Publications

In high-transmission regions, we expect parasite lineages within complex malaria infections to be unrelated due to parasite inoculations from different mosquitoes. This project was designed to test this prediction. We generated 485 single-cell genome sequences from fifteen P. falciparum malaria patients from Chikhwawa, Malawi-an area of intense transmission. Patients harbored up to seventeen unique parasite lineages. Surprisingly, parasite lineages within infections tend to be closely related, suggesting that superinfection by repeated mosquito bites is rarer than co-transmission of parasites from a single mosquito. Both closely and distantly related parasites comprise an infection, suggesting sequential transmission of complex infections between multiple hosts. We identified tetrads and reconstructed parental haplotypes, which revealed the inbred ancestry of infections and non-Mendelian inheritance. Our analysis suggests strong barriers to secondary infection and outbreeding amongst malaria parasites from a high transmission setting, providing unexpected insights into the biology and transmission of malaria.

Do mutations required for adaptation occur de novo, or are they segregating within populations as standing genetic variation? This question is key to understanding adaptive change in nature, and has important practical consequences for the evolution of drug resistance. We provide evidence that alleles conferring resistance to oxamniquine (OXA), an antischistosomal drug, are widespread in natural parasite populations under minimal drug pressure and predate OXA deployment. OXA has been used since the 1970s to treat Schistosoma mansoni infections in the New World where S. mansoni established during the slave trade. Recessive loss-of-function mutations within a parasite sulfotransferase (SmSULT-OR) underlie resistance, and several verified resistance mutations, including a deletion (p.E142del), have been identified in the New World. Here we investigate sequence variation in SmSULT-OR in S. mansoni from the Old World, where OXA has seen minimal usage. We sequenced exomes of 204 S. mansoni parasites from West Africa, East Africa and the Middle East, and scored variants in SmSULT-OR and flanking regions. We identified 39 non-synonymous SNPs, 4 deletions, 1 duplication and 1 premature stop codon in the SmSULT-OR coding sequence, including one confirmed resistance deletion (p.E142del). We expressed recombinant proteins and used an in vitro OXA activation assay to functionally validate the OXA-resistance phenotype for four predicted OXA-resistance mutations. Three aspects of the data are of particular interest: (i) segregating OXA-resistance alleles are widespread in Old World populations (4.29-14.91% frequency), despite minimal OXA usage, (ii) two OXA-resistance mutations (p.W120R, p.N171IfsX28) are particularly common (>5%) in East African and Middle-Eastern populations, (iii) the p.E142del allele has identical flanking SNPs in both West Africa and Puerto Rico, suggesting that parasites bearing this allele colonized the New World during the slave trade and therefore predate OXA deployment. We conclude that standing variation for OXA resistance is widespread in S. mansoni.

Introgression among parasite species has the potential to transfer traits of biomedical importance across species boundaries. The parasitic blood fluke Schistosoma haematobium causes urogenital schistosomiasis in humans across sub-Saharan Africa. Hybridization with other schistosome species is assumed to occur commonly, because genetic crosses between S. haematobium and livestock schistosomes, including S. bovis, can be staged in the laboratory, and sequencing of mtDNA and rDNA amplified from microscopic miracidia larvae frequently reveals markers from different species. However, the frequency, direction, age, and genomic consequences of hybridization are unknown. We hatched miracidia from eggs and sequenced the exomes from 96 individual S. haematobium miracidia from infected patients from Niger and the Zanzibar archipelago. These data revealed no evidence for contemporary hybridization between S. bovis and S. haematobium in our samples. However, all Nigerien S. haematobium genomes sampled show hybrid ancestry, with 3.3-8.2% of their nuclear genomes derived from S. bovis, providing evidence of an ancient introgression event that occurred at least 108-613 generations ago. Some S. bovis-derived alleles have spread to high frequency or reached fixation and show strong signatures of directional selection; the strongest signal spans a single gene in the invadolysin gene family (Chr. 4). Our results suggest that S. bovis/S. haematobium hybridization occurs rarely but demonstrate profound consequences of ancient introgression from a livestock parasite into the genome of S. haematobium, the most prevalent schistosome species infecting humans.

