ABSTRACT: Our previous research demonstrated that caffeine exposure on embryonic day (E) 8.5 increased body weight, altered cardiac morphology and function in adult male mice. However, these adverse effects were not observed in adenosine A1 receptor knockout (A1AR-/-) mice. Our hypothesis is that A1AR action mediates changes in DNA methylation patterns in mice exposed in utero to caffeine and that these changes in methylation lead to long-term effects in adult mice. To test this hypothesis, DNA Methylation 2.1M Deluxe Promoter Arrays (NimbleGen) were used to examine the methylation patterns of DNA isolated from adult left ventricles of the A1AR knockout line exposed to 20 mg/kg caffeine at E8.5 in utero. In A1AR+/+ mice, 4,896 hypermethylated and 2,823 hypomethylated regions were discovered in the left ventricles of the caffeine group compared to the normal saline group. In A1AR-/- mice, 1,024 hypermethylated and 1,757 hypomethylated regions were found in the caffeine group. The differentially methylated regions (DMRs) in the caffeine treated A1AR+/+ hearts mapped to 6,148 genes, 4,853 promoters, 4,111 primary transcripts, 816 CpG islands, and 98 miRNAs. Functional annotation clustering of genes revealed that many genes were involved in the development of hypertrophic cardiomyopathy and other heart diseases, which may explain our earlier findings of thickening of the left ventricular wall after in utero caffeine treatment. The methylation changes in several DMRs, mef2c, ins2, tnnt2, and myh6, were validated by bisulfite sequencing. In summary, in utero caffeine exposure caused DNA methylation changes in adult left ventricles and that these changes may be mediated by A1ARs. Pregnant mice were injected at embryonic day 8.5 with 0.9% NaCl (vehicle) or 20 mg/kg caffeine in vehicle i.p. Pups were born and raised until 8-10 weeks of age. Analysis was performed on the following groups vehicle A1AR+/+ (veh+/+), vehicle A1AR-/- (veh-/-), caffeine A1AR+/+ (caff+/+), and caffeine A1AR-/- (caff-/-). Genomic DNA from left ventricles of adult male mice was used. Two biological replicates were measured for each group.
