Advances in Vector Control Science: Rear-and-Release Strategies Show Promise… but Don’t Forget the Basics

Abstract

Both chikungunya and Zika viruses have recently swept from Africa across the Pacific to the Americas, causing major outbreaks of disease in humans. In the meantime, dengue epidemics continue throughout the tropics. Traditional vector control programs based on strategies from 1950s and 1960s have been relatively ineffective in combating recent epidemics. In response, new methods involving the rearing and releasing of large numbers of mosquitoes to eliminate or modify local Aedes populations are being developed, with several currently conducting field releases in high-risk countries. These advances, include the release of Wolbachia-infected Aedes aegypti and Aedes albopictus, for either its virus-blocking capabilities, sterilization by cytoplasmic incompatibility, or both; the release of Aedes carrying dominant lethal genes, such as the OX513A strain of A. aegypti; and other emerging techniques, such as advancing gene-drive technologies, are summarized, as well as current stages of development and primary operational and regulatory hurdles. Although these technologies show great promise, none are ready for widespread rollout for cities of millions of people. Thus, efforts should be made to avoid methods such as space sprays that have failed and improve existing technologies to increase their efficacy.

Many reasons have been given to explain the failure of Aedes aegypti–based dengue control programs. Generally, human activities, such as increased air travel that allows viruses to spread rapidly [1], and increased urbanization, have led to a proliferation of the viruses and their vectors in the tropics [2]. Climate change and international trade, particularly in used tires, have also allowed for the pole-ward expansion of competent vector species, such as Aedes albopictus [3]. Thus, vector control programs have faced an increasing onslaught of importations of viruses in elevated vector populations while often having to control several vector species with differing ecologies.

Traditional dengue control methods are also becoming generally ineffective owing to increasing levels of physiological resistance to the pesticides in mosquito populations and the reliance on ineffective insecticide applications that do not target the vector where it harbors [4]. For example, the images of outdoor fogging to control Zika have recently filled the airwaves, while the vector A. aegypti is quietly sequestered inside houses, safe from the temporary insecticidal cloud. Increasing urban sprawl, much of which consists of overcrowded slums with unscreened houses and unreliable water, also makes rapid scaling up of vector control to cover large areas an issue in itself. Finally, there have been renewed calls for the use of paramilitary style control programs like those used to effectively eradicate A. aegypti through the use of residual pesticides like DDT [5]. Although historically successful, we now know that such programs are not sustainable and that the use of DDT is not acceptable in today’s world [6, 7]. In response to these challenges and the recent failures of traditional vector control, new methods involving the rearing and releasing of large numbers of mosquitoes to eliminate or modify local Aedes populations are being developed (see Supplementary Table 1) [8].

REAR-AND-RELEASE METHODS

Using Males to Control the Vector

The use of “rear-and-release” methods to control insects dates back to the development of the sterile insect technique (SIT) to eliminate screwworm flies (Cochliomyia hominivorax) from the United States in the 1950s and 1960s [9]. Male flies were irradiated and sterilized. When released, they mated with wild females, resulting in infertile eggs and reduction and eventual elimination of the screwworm from the southeastern United States. SIT is currently used to control many agricultural pests, and has been developed for use in mosquitoes. Methods of male sterilization include radiation (typically of pupae), chemical methods (use of species-specific double stranded RNA to interfere with sperm production), genetic modification (genes for sperm production knocked out) and microbiological methods (use of the Wolbachia bacteria to induce male sterility).

Of these methods, irradiation has the longest history in vector control. Indeed, in the 1970s, the World Health Organization sponsored an SIT program designed to control A. aegypti and the filariasis vector Culex quinquefasciatus near New Delhi, India. Unfortunately, this program was disbanded [10], primarily due to poor community engagement and public relations, lessons that ring relevant in today’s Internet- and social media–connected society. One of the greatest hurdles in SIT programs is that females are produced in the rearing process and must be removed. This is no small task considering that males are often released at ratios of 10:1 or greater than the existing population. Thus if the sexing method is 99%, you would be releasing 1 female capable of biting and transmitting virus for every 10 females in the wild population. To overcome this issue, male and female pupae are separated mechanically, although this is imperfect and laborious. Furthermore, irradiation can affect male fitness, and reduce its competitiveness in finding and mating with wild females. This is why most SIT programs release “overwhelming” numbers of males to increase the likelihood of successful sterile matings. Modern irradiation methods have reduced the radiation used, and thus fitness loss in sterilized males, especially when combined with Wolbachia infection [11]. For a review of SIT applications for mosquito control, see Alphey et al [12].

