Development and applications of transgenesis in the yellow fever mosquito, Aedes aegypti

Abstract

Transgenesis technology has been developed for the yellow fever mosquito, Aedes aegypti. Successful integration of exogenous DNA into the germline of this mosquito has been achieved with the class II transposable elements, Hermes, mariner and piggyBac. A number of marker genes, including the cinnabar⁺ gene of Drosophila melanogaster, and fluorescent protein genes, can be used to monitor the insertion of these elements. The availability of multiple elements and marker genes provides a powerful set of tools to investigate basic biological properties of this vector insect, as well as the materials for developing novel, genetics-based, control strategies for the transmission of disease.

Introduction

The genus Aedes is a member of the order Diptera, suborder Nematocera, in the family Culicidae, subfamily Culicinae [1]. Aedes mosquitoes are found in tropical and subtropical zones throughout the world, and are responsible for the transmission of a number of viral and filarial human pathogens. The most well-known member of the genus, Aedes aegypti, is capable of transmitting viruses such as yellow fever (YF), dengue (serotypes 1–4) and chikungunya, all of which can cause severe morbidity and mortality. While a safe and effective vaccine for YF virus has been available for almost sixty years now, there is still epidemic transmission occurring in both Africa and South America [2]. Dengue viruses, which cause dengue fever (DF) and dengue hemmorrhagic fever (DHF), and for which there is currently no safe and effective vaccine, represent an even larger emerging problem. With between 50 and 100 million cases of DF and DHF annually worldwide, dengue viruses represent a huge burden on developing countries [3], [4]. Furthermore, 2001 has seen the return of epidemic dengue fever to Hawaii after a 60-year absence indicating that this disease is now emerging in the developed world as well [5].

In addition to viral pathogens, A. aegypti (and other Aedes spp.) can transmit a number of filarial parasites such as Wuchereria bancrofti, Brugia malayi and Brugia timori (although Culex spp. are a much more important disease vector of Brugia spp. [1]). Furthermore, A. aegypti has long been used in the laboratory to study malaria based on its ability to transmit the avian parasite, Plasmodium gallinaceum. This model system, which cycles the parasites between mosquitoes and chickens, is a simpler and safer way of studying the biology and genetics of a Plasmodium species. The underlying expectation is that information obtained by studying this model system will help understand a far more important pathogen, P. falciparum [6], which is responsible for the severest and most deadly forms of human malaria.

Other mosquitoes such as A. albopictus, A. pseudoscuttellaris and A. africanis are also capable of transmitting DF and YF viruses, however the behavior of these species makes them much less significant as vectors [7]. Domesticated A. aegypti has adapted to live in close contact with people, preferring to feed almost entirely on human blood, while breeding and depositing eggs in man-made containers such as trash and water storage bins [8], [9], [10]. It is for these reasons that A. aegypti is the most studied member of the genus, and the only one to be successfully transformed (A. triseriatus was successfully transformed [11], but this mosquito has since been reclassified to the genus Oclerotatus [12])

Attempts to transform Aedes mosquitoes began in the late 1980's using the P element from Drosophila melanogaster [13]. P elements belong to a group of transposable elements (designated class II, [14]) that mobilize (excise and integrate) via a DNA intermediate. These elements are small, usually 1–3 kilobases (kb) in length, contain characteristic inverted terminal repeat (ITR) DNA sequences at their ends, and an open reading frame that encodes a transposase that is capable of catalyzing the transposition of the element. Efforts to transform A. aegypti were based on the two plasmid system that was used successfully with the fruit fly. One plasmid, the donor, carries selectable or screenable marker genes flanked by the ITRs. The transposase gene has been removed so that this construct is not capable of mobilizing itself (non-autonomous). The second plasmid, the helper, contains the transposase gene cloned so as to be under the control of an inducible promoter, usually derived from one of the heat shock genes of D. melanogaster. Microinjection of both plasmids into the early embryos of the insect is followed by a brief heat shock. The transposase expressed from the helper plasmid acts in trans to mobilize the construct from the donor plasmid and integrate it into the germline chromosomes of insects. Injected insects (G₀) are grown to the adult stage, mated to members of the opposite sex, and the progeny (G₁) scored for evidence of integration of the marker gene. The presence of a phenotype associated with the marker gene in the G₁ progeny is taken to indicate that the transposable element has inserted into a chromosome that was contained in a future sperm or egg cell. The structure of the inserted element is verified by molecular analyses such as Southern hybridization or gene amplification [15]. P element constructs did integrate into A. aegypti, but these events were not mediated by the P transposase and occurred at a frequency far to low to be useful on a regular basis [13]. Efforts to modify transformation conditions and the element itself proved unsuccessful, and P was finally abandoned as a tool for generating transformed Aedes mosquitoes.

