Transmission and diversity of Schistosoma haematobium and S. bovis and their freshwater intermediate snail hosts Bulinus globosus and B. nasutus in the Zanzibar Archipelago, United Republic of Tanzania

Abstract

Background: The Zanzibar Archipelago (Pemba and Unguja islands) is targeted for the elimination of human urogenital schistosomiasis caused by infection with Schistosoma haematobium where the intermediate snail host is Bulinus globosus. Following multiple studies, it has remained unclear if B. nasutus (a snail species that occupies geographically distinct regions on the Archipelago) is involved in S. haematobium transmission on Zanzibar. Additionally, S. haematobium was thought to be the only Schistosoma species present on the Zanzibar Archipelago until the sympatric transmission of S. bovis, a parasite of ruminants, was recently identified. Here we re-assess the epidemiology of schistosomiasis on Pemba and Unguja together with the role and genetic diversity of the Bulinus spp. involved in transmission.

Methodology/principal findings: Malacological and parasitological surveys were conducted between 2016 and 2019. In total, 11,116 Bulinus spp. snails were collected from 65 of 112 freshwater bodies surveyed. Bulinus species identification were determined using mitochondrial cox1 sequences for a representative subset of collected Bulinus (n = 504) and together with archived museum specimens (n = 6), 433 B. globosus and 77 B. nasutus were identified. Phylogenetic analysis of cox1 haplotypes revealed three distinct populations of B. globosus, two with an overlapping distribution on Pemba and one on Unguja. For B. nasutus, only a single clade with matching haplotypes was observed across the islands and included reference sequences from Kenya. Schistosoma haematobium cercariae (n = 158) were identified from 12 infected B. globosus and one B. nasutus collected between 2016 and 2019 in Pemba, and cercariae originating from 69 Bulinus spp. archived in museum collections. Schistosoma bovis cercariae (n = 21) were identified from seven additional B. globosus collected between 2016 and 2019 in Pemba. By analysing a partial mitochondrial cox1 region and the nuclear ITS (1-5.8S-2) rDNA region of Schistosoma cercariae, we identified 18 S. haematobium and three S. bovis haplotypes representing populations associated with mainland Africa and the Indian Ocean Islands (Zanzibar, Madagascar, Mauritius and Mafia).

Conclusions/significance: The individual B. nasutus on Pemba infected with S. haematobium demonstrates that B. nasutus could also play a role in the local transmission of S. haematobium. We provide preliminary evidence that intraspecific variability of S. haematobium on Pemba may increase the transmission potential of S. haematobium locally due to the expanded intermediate host range, and that the presence of S. bovis complicates the environmental surveillance of schistosome infections.

Fig 1. Map of Pemba and Unguja islands (Zanzibar, United Republic of Tanzania) showing shehias where malacological surveys for Bulinus species were conducted in the current study and predicted distributions of Bulinus spp. and Schistosoma haematobium endemicity based on previous findings.

Bulinus spp. distribution inferred from Stothard et al. [23] and Pennance et al. [29,30]. Schistosoma haematobium infection distribution interpreted from Knopp et al. [18]. Digital shape files for Unguja and Pemba administrative regions were obtained from DIVA-GIS (https://www.diva-gis.org).

Fig 2. A: Inferred Bulinus globosus and B. nasutus distribution on Unguja and Pemba islands (Zanzibar, United Republic of Tanzania) as identified by mitochondrial cox1 sequences of a subset (n = 510) of Bulinus spp. collected. B: Highlighted South East region of Pemba, displaying human freshwater contact sites in four shehias (Matale, Pujini, Chambani, Ukutini) and the single freshwater body cohabited by B. globosus and B. nasutus (Puj11).

Digital shape files for Unguja and Pemba administrative regions were obtained from DIVA-GIS (https://www.diva-gis.org).

Fig 3. Bayesian inference of the partial mitochondrial cox1 haplotype dataset of Bulinus nasutus and B. globosus collected from Unguja and Pemba.

Reference data from East Africa (Kane et al. [42]). Tree produced using Bayesian inference using MrBayes v3.2.7A [51] under the HKY+I+G model (-lnl 2020.2144, AIC 4052.4287). Branches <0.95 posterior probability collapsed. The branch length scale bar indicates the number of substitutions per site. Text in red indicates Bulinus haplotypes generated from the current study as listed in S2 Table.

Fig 4. TCS haplotype network of Schistosoma spp. partial cox1 DNA sequences (750 bp).

Produced using PopArt [56]. Hatches represent SNP differences from joined nodes and size of nodes is scaled to the number of identical haplotypes listed. Schistosoma haplotype group 1 and 2 indicates whether cercariae were identified as mainland Africa (1) or Indian Ocean Island (2) haplotypes (as described in Webster et al. [27]). Schistosoma haematobium reference haplotypes (GU257334 –GU257360) from Webster et al. [28]. Schistosoma bovis reference haplotypes (OK484569, AY157212 and MH647141) from Pennance et al. [54], Lockyer et al. [53] and Djuikwo-Teukeng et al. [55], respectively. Schistosoma curassoni reference (AY157210) from Lockyer et al. [53].