DNA甲基化在孕妇砷暴露与新生儿出生体重间的中介作用分析

房瑞玲; 白文琳; 崔跃华; 王彤

doi:10.16462/j.cnki.zhjbkz.2024.09.015

DNA甲基化在孕妇砷暴露与新生儿出生体重间的中介作用分析

doi: 10.16462/j.cnki.zhjbkz.2024.09.015

房瑞玲^{1, 2},
白文琳²,
崔跃华³,
王彤^{1, 2, ,}

1.
山西医科大学公共卫生学院卫生统计学教研室，太原 030001
2.
煤炭环境致病与防治教育部重点实验室，太原 030001
3.
美国密西根州立大学统计与概率系，东兰辛 MI48824

基金项目:

国家自然科学基金 82204164

山西省应用基础研究计划 20210302124409

中国博士后科学基金 2021M691990

详细信息

通讯作者:
王彤，E-mail：tongwang@sxmu.edu.cn

中图分类号: R181.3;R174
计量
- 文章访问数: 63
- HTML全文浏览量: 8
- PDF下载量: 11
- 被引次数: 0
出版历程
- 收稿日期: 2023-12-13
- 修回日期: 2024-04-18
- 网络出版日期: 2024-10-24
- 刊出日期: 2024-09-10

The mediation analysis of DNA methylation in the relationship between maternal arsenic exposure and neonatal birth weight

FANG Ruiling^{1, 2},
BAI Wenlin²,
CUI Yuehua³,
WANG Tong^{1, 2
, ,}

1.
Department of Health Statistics, School of Public Health, Shanxi Medical University, Taiyuan 030001, China
2.
MOE Key Laboratory of Coal Environmental Pathogenicity and Prevention, Shanxi Medical University, Taiyuan 030001, China
3.
Department of Statistics and Probability, Michigan State University, East Lansing MI48824, USA

Funds:

National Natural Science Foundation of China 82204164

Applied Basic Research Program of Shanxi Province 20210302124409

China Postdoctoral Science Foundation 2021M691990

More Information

Corresponding author: WANG Tong, E-mail: tongwang@sxmu.edu.cn

摘要

摘要: 目的探讨DNA甲基化(DNA methylation, DNAm)水平的改变是否在孕期砷暴露与新生儿出生体重之间有中介作用。方法资料来自新罕布什尔州出生队列研究，研究对象为2012年2月―2013年9月入选该队列的343对母婴，数据可从高通量基因表达(gene expression omnibus, GEO)数据库获取，检索号为GSE71678，共有270对母婴纳入本研究。采用确定独立筛选(sure independence screening, SIS)策略和中介效应分析方法估计DNAm位点在胎盘砷暴露与出生体重之间的中介效应。结果调整孕妇年龄、孕前BMI、妊娠期糖尿病、胎龄、新生儿性别、胎盘组织细胞成分后，筛选出25个候选中介DNAm位点。中介效应分析结果显示，通过Bonferroni法校正后，识别出位于VENTX基因上的cg14900295在孕期砷暴露致低出生体重有中间的作用(ACME=0.057 6, 95% CI：0.017 6~0.106 4, P=0.000 8)；通过FDR方法校正，识别出18个胞嘧啶-磷酸-鸟嘌呤双核苷酸(cytosine-phosphate-guanine pairs of nucleotides, CpG)位点有中介作用，其中cg03348978、cg02435495、cg09463047、cg11862993均位于HNF1B基因。结论在妊娠期砷暴露导致低出生体重的过程中，共有18个特异性位点的DNAm水平改变起到了重要的中介作用，VENTX和HNF1B基因可能是其病因机制中的关键环节，为揭示砷暴露致低出生体重的遗传病因机制提供了参考依据。
- DNA甲基化 /
- 孕妇砷暴露 /
- 出生体重 /
- 中介效应
Abstract: Objective To investigate whether the changes of DNA methylation level mediate the relationship between prenatal arsenic exposure and neonatal birth weight. Methods Data of this study were derived from the New Hampshire Birth Cohort Study, with 343 mother-infant pairs enrolled in the cohort from February 2012 to September 2013 and available from the GEO database under accession number GSE71678. A total of 270 maternal and infant pairs were included in this study. The Sure independence screening (SIS) strategy and mediation analysis method were employed to estimate the mediation effect of DNA methylation sites between placental arsenic exposure and birth weight. Results After adjusting for maternal age, pre-pregnancy BMI, gestational diabetes mellitus, gestational age, infant gender, and placental tissue cell composition, 25 DNA methylation sites were selected as candidate mediators. The results of mediation analysis showed that after corrected by Bonferroni method. cg14900295 located on the VENTX gene was identified as a mediator between maternal arsenic exposure and low birth weight (ACME=0.057 6, 95% CI: 0.017 6-0.106 4, P=0.000 8). After corrected by FDR method, 18 CpG loci were identified as mediators, among which cg03348978, cg02435495, cg09463047 and cg11862993 were all located on the HNF1B gene. Conclusions This study identified 18 specific sites where changes in DNA methylation levels play a crucial mediating role in the process linking prenatal arsenic exposure to low birth weight. Genes VENTX and HNF1B may be the key factors in the etiological mechanism. Therefore, these findings can provide a reference for revealing the genetic etiology mechanism of arsenic-induced low birth weight.
- DNA methylation /
- Placental arsenic exposure /
- Birth weight /
- Mediation effect