With the failure of the P element, several other methods of transformation were explored. FLP-FRT, a high frequency, site-specific, recombination system adapted from the yeast, Saccharomyces cerevisiae, was tested for function in transient assays in A. aegypti embryos [16]. The FLP recombinase catalyzes recombination events between two FRT sites. The proposed strategy was to use a low frequency event to integrate an FRT site into the mosquito genome as a ‘docking site’, and then use FLP-mediated site-specific recombination to insert genes of interest. High levels of FLP-mediated recombination were detected among plasmids in mosquito embryos. The problem, of course, was that there was no good method of introducing the FRT docking sites into the mosquito chromosomes. Other methods of generating transgenics such as homologous recombination, non-homologous recombination, baculoviruses, retroviruses, were all tried with no success (Yardley and James, unpublished).

Transformation of A. aegypti and other mosquitoes was made possible by the isolation and characterization of additional class II transposable elements such as Hermes, mariner, Minos and piggyBac [17], [18], [19], [20]. These transposons were discovered first in insects, and all except piggyBac belong to large families of elements that appear widespread throughout eukaryotes. Having learned a lesson from the many failed experiments with P, researchers first tested these new elements for mobility with in vivo transposition assays. These experiments involve microinjecting embryos with three different plasmids, a donor that contains a marker gene flanked by the ITRs of the transposable element, a target that carries a number of marker genes distinct from the donor, and a helper that contains the appropriate transposase under the control of an inducible promoter [21]. Induction of the transposase mobilizes the marker gene construct from the donor to the target. Successful transposition assays in A. aegypti embryos were soon reported for Hermes [22], mariner [23], and piggyBac [24].

While the search for effective transformation systems were underway, researchers also were struggling to develop efficient selection and screening protocols. Selection based on resistance to antibiotics such as G418, while theoretically a powerful way to identify rare transformants by obviating the need to screen through large numbers of untransformed individuals, had its share of shortcomings. Antibiotic screening was targeted at selecting the immature mosquitoes, larvae, as they developed in their aqueous environment. Depending on the mosquito species, microgram to milligram quantities of antibiotics per milliliter were needed. A. aegypti were screened at concentrations as high as 1 mg ml⁻¹ [13]. This could be expensive when 250–500 ml of water are used for each group of 100–300 larval mosquitoes. Furthermore, the amount of antibiotic necessary to effect selection was difficult to set, not enough antibiotic, and too many untransformed larvae survived, too much antibiotic and all animals died. The difficulties with delivering a consistent, discriminatory amount of antibiotic make this method of selection impractical.

The need for marker genes was met first in A. aegypti by the discovery that a heterologous gene from D. melanogaster could rescue a visible mutant phenotype in a mosquito strain. The white-eye strain of Ae. aegypti is deficient in the expression of an enzyme, kynurenine hydroxylase, which is necessary for the formation of the pigments that are deposited into the eyes [25]. A wild-type copy of the gene, cinnabar, encoding the homologous protein in D. melanogaster, is capable of transient rescue of the mutant white-eye phenotype when microinjected into mosquito embryos [26]. This fruit fly gene was the first marker to be used successfully in A. aegypti transformation [23], [27] (Fig. 1). Since that time, a number of fluorescent marker genes based on the green fluorescent protein (GFP) and dsRed, have been used as markers in Aedes transformation [28] (Fig. 1).

One of the more powerful tools to use in adjunct to transposon-mediated integration of DNA is the recombinant Sindbis virus expression system, a method for transient expression of mRNAs (coding or non-coding). Sindbis virus is a positive-strand RNA virus that establishes a non-cytolytic, lifelong infection in mosquitoes. Recombinant Sindbis virus constructs have been developed that allow the high level expression of expressed portions of genes [29]. This saves a lot of time and effort when analyzing the effector portions of genes prior to their use in transgenic animals.