Dammering Felix, Martins Jade, Dittrich Katja, Czamara Darina, Rex-Haffner Monika, Overfeld Judith, de Punder Karin, Buss Claudia, Entringer Sonja, Winter Sibylle M, Binder Elisabeth B, Heim Christine
Charité - Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Institute of Medical Psychology, Berlin, Germany.
Dept. of Translational Research in Psychiatry, Max Planck Institute of Psychiatry, Munich, Germany.
Neurobiol Stress. 2021 Sep 11;15:100394. doi: 10.1016/j.ynstr.2021.100394. eCollection 2021 Nov.
Studies reporting accelerated ageing in children with affective disorders or maltreatment exposure have relied on algorithms for estimating epigenetic age derived from adult samples. These algorithms have limited validity for epigenetic age estimation during early development. We here use a pediatric buccal epigenetic (PedBE) clock to predict DNA methylation-based ageing deviation in children with and without internalizing disorder and assess the moderating effect of maltreatment exposure. We further conduct a gene set enrichment analysis to assess the contribution of glucocorticoid signaling to PedBE clock-based results.
DNA was isolated from saliva of 158 children [73 girls, 85 boys; mean age (SD) = 4.25 (0.8) years] including children with internalizing disorder and maltreatment exposure. Epigenetic age was estimated based on DNA methylation across 94 CpGs of the PedBE clock. Residuals of epigenetic age regressed against chronological age were contrasted between children with and without internalizing disorder. Maltreatment was coded in 3 severity levels and entered in a moderation model. Genome-wide dexamethasone-responsive CpGs were derived from an independent sample and enrichment of these CpGs within the PedBE clock was identified.
Children with internalizing disorder exhibited significant acceleration of epigenetic ageing as compared to children without internalizing disorder (F = 6.67, = .011). This association was significantly moderated by maltreatment severity (b = 0.49, 95% CI [0.073, 0.909], t = 2.322, = .022). Children with internalizing disorder who had experienced maltreatment exhibited ageing acceleration relative to children with no internalizing disorder (1-2 categories: b = 0.50, 95% CI [0.170, 0.821], t = 3.008, = .003; 3 or more categories: b = 0.99, 95% CI [0.380, 1.593], t = 3.215, p = .002). Children with internalizing disorder who were not exposed to maltreatment did not show epigenetic ageing acceleration. There was significant enrichment of dexamethasone-responsive CpGs within the PedBE clock (OR = 4.36, = 1.65*10-6). Among the 94 CpGs of the PedBE clock, 18 (19%) were responsive to dexamethasone.
Using the novel PedBE clock, we show that internalizing disorder is associated with accelerated epigenetic ageing in early childhood. This association is moderated by maltreatment severity and may, in part, be driven by glucocorticoids. Identifying developmental drivers of accelerated epigenetic ageing after maltreatment will be critical to devise early targeted interventions.
有关情感障碍或遭受虐待的儿童加速衰老的研究，一直依赖于源自成人样本的表观遗传年龄估算算法。这些算法在早期发育过程中进行表观遗传年龄估算时有效性有限。我们在此使用儿科口腔表观遗传（PedBE）时钟来预测有无内化障碍儿童基于DNA甲基化的衰老偏差，并评估虐待暴露的调节作用。我们还进行了基因集富集分析，以评估糖皮质激素信号传导对基于PedBE时钟结果的贡献。
从158名儿童（73名女孩，85名男孩；平均年龄（标准差）=4.25（0.8）岁）的唾液中分离DNA，这些儿童包括患有内化障碍和遭受虐待的儿童。基于PedBE时钟的94个CpG位点的DNA甲基化来估算表观遗传年龄。将表观遗传年龄与实际年龄进行回归分析得到的残差，在有和没有内化障碍的儿童之间进行对比。虐待被编码为3个严重程度级别，并纳入一个调节模型。全基因组地塞米松反应性CpG位点来自一个独立样本，并确定这些CpG位点在PedBE时钟内的富集情况。
与没有内化障碍的儿童相比，患有内化障碍的儿童表现出显著的表观遗传衰老加速（F=6.67，P=.011）。这种关联受到虐待严重程度的显著调节（b=0.49，95%置信区间[0.073，0.909]，t=2.322，P=.022）。与没有内化障碍的儿童相比，经历过虐待的内化障碍儿童表现出衰老加速（1 - 2级：b=0.50，95%置信区间[0.170，0.821]，t=3.008，P=.003；3级或以上：b=0.99，95%置信区间[0.380，1.593]，t=3.215，P=.002）。未遭受虐待的内化障碍儿童未表现出表观遗传衰老加速。在PedBE时钟内，地塞米松反应性CpG位点有显著富集（优势比=4.36，P=1.65×10⁻⁶）。在PedBE时钟的94个CpG位点中，18个（19%）对地塞米松有反应。
使用新型PedBE时钟，我们发现内化障碍与幼儿期表观遗传衰老加速有关。这种关联受到虐待严重程度的调节，并且可能部分由糖皮质激素驱动。确定虐待后表观遗传衰老加速的发育驱动因素对于制定早期针对性干预措施至关重要。
