.064)	＜.001	1.581 (1.188 to 2.105)	.002	0.005
Recessive model	3.178 (1.632 to 6.190)	＜.001	3.636 (1.650 to 8.010)	.001	0.012
Additive model	1.604 (1.307 to 1.970)	＜.001	1.599 (1.251 to 2.043)	＜.001	＜0.001
rs2287779
G/G	1	–	1	–	–
G/A	0.598 (0.473 to 0.756)	<.001	0.540 (0.406 to 0.718)	＜.001	＜0.001
A/A	0.482 (0.268 to 0.865)	.014	0.495 (0.248 to 0.991)	.047	0.125
Dominant model	0.584 (0.467 to0.732)	＜.001	0.535 (0.407 to 0.704)	＜.001	＜0.001
Recessive model	0.562 (0.314 to 1.005)	.049	0.602 (0.303 to 1.195)	.147	0.353
Additive model	0.632 (0.520 to 0.767)	＜.001	0.597 (0.472 to 0.756)	＜.001	＜0.001
rs10380
C/C	1	–	1	–	–
C/T	0.849 (0.659 to 1.092)	.203	0.823 (0.610 to 1.108)	.199	0.341
T/T	1.846 (0.768 to 4.436)	.171	1.349 (0.512 to 3.553)	.544	0.653
Dominant model	0.893 (0.698 to 1.141)	.364	0.852 (0.638 to 1.138)	.278	0.477
Recessive model	1.915 (0.798 to 4.593)	.139	1.411 (0.537 to 3.705)	.485	0.582
Additive model	0.952 (0.762 to 1.189)	.148	0.901 (0.694 to 1.169)	.432	0.518
rs16879334
C/C	1	–	1	–	–
C/G	0.537 (0.423 to 0.681)	<.001	0.454 (0.339 to 0.608)	＜.001	＜0.001
G/G	0.524 (0.297 to 0.923)	.025	0.579 (0.295 to 1.138)	.113	0.247
Dominant model	0.535 (0.426 to 0.672)	＜.001	0.468 (0.354 to 0.617)	＜.001	＜0.001
Recessive model	0.626 (0.357 to 1.099)	.100	0.732 (0.375 to 1.426)	.359	0.492
Additive model	0.602 (0.495 to 0.733)	＜.001	0.557 (0.439 to 0.706)	＜.001	＜0.001
rs1801394
A/A	1	–	1	–	–
A/G	0.870 (0.700 to 1.082)	.210	0.890 (0.685 to 1.157)	.385	0.578
G/G	0.901 (0.593 to 1.369)	.626	0.875 (0.536 to 1.430)	.595	0.679
Dominant model	0.875 (0.710 to 1.077)	.208	0.888 (0.692 to 1.140)	.352	0.521
Recessive model	0.958 (0.637 to 1.439)	.836	0.920 (0.571 to 1.483)	.731	0.731
Additive model	0.911 (0.771 to 1.077)	.446	0.915 (0.750 to 1.115)	.378	0.504
rs3776455
C/C	1	–	1	–	–
T/C	0.980 (0.786 to 1.222)	.857	0.997 (0.763 to 1.302)	.982	0.982
T/T	1.196 (0.844 to 1.696)	.315	1.555 (1.029 to 2.350)	.036	0.108
Dominant model	1.019 (0.826 to 1.57)	.589	1.092 (0.848 to 1.406)	.495	0.594
Recessive model	1.209 (0.869 to 1.681)	.260	1.557 (1.055 to 2.300)	.026	0.104
Additive model	1.055 (0.902 to 1.233)	.522	1.162 (0.964 to 1.402)	.115	0.230
rs2303080
T/T	1	–	1	–	–
T/A	0.746 (0.575 to 0.968)	.027	0.546 (0.393 to 0.757)	＜.001	＜0.001
A/A	0.460 (0.159 to 1.333)	.153	0.356 (0.092 to 1.382)	.136	0.272
Dominant model	0.728 (0.564 to 0.940)	.015	0.536 (0.388 to 0.740)	＜.001	＜0.001
Recessive model	0.489 (0.169 to 1.414)	.178	0.427 (0.112 to 1.630)	.213	0.390
Additive model	0.733 (0.578 to 0.929)	.035	0.554 (0.409 to 0.750)	＜.001	＜0.001
rs9332
G/G	1	–	1	–	–
A/G	0.809 (0.610 to 1.074)	.143	0.829 (0.593 to 1.160)	.274	0.438
A/A	1.403 (0.670 to 2.939)	.369	1.224 (0.548 to 2.738)	.622	0.679
Dominant model	0.858 (0.655 to 1.123)	.265	0.870 (0.634 to 1.196)	.391	0.521
Recessive model	1.465 (0.701 to 3.062)	.307	1.273 (0.571 to 2.836)	.556	0.607
Additive model	0.926 (0.733 to 1.169)	.202	0.930 (0.712 to 1.216)	.597	0.651