BACKGROUND: Parasite traits associated with transmission success, such as the number of infective stages released from the host, are expected to be optimized by natural selection. However, in the trematode parasite Schistosoma mansoni, a key transmission trait, i.e. the number of cercariae larvae shed from infected Biomphalaria spp. snails, varies significantly within and between different parasite populations and selection experiments demonstrate that this variation has a strong genetic basis. In this study, we compared the transmission strategies of two laboratory schistosome population and their consequences for their snail host.

METHODS: We infected inbred Biomphalaria glabrata snails using two S. mansoni parasite populations (SmBRE and SmLE), both isolated from Brazil and maintained in the laboratory for decades. We compared life history traits of these two parasite populations by quantifying sporocyst growth within infected snails (assayed using qPCR), output of cercaria larvae and impact on snail host physiological response (i.e. hemoglobin rate, laccase-like activity) and survival.

RESULTS: We identified striking differences in virulence and transmission between the two studied parasite populations. SmBRE (low shedder (LS) parasite population) sheds very low numbers of cercariae and causes minimal impact on the snail physiological response (i.e. laccase-like activity, hemoglobin rate and snail survival). In contrast, SmLE (high shedder (HS) parasite population) sheds 8-fold more cercariae (mean ± SE cercariae per shedding: 284 ± 19 vs 2352 ± 113), causes high snail mortality and has strong impact on snail physiology. We found that HS sporocysts grow more rapidly inside the snail host, comprising up to 60% of cells within infected snails, compared to LS sporocysts, which comprised up to 31%. Cercarial production is strongly correlated to the number of S. mansoni sporocyst cells present within the snail host tissue, although the proportion of sporocyst cells alone does not explain the low cercarial shedding of SmBRE.

CONCLUSIONS: We demonstrated the existence of alternative transmission strategies in the S. mansoni parasite consistent with trade-offs between parasite transmission and host survival: a "boom-bust" strategy characterized by high virulence, high transmission and short duration infections and a "slow and steady" strategy with low virulence, low transmission but long duration of snail host infections.

BACKGROUND: Competitive outcomes between co-infecting malaria parasite lines can reveal fitness disparities in blood stage growth. Blood stage fitness costs often accompany the evolution of drug resistance, with the expectation that relatively fitter parasites will be more likely to spread in populations. With the recent emergence of artemisinin resistance, it is important to understand the relative competitive fitness of the metabolically active asexual blood stage parasites. Genetically distinct drug resistant parasite clones with independently evolved sets of mutations are likely to vary in asexual proliferation rate, contributing to their chance of transmission to the mosquito vector.

METHODS: An optimized in vitro 96-well plate-based protocol was used to quantitatively measure-head-to-head competitive fitness during blood stage development between seven genetically distinct field isolates from a hotspot of emerging artemisinin resistance and the laboratory strain, NF54. These field isolates were isolated from patients in Southeast Asia carrying different alleles of kelch13 and included both artemisinin-sensitive and artemisinin-resistant isolates. Fluorescent labeled microsatellite markers were used to track the relative densities of each parasite throughout the co-growth period of 14-60 days. All-on-all competitions were conducted for the panel of eight parasite lines (28 pairwise competitions) to determine their quantitative competitive fitness relationships.

RESULTS: Twenty-eight pairwise competitive growth outcomes allowed for an unambiguous ranking among a set of seven genetically distinct parasite lines isolated from patients in Southeast Asia displaying a range of both kelch13 alleles and clinical clearance times and a laboratory strain, NF54. This comprehensive series of assays established the growth relationships among the eight parasite lines. Interestingly, a clinically artemisinin resistant parasite line that carries the wild-type form of kelch13 outcompeted all other parasites in this study. Furthermore, a kelch13 mutant line (E252Q) was competitively more fit without drug than lines with other resistance-associated kelch13 alleles, including the C580Y allele that has expanded to high frequencies under drug pressure in Southeast Asian resistant populations.

