出生前細懸浮微粒暴露與基因甲基化程度及孩童認知功能的關係探討

背景：孩童認知功能的發展涉及學習技能的建立，進一步能夠協助孩童建立決策的過程。而孩童認知功能的喪失可能會增加國內教育和醫療保健系統的負擔。越來越多的研究證據闡明了出生前細懸浮微粒暴露 (PM2.5) 與認知功能損害的關係，然而到目前為止，以更精確的易感受暴露時間窗口來探討出生前PM2.5暴露與孩童認知功能的研究仍是少數。此外，可以作為早期偵測環境暴露物質標記的DNA甲基化，有助於我們了解胎兒在子宮內暴露PM2.5之後所產生的反應。研究目的：旨在探討以週為PM2.5暴露時間單位下，出生前每週PM2.5暴露濃度的增加與DNA甲基化程度和認知功能間相關性的變化。方法：所有研究對象皆來自於台灣出生世代研究中，於西元2004至2005年期間出生於大台北地區的新生兒。於出生時蒐集了臍帶血樣本並透過焦磷酸測序來量測甲基化程度。並且台灣出生世代研究分別在剛出生、出生後第二年與出生後第七年分別進行了新生兒神經行為評估 (Neonatal Neurobehavioral Examination-Chinese Version, NNE-C)、兒童行為調查問卷 (Child Behavior Check List/1.5-5, CBCL/1.5-5) 以及魏氏兒童智力量表第四版 (Wechsler Intelligence Scale for Children-Fourth Edition, WISC-IV) 的評估，代表了孩童於不同時期的認知發展能力狀況。另外，我們結合來自環保署空氣品質監測站的PM2.5、氣象量測值、土地利用資訊以及時間數據，並應用極限梯度提升演算法 (Extreme gradient boosting, XGBoost) 來估算個人每日於居所暴露的PM2.5濃度。最後採用線性迴歸模型結合非線性遞延分配模型 (Distributed lag-nonlinear model, DLNM) 來分析出生前每週細懸浮微粒暴露的增加與NNE-C、CBCL/1.5-5和WISC-IV分數的相關性。並使用中介分析來確認基因甲基化程度是否為PM2.5暴露和孩童認知功能的中介因子。結果：懷孕中期 (約第14至25週) 增加PM2.5的暴露與新生兒神經行為能力降低有關，並且在同一時段發現IGF2 (insulin-like growth factor 2) 與PEG3 (paternally expressed 3) 兩基因的甲基化程度有顯著下降的狀況。從中介分析的結果中發現第14週PM2.5的暴露可能透過PEG3位點二的甲基化程度增加而影響新生兒身體交叉伸直的功能。其結果顯示平均因果中介效應 (average causal mediation effect, ACME) (aβ: -0.044 [-0.10, -0.0018], p值：0.034) 和平均直接效應 (average direct effect, ADE) (aβ: 0.14 [0.029, 0.24], p值：0.01) 皆為顯著。而懷孕的早中期 (第6至14週) 時，PM2.5暴露增加會與2歲時的內化行為問題相關，特別是情緒問題以及身體不適的抱怨，同時MEST (mesoderm specific transcript) CTCF (CCCTC-binding factor, CTCF) 結合位點的甲基化程度降低與IFNG (interferon gamma)甲基化程度增加也在此期間被發現 (第8至第11週)。另外，我們發現在母親教育程度為高中以下的孩童，懷孕早中期 (第6至15週) 的PM2.5暴露與其MEST啟動子區域甲基化增加、PEG3位點2甲基化增加以及MEST CTCF結合位點的甲基化下降有關，同時也在這群人中發現了顯著的2歲內化行為問題 (第6至16週)。結論：我們發現，產前PM2.5暴露對三個發展時程下量測地認知功能的易感受窗口都不同。在各自的易感受窗口中，觀察到IGF2和PEG3甲基化程度可能與新生兒神經行為有關。而MEST CTCF 結合位點和IFNG甲基化則與兩歲行為問題有相關。此外，我們發現PEG3位點二的甲基化程度增加為第14週PM2.5暴露影響新生兒身體交叉伸直功能的中介因子，並且PEG3位點二對出生後發展應持續地關注。此外我們也發現，母親教育程度為高中以下的孩童為內化行為問題的易感受族群，此結果提醒我們應更多地注意這群小孩的居住環境並且提供準父母提升支持環境的知識。