SNP: single nucleotide polymorphism; MTRR: Methionine synthase reductase; CHD: congenital heart disease; CI: confidence interval; FDR_P: False discovery rate p-value.

^aAdjustment was made for maternal age, maternal educational level, history of fetal death or stillbirth, history of hypertension of pregnancy, history of hypertension of pregnancy, History of gestational diabetes mellitus, family history of Consanguineous marriages, family history of Congenital malformations, active smoking before pregnancy, passive smoking before pregnancy, alcohol consumption before pregnancy, harmful substances near place of residence, situation of house decoration and folic acid consumption before or during pregnancy.

Linkage disequilibrium test and haplotype analysis

The linkage disequilibrium test results between SNPs loci of maternal MTRR gene polymorphisms are listed in Table 3. The linkage disequilibrium test showed a strong correlation (r² > 0.8) between rs9332 and rs10380 (r² = 0.892).

Table 3. Degree of linkage disequilibrium of maternal MTRR gene polymorphisms between CHD group and control group.

r^2	rs326120	rs1532268	rs162048	rs1802059	rs2287779	rs10380	rs16879334	rs1801394	rs3776455	rs2303080	rs9332
rs162036	0.026	0.008	0.027	0.018	0.030	0.670	0.034	0.038	0.069	0.019	0.697
rs326120	–	0.031	0.743	0.022	0.080	0.057	0.087	0.151	0.177	0.059	0.058
rs1532268	–	–	0.052	0.663	0.032	0.006	0.029	0.071	0.242	0.021	0.005
rs162048	–	–	–	0.046	0.078	0.065	0.095	0.150	0.221	0.063	0.068
rs1802059	–	–	–	–	0.026	0.016	0.033	0.078	0.259	0.022	0.011
rs2287779	–	–	–	–	–	0.019	0.788	0.057	0.091	0.507	0.024
rs10380	–	–	–	–	–	–	0.028	0.044	0.074	0.009	0.892
rs16879334	–	–	–	–	–	–	–	0.058	0.086	0.470	0.024
rs1801394	–	–	–	–	–	–	–	–	0.625	0.043	0.032
rs3776455	–	–	–	–	–	–	–	–	–	0.066	0.057
rs2303080	–	–	–	–	–	–	–	–	–	–	0.016

Haplotype analysis revealed that twelve SNPs formed four haplotype blocks, block 1 contained two SNPs rs1801394 and rs326120, block 2 contained three SNPs rs2303080, rs162036 and rs2287779, block 3 contained four SNPs rs16879334, rs162048, rs3776455 and rs10380, block 4 contained two SNPs rs1802059 and rs9332 (Figure 1). The result of haplotype frequencies of maternal MTRR gene polymorphisms for cases and controls was summarized in Table 4. Three haplotypes for the maternal MTRR gene of T-G-A (involving rs2303080, rs162036 and rs2287779; OR = 1.298, 95%CI = 1.109 to 1.518), C-A-C-C (involving rs16879334, rs162048, rs3776455 and rs10380; OR = 4.824, 95%CI = 3.024 to 7.697), and A-G (involving rs1802059 and rs9332; OR = 1.751, 95%CI = 1.395 to 2.197) were associated with increased risk of CHD in offspring. Nevertheless, four haplotypes for the maternal MTRR gene of A-A-A (involving rs2303080, rs162036 and rs2287779; OR = 0.773, 95%CI = 0.607 to 0.985), T-A-A (involving rs2303080, rs162036 and rs2287779; OR =0.557, 95%CI = 0.415 to 0.748), G-A-C-C (involving rs16879334, rs162048, rs3776455 and rs10380; OR = 0.598, 95%CI = 0.489 to 0.731) and G-C (involving rs1802059 and rs9332; OR = 0.740, 95%CI = 0.618 to 0.886) were associated with decreased risk of CHD in offspring.

