Article Text
Abstract
Background Systemic lupus erythematosus (SLE) is a heterogeneous disease in which epigenetic and environmental risk factors have been implicated. DNA methylation can be influenced by environmental exposures. Exposure to motor vehicle emissions has been linked to increase in overall mortality, asthma and increase incidence of rheumatoid arthritis. The focus of this study was to evaluate methylation changes in relationship to residential proximity to highways in patients with SLE.
Materials and methods We studied 307 patients with SLE who were previously enrolled in a Lupus Genetics Project. As a replication cohort, 225 participants from a gene-environment study of asthma were studied. Residence at the time of blood draw was recorded and geocoded. The distance to the nearest roads from the geocoded locations were calculated for the four major Tele Atlas Feature Class Codes (FCC) road classes. The Geographic Data Technology, Inc. (GDT) road network data were used for these calculations. Genome-wide methylation profiling was performed using the Illumina Infinium HumanMethylation 450 BeadChip.
Results Patients residing within a 300 metre radius from a major highway were defined as at high risk for significant hazardous health outcomes1. Thirty-eight patients (12.4%) were residing in a high risk area. Multivariate analysis did not reveal any statistically significant association between proximity to highways and disease phenotypes, however there was a trend for higher incidence of discoid rash (OR 1.5) neurological disorders (OR 1.4) and antiphospholipid antibodies (1.4). Analysis of genome-wide methylation data revealed 3 methylation sites that were significantly hypomethylated in patients who resided in a high risk zone (P value < 1.7 E-07, Table 1). These three sites belonged to a single gene, UBE2U, which encodes one of the E2 enzymes involved in the ubiquitination of proteins and histones, as well as DNA repair. To replicate these findings, CpG Sites for all the UBE gene family were analysed in a control group. Our 3 top hits were not replicated, however one CpG site (cg22352634) belonging to the first intron of the gene UBE2U was found hypomethylated in controls who resided in a high risk zone, with a p value of 0.0044.
Conclusions Hypomethylation of UBE2U was associated with residing close to a highway in our SLE patients, however this was not seen in a control cohort, and suggesting increased susceptibility for exposures in patients with SLE. Additional work is warranted to confirm these findings, examine other potentially relevant exposures, and determine whether these epigenetic changes are associated with increased UBE2U expression.