Association Between Variants of PRDM1 and NDP52 and Crohn's Disease, Based on Exome Sequencing and Functional Studies
David Ellinghaus,Hu Zhang,Sebastian Zeissig,Simone Lipinski,Andreas Till,Tao Jiang,Bjoern Stade,Yana Bromberg,Eva Ellinghaus,Andreas Keller,Manuel A. Rivas,Jurgita Skieceviciene,Nadezhda T. Doncheva,Xiao Liu,Qing Liu,Fuman Jiang,Michael Forster,Gabriele Mayr,Mario Albrecht,Robert Haesler,Bernhard O. Boehm,Jane Goodall,Carlo R. Berzuini,James Lee,Vibeke Andersen,Ulla Vogel,Limas Kupcinskas,Manfred Kayser,Michael Krawczak,Susanna Nikolaus,Rinse K. Weersma,Cyriel Y. Ponsioen,Miquel Sans,Cisca Wijmenga,David P. Strachan,Wendy L. McAardle,Severine Vermeire,Paul Rutgeerts,Jeremy D. Sanderson,Christopher G. Mathew,Morten H. Vatn,Jun Wang,Markus M. Noethen,Richard H. Duerr,Carsten Buening,Stephan Brand,Juergen Glas,Juliane Winkelmann,Thomas Illig,Anna Latiano,Vito Annese,Jonas Halfvarson,Mauro D'Amato,Mark J. Daly,Michael Nothnagel,Tom H. Karlsen,Suresh Subramani,Philip Rosenstiel,Stefan Schreiber,Miles Parkes,Andre Franke
DOI: https://doi.org/10.1053/j.gastro.2013.04.040
IF: 29.4
2013-01-01
Gastroenterology
Abstract:BACKGROUND & AIMS:Genome-wide association studies (GWAS) have identified 140 Crohn's disease (CD) susceptibility loci. For most loci, the variants that cause disease are not known and the genes affected by these variants have not been identified. We aimed to identify variants that cause CD through detailed sequencing, genetic association, expression, and functional studies.METHODS:We sequenced whole exomes of 42 unrelated subjects with CD and 5 healthy subjects (controls) and then filtered single nucleotide variants by incorporating association results from meta-analyses of CD GWAS and in silico mutation effect prediction algorithms. We then genotyped 9348 subjects with CD, 2868 subjects with ulcerative colitis, and 14,567 control subjects and associated variants analyzed in functional studies using materials from subjects and controls and in vitro model systems.RESULTS:We identified rare missense mutations in PR domain-containing 1 (PRDM1) and associated these with CD. These mutations increased proliferation of T cells and secretion of cytokines on activation and increased expression of the adhesion molecule L-selectin. A common CD risk allele, identified in GWAS, correlated with reduced expression of PRDM1 in ileal biopsy specimens and peripheral blood mononuclear cells (combined P = 1.6 × 10(-8)). We identified an association between CD and a common missense variant, Val248Ala, in nuclear domain 10 protein 52 (NDP52) (P = 4.83 × 10(-9)). We found that this variant impairs the regulatory functions of NDP52 to inhibit nuclear factor κB activation of genes that regulate inflammation and affect the stability of proteins in Toll-like receptor pathways.CONCLUSIONS:We have extended the results of GWAS and provide evidence that variants in PRDM1 and NDP52 determine susceptibility to CD. PRDM1 maps adjacent to a CD interval identified in GWAS and encodes a transcription factor expressed by T and B cells. NDP52 is an adaptor protein that functions in selective autophagy of intracellular bacteria and signaling molecules, supporting the role of autophagy in the pathogenesis of CD.