Figure 1. Linkage disequilibrium tests for maternal MTRR gene polymorphisms between CHD group and control group.

Table 4. Haplotype frequencies of maternal MTRR gene polymorphisms between case group and control group.

Haplotypes	Frequency	Control (%)	Case (%)	OR (95%CI)	p Value
Block1
A-A	0.402	40.5	39.9	0.990 (0.852–1.151)	.898
A-G	0.324	33.0	31.8	1.058 (0.904–1.239)	.480
G-A	0.265	25.9	27.0	0.950 (0.804–1.122)	.545
Block2
T-G-A	0.657	69.3	62.5	1.298 (1.109–1.518)	.001
T-G-G	0.154	15.6	15.3	1.039 (0.847–1.274)	.713
A-A-A	0.106	9.3	11.8	0.773 (0.607–0.985)	.037
T-A-A	0.072	5.2	9.1	0.557 (0.415–0.748)	＜.001
Block3
C-A-T-C	0.328	33.3	32.4	0.970 (0.775–1.213)	.789
C-G-C-C	0.323	33.0	31.6	1.061 (0.906–1.242)	.463
G-A-C-C	0.169	13.3	20.3	0.598 (0.489–0.731)	＜.001
C-A-C-T	0.123	12.2	12.5	1.035 (0.884–1.211)	.669
C-A-C-C	0.041	6.9	1.6	4.824 (3.024–7.697)	＜.001
Block4
G-G	0.737	0.709	0.762	0.740 (0.618–0.886)	.001
A-G	0.140	0.171	0.110	1.751 (1.395–2.197)	＜.001
G-A	0.118	0.113	0.123	0.907 (0.709–1.161)	.412

Block1: rs1801394-rs326120.

Block2: rs2303080- rs162036-rs2287779.

Block3: rs16879334-rs162048-rs3776455-rs10380.

Block4: rs1802059-rs9332.

Discussion

In this study, we systematically assessed the association of twelve SNPs of the maternal MTRR gene with the risk of CHD in offspring. We observed a significant association of five SNPs (rs1532268, rs1802059, rs2287779, rs16879334, and rs2303080) in the maternal MTRR gene with the susceptibility of CHD in offspring. Additionally, we also observed that three haplotypes for the maternal MTRR gene of T-G-A (involving rs2303080, rs162036 and rs2287779), C-A-C-C (involving rs16879334, rs162048, rs3776455 and rs10380), and A-G (involving rs1802059 and rs9332) were associated with increased risk of CHD in offspring; and four haplotypes of A-A-A (involving rs2303080, rs162036 and rs2287779), T-A-A (involving rs2303080, rs162036 and rs2287779), G-A-C-C (involving rs16879334, rs162048, rs3776455 and rs10380) and G-C (involving rs1802059 and rs9332) were associated with decreased risk of CHD in offspring. As far as we know, this is the first comprehensive and systematic study on the association between multiple genetic variants of the maternal MTRR gene and the risk of CHD in offspring, the findings emphasize how crucial a role genetic factors play in the development of CHD.

Congenital heart disease is a disease that is influenced by a combination of genetic, environmental, and lifestyle factors. Factors such as maternal pre-pregnancy smoking [20], alcohol consumption [21], history of disease [22], and exposure to environmental pollutants [23] all influence the development of CHD. To evaluate the independent effect of maternal MTRR gene polymorphisms on CHD pathogenesis, we controlled for these confounding factors with a priori knowledge and clinical basis, and with differences in demographic characteristics in our study.

