Common variation in the ABO glycosyltransferase is associated with susceptibility to severe Plasmodium falciparum malaria

Abstract

There is growing epidemiological and molecular evidence that ABO blood group affects host susceptibility to severe Plasmodium falciparum infection. The high frequency of common ABO alleles means that even modest differences in susceptibility could have a significant impact on the health of people living in malaria endemic regions. We performed an association study, the first to utilize key molecular genetic variation underlying the ABO system, genotyping >9000 individuals across three African populations. Using population- and family-based tests, we demonstrated that alleles producing functional ABO enzymes are associated with greater risk of severe malaria phenotypes (particularly malarial anemia) in comparison with the frameshift deletion underlying blood group O: case-control allelic odds ratio (OR), 1.2; 95% confidence interval (CI), 1.09-1.32; P = 0.0003; family-studies allelic OR, 1.19; 95% CI, 1.08-1.32; P = 0.001; pooled across all studies allelic OR, 1.18; 95% CI, 1.11-1.26; P = 2 x 10(-7). We found suggestive evidence of a parent-of-origin effect at the ABO locus by analyzing the family trios. Non-O haplotypes inherited from mothers, but not fathers, are significantly associated with severe malaria (likelihood ratio test of Weinberg, P = 0.046). Finally, we used HapMap data to demonstrate a region of low F(ST) (-0.001) between the three main HapMap population groups across the ABO locus, an outlier in the empirical distribution of F(ST) across chromosome 9 (approximately 99.5-99.9th centile). This low F(ST) region may be a signal of long-standing balancing selection at the ABO locus, caused by multiple infectious pathogens including P. falciparum.

Figure 1

Linkage disequilibrium around the ABO glycosyltransferase gene. Strong linkage disequilibrium (LD) exists between the four key functional SNPs (A) Yoruba HapMap parent-offspring trios from Ibadan in Nigeria. (B) 1320 Gambian parent-offspring trios. Both of these West-African populations have near perfect LD between the three nonsynonymous SNPs that differentiate A and B ABO alleles, and moderate but lower LD with the common deletion which generates the O allele. (C) r² values for the Yoruba population across the ABO gene. Phased genotypes (HapMap, July 2006) with additional genotyping for four functional polymorphisms in ABO. The ABO gene (total region shown chromosome 9, ~24kb, 133156822 – 133180999, NCBI Build 35) is illustrated 3′ to 5′ with exons 6 and 7 on the left. The SNPs differentiating A and B blood groups are in a high LD block, and indicated by three arrows under exon 7. The frameshift mutation responsible for the O allele (arrow under exon 6) is on the 3′ edge of another high LD block.

Figure 2

Estimated risk for ABO rs8176719 in severe malaria. For each study the figure represents the estimated odds ratios (circular marks) and 95% confidence interval (CI, horizontal lines) for the full-length allele, in severe malaria. The P-values were derived from regression analysis with covariates of ethnic group, gender and HbS genotype (for population-based studies), case-pseudo-control approach (for family studies) or UNPHASED analysis (pooled data across all studies). The interrupted lines represent the CI of the pooled data for all studies, both family- and population-based, using the UNPHASED application.

Figure 3

Extended haplotype homozygosity plots around functional SNPs in the ABO locus. Alleles that have risen rapidly in frequency due to recent positive or balancing selection (e.g. a partial selective sweep) can be surrounded by a region of similar haplotypes that can extend for hundreds of kilobases (45, 46). This occurs because recombination has had insufficient time to swap variation between the selected haplotype and others in the population. The decay of homozygosity (EHH) on phased haplotypes partitioned by the alleles of the functional ABO SNPs. (A) rs8176719, the frameshift mutation in exon 6 of the ABO gene, marking O haplotypes, which is associated with protection from severe malaria. (B) rs8176746, a nonsynonymous coding SNP in the N-terminal catalytic domain of ABO, one of the functional variants determining A/B glycosyltransferase activity. The lack of a pronounced EHH signal suggests that the balanced selection affecting variation at the ABO gene is longstanding.

Figure 4

A region of Low F_ST region around the ABO gene. (A) 400 Single SNP F_ST values for three HapMap populations (CEU, YRI and combined Asian) surrounding the ABO gene. F_ST drops to a level of −0.001 in an 85 SNP window across the gene and the noncoding sequence ~20-30kb upstream. (B) Histogram representing an empirical distribution of F_ST determined by screening similarly sized windows across chromosome 9. Three different window sizes were used based on either marker numbers (85 SNPs), genetic distance (0.054 cM) or physical distance (50kb), only windows containing more than 2 markers were included. By all three distributions the region around the ABO locus is a relative outlier ~99.5 – 99.9^th centile for low F_ST.