世代研究中環境暴露與表觀基因改變的相關性：以產前陶斯松暴露及環境溫度對成人的效應為例

Abstract

環境暴露與表觀基因改變的相關性 環境暴露會經由改變表觀基因調控造成不良健康效應，表觀基因標記可以在不影響 DNA 序列的狀況下調控基因表現，並進一步影響表現型或疾病感受性。我們以產前陶斯松暴露及環境溫度對成人的效應為例，探討在世代研究中環境暴露與表觀基因改變的相關性。 產前陶斯松暴露與胎盤 PPARγ 基因之 H3K4me3 調控及健康效應的相關性 背景：陶斯松是常見的殺蟲劑之一，他可以穿過胎盤影響胎兒生長及神經發展。peroxisome proliferator-activated receptor gamma (PPARγ) 基因的表觀基因調控是陶斯松暴露影響胎兒生長及發展的可能機轉，例如透過trimethylation of lysine 4 of H3 (H3K4me3) 之組蛋白修飾。本研究欲探討產前陶斯松暴露與胎盤PPARγ 基因之 H3K4me3 調控、新生兒健康及神經發展的相關性。 方法：本研究納入 2004 年 4 月至 2005 年 1 月間進入臺灣出生世代研究 (Taiwan Birth Panel Study) 的 425 對母嬰對，使用 online SPE-LC/HESI/MS/MS 測量臍帶血中陶斯松濃度，胎盤PPARγ 基因之 H3K4me3 調控程度以 ChIP-qPCR 測量，新生兒健康資料從病歷取得，使用嬰幼兒綜合發展測驗 (CDIIT) 計算兩歲孩童之神經發展。我們使用多變項迴歸模型校正潛在干擾因子。 結果：校正潛在干擾因子後，產前陶斯松暴露濃度每增加一個單位 (log-ng/mL) 和降低兩歲孩童的認知及語言表現有關，尤其是男童 (adjusted β (aβ) = -1.66, p = 0.016 及 aβ = -1.79, p = 0.023)。PPARγ 基因之 H3K4me3 調控程度和妊娠週數 (aβ = 0.16, p = 0.011) 、出生體重 (aβ = 30.52 (g), p = 0.013) 、出生身長 (aβ = 0.18 (cm), p = 0.003 及 aβ = 0.15 (cm), p = 0.042) 、兩歲粗大動作表現 (aβ = 1.67, p = 0.036) 存在正相關。 結論：本研究發現產前陶斯松暴露會影響認知及語言表現。 環境溫度暴露與成人 DNA 甲基化老化 背景：氣候變遷導致過去數十年環境溫度上升，暴露環境高溫會造成生物老化現象，但在溫暖氣候區的相關研究依然缺乏。本研究目標為分析環境溫度與熱指數暴露的年齡加速效應，並以副熱帶氣候的臺灣人為研究族群。 方法：本研究納入 2008 年至 2016 年參與臺灣人體生物資料庫共 2,084 名研究參與者。我們從氣象站蒐集每日氣溫及相對溼度資料，並以 ordinary kriging 推估每位參與者在居住地的暴露量，再計算參與者進入研究前 1 至 180 天不同時間區間的環境溫度及熱指數移動平均值，而後用以估計暴露之效應。DNA 甲基化年齡使用 Horvath’s 、 Hannum’s 、 Weidner’s 、 ELOVL2 、 FHL2 、 phenotypic (Pheno) 、 Skin & blood 、 GrimAge2 (Grim2) 等方法進行估計，年齡加速定義為 DNA 甲基化年齡減去實際年齡。本研究使用多變項線性迴歸模型、廣義加成模型 (GAMs)、遞延非線性模型 (DLNMs) 分析環境溫度與熱指數暴露對年齡加速的影響。 結果：暴露環境高溫及高熱指數存在年齡加速效應，此效應會隨著暴露時間增加而增強。每日最高溫熱指數以警告 (80-90°F) 、高度警告 (90-103°F) 、危險 (103-124°F) 、高度危險 (> 124°F) 計算熱壓力天數，我們發現熱壓力天數也具有年齡加速效應，尤其是高度危險天數。高度危險天數每增加一天平均會增加 Horvath’s 571.38 (95% CI: 42.63-1100.13) 天、Hannum’s 528.02 (95% CI: 36.16-1019.87) 天、 Weidner’s 43.9 (95% CI: 0.28-87.52) 天、Pheno 16.82 (95% CI: 2.36-31.28) 天、Skin & blood 15.52 (95% CI: 2.17-28.88) 天的年齡加速效應。此外我們發現環境高溫的年齡加速效應存在季節及地理變異。 結論：環境高溫及高熱指數和加速生物老化有關。 結論 我們發現在世代研究中環境暴露會改變表觀基因標記，而表觀基因標記又和健康效應有關，因此表觀基因標記在研究環境暴露對於人體的衝擊中扮演著重要的角色。