Our results indicated that the maternal MTRR gene polymorphisms were associated with the risk of CHD in offspring. The most common and studied polymorphisms in the MTRR gene are substituting A for G at nucleotide 66 (i.e. A66G, rs1801394), which results in the substitution of isoleucine by methionine. This mutation is located in the putative flavin mononucleotide-binging domain of the MTRR enzyme, which interacts with MTR, and thus, disrupts the binding domain of the MTR-cobalamin-complex and decreases the enzyme activity and the rate of homocysteine re-methylation [24,25]. A Chinese population-based study suggested that MTRR A66G polymorphisms increased the risk of CHD, which was also verified in other populations and subtypes of CHD [14,26,27]. On the contrary, some studies failed to find evidence of an association between the occurrence of CHDs and A66G polymorphisms of the MTRR gene in Caucasian Netherlands [17,18] and the Han Chinese population [28], and our study came to the same conclusion: that there is no association between the maternal MTRR gene A66G polymorphisms and CHD. A meta-analysis [15] based on 48 studies showed that the MTRR A66G polymorphisms might only affect in risk of CHD in Asians but not in Caucasians. Therefore, we speculate that ethnic background substantially affects susceptibility to different CHD in different populations. Yang et al.[29] explored the geographical distribution of MTRR A66G polymorphisms in China. Their study indicated that the prevalence of the MTRR A66G polymorphisms varies significantly between the Han population residing in different regions of China. Most of our study population comes from southern China, which might also be why it is different from other studies on the Han Chinese population. It is worth noting that Li et al.[25] suggested that MTRR 66AG + GG may not affect the serum folate level or the incidence of folate deficiency. Similarly, Feix et al.[30] have reported that the MTRR A66G polymorphism does not affect the total concentrations of homocysteine, folate, or vitamin B₁₂. Also, van Beynum [18] has reported that maternal MTRR 66 A > G polymorphism is not a risk factor for CHD, and maternal MTRR 66GG genotype with compromised vitamin B₁₂ status may result in increased CHD risk. As we all know, the MTRR enzyme is essential for adequate re-methylation of homocysteine. A66G polymorphism of the MTRR gene is a missense mutation, which is a leading element for increased plasma homocysteine levels [31], and increased maternal homocysteine may lead to an increased risk of CHD in offspring, which may be a maternal MTRR gene contributing to the development of CHD in the offspring the causes [32]. We speculate that the association between maternal MTRR A66G polymorphism and CHD may have not only ethnic influence but also have potentially unknown interactions with other factors in the folate metabolic pathway and environmental factors. This may provide a new research direction for the maternal MTRR gene and CHD study.

In addition, a few scholars [14,15,26,27,33] have studied the association between MTRR C524T (rs1532268) polymorphisms and CHD or its subtypes. All their studies indicated that rs1532268 can increase the risk of CHD, and it is the same conclusion in our study. So far, few studies have detected multiple SNPs simultaneously to comprehensively analyze the association between maternal MTRR gene polymorphism and the risk of CHD in offspring. Therefore, when evaluating the association between the maternal MTRR gene and CHD, it is inevitable to miss a lot of important genetic information. In this article, we focused not only on the above two SNPs but also on the other 10 SNPs of the MTRR gene (i.e. rs162036, rs326120, rs162048, rs1802059, rs2287779, rs10380, rs16879334, rs3776455, rs2303080, rs9332). We also found that the genetic variant at rs1802059, rs2287779, rs16879334, and rs2303080 were associated with the occurrence of CHD in offspring. The current study has not studied the association between these loci and CHD, therefore our study could provide clues for screening genetic susceptibility sites of the MTRR gene in the future. In our study, rs1802059 loci increase the risk of CHD, which is consistent with the hypothesis of this study. On the contrary, we observed that rs2287779, rs16879334, and rs2303080 polymorphisms had a protective effect on offspring CHD, but their mechanism of action in not yet clear and needs further study and analysis.

SNP haplotypes mean a linear arrangement of nucleotide disease bases on different SNPs. Haplotype affects phenotypes through influencing promoter activity and protein structure or through tagging nearby untyped causal variants [34]. In general, human diseases are often not caused by a single SNP, and a single polymorphism is not sufficient to reflect its association with the disease, whereas haplotype analysis is a study that integrates multiple SNP loci. Haplotype association takes into account allelic heterogeneity, where different mutations within a gene lead to similar phenotypes, which cannot be expressed by individual SNP analysis [34]. So the use of haplotypes for disease association analysis in complex diseases such as CHD is an effective means to pinpoint the causative gene [35]. To date, few studies have performed haplotype analysis of the association between the maternal MTRR gene and offspring CHD. We found three haplotypes that increased CHD risk and four haplotypes that decreased CHD risk, suggesting that the haplotype formed by SNP of the MTRR gene is associated with the occurrence of CHD. Shaw et al. [19] examined 118 SNP loci associated with folic acid and CHD risk and concluded that haplotypes formed by the MTRR gene were associated with the risk of developing spina bifida but not conotruncal heart defects. Li et al. [36] showed that haplotypes formed by the fetal MTRR gene prevented conotruncal heart defects, while maternal genotypes did not alter disease risk. Considering that haplotypes contain more linkage disequilibrium information than an SNP, association analysis using haplotypes in a complex disease such as CHD is conducive to a more accurate location of pathogenic sites and improves statistical efficiency.

