Genome-wide insights into adaptive hybridisation across the Schistosoma haematobium group in West and Central Africa

Abstract

Schistosomiasis remains a public health concern across sub-Saharan Africa; current control programmes rely on accurate mapping and high mass drug administration (MDA) coverage to attempt disease elimination. Inter-species hybridisation can occur between certain species, changing epidemiological dynamics within endemic regions, which has the potential to confound control interventions. The impact of hybridisation on disease dynamics is well illustrated in areas of Cameroon where urogenital schistosomiasis, primarily due to Schistosoma haematobium and hybrid infections, now predominate over intestinal schistosomiasis caused by Schistosoma guineensis. Genetic markers have shown the ability to identify hybrids, however the underlying genomic architecture of divergence and introgression between these species has yet to be established. In this study, restriction site associated DNA sequencing (RADseq) was used on archived adult worms initially identified as; Schistosoma bovis (n = 4), S. haematobium (n = 9), S. guineensis (n = 3) and S. guineensis x S. haematobium hybrids (n = 4) from Mali, Senegal, Niger, São Tomé and Cameroon. Genome-wide evidence supports the existence of S. guineensis and S. haematobium hybrid populations across Cameroon. The hybridisation of S. guineensis x S. haematobium has not been demonstrated on the island of São Tomé, where all samples showed no introgression with S. haematobium. Additionally, all S. haematobium isolates from Nigeria, Mali and Cameroon indicated signatures of genomic introgression from S. bovis. Adaptive loci across the S. haematobium group showed that voltage-gated calcium ion channels (Cav) could play a key role in the ability to increase the survivability of species, particularly in host systems. Where admixture has occurred between S. guineensis and S. haematobium, the excess introgressive influx of tegumental (outer helminth body) and antigenic genes from S. haematobium has increased the adaptive response in hybrids, leading to increased hybrid population fitness and viability.

Fig 1. Geographic locations and the initial species identification of the isolates analysed.

The map was created in Quantum GIS, version 3.14 (Pi), and the satellite layer sourced from ‘openlayers’ via the OpenLayers python plugin deriving from OpenStreetMaps Satellite data. Base-layer tileset were downloaded as follows: Mali Tile Bounds -8.30625, 12.45391, -3.97171, 14.69223; Senegal Tile Bounds -16.53588, 15.91597, -14.56542, 16.70207; Nigeria Tile Bounds 5.07816, 8.90230, 8.62072, 12.09504; Cameroon Tile Bounds 9.22728, 3.66772, 10.32132, 4.91165; Sao Tome and Principe Tile Bounds 6.07606, -0.47990, 9.78499, 1.98712; West African Tile Bounds -17.40185,-1.58269,14.02316,16.63042.

Fig 2. Output from STRUCTURE analysis using K = 3, see S3 Fig for comparisons with K = 1–10.

Each bar represents an individual and the relative proportion of its genotype belonging to each genetic cluster. The analysis was performed excluding differentially fixed SNPs, outlier loci and adjusted for computational power. Total number of loci used in the analysis was 10,000.

Fig 3

A) Multivariate analysis illustrating the genetic differentiation between the isolates. Analysis was performed with 19 isolates showing relative percentages of variance explained by the first eight principal components B) PCoA plot of each species with average Euclidean distance of clusters in top right hand corner.

Fig 4. Plot of F_st outliers and genome scan of all identified loci that match to the indexed S. mansoni genome.

On the left F_st values obtained from BAYESCAN, each plot figure represents outliers in each simulated comparison: A) S. haematobium x S. bovis, B) S. haematobium x S. guineensis, C) S. guineensis x S. bovis. Loci that fall to the right of the vertical line have a probability greater than 50% of being designated outlier loci. Genome scan plots on the right show genomic location of outlier loci across the genome.

Fig 5

Introgression of A) Autosomal, B) Sex-linked and C) Antigen derived candidate gene sets. Genomic clines of candidate genes with excess ancestry, the grey shaded region represent the neutral genome background and are bounded by the positive and negative means of α. The (blue) colour lines represent outlier locus clines with evidence of S. guineensis (top left) and S. haematobium (bottom right) ancestry. Green lines represent genes that present heterozygous advantage, with shared ancestry between two parental species. When both α_ij (genomic cline center) and β_ij (genomic cline rate) are both set to 0, this describes a on-to-one relationship of the hybrid index, describing neutral ancestry expectations.