A new approach to SIT involves the use of the bacteria Wolbachia to effectively sterilize male mosquitoes. Female mosquitoes that mate with male mosquitoes infected with the bacterium Wolbachia pipientis (wPip) produce eggs that fail to fully develop and hatch (termed cytoplasmic incompatibility; CI), and this method is referred to as the incompatible insect technique (IIT) [13, 14]. Thus, Wolbachia-infected males are effectively sterile. For population elimination it is critical for sex sorting to be very high, >99%, because any Wolbachia-infected females that escape are not sterilized when mating with Wolbachia-infected males. Thus, repeated releases of even small numbers of Wolbachia-infected females can result in establishment of the Wolbachia infection, effectively eliminating the CI sterilization method. To counter this, some groups expose Wolbachia-infected mosquito pupae to a light radiation that does not harm males but sterilizes released females (SIT/IIT method)[15].

In the first successful proof-of-principle IIT release of wPip A. albopictus (wPip), adult abundance was reduced >60% in a single release site in a suburban neighborhood in the United States [16]. These results support IIT has an effective means of population suppression for a nuisance species. Furthermore, some strains of Wolbachia (wMel in A. aegypti) interfere with virus (eg, dengue, chikungunya, and Zika) replication in the mosquito, significantly reducing transmission risk. Using a Wolbachia-based SIT/IIT reduces the risk associated with the release of a small percentage of females that have a compromised ability to spread viruses [11].

A. aegypti has also been genetically modified (GM) to produce males that effectively are sterile due to the presence of a foreign dominant lethal gene [17, 18]. In short, transgenic males harboring the lethal gene are released into the wild to mate with wild females following similar release ratios of traditional SIT programs. Owing to the dominant nature of the inserted gene, all offspring from a mated female will express lethality and die as larvae or pupae, resulting in population reduction. This lethality also makes it a self-limiting system, because the population carrying this gene will eventually collapse. This method is referred to as release of insects carrying a dominant lethal (RIDL). Because RIDL releases follow protocols similar to those of standard SIT releases, they face the same sex-sorting, quality control, and scalability issues. The first, and currently only, RIDL strain to undergo field releases is the OX513A strain of A. aegypti [19, 20]. Field trials have shown that OX513A males can compete equally for mates and have comparable fitness relative to wild populations [19, 21], and sustained releases of OX513A males have been shown to suppress field populations by >90% [20, 22]. These results demonstrate that the release of GM mosquitoes can significantly reduce vector populations and support further development and refinement of this and other GM control methods.

Reducing Mosquitoes’ Ability to Transmit Viruses

The beauty of the use of Wolbachia to control virus transmission in A. aegypti lies in the ability of CI to drive Wolbachia into populations as it effectively sterilizes males by resulting in early embryonic death when Wolbachia-infected males mate with uninfected females [14]. Thus, if Wolbachia-infected males were released at a ratio of 1:1 with wild populations, and there was no fitness loss nor mating bias, then 50% of the wild females would mate with Wolbachia-infected males, and lay effectively sterile eggs. Within 1 generation, 75% of the offspring are infected with Wolbachia, and 87.5% after 2 generations. This powerful drive mechanism allows Wolbachia to spread to fixation from relatively small number of releases of moderate mosquito numbers.

However, there are fitness constraints (fecundity, longevity, etc) caused by the bacteria, and these do affect release numbers and the ability of Wolbachia to spread and persist in the population. The original Wolbachia-based A. aegypti control program relied on the life-shortening abilities of the virulent wMelPop strain of W. pipientis to reduce the capacity of A. aegypti to transmit dengue and other arboviruses. This virulent strain of Wolbachia overreplicated inside mosquitoes’ cells, inducing a range of disease and early death, even in eggs [23, 24]. Efforts to establish wMelPop in A. aegypti in 3 populations (2 in Cairns, Australia; 1 in Tri Nguyen Island, Vietnam) were unsuccessful, and indeed the infection did not persist beyond a few months after releases ceased [25]. However, it was discovered that mosquitoes infected with wMelPop had a dramatically reduced virus infection that nullified their capacity to transmit virus [26]. Further investigation showed that “virus blocking” also extended to wMel-infected A. aegypti, a less virulent strain of Wolbachia that should be easier to establish and maintain in field populations [27]. This virus blocking was not limited to dengue, and has been shown to occur with chikungunya, yellow fever, and Zika virus [28, 29]. The mechanism of viral interference is still unclear, although Wolbachia-induced innate immune system priming [30], competitive interference for intracellular nutrients [31], and suppression of host cellular factors that are up-regulated during viral infection [32] are thought to decrease viral replication.