Some limitations should be considered. First, cases and controls were recruited from the same hospital, albeit in a different department, and this convenient sampling rather than random sampling would have resulted in relatively poor representativeness and thus limited extrapolation of results. Second, due to the nature of this study, participants had to recall some past information, such as contact history, past habits, etc., the accuracy of which was highly dependent on their memory. Inevitably, information bias occurred. Third, although potential known confounding factors were controlled for when assessing the association of genetic model and haplotype blocks with CHD, there are undoubtedly other unknown confounding factors beyond our knowledge. Fourth, because of the limited sample size in this study, we did not assess the risk for specific CHD phenotypes. Fifth, our study was limited to the Han Chinses population, and because of the various ethical and regional differences in MTRR gene polymorphisms, there is a need to investigate this in a large population of different ethnicities.

Our study found a few SNP loci and haplotypes on the maternal MTRR gene that may influence the risk of CHD development, and these SNP loci and haplotypes may provide clues for future etiological studies of CHD by purposefully targeting the mechanisms of the relevant SNP loci. Likewise, it reveals that maternal genetic factors can be used as potential biomarkers for predicting the incidence of CHD in offspring. Therefore, in future studies, the association of other key enzyme gene polymorphisms in the maternal folate metabolism pathway with offspring CHD should be further analyzed. Based on the SNPs in the entire pathway, their association with the offspring CHD should be analyzed from a holistic perspective. At the same time, biochemical indicators such as folate metabolism-related products and Hcy serological concentrations can be included in the analysis to reveal the mechanism of CHD, in combination with gene function studies and epigenetic studies. In addition, a prospective, large-sample cohort study could be conducted to include more participants and accumulate a sufficient sample size for the analysis of CHD subtypes, as well as consider the influence of paternal genetic factors on CHD.

Conclusion

In this study, we systematically examined the association of twelve SNPs in the maternal MTRR gene with offspring CHD risk. After adjusting for potential confounding factors, we found that five maternal MTRR loci, rs1532268, rs1802059, rs2287779, rs16879334, and rs2303080, were associated with the risk of CHD, which provided clues for further exploration of the association between maternal MTRR gene and CHD prevalence and for the search of genetic susceptibility loci. In addition, we found that three significant haplotype blocks in haplotype analysis were both associated with CHD development. However, given the complexity and heterogeneity of CHD, the mechanisms by which these factors influence offspring cardiac development remain unknown, and studies in larger samples in an ethnically diverse population are needed.

Ethical approval

The study received approval from the ethics committee of Xiangya School of Public Health Central South University (No: XYGW-2018-36) and was conducted based on the Helsinki declaration. The study protocol has been registered with the Chinese Clinical Trial Registry (No: ChiCTR1800018492). We obtained written informed consent from all mothers.

Author contributions

JBQ and TTW conceptualized and designed the study, coordinated all study phases. JHW and TTW analyzed the data and wrote the paper. XLS, YPL, JS, and MTS performed the experiments. JHW, JYD, JQL, YHL, LTC and SMZ collected and check the data. JHW, TTW, and JBQ controlled the quality of this study. All authors reviewed and approved the paper.

Acknowledgments

The authors thank all pediatricians and patients who participated in the study.

Disclosure statement

No potential conflict of interest was reported by the author(s). No financial or non-financial benefits have been received or will be received from any party related directly or indirectly to the subject of this article.

Additional information

Funding

Funding for this study was provided by the National Natural Science Foundation Program of China [82073653 and 81803313], Hunan Outstanding Youth Fund Project [2022JJ10087], National Key Research and Development Project [2018YFE0114500], China Postdoctoral Science Foundation [2020M682644], Hunan Provincial Science and Technology Talent Support Project [2020TJ-N07], Hunan Provincial Key Research and Development Program [2018SK2063], Open Project from NHC Key Laboratory of Birth Defect for Research and Prevention [KF2020006], Natural Science Foundation of Hunan Province [2018JJ2551], Natural Science Foundation of Hunan Province of China [2022JJ40207], and Changsha Municipal Natural Science Foundation [kq2202470].