The announcement of the Grand Challenges in Global Health in October 2003 elevates the development of a ‘genetic strategy to deplete or incapacitate’ disease-transmitting insect populations to the centre of future public health measures aimed at vector-borne diseases (Varmus et al. 2003). However, for those concerned with the day-to-day fight against diseases such as malaria and dengue fever, genetic control strategies may seem little more than elegant theories or laboratory curiosities rather than realistic control alternatives. Moreover, confusion and ignorance over the meaning of a ‘genetic control strategy’ coupled with fears over the safety, or even desirability, of a genetic approach reinforce those sceptics who argue such systems are a dangerous distraction from the problem in hand.
Across the developing world, the public health burden due to vector-borne diseases continues to grow as current control measures fail to cope. There is an urgent need to identify improved control strategies that will remain effective even in the face of growing insecticide and drug resistance. The intolerable burden of malaria across sub-Saharan and failures of dengue control systems across the globe, even in economically advanced settings, illustrate the scale of the problem and the urgent need for viable solutions. Given the enormous scale of the problem, serious consideration should be given to the range of potential solutions offered by genetic control strategies aimed at vector populations.
What is meant by a genetic control strategy against vector-borne disease? Genetic control is not a single strategy but rather a range of potential techniques that may be subdivided into two main categories: population suppression and population replacement. The common factor is that the controlling mechanisms are genetic traits introduced into the wild population by mating. Both ideas are well established, dating back over several decades (Serebrovskii 1940; Vanderplank 1948; Knipling 1955; Curtis 1968, 2002). Population suppression, most prominent in the form of the Sterile Insect Technique (SIT), is used around the world, primarily against agricultural pest insects. Genetic control of pest insect populations is, therefore, already with us today and, with modest improvement through the use of currently available transgenic technologies, could become a mainstay for public health control of specific vector-borne diseases.
SIT is probably the most familiar genetic control strategy, although it may not be familiar to think of it as a genetic system. SIT is a species-specific and environmentally non-polluting means of controlling insect populations, with a proven track record in eliminating a range of agricultural pests and disease vectors (including the Mediterranean fruit fly, the screwworm fly and the tsetse fly) over large areas. First conceived in the 1930s, SIT relies on the mass rearing, sterilization and release of large numbers of the sterile males over the target area (Knipling 1955). Released sterile males mate with wild females, reducing the reproductive potential of the wild population and so causing a reduction in the wild population in the subsequent generation. If enough sterile males continue to be released for a sufficient time, the target population will collapse, leading to its suppression, or even total elimination, over the release area.
SIT has many advantages over conventional vector control strategies: (1) it is highly species-specific and avoids detrimental effects of insecticides on non-target species, (2) it uses the highly evolved, and highly efficient mate-seeking behaviour of male insects rather than relying on human operatives and (3) it works better as wild insect densities are reduced, so that the ratio of wild males relative to released sterile males tilts progressively in favour of the latter. The ability to drive populations to local extinction distinguishes SIT from all other current control options in almost all circumstances.
Based on these principles, there was great optimism in the late 1960s and the early 1970s regarding the development of SIT as a control strategy against mosquito vectors of malaria and other diseases (Curtis 2002). However, in practice, there are two main technological shortfalls that prevent SIT being harnessed as a major control strategy for vector-borne disease: (1) an efficient method of separating male and female insects; (2) the use of irradiation or chemosterilants as a means of sterilizing male insects. Recent advances in genetic engineering offer solutions to both these constraints.
The mass-rearing process in SIT initially produces males and females of the target species in equal numbers. For mosquito-borne diseases, the female mosquito is exclusively responsible for pathogen transmission through feeding on humans; males do not to bite. It is unacceptable for a control programme to release female vectors that have the potential of transmitting disease. Moreover, the release of females may also be detrimental in that they may divert released sterile males from searching for wild female mates, thereby reducing the overall efficiency of the SIT programme. Mechanical sex separation methods using, for example, pupal size, which works for Culex and Aedes mosquitoes, are not effective for the Anopheles vectors of malaria. Current genetic sexing mechanisms, based on chromosome aberration-based systems, tend to be unstable (which is a major problem when insects are mass-produced in their millions) and may confer some fitness costs, in terms of competitiveness with wild males, thereby impacting on the success of the SIT programme. A novel genetic sexing mechanism, applicable across different target species, is highly desirable and has been identified as a major area of research that would help lead to the application of SIT for the control or eradication of mosquitoes (International Atomic Energy Agency, Board of Governors 2001).
The second constraint is the means of sterilization. Irradiation is currently the standard method, inducing dominant lethal mutations in the sperm of the mass-reared males. However, irradiation causes unwanted collateral damage to the chromosomes in cells other than the germ line. The deleterious effects of radiation impose a fitness penalty, rendering the sterile males less competitive relative to wild males and so a less effective agent of SIT. The effects of irradiation vary between species. Importantly, for many mosquitoes, irradiating pupal stages of mosquitoes causes severe abnormalities in adults making them non-viable for SIT (Andreasen & Curtis 2004). Other sterilization strategies, including irradiating adults or chemosterilization, are either impractical on an industrial scale or require handling mutagenic chemosterilants and release of males with chemically detectable residues (Curtis 2002).