Semifield cage and open-field releases demonstrated that wMel-infected A. aegypti could drive Wolbachia into wild populations [27, 33]. wMel was introgressed into 2 populations near Cairns after 10 weeks of releases and has persisted at an infection rate of >90% infection for 5 consecutive years [33]. Further expansion of wMel releases has occurred in Cairns and Townsville, Australia. Coincidentally, no locally acquired dengue has occurred in the release areas [34]. The Eliminate Dengue Wolbachia release program is an advanced program, having established releases in Australia, Vietnam, Indonesia, Brazil, and Colombia. The novel end point of Wolbachia releases (reduction in disease transmission with little change in vector populations) indicates that epidemiological disease transmission data, preferable via randomized control trials, will be necessary to measure the efficacy of the intervention.

EMERGING REAR-AND-RELEASE TECHNOLOGY: GENE DRIVE

In addition to Wolbachia, genetic approaches that result in altering vector populations in such a way to eliminate pathogen transmission are advancing rapidly. Some of the most promising advances are being made with gene drive systems [35, 36]. These “selfish” genetic elements enable supermendelian inheritance of a transgene using site-specific endonucleases to spread desired traits into a population rapidly and efficiently. The use of such elements allows them to spread much faster than current technologies as the genetic elements spread within populations even when they provide no benefit to the host organism or even reduce the chance that an individual will reproduce [35]. With these characteristics, gene drives have enormous potential to modify insect populations over a time frame of months to a few years, depending on the scale of initial releases, the size of the targeted population, and the reproductive cycle of the target organism.

Although several genetic elements for gene drive exist, such as transposable elements, meiotic drive, and maternal effect dominant embryonic arrest mechanisms, the most promising, at least in the short term, are those using autonomous clustered regularly interspaced short palindromic repeats (CRISPR)–associated protein 9 (Cas9)–mediated gene editing complexes [35, 37, 38]. CRISPR-Cas9 is an endonuclease that acts as a molecular “scissors” for snipping genes and allows researchers to target, modify, and regulate genomic loci quickly and efficiently. The CRISPR-Cas9 system has been used to successfully sterilize female Anopheles gambiae [38] and to spread antimalarial genes, effectively immunizing the mosquito against malaria, in Anopheles stephensi [37] under laboratory settings. Similar systems could be used to control or block pathogen transmission in A. aegypti [39] and other disease vectors. Although highly promising, there are significant regulatory and safety issues to consider because gene drives may have unanticipated effects, such as potential extinction of the targeted vector, spread of the gene drive outside target area, or traverse into another species [40, 41].

Thus, several criteria need to be considered in selecting candidate gene drive systems. These criteria, outlined by Sinkins and Gould [36], are as follows: (1) the system needs to be powerful enough to spread to fixation during operationally relevant time scales (months to a few years); (2) it must be robust to loss of linkage between the drive mechanism and the effector gene; (3) it must have the ability to spread or modify new target genes in the event of loss of linkage, mutational inactivation, or development of resistance or evasion of the pathogen or vector; and, finally, (4) candidate gene drives should be as safe as possible, without risk of causing undesirable side effects in the target vector or of causing ecological damage in nontarget species and, ideally, would allow removal of the gene drive in the case of unanticipated negative effects. Although the field has advanced substantially since Sinkins and Gould published their review, the basic criteria extend to new technologies not available at the time, including CRISPR-Cas9.

KEY ISSUES AFFECTING REAR-AND-RELEASE INTERVENTIONS

The concept of public health staff releasing large numbers of mosquitoes in residential areas is incongruous and even alarming. Thus, community engagement is critical. This must be used to educate and inform the public to the point that they have acceptance and support. Political and civic leader support is also critical. The Eliminate Dengue program developed a stakeholder group of key community members who galvanized support and served as a conduit for community attitudes [42]. Media must also be supportive and be used to drive key messaging. One only has to read the account of the failure of the 1970s India SIT program to realize the importance of community engagement and media support. GM mosquitoes, especially gene drive systems, have a steeper climb than Wolbachia and other methods of SIT. Many citizens are hesitant or even hostile to releases of GM mosquitoes [43].

Government regulatory approvals are also required for release of insects for population reduction or modification. Novel insect releases are often placed in a gray zone of “not pesticides, not genetically modified, and not classic biocontrol.” Many jurisdictions also have laws against mosquito breeding, so the concept of releasing millions of adult mosquitoes may be alarming to local public health officials and residents. Confused regulators may simply delay making a decision or pass the issue to another agency. Release of GM mosquitoes has been problematic for many regulators [44], although repeated permits have been granted in Brazil, and the World Health Organization has endorsed pilot trials. Once initial permits are given, precedent has been established, and subsequent approvals are much simpler and quicker.

Another issue with male release approaches is the need to sustain releases for extended periods. Although releases of GM male A. aegypti have reduced several A. aegypti populations by >90% [45], once releases stop the populations can rebound or be reintroduced. Thus, strategies to maintain low populations or eliminate populations are needed, and ideally, they can be integrated within existing vector control programs. The maintenance phase of Wolbachia based population modification should be less intensive and expensive owing to the persistence of the Wolbachia infection. Finally, sex separation must be very high, nearly perfect, to avoid establishing Wolbachia in the population and effectively cancelling out the IIT approach.