HTML全文

图 1 NHBCS研究对象筛选流程图

NHBCS：新罕布什尔州出生队列研究。

Figure 1. Flow chart of subject screening in the NHBCS

NHBCS: New Hampshire birth cohort study.

下载: 全尺寸图片幻灯片

图 2 中介效应模型图

Figure 2. Mediation effect model diagram

下载: 全尺寸图片幻灯片

表 1 NHBCS研究对象基线特征

Table 1. Baseline characteristics of NHBCS subjects

变量 Variables	NHBCS研究对象 NHBCS subjects^①
母亲孕龄/岁 Maternal age/years	31.51(28.02, 34.93)
孕前母亲BMI Pre-pregnancy maternal BMI/(kg·m^-2)	25.16(22.31, 29.10)
妊娠期糖尿病 Gestational diabetes mellitus
是 Yes	28(10.37)
否 No	242(89.63)
胎龄/周 Gestational age / weeks	39.57(39.00, 40.43)
新生儿性别 Infant gender
男性 Male	139(51.48)
女性 Female	131(48.52)
新生儿出生体重 Infant birth weight/g	3440(3130, 3709)
胎盘砷含量 Placental arsenic levels/(ng·g^-1)	0.8132(0.5468, 1.2277)
注：NHBCS，新罕布什尔州出生队列研究。 ①以人数(占比/%)或M(P₂₅, P₇₅)表示。 Note: NHBCS, New Hampshire birth cohort study. ① Number of people (proportion/%) or M(P₂₅, P₇₅).