CONCLUSIONS: This optimized competitive growth assay can be employed for assessment of relative growth as an index of fitness during the asexual blood stage growth between natural lines carrying different genetic variants associated with artemisinin resistance. Improved understanding of the fitness costs of different parasites proliferating in human blood and the role different resistance mutations play in the context of specific genetic backgrounds will contribute to an understanding of the potential for specific mutations to spread in populations, with the potential to inform targeted strategies for malaria therapy.

Malaria parasites have small extremely AT-rich genomes: microsatellite repeats (1-9 bp) comprise 11% of the genome and genetic variation in natural populations is dominated by repeat changes in microsatellites rather than point mutations. This experiment was designed to quantify microsatellite mutation patterns in Plasmodium falciparum. We established 31 parasite cultures derived from a single parasite cell and maintained these for 114-267 days with frequent reductions to a single cell, so parasites accumulated mutations during ∼13,207 cell divisions. We Illumina sequenced the genomes of both progenitor and end-point mutation accumulation (MA) parasite lines in duplicate to validate stringent calling parameters. Microsatellite calls were 99.89% (GATK), 99.99% (freeBayes), and 99.96% (HipSTR) concordant in duplicate sequence runs from independent sequence libraries, whereas introduction of microsatellite mutations into the reference genome revealed a low false negative calling rate (0.68%). We observed 98 microsatellite mutations. We highlight several conclusions: microsatellite mutation rates (3.12 × 10-7 to 2.16 × 10-8/cell division) are associated with both repeat number and repeat motif like other organisms studied. However, 41% of changes resulted from loss or gain of more than one repeat: this was particularly true for long repeat arrays. Unlike other eukaryotes, we found no insertions or deletions that were not associated with repeats or homology regions. Overall, microsatellite mutation rates are among the lowest recorded and comparable to those in another AT-rich protozoan (Dictyostelium). However, a single infection (>1011 parasites) will still contain over 2.16 × 103 to 3.12 × 104 independent mutations at any single microsatellite locus.

Fitness costs are key determinants of whether drug resistance alleles establish and how fast they spread within populations. More than 125 different kelch13 alleles, each containing a different amino acid substitution, have arisen in Southeast Asian malaria parasite (Plasmodium falciparum) populations under artemisinin selection over the past 15 years in a dramatic example of a soft selective event. However, just one of these alleles (C580Y) is now outcompeting other alleles in multiple different countries and is spreading toward fixation. Here we examine the fitness consequences of C580Y, relative to another less successful kelch13 mutation (R561H), to try to explain the distinctive dynamics of C580Y. We hypothesized that C580Y will show lower fitness costs than other kelch13 substitutions in the absence of artemisinin treatment. We used CRISPR/Cas9 methods to introduce single mutations (C580Y or R561H) or synonymous control edits into a wild-type parasite isolated on the Thailand-Myanmar border, conducted replicated head-to-head competition assays, and determined the outcome of competition using deep sequencing of kelch13 amplicons. Contrary to our predictions, these experiments reveal that C580Y carries higher fitness costs (s [selection coefficient] = 0.15 ± 0.008 [1 standard error {SE}]) than R561H (s = 0.084 ± 0.005). Furthermore, R561H outcompetes C580Y in direct competition (s = 0.065 ± 0.004). We conclude that fitness costs of C580Y in isolation are unlikely to explain the rapid spread of this substitution.

Adult schistosomes live in the blood vessels and cannot easily be sampled from humans, so archived miracidia larvae hatched from eggs expelled in feces or urine are commonly used for population genetic studies. Large collections of archived miracidia on FTA cards are now available through the Schistosomiasis Collection at the Natural History Museum (SCAN). Here we describe protocols for whole genome amplification of Schistosoma mansoni and Schistosome haematobium miracidia from these cards, as well as real time PCR quantification of amplified schistosome DNA. We used microgram quantities of DNA obtained for exome capture and sequencing of single miracidia, generating dense polymorphism data across the exome. These methods will facilitate the transition from population genetics, using limited numbers of markers to population genomics using genome-wide marker information, maximising the value of collections such as SCAN.