The effects of subcurative praziquantel treatment on life-history traits and trade-offs in drug-resistant Schistosoma mansoni

Abstract

Natural selection acts on all organisms, including parasites, to maximize reproductive fitness. Drug resistance traits are often associated with life-history costs in the absence of treatment. Schistosomiasis control programmes rely on mass drug administration to reduce human morbidity and mortality. Although hotspots of reduced drug efficacy have been reported, resistance is not widespread. Using Bayesian state-space models (SSMs) fitted to data from an in vivo laboratory system, we tested the hypothesis that the spread of resistant Schistosoma mansoni may be limited by life-history costs not present in susceptible counterparts. S. mansoni parasites from a praziquantel-susceptible (S), a praziquantel-resistant (R) or a mixed line of originally resistant and susceptible parasites (RS) were exposed to a range of praziquantel doses. Parasite numbers at each life stage were quantified in their molluscan intermediate and murine definitive hosts across four generations, and SSMs were used to estimate key life-history parameters for each experimental group over time. Model outputs illustrated that parasite adult survival and fecundity in the murine host decreased across all lines, including R, with increasing drug pressure. Trade-offs between adult survival and fecundity were observed in all untreated lines, and these remained strong in S with praziquantel pressure. In contrast, trade-offs between adult survival and fecundity were lost under praziquantel pressure in R. As expected, parasite life-history traits within the molluscan host were complex, but trade-offs were demonstrated between parasite establishment and cercarial output. The observed trade-offs between generations within hosts, which were modified by praziquantel treatment in the R line, could limit the spread of R parasites under praziquantel pressure. Whilst such complex life-history costs may be difficult to detect using standard empirical methods, we demonstrate that SSMs provide robust estimates of life-history parameters, aiding our understanding of costs and trade-offs of resistant parasites within this system and beyond.

Keywords: Schistosoma; drug resistance; establishment; fecundity; fitness costs; praziquantel; reproduction; state‐space models; survival.

Figure 1

Simplified experimental design and life cycle of S. mansoni (as modelled). (a) Each parasite line (S, R and RS) was maintained in the laboratory through passage in four mice and 50 snails in each generation for each praziquantel treatment regime. In generation 1 (G1), mice were infected with 220 cercariae, and in subsequent generations (G2–G4), 110 cercariae were used to infect mice. In all generations, six miracidia were used to infect snails. (b) The life cycle of schistosomes was modelled using particular life stages. Cercariae (C) enter the definitive mouse host and establish themselves as adults (A). Those adults that survive (with survival rate s _a) will sexually reproduce and lay eggs with a rate λ, which hatch into miracidia (M). The miracidia that survive (with a rate s _m) will then leave the mouse host and enter the snail. The miracidia (M) establish inside the intermediate snail host with a rate θ and will start shedding sporocysts that become cercariae (C) at a rate ϕ. A new generation will begin as the new cercariae enter the definitive host. The parameters in red correspond to those which can be impacted by praziquantel. The dotted arrows indicate steps that occurred through experimental manipulation. The schistosome images are modified with permission from Genome Research Limited (https://www.yourgenome.org/facts/what-is-schistosomiasis)

Figure 2

Raw data counts of Schistosoma mansoni in the different life cycle stages for each parasite line (symbols: S susceptible; R: resistant; RS: resistant–susceptible) and praziquantel treatment group (colours). Average number (±standard deviation) of (a) adult worms per cercaria and (b) miracidia per adult worm/day and (c) the weekly number of cercariae produced per miracidium (right) are estimated across the four generations for each line/treatment group

Figure 3

Untreated control (baseline) life‐history traits and trade‐offs in the definitive (a–c) and intermediate (d–i) hosts. For each Schistosoma mansoni line, susceptible (S; left), mixed resistant and susceptible (RS; middle) and resistant (R; right), the median (with 95% CIs) life‐history traits (x‐ and y‐axis) are shown for each generation. The trade‐offs between traits (black line and associated slope, β) were obtained from a Poisson GLM of the posterior distributions of survival (a–c) or establishment (d–i) rates against those of fecundity (a–c) or shedding (d–i) rates, respectively. All regressions were significant with p<.001. p‐values and standard deviations shown in Appendix S3 TableS2

Figure 4

Sexual life‐history traits and trade‐offs under dose‐dependent praziquantel pressure. For each S. mansoni line, susceptible (S; left), mixed resistant and susceptible (RS; middle) and resistant (R; right), and praziquantel (PZQ) dose (rows), the median (with 95% CIs) survival (x‐axis) and fecundity (y‐axis) rates are shown for each generation. The trade‐offs between these traits (black line and associated slope, β) were obtained from a Poisson GLM of the posterior distributions of survival and fecundity. All regressions were significant with p<.001, except R low and R and RS high PZQ. p‐values and standard deviations shown in Appendix S3Table S2

Figure 5

Life‐history traits quantify an organism's investment into components of fitness, such as reproduction (y‐axis) and survivorship (x‐axis). (a) There are recognized trade‐offs between life‐history traits (solid curve), so that an increase in fecundity is associated with decreased survivorship (orange) or conversely higher survivorship leads to lower fecundity (green). These life‐history trade‐offs arise from limited resources available for an organism's growth, reproduction and survivorship. (b) However, trade‐off curves can shift as resource (energy) availability for certain traits change—this may be due to other resource allocation needs in the organism or a decline in resource availability. When this occurs, negative trade‐offs are difficult to observe and a decline in one trait (orange) may not be compensated by a second observed trait (green). [Figure adapted from work by Noordwijk and de Jong (1986)]