Abstract
Background: Allopurinol-induced severe cutaneous adverse reactions (SCARs), including drug rash with eosinophilia and systemic symptoms (DRESS), Stevens-Johnson syndrome (SJS) and toxic epidermal necrosis (TEN), are life-threatening autoimmune reactions. Evidence is growing that epigenetic variation, particularly DNA methylation, is associated with autoimmune diseases. However, the potential role of aberrant DNA methylation in allopurinol-SCARs is largely unknown. Objective: To address the knowledge gap between allopurinol-SCARs and DNA methylation, we studied the DNA methylation profiles in peripheral blood cells from allopurinol-SCARs and allopurinol-tolerant subjects. Methods: A genome-scale DNA methylation profiling was conducted using the Illumina Infinium HumanMethylation450 (HM450) platform on 15 patients with allopurinol-SCARs (3 TEN, 2 SJS/TEN overlap and 10 SJS) and 20 age- and gender-matched allopurinol-tolerant controls at disease onset. Pyrosequencing was used to validate the candidate CpG (cytosine-guanine dinucleotide) sites in an independent cohort of 40 allopurinol-SCARs and 48 allopurinol-tolerants. Results: After bioinformatics analysis of methylation data obtained from HM450 BeadChip, we identified 41 differentially methylated CpG loci (P < 0.05) annotated to 26 genes showing altered DNA methylation between allopurinol-SCARs and allopurinol-tolerants. Among these genes, significant hypomethylation of PSORS1C1 (cg24926791) was further validated in a larger sample cohort, showing significant difference between DRESS and controls (P = 0.00127), ST (SJS and TEN) and controls (P = 3.75 × 10(-13)), and SCARs and controls (P = 5.93 × 10(-15)). Conclusions: Our data identified differentially methylated genes between allopurinol-SCARs and allopurinol-tolerant controls and showed that PSORS1C1 hypomethylation was associated with allopurinol-SCARs (OR = 30.22, 95%CI = 4.73-192.96) during disease onset, suggesting that aberrant DNA methylation may be a mechanism of allopurinol-SCARs. Limitations: Firstly, the data come from whole blood samples known to possess epigenetic heterogeneity, i. e., blood samples comprise a heterogeneous cell population with varying proportions of distinct cell-types with different DNA methylation patterns. Consequently, the interpretation of DNA methylation results should be performed with great caution due to the heterogeneous nature of the sample. Secondly, whether the identified disease-associated changes of epigenome precede disease onset, or result from the disease progression, needs further investigation. Comparing the methylation status before patients develop allopurinol-SCARs and after may help examine methylation levels from disease onset to disease progression.