Image: An Epstein-Barr virus budding in a B cell (Photo courtesy of Kenyon College).
Epstein-Barr Virus (EBV) infection is common and more than 90% of people in the USA become infected by age 20. While resulting in mild disease or no symptoms in children, EBV causes mononucleosis in adolescents and young adults, with symptoms that include fever and extreme fatigue.
EBV had already been implicated in systemic lupus erythematosus (SLE), increasing the risk for the disease by up to 50-fold in children, but the molecular mechanism by which the virus contributes to SLE had been unknown. In addition, genome-wide association studies had previously identified more than 50 SLE risk loci in cohorts with European ancestry.
Scientists at the Cincinnati Children’s Hospital Medical Center (Cincinnati, OH, USA) developed the Regulatory Element Locus Intersection (RELI) algorithm, which looks for overlap between disease-associated genetic variants and transcription factor binding sites, as determined by ChIP-seq studies. They investigated whether RELI could recapitulate already known relationships between transcription factors and diseases and found that it correctly identified links between the androgen receptor and prostate cancer, between GATA3 and breast cancer, and between EBV-encoded EBNA2 and multiple sclerosis.
They then analyzed 53 SLE risk loci in European-ancestry individuals together with ChIP-seq data for several EBV-encoded transcription factors, gathered from EBV-infected B cells, and found that EBNA2 binding sites overlap significantly with SLE risk loci, whereas other EBV protein binding sites do not. Next, they applied RELI to a large number of human transcription factor ChIP-seq datasets and found that 60 transcription factors also intersected with SLE loci. Further studies showed that several of these transcription factors bound SLE risk loci only in the presence of EBV and suggested a role for EBV-infected B cells in the disease.
The team went on to run RELI on 213 other diseases and phenotypes and they identified strong associations between EBNA2-binding and risk loci for nine additional conditions or phenotypes: rheumatoid arthritis, inflammatory bowel syndrome, type 1 diabetes, juvenile idiopathic arthritis, celiac disease, chronic lymphocytic leukemia, Kawasaki disease, ulcerative colitis, and immunoglobulin glycosylation. Applying RELI to GWAS data for 213 other phenotypes, they discovered additional associations between transcription factors and diseases that were independent of EBNA2, for example between hematopoietic phenotypes and blood cell regulators; between GATA3, FOXA1, and TCF7L2 and breast cancer; and between AR, NR3C1, and EZH2 and prostate cancer. The study was published on April 16, 2018, in the journal Nature Genetics.
Cincinnati Children’s Hospital Medical Center