AUTHOR=Reimann Brigitte , Janssen Bram G. , Alfano Rossella , Ghantous Akram , Espín-Pérez Almudena , de Kok Theo M. , Saenen Nelly D. , Cox Bianca , Robinson Oliver , Chadeau-Hyam Marc , Penders Joris , Herceg Zdenko , Vineis Paolo , Nawrot Tim S. , Plusquin Michelle 

TITLE=The Cord Blood Insulin and Mitochondrial DNA Content Related Methylome

JOURNAL=Frontiers in Genetics

VOLUME=10

YEAR=2019

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2019.00325

DOI=10.3389/fgene.2019.00325

ISSN=1664-8021

ABSTRACT=<p>Mitochondrial dysfunction seems to play a key role in the etiology of insulin resistance. At birth, a link has already been established between mitochondrial DNA (mtDNA) content and insulin levels in cord blood. In this study, we explore shared epigenetic mechanisms of the association between mtDNA content and insulin levels, supporting the developmental origins of this link. First, the association between cord blood insulin and mtDNA content in 882 newborns of the ENVIR<italic>ON</italic>AGE birth cohort was assessed. Cord blood mtDNA content was established via qPCR, while cord blood levels of insulin were determined using electrochemiluminescence immunoassays. Then the cord blood DNA methylome and transcriptome were determined in 179 newborns, using the human 450K methylation Illumina and Agilent Whole Human Genome 8 × 60 K microarrays, respectively. Subsequently, we performed an epigenome-wide association study (EWAS) adjusted for different maternal and neonatal variables. Afterward, we focused on the 20 strongest associations based on <italic>p</italic>-values to assign transcriptomic correlates and allocate corresponding pathways employing the R packages ReactomePA and RDAVIDWebService. On the regional level, we examined differential methylation using the DMRcate and Bumphunter packages in R. Cord blood mtDNA content and insulin were significantly correlated (<italic>r</italic> = 0.074, <italic>p</italic> = 0.028), still showing a trend after additional adjustment for maternal and neonatal variables (<italic>p</italic> = 0.062). We found an overlap of 33 pathways which were in common between the association with cord blood mtDNA content and insulin levels, including pathways of neurodevelopment, histone modification, cytochromes P450 (CYP)-metabolism, and biological aging. We further identified a DMR annotated to Repulsive Guidance Molecule BMP Co-Receptor A (<italic>RGMA</italic>) linked to cord blood insulin as well as mtDNA content. Metabolic variation in early life represented by neonatal insulin levels and mtDNA content might reflect or accommodate alterations in neurodevelopment, histone modification, CYP-metabolism, and aging, indicating etiological origins in epigenetic programming. Variation in metabolic hormones at birth, reflected by molecular changes, might via these alterations predispose children to metabolic diseases later in life. The results of this study may provide important markers for following targeted studies.</p>