Sanchez-Guerra Marco, Zheng Yinan, Osorio-Yanez Citlalli, Zhong Jia, Chervona Yana, Wang Sheng, Chang Dou, McCracken John P, Díaz Anaite, Bertazzi Pier Alberto, Koutrakis Petros, Kang Choong-Min, Zhang Xiao, Zhang Wei, Byun Hyang-Min, Schwartz Joel, Hou Lifang, Baccarelli Andrea A
a Department of Environmental Health; Harvard T.H. Chan School of Public Health ; Boston , MA , USA.
Epigenetics. 2015;10(7):633-42. doi: 10.1080/15592294.2015.1050174.
Previous studies have reported epigenetic changes induced by environmental exposures. However, previous investigations did not distinguish 5-methylcytosine (5mC) from a similar oxidative form with opposite functions, 5-hydroxymethylcytosine (5hmC). Here, we measured blood DNA global 5mC and 5hmC by ELISA and used adjusted mixed-effects regression models to evaluate the effects of ambient PM10 and personal PM2.5 and its elemental components-black carbon (BC), aluminum (Al), calcium (Ca), potassium (K), iron (Fe), sulfur (S), silicon (Si), titanium (Ti), and zinc (Zn)-on blood global 5mC and 5hmC levels. The study was conducted in 60 truck drivers and 60 office workers in Beijing, China from The Beijing Truck Driver Air Pollution Study at 2 exams separated by one to 2 weeks. Blood 5hmC level (0.08%) was ∼83-fold lower than 5mC (6.61%). An inter-quartile range (IQR) increase in same-day PM10 was associated with increases in 5hmC of 26.1% in office workers (P = 0.004), 20.2% in truck drivers (P = 0.014), and 21.9% in all participants combined (P < 0.001). PM10 effects on 5hmC were increasingly stronger when averaged over 4, 7, and 14 d preceding assessment (up to 132.6% for the 14-d average in all participants, P < 0.001). PM10 effects were also significant after controlling for multiple testing (family-wise error rate; FWER < 0.05). 5hmC was not correlated with personal measures of PM2.5 and elemental components (FWER > 0.05). 5mC showed no correlations with PM10, PM2.5, and elemental components measures (FWER > 0.05). Our study suggests that exposure to ambient PM10 affects 5hmC over time, but not 5mC. This finding demonstrates the need to differentiate 5hmC and 5mC in environmental studies of DNA methylation.
先前的研究报告了环境暴露引起的表观遗传变化。然而，先前的调查并未区分5-甲基胞嘧啶（5mC）和具有相反功能的类似氧化形式5-羟甲基胞嘧啶（5hmC）。在此，我们通过酶联免疫吸附测定（ELISA）测量血液DNA中的整体5mC和5hmC，并使用调整后的混合效应回归模型来评估环境空气中细颗粒物（PM10）、个人可吸入颗粒物（PM2.5）及其元素成分——黑碳（BC）、铝（Al）、钙（Ca）、钾（K）、铁（Fe）、硫（S）、硅（Si）、钛（Ti）和锌（Zn）——对血液中整体5mC和5hmC水平的影响。该研究在中国北京的60名卡车司机和60名办公室职员中进行，这些参与者来自“北京卡车司机空气污染研究”，在间隔1至2周的两次检查中进行采样。血液中5hmC水平（0.08%）比5mC（6.61%）低约83倍。当日PM10每增加一个四分位数间距（IQR），办公室职员的5hmC增加26.1%（P = 0.004），卡车司机增加20.2%（P = 0.014），所有参与者综合增加21.9%（P < 0.001）。在评估前平均4天、7天和14天的PM10对5hmC的影响越来越强（所有参与者14天平均值时高达132.6%，P < 0.001）。在控制多重检验后（家族性错误率；FWER < 0.05），PM10的影响仍然显著。5hmC与PM2.5及其元素成分的个人测量值无相关性（FWER > 0.05）。5mC与PM10、PM2.5及其元素成分测量值均无相关性（FWER > 0.05）。我们的研究表明，长期暴露于环境PM10会影响5hmC，但不会影响5mC。这一发现表明，在DNA甲基化的环境研究中需要区分5hmC和5mC。