Clearly, careful monitoring of mosquito release sites must be done before, during, and after release. Typically, this involves monitoring of adult Aedes populations. Aedes SIT and Wolbachia programs have used ovitraps and BG sentinel traps to monitor Wolbachia infections in mosquitoes [20, 25]. Inexpensive passive ovitraps can also be used to sample adult Aedes [46, 47], and sound lures can be added to increase captures of male A. aegypti [48]. For Wolbachia programs, monitoring of established populations should be conducted to confirm infection, vector competence, and population impacts (reduction via IIT).

All rear-and-release programs face the difficult task of achieving the production capacity to effectively treat large urban areas. Indeed, a limited ability to rapidly scale up has compromised the efficacy of insecticidal-based dengue control programs. Insecticidal fogging from vehicles and even planes is scale-up but has proved ineffective. So, how does one scale up the production and release of mosquitoes to cities of several million inhabitants? Efforts are underway to address the challenge of such scaling up to cities of several million inhabitants. The International Atomic Energy Agency SIT program (sterilization via radiation) has developed larval rearing trays and a racking system that can produce 100000 male A. albopictus pupae per week in just 2 m² of laboratory space [49], and a rearing facility in China claimed to be able to produce 5 million by September 2016 [50].

Current release strategies involve releasing adults from vans or trucks, aerially or even by drones. Pupal and egg releases have been conducted. For instance, the Eliminate Dengue program distributes egg strips in take-away food boxes to which residents add water and fish food for home rearing of wMel-infected A. aegypti [51]. This program can be facilitated by community groups (eg, schoolchildren) to increase reach and engage the community. Because the Eliminate Dengue program only has to release until fixation is achieved (typically 8–12 weeks), releases can be rolled out stepwise across a city. Ultimately, mosquito rear-and-release methods will be mechanized, with robotic mosquito rearing factories and potentially automated release systems, to produce the billions of mosquitoes needed for treatment of large urban areas. Development of these systems will take time.

Finally, questions of cost effectiveness and cost-benefit will inevitably arise. Although no data are available on the contemporary costs of mosquito-based SIT, in practice its cost-benefit profile is advantageous compared with traditional vector control. This is because the costs of traditional insecticide control are proportional to the area treated, given that a given amount of pesticide must be used regardless of whether the target population in a given area is large or small, resulting in decreased efficiency (fewer insects killed) [12, 52]. But, for SIT programs, the number of sterile males released and subsequent costs are proportional to the size of the target population in a given area. Thus, the cost-benefit ratio of SIT programs decreases with a decrease in the size of the target population, regardless of the size of the area treated. This allows SIT programs to reduce costs while increasing efficacy, because such efforts are better at getting the last vectors (small population) than the first ones (large population) [12]. The cost-benefit profile of Wolbachia-based population modification programs is potentially greater because they do not require continued releases after fixation. Ultimately, whatever the costs may be, they will be weighed against the economic, social, and healthcare costs of the diseases they target, which in case of dengue is currently $8.9 billion annually [53]. Thus, if programs like Eliminate Dengue are effective, they would represent extremely good value for the money.

IMPROVING THE BASICS

Because rear-and-release strategies are not ready for widespread implementation in large cities, renewed effort should be made to improve existing vector control tools. To begin, existing programs need to avoid methods that have failed, such as space sprays, and improve existing technologies to increase their efficacy [54], as well as building up budgetary and infrastructure support for local control programs. In essence, ineffective outdoor thermal and ultralow volume fogging should be phased out, while indoor spraying with residual insecticides should be increased [55–57], using insecticides to which vectors are susceptible [4]. New trapping technologies using gravid traps can also reduce vector density and virus transmission [58, 59] and could be useful, provided cost and logistics are reduced. Insecticide-treated screens, harborage spraying (for A. albopictus [60–62]), and improved sanitation and access to clean water are also known to reduce vector populations [63, 64]. Finally, there is a growing consensus that eliminating diseases like dengue will be achieved only by integrating vector control with vaccines [4, 65]. Thus, prioritizing vector interventions based on their potential to prevent disease in combination with vaccine distribution is essential. To this end, future assessments of emerging and existing tools should include epidemiological as well as entomological outcomes. It is likely that rear-and-release strategies will become viable interventions for combination with vaccine distribution to reduce disease, but until further evidence is available, the current focus should remain on improving existing interventions.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Financial support. This work was in part supported by the National Health and Medical Research Council of Australia and the Queensland State Government (Accelerate Fellowship to D.A.M.).

Potential conflicts of interest. Both authors: no reported conflicts.

Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References

Author notes