下载: 导出CSV

表 2 25个候选CpG位点集的中介效应分析结果

Table 2. Mediation analysis results of 25 candidate CpG sites

CpGs	CHR	基因 Gene	P_α	P_β	P_max	P_ME	P_{ME, Bon}	P_{ME, FDR}
cg14900295	chr10	VENTX	0.000 9	0.000 3	0.000 9	0.000 8	0.020 0 ^①	0.020 0 ^①
cg03348978	chr17	HNF1B	0.001 9	0.000 4	0.001 9	0.006 0	0.150 0	0.028 8 ^①
cg02435495	chr17	HNF1B	0.003 4	0.000 2	0.003 4	0.007 8	0.195 0	0.028 8 ^①
cg13814374	chr18	APCDD1	0.005 2	0.003 5	0.005 2	0.009 6	0.240 0	0.028 8 ^①
cg10205979	chr3	VPRBP	0.003 3	0.000 6	0.003 3	0.010 0	0.250 0	0.028 8 ^①
cg18676393	chr2	GEMIN6	0.002 4	0.001 5	0.002 4	0.010 8	0.270 0	0.028 8 ^①
cg18829186	chr14	TMED10	0.001 4	0.001 2	0.001 4	0.011 0	0.275 0	0.028 8 ^①
cg17200564	chr11	EIF3F	0.004 9	0.004 8	0.004 9	0.011 2	0.280 0	0.028 8 ^①
cg11939075	chr4	COPS4	0.000 4	0.000 9	0.000 9	0.012 0	0.300 0	0.028 8 ^①
cg02216727	chr17	GJD3	0.004 1	0.002 3	0.004 1	0.012 4	0.310 0	0.028 8 ^①
cg22863744	chr12	NCOR2	0.002 4	0.001 2	0.002 4	0.012 8	0.320 0	0.028 8 ^①
cg09463047	chr17	HNF1B	0.001 0	0.002 0	0.002 0	0.014 0	0.350 0	0.028 8 ^①
cg11862993	chr17	HNF1B	0.001 8	0.000 9	0.001 8	0.015 0	0.375 0	0.028 8 ^①
cg24138419	chr18	HDHD2	0.002 2	0.002 3	0.002 3	0.017 6	0.440 0	0.031 4 ^①
cg06400704	chr1	FMN2	0.001 9	0.004 7	0.004 7	0.024 2	0.605 0	0.040 3 ^①
cg04290595	—	—	0.002 7	0.005 3	0.005 3	0.032 4	0.810 0	0.048 6 ^①
cg07249742	chr3	ITGB5	0.000 1	0.001 5	0.001 5	0.034 2	0.855 0	0.048 6 ^①
cg02908438	chr7	CPA1	0.002 3	0.001 1	0.002 3	0.035 0	0.875 0	0.048 6 ^①
cg00834223	chr2	AGAP1	0.003 8	0.000 6	0.003 8	0.056 4	1.000 0	0.074 2
cg17606548	chr5	MAST4	0.004 7	0.000 4	0.004 7	0.089 8	1.000 0	0.111 0
cg15827101	chr14	KLHDC2	0.002 8	0.005 0	0.005 0	0.093 2	1.000 0	0.111 0
cg20916535	chr6	VGLL2	0.000 1	0.004 8	0.004 8	0.111 2	1.000 0	0.126 4
cg27177983	chr4	LGI2	0.003 3	0.004 6	0.004 6	0.167 6	1.000 0	0.182 2
cg17156276	chr1	KDM4A	0.003 9	0.001 0	0.003 9	0.245 2	1.000 0	0.255 4
cg22763407	chr12	ELK3	0.005 1	0.004 9	0.005 1	0.299 2	1.000 0	0.299 2
注：CpG，胞嘧啶-磷酸-鸟嘌呤双核苷酸；P_α为中介模型中胎盘砷暴露的回归系数显著性；P_β为结局模型中DNA甲基化位点的回归系数显著性；P_max为P_α与P_β二者的较大值；P_ME为因果中介效应的显著性；P_{ME, Bon}为对采用Bonferroni法校正P值；P_{ME, FDR}为对采用FDR法校正P值; “—”表示无法获取。 ①表示经校正后P < 0.05。 Note: CpG, cytosine-phosphate-guanine pairs of nucleotides; P_α means the regression coefficient significance of placental arsenic exposure in the mediation model; P_β means the regression coefficient significance of DNAm in the outcome model; P_max means the maximum value of P_α and P_β; P_ME means the significance of causal mediation effect; P_{ME, Bon} means the p values corrected by Bonferroni method; P_{ME, FDR} means the p values corrected by FDR method; "—" indicates that it cannot be obtained. ① Means the corrected P < 0.05.

下载: 导出CSV

表 3 VENTX和HNF1B基因上5个CpG位点的中介分析结果

Table 3. Mediation analysis results of 5 CpG sites on the VENTX and HNF1B genes

CpGs	基因 Gene	ACME		ADE		TE
CpGs	基因 Gene	估计值Estimate (95% CI)	P值 value	估计值Estimate (95% CI)	P值 value	估计值Estimate (95% CI)	P值 value
cg14900295	VENTX	0.057 6(0.017 6, 0.106 4)	0.000 8	-0.091 3(-0.208 5, 0.066 7)	0.256 8	-0.033 7(-0.159 2, 0.119 4)	0.684 2
cg03348978	HNF1B	0.050 2(0.012 2, 0.088 3)	0.006 0	-0.076 7(-0.191 9, 0.080 4)	0.367 2	-0.026 5(-0.155 8, 0.123 5)	0.758 2
cg02435495	HNF1B	0.050 2(0.010 8, 0.092 8)	0.007 8	-0.077 1(-0.190 7, 0.079 5)	0.362 8	-0.026 9(-0.157 5, 0.123 1)	0.754 2
cg09463047	HNF1B	0.047 6(0.008 1, 0.090 9)	0.014 0	-0.072 4(-0.192 5, 0.088 5)	0.416 4	-0.024 8(-0.156 5, 0.125 8)	0.766 2
cg11862993	HNF1B	0.048 4(0.007 3, 0.089 5)	0.015 0	-0.071 6(-0.189 4, 0.089 4)	0.423 6	-0.023 1(-0.155 8, 0.125 6)	0.771 4
注：CpG，胞嘧啶-磷酸-鸟嘌呤双核苷酸；ACME，平均因果中介效应；ADE，平均直接效应；TE，总效应。 Note: CpG, cytosine-phosphate-guanine pairs of nucleotides; ACME, average causal mediation effect; ADE, average direct effect; TE, total effect.