Environmental exposures in association with epigenetic alterations Environmental exposures can cause adverse health outcomes through the alteration of epigenetic regulation. Epigenetic marks can regulate gene expression without affecting DNA sequence, which may further influence phenotypes or susceptibility to diseases. We used prenatal chlorpyrifos exposure and ambient temperature effects in adults as examples to demonstrate environmental exposures in association with epigenetic alterations in cohort studies. Prenatal chlorpyrifos exposure in relation to placental PPARγ H3K4me3 modifications and health outcomes Background: Chlorpyrifos, one of the most widely used pesticides, can penetrate the placenta and affect fetal growth and neurodevelopment. Epigenetic regulation of peroxisome proliferator-activated receptor gamma (PPARγ), such as trimethylation of lysine 4 of H3 (H3K4me3), may provide a potential mechanism for how fetal growth and development are impacted by chlorpyrifos exposure. The aims of the study were to investigate whether prenatal chlorpyrifos exposure was associated with H3K4me3 levels of the PPARγ gene in the placenta and the related effects on birth outcomes and neurodevelopment. Methods: Among 425 mother-infant pairs enrolled from April 2004 to January 2005 in the Taiwan Birth Panel Study, chlorpyrifos levels were measured in cord blood by using online SPE-LC/HESI/MS/MS; placental PPARγ H3K4me3 levels were measured by ChIP-qPCR; the neonates’ health outcomes were extracted from the medical records; and childhood neurodevelopment was evaluated by using the Comprehensive Developmental Inventory for Infants and Toddlers in 2-year-old children. Multivariable regression models were used to adjust for potential confounders. Results: After controlling for potential confounders, each unit increase in the natural log-transformed prenatal chlorpyrifos exposure level was associated with poorer performance in the cognitive and language domains at 2 years old, especially in boys (adjusted β (aβ) = -1.66, p = 0.016, and aβ = -1.79, p = 0.023, respectively). PPARγ H3K4me3 levels were positively associated with gestational age (aβ = 0.16 (week), p = 0.011), birth weight (aβ = 30.52 (g), p = 0.013), birth length (aβ = 0.18 (cm), p = 0.003 and aβ = 0.15, p = 0.042), and gross-motor performance (aβ = 1.67, p = 0.036). Conclusions: Our findings suggested that prenatal chlorpyrifos exposure affected performance in the cognitive and language domains. Ambient temperature exposures and DNA methylation aging in adults Background: Climate change caused an increase in ambient temperature in the past decades. Exposure to high ambient temperature could result in biological aging, but relevant studies in a warm environment were lacking. We aimed to study the exposure effects of ambient temperature and heat index (HI) in relation to age acceleration in Taiwan, a subtropical island in Asia. Methods: The study included 2,084 participants recruited from 2008 to 2016 in Taiwan Biobank. Daily temperature and relative humidity data were collected from weather monitoring stations. Individual residential exposure was estimated by ordinary kriging. Moving averages of ambient temperature and HI from 1 to 180 days prior to enrollment were calculated to estimate the exposure effects in multiple time periods. Age acceleration was defined as the difference between DNA methylation age and chronological age. DNA methylation age was calculated by the Horvath’s, Hannum’s, Weidner’s, ELOVL2, FHL2, phenotypic (Pheno), Skin & blood, and GrimAge2 (Grim2) DNA methylation age algorithms. Multivariable linear regression models, generalized additive models (GAMs), and distributed lag non-linear models (DLNMs) were conducted to estimate the effects of ambient temperature and HI exposures in relation to age acceleration. Results: Exposure to high ambient temperature and HI were associated with increased age acceleration, and the associations were stronger in prolonged exposure. The heat stress days with maximum HI in caution (80-90°F), extreme caution (90-103°F), danger (103-124°F), and extreme danger (> 124°F) were also associated with increased age acceleration, especially in the extreme danger days. Each extreme danger day was associated with 571.38 (95% CI: 42.63-1100.13), 528.02 (95% CI: 36.16-1019.87), 43.9 (95% CI: 0.28-87.52), 16.82 (95% CI: 2.36-31.28) and 15.52 (95% CI: 2.17-28.88) days increase in the Horvath’s, Hannum’s, Weidner’s, Pheno, and Skin & blood age acceleration, respectively. In addition, seasonal and geographical variations were performed in the relationships between high ambient temperature exposure and increasing age acceleration. Conclusion: High ambient temperature and HI may accelerate biological aging. Conclusion Environmental exposures can alter epigenetic marks, and epigenetic marks are associated with health outcomes in cohort studies. For this reason, epigenetic marks play important roles in studying the impact of environmental exposures on humans.