Molecular Genetic Manipulation of Vector Mosquitoes

Abstract

Genetic strategies for reducing populations of vector mosquitoes or replacing them with those that are not able to transmit pathogens benefit greatly from molecular tools that allow gene manipulation and transgenesis. Mosquito genome sequences and associated EST (Expressed Sequence Tags) databases enable large-scale investigations to provide new insights into evolutionary, biochemical, genetic, metabolic and physiological pathways. Additionally, comparative genomics reveals the bases for evolutionary mechanisms with particular focus on specific interactions between vectors and pathogens. We discuss how this information may be exploited for the optimization of transgenes that interfere with the propagation and development of pathogens in their mosquito hosts.

INTRODUCTION

We have known for more than a century that mosquitoes are vectors of pathogens that cause human disease. Still the mortality and morbidity rates associated with these pathogens remain high in low-income countries in tropical and sub-tropical regions (Guinovart et al., 2006). In total, more than half of the global population is affected by mosquito-borne diseases resulting in millions of deaths and hundreds of millions of cases every year. These statistics provide the impetus to study mosquitoes with the expectation that new knowledge could contribute to the alleviation of this disease burden.

The application of genetic analyses and molecular biological techniques for research on mosquitoes provides opportunities for the development of new disease control strategies (Hill et al., 2005). Among these opportunities are novel vector control methods for population reduction or replacement (Curtis and Graves, 1988). Population reduction seeks to decrease the absolute number of mosquitoes, and therefore, lower the probability of contact between mosquitoes and their human hosts. Population replacement strategies are designed to replace susceptible mosquitoes (can transmit a pathogen) with refractory mosquitoes (cannot transmit a pathogen) and such strategies do not require changes in mosquito population densities (Figure 1). For any of these strategies to be effective, it is important to reduce the number of infectious mosquitoes below a threshold level so that the probability of transmission falls to a point where the parasite population declines steeply and irreversibly.

Figure 1.

Theoretical display of transgenesis strategy impact on vector populations. In mosquito population reduction, the total population size decreases and thereby the number of infected vectors decreases as well. In mosquito population replacement, the susceptible vectors (able to transmit the pathogen) are replaced by refractory vectors that cannot transmit. The transmission threshold indicates the lowest level of transmission of infectious agents for a disease to be sustained in a population of humans or animals.

One population replacement strategy has the goal to genetically-modulate vector competence and is based on the hypothesis that an increased frequency in a vector population of a gene that interferes with a pathogen will result in the reduction or elimination of transmission of that pathogen (Collins and James, 1996). Testing this hypothesis spurred the development of molecular genetic technologies for gene analysis and manipulation in mosquitoes. A key objective was to establish routine methods for generating transgenic mosquitoes and this was achieved with a number of species using Class II transposable elements (Allen et al., 2001; Grossman et al., 2001; Jasinskiene et al., 1998). These successes stimulated debate and research to measure the impact of introduced genes on mosquito fitness. For example, integration of a transgene may disrupt or alter expression characteristics of endogenous genes, transgene products may be toxic, or transgene transcription and translation may usurp resources needed for normal survival or reproductive functions. Accordingly, the effects of transgenes integration and expression on mosquito fitness vary (Catteruccia et al., 2003; Irvin et al., 2004; Marrelli et al., 2006). Most genetic approaches require the transformed insects to exhibit as low a fitness cost as is possible (Lambrechts et al., 2008). Therefore, the expression of a transgene should be limited to a specific tissue and time in the mosquito to achieve the maximum effect on the pathogen, while minimizing the potential load on the vector. Recent advances in the areas of genomics and genetic engineering are expected to enable the design and production of mosquitoes expressing anti-pathogen effector molecules under the control of synthetic or hybrid (chimeric or mosaic) promoter-regulatory DNA to achieve optimum performance. Functional synthetic plant promoters that combine a collection of DNA sequence elements from pathogen-activated genes (Rushton et al., 2002) serve as a conceptual model for similar developments in mosquitoes.