關鍵字

表觀遺傳學； DNA甲基化；出生前細懸浮微粒暴露；非線性遞延分配模型；孩童認知功能

並列摘要

Background: Cognitive development in children involves the building process of learning skill further help children to make decision. The loss of cognitive performance may increase the burden of domestic education and healthcare system. Increasing evidences elucidated the relationship between early-life PM2.5 exposure and cognitive impairment. So far, limited researches have explored more refined sensitive time windows of in-utero PM2.5 exposure effect on children cognition. Besides, as an early marker of toxicity, DNA methylation may help us to understand what happened after PM2.5 exposure in uterus. Objective: Our study aims to explore weekly gestational PM2.5 exposure effects in association with DNA methylation and cognitive function. Material and methods: Study subjects came from Taiwan Birth Panel Study (TBPS), who born in 2004 to 2005, lived in Great Taipei metropolitan area. Cord blood samples collected at delivery were used to measure DNA methylation by pyrosequencing. To evaluate children cognitive function, Neonatal Neurobehavioral Examination-Chinese Version (NNE-C), Child Behavior Check List/1.5-5 (CBCL/1.5-5) and Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) were measured at birth, the second year and the seventh year, respectively. Extreme gradient boosting (XGBoost) algorithm with PM2.5 measurements from air quality monitoring stations, meteorological, land use and time data were applied to estimate individual daily PM2.5 exposure at residential address. Linear regression model combined with distributed lag non-linear model (DLNM) were used to analyze the association between prenatal weekly PM2.5 exposure and scores of NNE-C, CBCL/1.5-5 and WISC-IV. Mediation analysis was applied to elucidate if gene methylation level was a mediator between prenatal PM2.5 exposure and cognition. Results: In general, PM2.5 exposure during mid-pregnancy (the 14th to 25th weeks) was link to impaired neonatal neurobehavior and decreasing methylated IGF2 (insulin-like growth factor 2) and PEG3 (paternally expressed 3) were link to the same period. However, we found methylation level of PEG3 site 2 was a mediator for PM2.5 at the 14th week on neonatal crossed extensor. The average causal mediation effect (ACME) (aβ: -0.044 [-0.10, -0.0018], p-value: 0.034) and average direct effects (ADE) (aβ: 0.14 [0.029, 0.24], p-value: 0.01) were significant. Early-mid pregnancy (the 6th to 14th weeks) was critical for internalizing behavioral problems at 2 years old especially for emotion and somatic complaints, while reducing methylated MEST (mesoderm specific transcript) CTCF (CCCTC-binding factor, CTCF) binding sites and increasing methylated IFNG (interferon gamma) were observed (the 8th to 11th weeks). Besides, higher methylated MEST promoter region, PEG3 site 2 and site 4 and reducing methylated MEST CTCF binding sites were found in the group of maternal education level below senior high school (the 6th to 15th weeks), which were link to internalizing problems (the 6th to 16th weeks) during early-middle pregnancy. Conclusion: For cognitive function in three life-stage, different susceptible windows for prenatal PM2.5 were found in each other. In separate crucial exposure windows, IGF2 and PEG3 methylation levels were related to neonatal neurobehavior, MEST CTCF binding sites and IFNG were associated with children behavioral problems in 2 years old. In addition, we found that increasing PEG3 site 2 methylation level was a mediator for PM2.5 on crossed extensor at the 14th gestational week and the influence on postnatal development should be concerned. Besides, we discovered that children with lower maternal education level was a vulnerable population for internalizing problems. It supposed us to concern residential area for these people and provide knowledge about raising parenting support environment.

並列關鍵字

Epigenetic ； DNA methylation ； prenatal PM2.5 ； DLNM ； cognition

參考文獻

Achenbach, T. M. (2011). Child Behavior Checklist. In J. S. Kreutzer, J. DeLuca, B. Caplan (Eds.), Encyclopedia of Clinical Neuropsychology (pp. 546-552). New York, NY: Springer New York.

Google Scholar

Agency, U. S. E. P. (2020, APRIL 5, 2022). NAAQS Table.

Google Scholar

Anderson, G. B., Bell, M. L., Peng, R. D. (2013). Methods to calculate the heat index as an exposure metric in environmental health research. Environmental health perspectives, 121(10), 1111-1119. doi: 10.1289/ehp.1206273

Google Scholar

Ballester, F., Iniguez, C. (2011). Air Pollution Exposure During Pregnancy and Reproductive Outcomes: InTech.

Google Scholar

Bansal, E., Hsu, H.-H., de Water, E., Martínez-Medina, S., Schnaas, L., Just, A. C., . . . Wright, R. O. (2021). Prenatal PM2.5 exposure in the second and third trimesters predicts neurocognitive performance at age 9–10 years: A cohort study of Mexico City children. Environmental Research, 202, 111651. doi: https://doi.org/10.1016/j.envres.2021.111651

Google Scholar

主題瀏覽