A cord blood multi-omic analysis of birthweight reveals new underlying mechanisms related to cholesterol metabolism [gene expression]

Birthweight reflects in utero exposures and later health evolution. Despite existing studies employing high-dimensional molecular measurements, the understanding of underlying mechanisms of birthweight remains limited. To investigate the systems biology of birthweight, we cross-sectionally integrated the methylome, the transcriptome, the metabolome and a set of inflammatory proteins measured in cord blood samples, collected from four birth-cohorts (n=489). We focused on two sets of 68 metabolites and 903 CpGs previously related to birthweight and investigated the correlation structures existing between these two sets and all other omic features via bipartite Pearson correlations. This dataset revealed that the set of metabolome and methylome signatures of birthweight have seven signals in common, including three metabolites [PC(34:2), plasmalogen PC(36:4)/PC(O-36:5), and a compound with m/z of 781.0545], two CpGs (on the DHCR24 and SC4MOL gene), and two proteins (periostin and CCL22). CCL22, a macrophage-derived chemokine has not been previously identified in relation to birthweight. Since the results of the omics integration indicated the central role of cholesterol metabolism, we explored the association of cholesterol levels in cord blood with birthweight in the ENVIRONAGE cohort (n=1,097), finding that higher birthweight was associated with increased high-density lipoprotein cholesterol and that high-density lipoprotein cholesterol was lower in small versus large for gestational age newborns. Our data suggests that an integration of different omic-layers in addition to single omics studies is a useful approach to generate new hypotheses regarding biological mechanisms. CCL22 and cholesterol metabolism in cord blood play a mechanistic role in birthweight. epidemiologic study (200 cord blood samples). 165 samples were analyzed. 35 samples were excluded due to failed RNA extraction, processing or missing covariates. Raw data for all samples are included in this series.

出生体重能够反映宫内暴露情况与后续健康演化轨迹。尽管已有研究采用高维度分子检测手段，但学界对出生体重背后的调控机制仍缺乏充分认知。为探究出生体重的系统生物学特征，本研究对4个出生队列（n=489）的脐带血样本开展横断面整合分析，涵盖甲基化组（methylome）、转录组（transcriptome）、代谢组（metabolome）以及一组炎症蛋白检测数据。本研究聚焦于此前已被报道与出生体重相关的两组特征：68种代谢物与903个CpG位点，并通过二部Pearson相关分析，探究这两组特征与其余组学特征间的关联结构。本数据集显示，与出生体重相关的代谢组与甲基化组特征存在7处共有信号，包括3种代谢物[PC(34:2)、缩醛磷脂PC(36:4)/PC(O-36:5)以及质荷比为781.0545的化合物]、2个CpG位点（分别位于DHCR24与SC4MOL基因上）以及2种蛋白（骨膜蛋白（periostin）与CCL22）。其中，CCL22作为一种巨噬细胞趋化因子，此前从未被发现与出生体重存在关联。鉴于组学整合分析结果提示胆固醇代谢发挥核心调控作用，本研究进一步在ENVIRONAGE队列（n=1097）中探究脐带血胆固醇水平与出生体重的关联，结果显示：出生体重越高，脐带血高密度脂蛋白胆固醇水平也越高；且小于胎龄儿的高密度脂蛋白胆固醇水平显著低于大于胎龄儿。本研究数据表明，相较于单一组学研究，整合多组学层面数据是针对生物学机制提出全新假说的有效策略。脐带血中的CCL22与胆固醇代谢在出生体重的调控中发挥了机制性作用。本研究为流行病学研究，初始纳入200份脐带血样本，最终165份样本完成分析，另有35份样本因RNA提取失败、实验流程异常或协变量缺失被排除。本数据集包含所有样本的原始数据。