Suderman Matthew, Pappas Jane J, Borghol Nada, Buxton Jessica L, McArdle Wendy L, Ring Susan M, Hertzman Clyde, Power Chris, Szyf Moshe, Pembrey Marcus
MRC Integrative Epidemiology Unit (IEU), School of Social and Community Medicine, University of Bristol, Bristol, UK,
Department of Pharmacology and Therapeutics, McGill University, Montreal, Quebec, Canada, Department of Physiology, University of Toronto, Toronto, Ontario, Canada.
Int J Epidemiol. 2015 Aug;44(4):1331-40. doi: 10.1093/ije/dyv168. Epub 2015 Sep 8.
Some cohort studies bank lymphoblastoid cell lines (LCLs) as a renewable source of participant DNA. However, although LCL DNA has proved valuable for genetic studies, its utility in epigenetic epidemiology research is unknown.
To assess whether LCL DNA can be used for life-course environmental epigenomics, we carried out a pilot methylomic study (using the Illumina Infinium Human Methylation 450 BeadChip) of nil-passage, Epstein-Barr virus (EBV)-transformed LCLs (n = 42) and 28 matched whole-blood (WB) samples. These were from adult male participants of the British 1958 birth cohort, selected for extremes of social economic position (SEP) in childhood and adulthood, with additional information available on childhood abuse and prenatal tobacco exposure.
We identified a small number of weak associations between these exposures and methylation levels of both individual CpG sites and genomic regions in WB and LCLs. However, only one of the regional, and none of the individual CpG site associations were common to both sample types. The lack of overlap between the associations detected in LCL compared with those found in WB could either be due to the EBV-transformation process, or to the fact that, unlike WB, LCLs are essentially a single (CD19+) cell type. We provide evidence that the latter is the more potent explanation, by showing that CpG sites known to be differentially methylated between different types of blood cell have significantly lower correlations (R = 0.11) than average (R = 0.2) between WB and LCLs in our datasets, whereas sites known to be affected by EBV-transformation have significantly higher correlations (R = 0.3).
This small pilot study suggests that the DNA methylation profile of LCLs is more closely related to that of B cells than WB and, additionally, that LCLs may nevertheless be useful for life-course environmental epigenomics.
一些队列研究将淋巴母细胞系(LCL)作为参与者DNA的可再生来源储存起来。然而,尽管LCL DNA已被证明对基因研究有价值,但其在表观遗传流行病学研究中的效用尚不清楚。
为了评估LCL DNA是否可用于生命历程环境表观基因组学研究,我们开展了一项试点甲基化组研究(使用Illumina Infinium Human Methylation 450 BeadChip),研究对象为零代、爱泼斯坦-巴尔病毒(EBV)转化的LCL(n = 42)和28份匹配的全血(WB)样本。这些样本来自英国1958年出生队列的成年男性参与者,根据童年和成年时期的社会经济地位(SEP)极端情况进行选择,并提供了关于童年虐待和产前烟草暴露的额外信息。
我们在这些暴露因素与WB和LCL中单个CpG位点及基因组区域的甲基化水平之间发现了少量微弱关联。然而,两种样本类型中只有一个区域关联相同,单个CpG位点关联则无相同情况。与WB中发现的关联相比,LCL中检测到的关联缺乏重叠,这可能是由于EBV转化过程,或者是因为与WB不同,LCL本质上是单一(CD19 +)细胞类型。我们通过表明在我们的数据集中,已知在不同类型血细胞之间存在差异甲基化的CpG位点,其WB和LCL之间的相关性(R = 0.11)显著低于平均水平(R = 0.2),而已知受EBV转化影响的位点具有显著更高的相关性(R = 0.3),从而提供证据表明后者是更有力的解释。
这项小型试点研究表明,LCL的DNA甲基化谱与B细胞的甲基化谱比与WB的更密切相关,此外,LCL可能仍然对生命历程环境表观基因组学有用。