下载: 导出CSV

参考文献(17)

[1]	Barker DJ. The fetal and infant origins of adult disease[J]. BMJ, 1990, 301(6761): 1111. DOI: 10.1136/bmj.301.6761.1111.
[2]	Wadhwa PD, Buss C, Entringer S, et al. Developmental origins of health and disease: brief history of the approach and current focus on epigenetic mechanisms[J]. Semin Reprod Med, 2009, 27(5): 358-368. DOI: 10.1055/s-0029-1237424.
[3]	Gluckman PD, Hanson MA, Cooper C, et al. Effect of in utero and early-life conditions on adult health and disease[J]. N Engl J Med, 2008, 359(1): 61-73. DOI: 10.1056/NEJMra0708473.
[4]	Jamnik T, Flasch M, Braun D, et al. Next-generation biomonitoring of the early-life chemical exposome in neonatal and infant development[J]. Nat Commun, 2022, 13(1): 2653. DOI: 10.1038/s41467-022-30204-y.
[5]	Shih YH, Scannell Bryan M, Argos M. Association between prenatal arsenic exposure, birth outcomes, and pregnancy complications: an observational study within the national children's study cohort[J]. Environ Res, 2020, 183: 109182. DOI: 10.1016/j.envres.2020.109182.
[6]	Green BB, Karagas MR, Punshon T, et al. Epigenome-wide assessment of DNA methylation in the placenta and arsenic exposure in the new Hampshire birth cohort study (USA)[J]. Environ Health Perspect, 2016, 124(8): 1253-1260. DOI: 10.1289/ehp.1510437.
[7]	Dye CK, Domingo-Relloso A, Kupsco A, et al. Maternal DNA methylation signatures of arsenic exposure is associated with adult offspring insulin resistance in the strong heart study[J]. Environ Int, 2023, 173: 107774. DOI: 10.1016/j.envint.2023.107774.
[8]	Küpers LK, Monnereau C, Sharp GC, et al. Meta-analysis of epigenome-wide association studies in neonates reveals widespread differential DNA methylation associated with birthweight[J]. Nat Commun, 2019, 10(1): 1893. DOI: 10.1038/s41467-019-09671-3.
[9]	Yuan V, Hui D, Yin YF, et al. Cell-specific characterization of the placental methylome[J]. BMC Genomics, 2021, 22(1): 6. DOI: 10.1186/s12864-020-07186-6.
[10]	Houseman EA, Accomando WP, Koestler DC, et al. DNA methylation arrays as surrogate measures of cell mixture distribution[J]. BMC Bioinformatics, 2012, 13: 86. DOI: 10.1186/1471-2105-13-86.
[11]	Fan JQ, Lyu JC. Sure independence screening for ultrahigh dimensional feature space[J]. J R Stat Soc Ser B Stat Methodol, 2008, 70(5): 849-911. DOI: 10.1111/j.1467-9868.2008.00674.x.
[12]	Tingley D, Yamamoto T, Hirose K, et al. Mediation: R package for causal mediation analysis[J]. J Stat Soft, 2014, 59(5): 1-38. DOI: 10.18637/jss.v059.i05.
[13]	Kumar S, Kumar V, Li WC, et al. Ventx family and its functional similarities with nanog: involvement in embryonic development and cancer progression[J]. Int J Mol Sci, 2022, 23(5): 2741. DOI: 10.3390/ijms23052741.
[14]	Wang GY, Yu J, Yang YW, et al. Whole-transcriptome sequencing uncovers core regulatory modules and gene signatures of human fetal growth restriction[J]. Clin Transl Med, 2020, 9(1): 9. DOI: 10.1186/s40169-020-0259-0.
[15]	Bozack AK, Rifas-Shiman SL, Coull BA, et al. Prenatal metal exposure, cord blood DNA methylation and persistence in childhood: an epigenome-wide association study of 12 metals[J]. Clin Epigenetics, 2021, 13(1): 208. DOI: 10.1186/s13148-021-01198-z.
[16]	Wang X, Wu H, Yu WH, et al. Hepatocyte nuclear factor 1b is a novel negative regulator of white adipocyte differentiation[J]. Cell Death Differ, 2017, 24(9): 1588-1597. DOI: 10.1038/cdd.2017.85.
[17]	Bozack AK, Domingo-Relloso A, Haack K, et al. Locus-specific differential DNA methylation and urinary arsenic: an epigenome-wide association study in blood among adults with low-to-moderate arsenic exposure[J]. Environ Health Perspect, 2020, 128(6): 67015. DOI: 10.1289/EHP6263.