Fitness components and natural selection: why are there different patterns on the emergence of drug resistance in Plasmodium falciparum and Plasmodium vivax?

Abstract

Background: Considering the distinct biological characteristics of Plasmodium species is crucial for control and elimination efforts, in particular when facing the spread of drug resistance. Whereas the evolutionary fitness of all malarial species could be approximated by the probability of being taken by a mosquito and then infecting a new host, the actual steps in the malaria life cycle leading to a successful transmission event show differences among Plasmodium species. These "steps" are called fitness components. Differences in terms of fitness components may affect how selection imposed by interventions, e.g. drug treatments, differentially acts on each Plasmodium species. Thus, a successful malaria control or elimination programme should understand how differences in fitness components among different malaria species could affect adaptive evolution (e.g. the emergence of drug resistance). In this investigation, the interactions between some fitness components and natural selection are explored.

Methods: A population-genetic model is formulated that qualitatively explains how different fitness components (in particular gametocytogenesis and longevity of gametocytes) affect selection acting on merozoites during the erythrocytic cycle. By comparing Plasmodium falciparum and Plasmodium vivax, the interplay of parasitaemia and gametocytaemia dynamics in determining fitness is modelled under circumstances that allow contrasting solely the differences between these two parasites in terms of their fitness components.

Results: By simulating fitness components, it is shown that selection acting on merozoites (e.g., on drug resistant mutations or malaria antigens) is more efficient in P. falciparum than in P. vivax. These results could explain, at least in part, why resistance against drugs, such as chloroquine (CQ) is highly prevalent in P. falciparum worldwide, while CQ is still a successful treatment for P. vivax despite its massive use. Furthermore, these analyses are used to explore the importance of understanding the dynamic of gametocytaemia to ascertain the spreading of drug resistance.

Conclusions: The strength of natural selection on mutations that express their advantage at the merozoite stage is different in P. vivax and P. falciparum. Species-specific differences in gametocytogenesis and longevity of gametocytes need to be accounted for when designing effective malaria control and elimination programmes. There is a need for reliable data on gametocytogenesis from field studies.

Figure 1

Intra-host and population-level dynamics resulting from fitness components. A) Population-average dynamics of asexual parasitaemia as a function of time after the erythrocytic cycle’s onset. They dynamics are shown separately for the resistant and sensitive parasites following equation (2). The y-axes on the left and right side correspond to the P. falciparum and P. vivax respectively. The dynamics are identical up to a scaling constant. On day seven drugs are administered to eliminate parasites. Resistant parasites are eliminated slightly less efficiently. B) Population-average gametocytaemia derived from the parasitaemia dynamics in A). Whereas, gametocytes are produced readily in P. vivax, they appear much later in P. falciparum. The dynamics are according to Equations (3) and (4). Vertical lines in A) and B) mark important events in a clinical episode. C) Spread of a resistant mutation - according to equation (1) - in P. falciparum and P. vivax is determined by the ratio of the areas under the curve of the gametocytaemia dynamics of resistant over sensitive parasites derived from B). Shown is the frequency of the resistant mutation as a function of time measured in generations corresponding to transmission cycles. The horizontal lines show the fitness ratio of resistant over sensitive parasites with corresponding right-hand side y-axis. Parameters are as described in Methods.

Figure 2

Intra-host dynamics with refined gametocytaemia. A) As Figure 1A with parameters described in Methods. B) As compared to Figure 1B, gametocytes are not just assumed to be proportional to merozoites but their longevity is incorporated – according to Equations (5) and (6) respectively. The grey shaded area indicates a period of isolation in which no transmission can occur, and is only relevant for figure D. C) as Figure 1C but fitness is derived from Figure 2B and (7). D) As in Figure C, but the two-day isolation (indicated in grey in Figure B) according to equation (8) is taken into account. Hence, the gametocytaemia dynamics in the grey shaded area are disregarded for the derivation of fitness.