Confinement of gene drive systems to local populations: A comparative analysis

Abstract

Mosquito-borne diseases such as malaria and dengue fever pose a major health problem through much of the world. One approach to disease prevention involves the use of selfish genetic elements to drive disease-refractory genes into wild mosquito populations. Recently engineered synthetic drive systems have provided encouragement for this strategy; but at the same time have been greeted with caution over the concern that transgenes may spread into countries and communities without their consent. Consequently, there is also interest in gene drive systems that, while strong enough to bring about local population replacement, are unable to establish themselves beyond a partially isolated release site, at least during the testing phase. Here, we develop simple deterministic and stochastic models to compare the confinement properties of a variety of gene drive systems. Our results highlight several systems with desirable features for confinement—a high migration rate required to become established in neighboring populations, and low-frequency persistence in neighboring populations for moderate migration rates. Single-allele underdominance and single-locus engineered underdominance have the strongest confinement properties, but are difficult to engineer and require a high introduction frequency, respectively. Toxin–antidote systems such as Semele, Merea and two-locus engineered underdominance show promising confinement properties and require lower introduction frequencies. Killer-rescue is self-limiting in time, but is able to disperse to significant levels in neighboring populations. We discuss the significance of these results in the context of a phased release of transgenic mosquitoes, and the need for characterization of local ecology prior to a release.

Introduction

Mosquito-borne diseases such as dengue fever, chikungunya and malaria continue to pose a major health problem through much of the world. Treatments for dengue fever and chikungunya remain elusive, and malaria is proving exceptionally difficult to control in highly endemic areas with insecticide-treated nets, indoor residual spraying and antimalarial drugs (Griffin et al., 2010, World Health Organization, 2010). The failure of currently available methods to control these diseases has renewed interest in approaches to disease prevention that involve the use of gene drive systems to spread disease-refractory genes into wild mosquito populations (Alphey et al., 2002, Marshall and Taylor, 2009). Such strategies are attractive because they are self-perpetuating and are expected to result in disease suppression far beyond the release site.

A number of gene drive systems have been proposed, including naturally occurring selfish genetic elements such as transposable elements (TEs), post-meiotic segregation distorters, Medea elements, homing endonuclease genes (HEGs), and the intracellular bacterium Wolbachia (Braig and Yan, 2001, Gould and Schliekelman, 2004, Sinkins and Gould, 2006). Another set of approaches to bringing about population replacement involves creating insects in which genes of interest are linked to engineered chromosomes: compound chromosomes or translocations (Curtis, 1968, Foster et al., 1972), or pairs of unlinked lethal genes, each of which is associated with a repressor of the lethality induced by expression of the other lethal gene—a system known as engineered underdominance (Davis et al., 2001, Magori and Gould, 2006).

An essential feature of population replacement is the ability of released transgenic insects to spread disease-refractory genes through a wild population on a human timescale. The observation that several naturally occurring selfish elements have spread over wide geographical areas – a TE in Drosophila melanogaster (Kidwell, 1983), Wolbachia in Drosophila simulans (Turelli and Hoffmann, 1991), and Medea in Tribolium (Beeman and Friesen, 1999, Lorenzen et al., 2008) – provides encouragement for this strategy. However, these observations also highlight the potential for gene drive systems to spread transgenes into countries before they have agreed to their introduction (Knols et al., 2007). The Cartagena Protocol – the United Nations protocol on the international movement of genetically modified organisms (GMOs) – is prohibitive of a release of GMOs capable of self-propagating across national borders in the absence of a multilateral international agreement (Marshall, 2010). It also allows nations to decide for themselves how they would like to regulate the import of GMOs.

The movement of mosquitoes carrying transgenes also requires consideration at the community level. A recent survey of public attitudes to a release of malaria-refractory mosquitoes in Mali suggests that a small number of people would be happy for the first release to be conducted in their community; however, most people would first like to see the results of a release in an isolated community where malaria prevalence has been shown to decrease in the absence of side-effects (Marshall et al., 2010). These regulations and societal views highlight the importance of confining the first releases of transgenic mosquitoes to isolated locations, particularly when gene drive systems are involved. This is in agreement with the principle of scientific risk management, which recommends that the magnitude of potential risks be minimized by spatially limiting the release and providing mechanisms for removal of transgenes in the event of adverse effects (Beech et al., 2009).

Consequently, the strategy of population replacement is left with two competing mandates, at least during the testing phase. Transgenes must be able to spread to high frequencies locally, in order to provide an adequate test of the technology's ability to prevent disease (Boete and Koella, 2002, Boete and Koella, 2003); but they must not spread to significant levels in neighboring populations. The use of trap crops, vegetation-free zones and spatial isolation have been discussed in the context of confining an accidental release of transgenic mosquitoes from field cages (Benedict et al., 2008). Additionally, studies of mosquito ecology on islands off the coast of Africa have begun with an isolated transgenic release in mind (Pinto et al., 2003). These safeguards are effective at reducing the migration rate of mosquitoes to nearby locations; however, a small number of escapees are inevitable and, if these escapees carry invasive gene drive systems, this may be sufficient for transgenes to spread beyond their release site. The release of mosquitoes carrying only transgenes conferring disease refractoriness has been proposed (Benedict et al., 2008) and is an important step towards assessing the behavior of the refractory gene in the wild. However, gene drive systems will ultimately be required to achieve the transgenic frequencies necessary for disease control without prohibitive release sizes.

Underdominant systems provide a potential solution to these competing mandates (Altrock et al., 2010). The mosquito species that transmit human diseases are largely anthropophilic, forming subpopulations around human settlements as discrete blood sources (Service, 1993). Abstract two-population models have shown that an underdominant allele can become established in one population while persisting in a neighboring population at low frequencies provided that the migration rate between populations is sufficiently low and the strength of selection against heterozygotes is sufficiently strong (Karlin and McGregor, 1972, Lande, 1985, Altrock et al., 2010). The single underdominance allele described in these models is particularly difficult to engineer; however several genetic systems displaying similar properties have been proposed, including translocations (Curtis, 1968), engineered underdominance (Davis et al., 2001), and a variety of single-locus toxin–antidote systems such as Semele (Marshall et al., 2011), inverse Medea (Marshall and Hay, 2011a) and Medea with a recessive antidote (Merea).

A proper assessment of the ability to confine these systems to discrete populations will require a detailed ecological analysis, taking into account phenomenological features of the mosquito populations of interest – including seasonally fluctuating population sizes and migration rates as well as seasonally varying chromosomal form and species make-ups (Lanzaro et al., 1998, Taylor et al., 2001, Tripet et al., 2005). The present analysis, however, focuses on a comparison of the gene drive systems currently being considered, and their relative ability to be confined to partially isolated populations. For the purposes of comparison, we consider the simplest possible model of population structure—a two-population model in which mosquitoes with gene drive systems are introduced into one population, which exchanges migrants with a neighboring population. This differs from the analysis of Marshall (2009), which uses a branching process framework to compare the ability of gene drive systems to persist in the environment following an accidental release from a contained facility. Here, we use a simple difference equation framework to see whether there is a realistic set of parameters for which a transgene can become established at its release site without spreading into neighboring populations. We discuss the relevance of these findings to a phased release of transgenic mosquitoes intended for the control of vector-borne diseases, and to the strategy of population replacement in general.