Whole genome resequencing of the human parasite Schistosoma mansoni reveals population history and effects of selection

Abstract

Schistosoma mansoni is a parasitic fluke that infects millions of people in the developing world. This study presents the first application of population genomics to S. mansoni based on high-coverage resequencing data from 10 global isolates and an isolate of the closely-related Schistosoma rodhaini, which infects rodents. Using population genetic tests, we document genes under directional and balancing selection in S. mansoni that may facilitate adaptation to the human host. Coalescence modeling reveals the speciation of S. mansoni and S. rodhaini as 107.5-147.6KYA, a period which overlaps with the earliest archaeological evidence for fishing in Africa. Our results indicate that S. mansoni originated in East Africa and experienced a decline in effective population size 20-90KYA, before dispersing across the continent during the Holocene. In addition, we find strong evidence that S. mansoni migrated to the New World with the 16-19th Century Atlantic Slave Trade.

Figure 1. Maximum likelihood phylogeny of Schistosoma isolates based on whole-genome SNP data.

Values on nodes are the number of bootstrap replicates supporting the split induced by that node. S. rodhaini has been set as the outgroup.

Figure 2. Principal component analyses of isolate SNP data.

(A) including both S. mansoni and S. rodhaini samples (Variance represented by PC1: 72.5%, PC2: 9.9%) (B) S. mansoni samples only, showing more details of the relatedness of isolates within this species (Variance represented by PC1: 36.0%, PC2: 13.7%).

Figure 3. Population genetic evidence for balancing and directional selection on S. mansoni genes.

McDonald-Kreitman Skew (MKS) and Hudson-Kreitman-Aguade ratio (HKAr) performed on 4147 coding sequences. Genes with a significant MKS (Fisher’s exact test, p < 0.05) are shown in red. We have selected a cutoff of HKAr > 0.25 and MKS > 1 to identify genes under balancing selection (hashed lines). Significant genes with a high HKAr and positive MKS are (A) Smp_035200- hsp40 protein (B) Smp_157450- uncharacterised protein (C) Smp_145220- Homeobox protein (D) Smp_160080- Elongator Complex Protein 6.

Figure 4. PSMC estimates of effective population size variation through time.

PSMC estimates of N_e are shown for all African S. mansoni isolates for dates between 1,000,000–1,000 years before present. Bootstrap confidence intervals for these lines are show in Supplementary Fig. S3.

Figure 5. Date of S. mansoni – S. rodhaini divergence and archaeological evidence for fishing in Africa.

The posterior probability distribution for the date of the last common ancestor of S. mansoni and S. rodhaini from G-PHoCS (in grey), with the 95% highest posterior density confidence interval in darker shading. Coloured lines show central estimates for the dates of three early fishing sites in Africa, at (A) Upper Semliki Valley, Democratic Republic of Congo (dated to between 74 and 111KYA), (B) Red Sea Basin, Eritrea (dated to between 118 and 132KYA), (C) Mossel Bay, South Africa (dated to between 152–176KYA).

Figure 6. Summary of Schistosoma mansoni population history.

Widths of branches represent approximate log N_e (effective population size. Note timescale is drawn on a log axis.