Cord blood DNA methylation reflects cord blood C-reactive protein levels but not maternal levels: a longitudinal study and meta-analysis

Yeung, Edwina H.; Guan, Weihua; Zeng, Xuehuo; Salas Díaz, Lucas Andrés, 1980-; Mumford, Sunni L.; Prado Bert, Paula de; van Meel, Evelien R.; Malmberg, Anni; Sunyer Deu, Jordi; Duijts, Liesbeth; Felix, Janine Frédérique; Czamara, Darina; Hämäläinen, Esa; Binder, Elisabeth B.; Räikkönen, Katri; Lahti, Jari; London, Stephanie J.; Silver, Robert M.; Schisterman, Enrique F.

doi:http://dx.doi.org/10.1186/s13148-020-00852-2

Cord blood DNA methylation reflects cord blood C-reactive protein levels but not maternal levels: a longitudinal study and meta-analysis

Citation

Yeung EH, Guan W, Zeng X, Salas LA, Mumford SL, de Prado Bert P, van Meel ER, Malmberg A, Sunyer J, Duijts L, Felix JF, Czamara D, Hämäläinen E, Binder EB, Räikkönen K, Lahti J, London SJ, Silver RM, Schisterman EF. Cord blood DNA methylation reflects cord blood C-reactive protein levels but not maternal levels: a longitudinal study and meta-analysis. Clin Epigenetics. 2020; 12(1):60. DOI: 10.1186/s13148-020-00852-2

Abstract

Background: Prenatal inflammation has been proposed as an important mediating factor in several adverse pregnancy outcomes. C-reactive protein (CRP) is an inflammatory cytokine easily measured in blood. It has clinical value due to its reliability as a biomarker for systemic inflammation and can indicate cellular injury and disease severity. Elevated levels of CRP in adulthood are associated with alterations in DNA methylation. However, no studies have prospectively investigated the relationship between maternal CRP levels and newborn DNA methylation measured by microarray in cord blood with reasonable epigenome-wide coverage. Importantly, the timing of inflammation exposure during pregnancy may also result in different effects. Thus, our objective was to evaluate this prospective association of CRP levels measured during multiple periods of pregnancy and in cord blood at delivery which was available in one cohort (i.e., Effects of Aspirin in Gestation and Reproduction trial), and also to conduct a meta-analysis with available data at one point in pregnancy from three other cohorts from the Pregnancy And Childhood Epigenetics consortium (PACE). Secondarily, the impact of maternal randomization to low dose aspirin prior to pregnancy on methylation was assessed. Results: Maternal CRP levels were not associated with newborn DNA methylation regardless of gestational age of measurement (i.e., CRP at approximately 8, 20, and 36 weeks among 358 newborns in EAGeR). There also was no association in the meta-analyses (all p > 0.5) with a larger sample size (n = 1603) from all participating PACE cohorts with available CRP data from first trimester (< 18 weeks gestation). Randomization to aspirin was not associated with DNA methylation. On the other hand, newborn CRP levels were significantly associated with DNA methylation in the EAGeR trial, with 33 CpGs identified (FDR corrected p < 0.05) when both CRP and methylation were measured at the same time point in cord blood. The top 7 CpGs most strongly associated with CRP resided in inflammation and vascular-related genes. Conclusions: Maternal CRP levels measured during each trimester were not associated with cord blood DNA methylation. Rather, DNA methylation was associated with CRP levels measured in cord blood, particularly in gene regions predominately associated with angiogenic and inflammatory pathways.