DNA methylation in inflammatory genes among children with obstructive sleep apnea

Abstract

Background: Pediatric obstructive sleep apnea (OSA) leads to multiple end-organ morbidities that are mediated by the cumulative burden of oxidative stress and inflammation. Because not all children with OSA exhibit increased systemic inflammation, genetic and environmental factors may be affecting patterns of DNA methylation in genes subserving inflammatory functions.

Methods: DNA from matched children with OSA with and without high levels of high-sensitivity C-reactive protein (hsCRP) were assessed for DNA methylation levels of 24 inflammatory-related genes. Primer-based polymerase chain reaction assays in a case-control setting involving 47 OSA cases and 31 control subjects were conducted to confirm the findings; hsCRP and myeloid-related protein (MRP) 8/14 levels were also assayed.

Measurements and main results: Forkhead box P3 (FOXP3) and interferon regulatory factor 1 (IRF1) showed higher methylation in six children with OSA and high hsCRP levels compared with matched children with OSA and low hsCRP levels (P < 0.05). In the case-control cohort, children with OSA and high CRP levels had higher log FOXP3 DNA methylation levels compared with children with OSA and low CRP levels and control subjects. IRF1 did not exhibit significant differences. FOXP3 DNA methylation levels correlated with hsCRP and MRP 8/14 levels and with apnea-hypopnea index (AHI), BMI z score, and apolipoprotein B levels. A stepwise multiple regression model showed that AHI was independently associated with FOXP3 DNA methylation levels (P < 0.03).

Conclusions: The FOXP3 gene, which regulates expression of T regulatory lymphocytes, is more likely to display increased methylation among children with OSA who exhibit increased systemic inflammatory responses. Thus, epigenetic modifications may constitute an important determinant of inflammatory phenotype in OSA, and FOXP3 DNA methylation levels may provide a potential biomarker for end-organ vulnerability.

Figure 1.

Semi-DNA methylation array data. Heatmap of DNA methylation profile that includes 24 inflammatory related genes in children with obstructive sleep apnea (OSA) and high levels of high-sensitivity C-reactive protein (hsCRP) (n = 6) and matched children with OSA and low hsCRP levels (n = 6) (see Table E1).

Figure 2.

Prevalence of Forkhead box P3 (FOXP3) DNA hypermethylation in children with obstructive sleep apnea (OSA) and control subjects. *FOXP3 DNA hypermethylation is defined as > 24%. hsCRP = high-sensitivity C-reactive protein.

Figure 3.

Prevalence of Forkhead box P3 (FOXP3) DNA hypermethylation in children according to the categorical severity of obstructive sleep apnea (OSA). *FOXP3 DNA hypermethylation was defined as > 24%.

Figure 4.

Correlation between log high-sensitivity C-reactive protein (hsCRP), myeloid-related protein (MRP) 8/14, and apolipoprotein B levels and log Forkhead box P3 (FOXP3) DNA methylation levels in children with obstructive sleep apnea and control subjects. (A) Correlation between serum log hsCRP levels and log FOXP3 DNA methylation levels. (B) Correlation between plasma log MRP 8/14 levels and log FOXP3 DNA methylation levels. (C) Correlation between serum apolipoprotein B and log FOXP3 DNA methylation levels.

Figure 5.

Correlation between apnea-hypopnea index (AHI), body mass index (BMI) z score, and Forkhead box P3 (FOXP3) DNA methylation levels in children. (A) Correlation between AHI and FOXP3 DNA methylation levels. (B) Correlation between BMI z score and FOXP3 DNA methylation levels.