Modeling Gene Drives in Africa – Seasonal, Spatial and Epidemiological Considerations

Anopheles gambiae

Anopheles gambiae

Gene drive as a tool for fighting malaria has received much attention because of promising results in the laboratory. Moving these technologies to the field is complicated and Eckoff et al (2016) in the Proceedings of the National Academy of Sciences report on their modeling efforts in which they attempt to include spatial and seasonality data from Africa to try and capture a level of complexity that is intended to provide meaningful insights into if and how mosquito gene drive systems will be the tools they are being imagined.

There are three approaches being considered and developed for using gene drive in the context of malaria. First, a drive system that targets genes that impact fertility and leading to population suppression or collapse. Second, a drive system that moves an antiparasite transgene through the population, rendering it refractory to Plasmodium infection or ineffective at transmission. Third, a Y-drive system in which X chromosome containing gametes are eliminated leading eventually to collapse.

The Epidemiological Modeling (EMOD) software provides an infrastructure that is designed primarily for use by disease modelers, researchers, epidemiologists, and public health professionals seeking to simulate infectious disease conditions and evaluate the effectiveness of eradication or mitigation approaches. The software is an agent-based, discrete time, Monte Carlo simulator used for disease modeling. The architecture provides disease transmission for environmental, sexual, vector based and air borne diseases and may be adapted to support additional infectious diseases. The binary software or source files are available for download. Here for details

Eckoff et al. use the epidemiological modeling model (EMOD). This model is used to simulate effects on mosquito populations and malaria transmission. This is a stochastic model that is individual based with explicit representation of vector populations, adult feeding behavior, egg laying and larval dynamics. At a spatial level the model is patch-based – allowing movements of humans and vectors to be defined.  (malaria model here)

An interesting aspect of this modeling effort was the inclusion of realistic seasonality data. They used actual historical rainfall data from Tanzania and Nigeria.

Simulations went something like this – they was run for 2 (simiulated ) years and then 500 gene drive males were releases after the start of the high season and the model was run for at least 8 (simulated) years.

One thousand combinations of parameters for homing and fecundity reduction were simulated.

There were four categorical outcomes to the simulations – 1) population persists and construct is lost, 2) population persists but no wild-types, 3) population persists with wild-type and gene drive mosquitoes present 4) population extinction.

The details of the modeling outcomes and under what specific parameter conditions are very interesting and worth looking at carefully in the original paper and are beyond summarizing here.

A selfish gene (blue) spreading through a mosquito population

A selfish gene (blue) spreading through a mosquito population

One particularly interesting observation was the impact of seasonality. For example, if seasonality is extreme and results in disconnected vector populations for a significant amount of time then it becomes difficult to come up with release strategies to deal with this. Aestivation behavior can also impact success.

The authors also consider resistance to gene drive through NHEJ which destroys target sites. Resistance is a considerable problem and it will require targeting multiple sites and using multiple drive constructs.

Overall, this is an interesting paper that takes modeling efforts a bit further than before and despite the complexities of executing a gene drive control or eradication program against a major malaria vector, there seems to be ways of dealing with them and the authors are optimistic about the prospects of these tools.

Philip A. Eckhoff, Edward A. Wenger, H. Charles J. Godfray, and Austin Burt (2016) Impact of mosquito gene drive on malaria elimination in a computational model with explicit spatial and temporal dynamics PNAS 2016 ; published ahead of print December 27, 2016, doi:10.1073/pnas.1611064114



Post a Comment

Your email address will not be published. Required fields are marked *