Comparative genomics revealed adaptive admixture in Cryptosporidium hominis in Africa

Abstract

Cryptosporidiosis is a major cause of diarrhoeal illness among African children, and is associated with childhood mortality, malnutrition, cognitive development and growth retardation. Cryptosporidium hominis is the dominant pathogen in Africa, and genotyping at the glycoprotein 60 (gp60) gene has revealed a complex distribution of different subtypes across this continent. However, a comprehensive exploration of the metapopulation structure and evolution based on whole-genome data has yet to be performed. Here, we sequenced and analysed the genomes of 26 C. hominis isolates, representing different gp60 subtypes, collected at rural sites in Gabon, Ghana, Madagascar and Tanzania. Phylogenetic and cluster analyses based on single-nucleotide polymorphisms showed that isolates predominantly clustered by their country of origin, irrespective of their gp60 subtype. We found a significant isolation-by-distance signature that shows the importance of local transmission, but we also detected evidence of hybridization between isolates of different geographical regions. We identified 37 outlier genes with exceptionally high nucleotide diversity, and this group is significantly enriched for genes encoding extracellular proteins and signal peptides. Furthermore, these genes are found more often than expected in recombinant regions, and they show a distinct signature of positive or balancing selection. We conclude that: (1) the metapopulation structure of C. hominis can only be accurately captured by whole-genome analyses; (2) local anthroponotic transmission underpins the spread of this pathogen in Africa; (3) hybridization occurs between distinct geographical lineages; and (4) genetic introgression provides novel substrate for positive or balancing selection in genes involved in host-parasite coevolution.

Keywords: Africa; Cryptosporidium hominis; genetic introgression; population structure; recombination; whole-genome sequencing.

Fig. 1.

Phylogenetic relationships among C. hominis African isolates. Isolates are coloured according to their geographical origin in panels (a–f). (a) Principal component analysis (PCA), where PC1 and PC2 account for variability among C. hominis African isolates, differentiating them by physical boundaries. (b) Phylogenetic analysis inferred by Bayesian inference (MrBayes) using a concatenated set of 3128 genomic SNPs. (c) A cloudogram of 2130000 trees obtained by DensiTree. The consensus tree is represented in dark blue. Note the transparent green branches (indicated by the arrow), which reveals that a small part of the Madagascar genomes has a different phylogenetic origin. (d) A phylogenetic network based on all genomic SNPs. The network shows geographical clustering of African isolates by the country of origin. The loops in the network are suggestive of recombination events both within and between isolates from different countries. (e) Network of the gp60 gene reveals admixture of samples from Madagascar and Tanzania. (f) Network of the CHUDEA6_5260 gene shows no evidence of admixture.

Fig. 2.

Population structure and recombination analyses. (a) structure plot representing the percentage of shared ancestry among the four African C. hominis metapopulation (for K=6). (b) HybridCheck plot illustrating the sequence similarity among the three isolates (Madagascar Afr12, Ghana Afr9 and Tanzania Afr14) involved in recombination at the gp60 locus.

Fig. 3.

Polymorphism and linkage analyses. Pa(a) Nucleotide variation at chromosome 6 is significantly higher than that of all other chromosomes (GLM: F _1,32=25.16, P=5.3e-7), and coding regions show a higher diversity than con-coding regions (GLM: F _1,72=56.65, P=5.4e-14). (b) LD decay versus the mean distance between SNPs calculated across the genome for the African metapopulation and the Bangladeshi population. Also shown is the LD decay for chromosome 6 and all other chromosomes in the African metapopulation. The LD decays more rapidly in the African sample compared to the Bangladeshi, and the LD decay is particularly rapid for SNPs in chromosome 6. (c) Nucleotide diversity (π) of each polymorphic gene. The 37 genes above the threshold (dashed line) are considered to be outliers, and genes with π>0.002 are labelled. (d) Tajima’s D values for each polymorphic gene. Genes with >3 Tajima’s D values are labelled. In (c) and (d) each chromosome is represented with a different colour.

Fig. 4.

Distribution and density plots. (a) Density of K _a/K _s ratio of polymorphic genes. K _a/K _s ratios of the CHUDEA2_430 (red arrow), CHUDEA2_440 (blue arrow) and CHUDEA2_450 (green arrow) genes are labelled along with their ranks. The CHUDEA2_430 gene has the highest K _a/K _s ratio (K _a/K _s >1) and hence acts as an outlier. Panels (b) and (c) represent density plots of polymorphic genes in recombinant and recombinant-free regions. (b) Nucleotide diversity. (c) Haplotype diversity. Green and red arrows indicate values for the CHUDEA6_1070 and CHUDEA6_1080 (gp60) genes, respectively, in each panel.