A comparison of DNA methylation in newborn blood samples from infants with and without orofacial clefts

Zongli XuView ORCID ID profile,
Rolv T. Lie,
Allen J. Wilcox,
Ola Didrik Saugstad and
Jack A. TaylorEmail author

Clinical Epigenetics201911:40

https://doi.org/10.1186/s13148-019-0638-9

Received: 29 October 2018
Accepted: 20 February 2019
Published: 4 March 2019

Abstract

Background

Isolated orofacial clefts are among the most common congenital birth defects. Although the underlying biological mechanisms remain largely unknown, clefts are thought to be complex disorders influenced by genetic, environmental, and potentially epigenetic factors.

Methods

In blood samples from 2- to 3-day-old infants (n = 747) collected in a nationwide population-based study of orofacial clefts in Norway, we measured DNA methylation profiles for more than 450,000 CpGs and then conducted epigenome-wide association analyses (EWAS). We tested methylation profile difference at each CpG between controls (n = 436) and each of the cleft subtypes (92 cleft lip only, CLO; 84 cleft palate only, CPO; 132 cleft lip and palate, CLP). We also compared controls to various combinations of case groups and compared case subtypes to each other. Finally, using the EWAS results, we searched for larger differentially methylated regions (DMRs) associated with orofacial clefts.

Results

In EWAS comparing controls to individual cleft subtypes, we found no significant associations at a Bonferroni P value threshold of 10−7. After pooling case groups, we found two significantly differentially methylated CpGs: cg09696939 near gene BICC1 is associated with CLO+CLP (P = 9.58 × 10−8); cg26985354 in gene CLASRP is associated with CPO+CLP (P = 7.38 × 10−8). In DMR analysis, we identified a total of 56 significant regions when comparing controls to individual cleft subtypes (10 for CLO, 6 for CPO, 41 for CLP). Only one DMR is shared among the three cleft groups. In combined case group analysis, we found 26 DMRs for CLP+CLO, 31 for CLP+CPO, and 37 when all subtypes are combined. Finally, in case-case comparisons of subtypes, we identified 10 DMRs when comparing CLP to CPO, 9 in CLP compared to CLO, and 13 in CLP compared to CPO.

Conclusions

We identified two individual CpGs and multiple DMRs that differ between controls and cleft case subtypes. Although we find some evidence for the possible role of DNA methylation in etiology of orofacial clefts, our study does not support previous reports of widespread differences in blood DNA methylation between babies with and without facial clefts.