Development of a genetic system that induces repressible female-specific lethality in target vector species is one proposed genetic control strategy that would overcome the double constraint of sex separation and irradiation, allowing the full potential of SIT in public health to be realized. A genetic system known as RIDLTM (Release of Insects with a Dominant Lethal; Alphey & Andreasen 2002) is one candidate method.
RIDL here describes a system in which the mass-reared insects carry a repressible female-specific lethal gene, enabling the production of a male only population simply by removal of the repressor (Alphey & Andreasen 2002). RIDL also bypasses the need for irradiation and so avoids the associated cost and damage. RIDL males are, strictly speaking, not sterile. If released into the wild, they will mate with wild females. Such mating will result in viable male offspring, but no viable daughters as the dominant lethality would be expressed in the natural environment free from the dietary additive used to suppress the system in the rearing facility. Many vectors, including mosquitoes, need only be sterile in the sense that they produce no daughters; any males produced by released RIDL males would in turn carry the RIDL construct, in heterozygous form, and contribute to the elimination of females in later generations.
The RIDL system has so far been exemplified in Drosophila (Alphey & Andreasen 2002) and is under development for Aedes aegypti, the vector of dengue fever and urban yellow fever. Coupled with a traditional mechanical sexing mechanism, as developed for Culex and Aedes SIT trials in India in the 1970s and shown to be effective with pupae reared by the hundreds of thousands (Ansari et al. 1977), a non-sex specific version of RIDL in Aedes mosquitoes could potentially be used to control dengue fever. This seems to bring the prospect of an effective SIT-like genetic control method for Aedes tantalisingly close.
However, there are broader concerns about SIT as a viable means of controlling vector populations, other than the technological limitations of the traditional approach. SIT has been successfully used to eradicate tsetse flies, the vector of African trypanosomiasis in both humans and livestock, from the island of Zanzibar. This success has led to calls for the adoption of SIT, along with conventional control measure, to control tsetse flies across mainland Africa (Kabayo 2002). The widespread distribution of tsetse flies across rural continental Africa and the associated reinvasion pressures of tsetse from uncontrolled regions into areas under SIT control, the role of multiple tsetse fly species in trypanosomiasis transmission, and the scale of the investment and infrastructure required to undertake SIT programmes have all been cited as reasons why funds should not be diverted from traditional control activities (Rogers & Randolph 2002).
The utility of a population suppression approach to vector population control is critically dependent on the population ecology of the target vector species, particularly the size of the wild population, the area over which mass release is conducted, the population genetic structure of the wild populations (e.g. spatial or ecological barriers to breeding) and reinvasion pressures from non-target areas. Furthermore, the effectiveness of vector population control as a strategy for disease control itself depends on the epidemiology of the disease in the target region. It is clear that SIT, even with the improvements afforded by a heritable repressible dominant lethal genetic system, is not a panacea for the problem of vector-borne diseases.
Nevertheless, a genetic system such as RIDL does have the potential to extend the utility of SIT to a wider range of species and populations than at present, the most immediate being the urban vector of dengue fever Ae. aegypti. There currently exists no viable means of controlling dengue fever, so the necessity to identify a new control option is paramount. Other suitable targets may include isolated populations of Anopheles species such as An. stephansi or An. arabiensis.
In the case of malaria, over the majority of its geographic range, however, one has to turn to the second category of genetic strategies, population replacement. This strategy requires two fundamental steps: (1) creation in the laboratory of a genetically engineered strain of the target vector that is refractory to the relevant pathogen and (2) develop-ment of a system whereby the refractory trait may be spread throughout the wild vector population. Ultimately, if the frequency of the refractory trait in the vector population can be increased to a critical level at which pathogen transmission no longer persists, the disease burden will be eliminated. Like genetics-based population suppression, the idea of driving a genetic construct for refractoriness to a human pathogen into a wild vector population has a long history (Curtis 1968). Recently, major advances in insect molecular biology have renewed interest in this form of genetic control. These advances, including the production of stable Aedes and Anopheles transgenic lines and the sequencing of the An. gambiae genome, together with the development of powerful new technologies like RNA interference, have helped elucidate the steps required to produce stable laboratory lines that are refractory for human pathogens, particularly malaria parasites and dengue fever (Olson et al. 2002; Riehle et al. 2003). By contrast, there has been relatively little progress in the development of effective mechanisms for driving the refractory constructs into wild vector populations, a so-called ‘gene driver’. This second step is pivotal in the realizing any genetic control system as an effective tool to reduce disease burden.
