Using infective mosquitoes to challenge monkeys with Plasmodium knowlesi in malaria vaccine studies

Abstract

Background: When rhesus monkeys (Macaca mulatta) are used to test malaria vaccines, animals are often challenged by the intravenous injection of sporozoites. However, natural exposure to malaria comes via mosquito bite, and antibodies can neutralize sporozoites as they traverse the skin. Thus, intravenous injection may not fairly assess humoral immunity from anti-sporozoite malaria vaccines. To better assess malaria vaccines in rhesus, a method to challenge large numbers of monkeys by mosquito bite was developed.

Methods: Several species and strains of mosquitoes were tested for their ability to produce Plasmodium knowlesi sporozoites. Donor monkey parasitaemia effects on oocyst and sporozoite numbers and mosquito mortality were documented. Methylparaben added to mosquito feed was tested to improve mosquito survival. To determine the number of bites needed to infect a monkey, animals were exposed to various numbers of P. knowlesi-infected mosquitoes. Finally, P. knowlesi-infected mosquitoes were used to challenge 17 monkeys in a malaria vaccine trial, and the effect of number of infectious bites on monkey parasitaemia was documented.

Results: Anopheles dirus, Anopheles crascens, and Anopheles dirus X (a cross between the two species) produced large numbers of P. knowlesi sporozoites. Mosquito survival to day 14, when sporozoites fill the salivary glands, averaged only 32% when donor monkeys had a parasitaemia above 2%. However, when donor monkey parasitaemia was below 2%, mosquitoes survived twice as well and contained ample sporozoites in their salivary glands. Adding methylparaben to sugar solutions did not improve survival of infected mosquitoes. Plasmodium knowlesi was very infectious, with all monkeys developing blood stage infections if one or more infected mosquitoes successfully fed. There was also a dose-response, with monkeys that received higher numbers of infected mosquito bites developing malaria sooner.

Conclusions: Anopheles dirus, An. crascens and a cross between these two species all were excellent vectors for P. knowlesi. High donor monkey parasitaemia was associated with poor mosquito survival. A single infected mosquito bite is likely sufficient to infect a monkey with P. knowlesi. It is possible to efficiently challenge large groups of monkeys by mosquito bite, which will be useful for P. knowlesi vaccine studies.

Figure 1

Survival of heavily and lightly infected mosquitoes.Anopheles dirus mosquitoes were put into several pint containers and allowed to feed on a P. knowlesi infected monkey on consecutive nights, and survival to day 15 was monitored. For each group between 140 and 160 mosquitoes were fed. Figure 1A shows the fraction of mosquitoes surviving from the first night feeds, either on an uninfected monkey or a P. knowlesi-infected monkey (parasitaemia of 0.2%). Figure 1B shows survival of mosquitoes fed on the same two monkeys one day later, when the parasitaemia in the P. knowlesi-infected monkey was 2.6%. Because there was a strong day effect (logrank tests showed significant differences between control groups by day), we used a Cox frailty model using only parasitaemia level of the donor monkey, and with a separate frailty parameter for the four groups. There was a significant difference between the 2.6% group and the combined controls (p = 0.038), but no significant difference between the 2.6% group and the 0.2% group (p = 0.18).

Figure 2

Effect of number of infected mosquitoes feeding on day parasites detected in blood. Thirty monkeys were infected with P. knowlesi by the bite of from 1 to 10 mosquitoes which had sporozoites in the salivary glands and had ingested blood. The day on which the first P. knowlesi parasites were found in the blood is plotted against the number of infectious bites. Increasing numbers of bites lead to earlier detection of parasites in the blood. Cox regression analysis shows the number of infected mosquito bites increased the hazard of parasites appearing in the blood (p = 0.01).