Abstract LB1.6
Genetic determinants of multiple sclerosis in African Americans Print
N Isobe1, L Madireddy1, P Khankhanian1, T Matsushita1,2, SJ Caillier1, JM More1, P-A Gourraud1, JL Mccauley3, S Onengut-Gumuscu4, SS Rich4, SL Hauser1, S Sawcer5, JR Oksenberg1, International Multiple Sclerosis Genetics Consortium
1University of California, San Francisco, Neurology, San Francisco, CA, United States, 2Kyushu University, Neurology, Fukuoka, Japan, 3University of Miami, Miller School of Medicine, John P. Hussman Institute for Human Genomics, Miami, FL, United States, 4University of Virginia, Center for Public Health Genomics, Charlottesville, VA, United States, 5University of Cambridge, Clinical Neurosciences, Cambridge, United Kingdom
Background: While substantial progress has been made in the past decade to identify genetic variants associated with multiple sclerosis (MS) risk in populations of European descent, it remains unclear whether these variants also influence disease susceptibility in non-European populations with lower rates of prevalence.
Objectives: We assess whether the most updated non-MHC MS-associated single nucleotide polymorphisms (MSSNPs) discovered in individuals with European ancestry (Nature Genet, 2013; 45:1353-60) also influence susceptibility in African Americans. We also search for novel African American MS-associated loci.
Methods: With the ImmunoChip custom genotyping array, we analyzed 803 African American MS cases and 1,516 controls across a total of 130,135 SNPs that passed quality controls. An association analysis was conducted with rigorous adjustments for admixture and stratification, and the replication status for the European MSSNPs was evaluated. Significantly associated SNPs (p< 10-4) outside the established risk loci (1Mb centromeric and 1Mb telomeric from the designated MSSNPs) were defined as candidates tagging potential novel MS loci in African Americans and were further tested in an independent cohort consisting of 620 African American cases and 1,565 controls.
Results: Out of 96 European non-MHC MSSNPs available in the study dataset, 69 (71.9%) associated in the same direction in African Americans as in Europeans, demonstrating an excess of concordance (binomial test p=1.1x10-5). However, only 21 MSSNPs (21.9%) were formally replicated in African Americans (one-tailed test p< 0.05), which was within the range of expectation considering the average statistical power of this dataset (binomial test p=0.8941). Among the 86 MS loci where European MSSNPs were not replicated or not available, eight regions included significant risk-tagging SNPs within the LD blocks of the European MSSNPs, suggesting occasional success in narrowing MS-associated loci. In addition seven novel candidate MS loci were identified, one of which (top SNP: rs2702177) on chromosome 1 was replicated in the independent sample set (replication one-tailed p=0.034, joint p=6.3x10-5).
Conclusions: These results show a partial replication of European MS variants in African Americans, suggesting shared genetic contributions to MS risk. At the same time, in some MS-associated loci, different SNPs may more appropriately tag risk in non-European populations.

Assigned speakers:
MD, PhD Noriko Isobe , University of California, San Francisco , San Francisco , US

Assigned in sessions:
13.09.2014, 08:30-10:00, Late breaking news, LB1, Late Breaking News (LB1.1-LB1.7), Auditorium