Why is the development of a gene driver so important? It is highly likely that any genetic construct that reduces the capacity of vector species to transmit pathogens will have a fitness cost associated with it. This means the genetically altered vector will be at a selective disadvantage relative to the wild type that it is intended to replace. Thus, if the refractory strain were simply released into the field it would be selected against, and so the refractoriness trait would not spread. While it is possible that the refractoriness trait itself may confer a selective advantage by allowing the engineered vectors to avoid fitness costs associated with carrying pathogens, this is unlikely to exceed in magnitude the cost of the physiological abnormality required for refractoriness. Insects and parasites have generally co-evolved such that parasite infections are not a major fitness burden to the insect, while the proportion of a vector population infected by vector-borne pathogens is characteristically very small. Moreover, the fact that naturally occurring refractoriness traits have not been naturally selected for in field populations of the insects that we recognize as vectors suggests refractoriness is not at a net selective advantage. Thus, an engineered refractory construct is unlikely to spread through wild populations unaided. Rather, an effective system for driving the construct into wild vector populations is essential in order to bring the ‘refractory insect’ strategy to practical utility without the need for massive releases throughout the huge and almost continuous rural African areas where malaria remains a problem.
A variety of possible mechanisms, each with strengths and weaknesses, has been proposed for gene driving, including the use of intracellular symbionts, meiotic drive systems, transposable elements and under-dominance systems (Braig & Yan 2001; Gould & Schliekelman 2004). An ideal drive system would have most or all of the following features (adapted from Braig & Yan 2001), each one representing major scientific and technical challenges: (1) maintenance of a perfect linkage between the driver mechanism and refractory gene (Curtis 1968, 2002) so that the driver system, unshackled from the fitness cost of the refractory construct, would not spread at the expense of the beneficial genotype; (2) ability to achieve complete population replacement to the desired genotype, so that the disease burden reduction would be realized in high transmission settings and for diseases in which even at low levels of transmission disease burden is still relatively high (Boete & Koella 2002); (3) controllable spread of the trait to specific target populations thus avoiding the unchecked run away of a system that could, in theory, transform an entire species globally; (4) recallable, so that if the trait were found to be unexpectedly harmful the modified vectors could be eventually eliminated from the population; (5) replaceable, re-useable and generic so that the same genetic driving mechanism could be employed to drive different types of refractory constructs, including future improved beneficial traits, through the same target populations; (6) not transferable to non-target species.
The development of population replacement strategies for field use is generally regarded to be further in the future than a RIDL-based population suppression strategy. However, whether based on population suppression or population replacement, all genetic control strategies require the release of transgenic insects into the environment, raising immediate concerns about safety. Such concerns need to be addressed at an early stage, with input from parties both favourable and hostile to the technology.
The history of genetic control systems aimed at mosquito populations shows the potential problems that future strategies may face. In the 1970s, work on SIT and population replacement strategies against Ae. aegypti in India was abandoned after unfounded accusations of ‘bio-warfare’ in the press (see Curtis 2002). The current debate of genetically modified technology in the agricultural sector, although largely focussing on unrelated concerns, will also influence personal, journalistic and political opinions. The potential benefits and any possible risks must be fully explained to politicians, journalists, health professionals, NGOs and the general public in both the endemic countries as well as the wider international community. It is important to educate people in some of the details of the technology being promoted so as to address commonly held misconceptions, such as differentiating between non-autonomous transposons used in transgenesis and autonomous transposons for genetic drive systems.
Any potential strategy would need to be developed through a series of confined laboratory and field trials to address issues of safety and efficacy, to meet the permitting requirements of relevant regulatory bodies and international standards (such as the Cartagena Protocol on Biosafety, which covers the international shipment of genetically modified organisms). An important component of this procedure will include an environmental impact assessment. However, there are very few precedents to follow for genetically modified arthropods. It is therefore essential to address areas of concern such as horizontal transfer, toxicity of the genetically modified mosquitoes to non-target species (e.g. spiders), the ecological effects of removing mosquitoes from the food chain and the competence of genetically modified mosquitoes to transmit other non-target pathogens. The development of RIDL-based technologies, with the possibility of caged trials of genetically modified Aedes in the near future, will act as important regulatory ‘trail blazer’ for genetic control strategies.
The areawide nature of the proposed interventions (e.g. gene drivers propagating a trait in a runaway fashion through Anopheles populations or the mass release of RIDL-transformed Aedes) is another aspect of genetic control strategies that stands in contrast to the current trend to promote community or individual-based control strategies. While this may be considered as a strength, given the difficulties in sustaining community-based strategies, e.g. against dengue fever, it does raise potential implications, which must be identified and addressed. One immediate consideration is how to conduct trials of interventions, with full ethical approval and informed consent, that operate over wide areas or which are designed to have runaway effects once implemented.
It is also essential not to focus exclusively on the genetic modification technology of the novel interventions, and thereby neglect the need to consider genetic control strategies as one would any other novel intervention (such as a new vaccine, drug or use of treated bed-nets), in terms of standard measures including cost-effectiveness, affordability, acceptability, equity, accessibility and sustainability. It is important to assess whether the proposed genetic control strategies are likely to compare favourably with existing interventions (Goodman et al. 1999).
The regulatory and social issues, as much as any scientific or technological hurdles, will be critical in determining the future success of genetic control systems. Recent advances in mosquito genetic transformation, combined with information from the Anopheles and Aedes genome projects, now make the development of genetic control strategies against mosquito populations timely and attainable. Their potential as a viable control option against vector-borne diseases needs to be widely recognized and embraced if their full public health potential is to be realized.