Admixture mapping reveals evidence of differential multiple sclerosis risk by genetic ancestry

Abstract

Multiple sclerosis (MS) is an autoimmune disease with high prevalence among populations of northern European ancestry. Past studies have shown that exposure to ultraviolet radiation could explain the difference in MS prevalence across the globe. In this study, we investigate whether the difference in MS prevalence could be explained by European genetic risk factors. We characterized the ancestry of MS-associated alleles using RFMix, a conditional random field parameterized by random forests, to estimate their local ancestry in the largest assembled admixed population to date, with 3,692 African Americans, 4,915 Asian Americans, and 3,777 Hispanics. The majority of MS-associated human leukocyte antigen (HLA) alleles, including the prominent HLA-DRB1*15:01 risk allele, exhibited cosmopolitan ancestry. Ancestry-specific MS-associated HLA alleles were also identified. Analysis of the HLA-DRB1*15:01 risk allele in African Americans revealed that alleles on the European haplotype conferred three times the disease risk compared to those on the African haplotype. Furthermore, we found evidence that the European and African HLA-DRB1*15:01 alleles exhibit single nucleotide polymorphism (SNP) differences in regions encoding the HLA-DRB1 antigen-binding heterodimer. Additional evidence for increased risk of MS conferred by the European haplotype were found for HLA-B*07:02 and HLA-A*03:01 in African Americans. Most of the 200 non-HLA MS SNPs previously established in European populations were not significantly associated with MS in admixed populations, nor were they ancestrally more European in cases compared to controls. Lastly, a genome-wide search of association between European ancestry and MS revealed a region of interest close to the ZNF596 gene on chromosome 8 in Hispanics; cases had a significantly higher proportion of European ancestry compared to controls. In conclusion, our study established that the genetic ancestry of MS-associated alleles is complex and implicated that difference in MS prevalence could be explained by the ancestry of MS-associated alleles.

Fig 1. Multidimensional scaling analysis of study subjects with HGDP reference samples.

(A) Study subjects from Northern California Kaiser Permanente, Southern California Kaiser Permanente, U.S. Pediatric MS Network, Genetic Epidemiology Research on Aging datasets, and (B) IMSGC Immunochip.

Fig 2. Global admixture proportions of study subjects.

Global admixture proportion estimates by fastSTRUCTURE with HGDP reference samples. Proportions are shown by case/control status and with cases and controls combined for (A) African Americans, (B) Hispanics, and (C) Asian Americans. The x-axis label ‘All’ denotes admixture proportions for cases and controls combined. See Table 1 for the sample numbers corresponding to each admixed population.

Fig 3. Deviation of local from global European ancestry.

The difference between local and global ancestry at the major histocompatibility complex (MHC) region, plotted for cases and controls. The red vertical bars denote borders of class I, II, and III of the MHC. Local and global ancestries are estimated with RFMix. For both (A) African Americans and (B) Hispanics, cases tended to have higher European ancestry than controls at the MHC. For (C) Asian Americans, cases tended to have lower European ancestry than controls at the MHC.

Fig 4. Comparison of MS-associated HLA alleles across populations.

P-value heat map for HLA alleles that reached Bonferroni significance in either Asian Americans, African Americans, Hispanics, or White, non-Hispanic individuals. P-values of HLA alleles associated with MS in White, non-Hispanic individuals were taken from previous work[11]. The HLA allele HLA-DRB1*15:01 was most consistently associated with MS across all four populations, followed by HLA-DRB1*03:01. Gray (NA) denotes an HLA allele that is missing due to not being present in the population or failed to pass HLA imputation QC.

Fig 5. Admixture of HLA alleles associated with MS.

Ancestry was inferred for MS-associated HLA alleles that passed QC using RFMix. MS-associated alleles were significant at the nominal level (p < 0.05), had imputation score R² > 0.80, and had minor allele frequency greater than 0.005. Other than HLA-B*55:01 in (A) African Americans, HLA-DRB1*15:01, HLA-DRB1*16:02, HLA-DRB1*01:01, HLA-DRB1*14:02, HLA-A*01:01 in (B) Hispanics, and HLA-B*27:05 and HLA-C*03:04 in (C) Asian Americans, HLA alleles associated with MS are cosmopolitan.

Fig 6. European and African HLA-DRB1*15:01 subsequence comparison.

Comparison of SNP and AA subsequences imputed by SNP2HLA for European and African HLA-DRB1*15:01 alleles in African Americans. Note: subsequence implies the SNPs and AAs are not necessarily contiguous. The subsequences were aligned by position (GRCh38) with the UCSC genome browser NCBI gene track; the imputed positions correspond to exons 1–5 of HLA-DRB1. The top frequent (94.1%) SNP and amino acid subsequence for European HLA-DRB1*15:01 was compared against the top two frequent (59.4% and 27.8%) subsequences for African HLA-DRB1*15:01. Red indicates a mismatch between any two given positions between an African and European allele. AA = amino acid; EUR = European; AFR = African.

Fig 7. Admixture of non-HLA MS risk variants.

Local ancestry estimates from RFMix for the non-HLA risk variants that passed QC, sorted in order of increasing European ancestry. The admixture proportions of risk variants were estimated separately in (A) African American cases, (B) African American controls, (C) Hispanic cases, (D) Hispanic controls, (E) Asian American cases, and (F) Asian American controls. The ancestry proportions of risk variants in cases and controls were largely reflective of global admixture proportions in cases and controls, respectively.

Fig 8. Genome-wide association of European ancestry with MS.

P values from testing of association between European ancestry and MS using non-parametric test statistic proposed by Montana and Pritchard, as described in Methods. One locus was selected from each 0.2 cM window used by RFMix for ancestry inference to reduce the burden of multiple hypothesis testing, resulting in 15,282 tests. The red horizontal line indicates the negative log of the Bonferroni P value (p = 3.27 × 10⁻⁶) for establishing significance. (A) None of the loci tested for African Americans demonstrated evidence for significant association. (B) In Hispanics, a region spanning from 2Mb to 3Mb on chromosome 8 showed evidence for a significant association. (C) None of the loci tested for Asian Americans were significantly associated.