A theoretical approach to predicting the success of genetic manipulation of malaria mosquitoes in malaria control

Abstract

Background: Mosquitoes that have been genetically modified to better encapsulate the malaria parasite Plasmodium falciparum are being considered as a possible tool in the control of malaria. Hopes for this have been raised with the identification of genes involved in the encapsulation response and with advances in the tools required to transform mosquitoes. However, we have only very little understanding of the conditions that would allow such genes to spread in natural populations.

Methods: We present here a theoretical model that combines population genetical and epidemiological processes, thereby allowing one to predict not only these conditions (intensity of transmission, evolutionary cost of resistance, tools used to drive the genes) but also the impact of the spread of refractoriness on the prevalence of the disease.

Results: The main conclusions are 1) that efficient transposons will generally be able to drive genes that confer refractoriness through populations even if there is a substantial (evolutionary) cost of refractoriness, but 2) that this will decrease malaria prevalence in the human population substantially only if refractoriness is close to 100% effective.

Conclusions: If refractoriness is less than 100% effective (because of, for example, environmentally induced variation in the effectiveness of the mosquito's immune response), control programmes based on genetic manipulation of mosquitoes will have very little impact on the epidemiology of malaria, at least in areas with intense transmission.

Figure 1

Spread of refractoriness in the absence of a transposon as a genetic drive mechanism, and its effect on the prevalence of malaria in the human population at equilibrium under the assumption of a fixed maintenance cost of refractoriness. The parameters used to simulate the equations were: biting rate a = 0.5 day^-1, mortality μ = 0.1 day^-1, virulence = 0.1 and dominance h = 0.9. In panels (a) and (b) the efficacy of refractoriness was held at s = 1, and the cost of refractoriness and the initial intensity of transmission (i.e. before the spread of refractoriness) were varied. In panels (c) and (d) the cost of refractoriness was held at c = 0.01, and the efficacy of refractoriness and the initial intensity of transmission were varied.

Figure 2

Spread of refractoriness in the absence of a transposon as a genetic drive mechanism, and its effect on the prevalence of malaria in the human population at equilibrium under the assumption of a conditional cost of refractoriness that is induced by the response to parasite infection. The parameters used to simulate the equations were: biting rate a = 0.5 day^-1, mortality μ = 0.1 day^-1, virulence = 0.1 and dominance h = 0.9. In panels (a) and (b) the efficacy of refractoriness was held at s = 1, and the cost of refractoriness and the initial intensity of transmission (i.e. before the spread of refractoriness) were varied. In panels (c) and (d) the cost of refractoriness was held at c = 0.05, and the efficacy of refractoriness and the initial intensity of transmission were varied.

Figure 3

Spread of refractoriness linked to a transposon as a genetic drive mechanism and its effect on the prevalence of malaria in the human population at equilibrium. Panels (a)-(d) show the efficacy of the genetic drive mechanism (i.e. the proportion of heterozygotes that turn homozygous) that is required for refractoriness to spread to fixation. Panels (a) and (c) assume that the cost refractoriness is fixed, panels (b) and (d) assume a conditional cost induced by parasite infection. Panel (e) shows the effect of fixation of the allele coding for refractoriness on the prevalence of malaria. The parameters used to simulate the equations were: biting rate a = 0.5 day^-1, mortality μ = 0.1 day^-1, virulence = 0.1 and dominance h = 0.9. In panels (a) and (c) the efficacy of refractoriness was held at s = 1, the initial intensity of was set to R_0,init = 30 and the cost of refractoriness and the cost of the transposon (i.e. before the spread of refractoriness) were varied. In panels (b) and (d) the cost of refractoriness was held at c = 0.2, the cost of the transposon was held at c_T = 0.05, and the efficacy of refractoriness and the initial intensity of transmission were varied. In panel (e) the parameters not mentioned above were chosen to allow fixation of refractoriness.