GENOMIC TOOLS FOR THE DISCOVERY OF CIS-REGULATORY ELEMENTS FOR DIRECTING TRANSGENE EXPRESSION

The genomes of the malaria mosquito, Anopheles gambiae, the Yellow Fever mosquito, Aedes aegypti, and the common house mosquito, Culex quinquefasciatus, are available in publicly-accessible databases (Table 1). These resources provide both a wealth of information and challenges for the design and implementation of transgenic mosquitoes. It is now feasible to use these resources to better understand gene regulation in mosquitoes, and apply this knowledge to either identify or design promoter DNA capable of directing expression of effector molecules at relevant time points after infection and within tissues in which there are significant interactions with the pathogen, such as the midgut, salivary glands, and fat body.

Table 1.

Mosquito genomics resources.

Sequence databases
Vectorbase http://www.vectorbase.org at NIAID, USA, is a genomic resource center for invertebrate vectors of human pathogens. It contains the genomes for Aedes aegypti, Anopheles gambiae and Culex quinquefasciatus and several other vector insects.
NCBI http://www.ncbi.nlm.nih.gov at NLM, USA, is a general resource for bioinformatics tools which includes a public repository of DNA and protein sequences (GenBank). In addition to the complete genomes of Ae. aegypti, An. gambiae, and Cx. quinquefasciatus, this database includes sequence information on selected genes, mRNA and proteins of many other mosquito species.
TEfam http://tefam.biochem.vt.edu/tefam/index.php at Virginia Tech, USA, is a relational database for submission, retrieval, and analysis of transposable elements (TEs). Currently the focus is on Ae. aegypti and An. gambiae.
AnoBase http://www.anobase.org at IMBB, Greece, contains genomic/biological and other information on anopheline mosquitoes, with an emphasis on An. gambiae.
Exon http://exon.niaid.nih.gov/transcriptome_anopheles_gambiae.html at NIAID, USA, provides transcriptome data and related scientific papers for An. gambiae. Included are: 1) Clusterized set of more than 200,000 publicly available ESTs; 2) Anogold - containing more than 3,000 full length CDS, or truncated CDS containing either starting Met or Stop codon of genes; 3) In silico microsatellite database – a database of Primer3-designed primers flanking genomic microsatellite sequences; 4) Ag-Glimmer-Xcel – An. gambiae proteins obtained by combinational algorithm of Glimmer trained on the Anogold set (from 2 above) and ENSEMBL protein predictions, and AnoXcel – a database of ENSEMBL An. gambiae proteins, with various BLAST links and 5) Affy-Xcel – An annotated version of the probe sets included in the Plasmodium/Anopheles Affymetrix chip.
(Prote/Transcript)-omes http://exon.niaid.nih.gov/transcriptome at NIAID, USA, contains links to the transcriptomes of vector arthropods, as well as Excel formatted data sets of whole proteomes. Includes data on the salivary glands transcriptomes of An. darling, An. funestus, An. gambiae, An. stephensi, Aedes aegypti, Ae. albopictus, and Cx. quinquefasciatus.
FlyMine http://www.flymine.org at Cambridge University, United Kingdom, is an integrated database for Drosophila and Anopheles genomics with lists of orthologues for Drosophila melanogaster, Ae. aegypti and An. gambiae genes.
The Alternative Splicing Annotation Project (ASAP) database http://www.bioinformatics.ucla.edu/ASAP2 at UCLA, USA, enables biologists to access and mine alternative splicing information from genomics and proteomics.
Gene expression databases
The Anopheles gambiae Gene Expression database at UC Irvine http://www.angagepuci.bio.uci.edu is a relational database that combines data from microarray experiments, functional annotation, and the An. gambiae genome project.
SEBIDA http://141.61.102.16:8080/sebida/index.php at Max Planck Institute of Biochemistry, Germany, is a database for the functional and evolutionary analysis of sex-biased genes which integrates data from multiple microarray studies comparing male versus female gene expression in D. melanogaster, D. simulans, and An. gambiae.
FlyMine http://www.flymine.org shows gene expression data for a particular An. gambiae developmental stage and filters the data by mean signal ratio.
VectorBase http://www.vectorbase.org/ExpressionData provides a repository for microarray data produced by the insect vector community.
Genome browsers
VectorBase http://www.vectorbase.org/ExpressionData provides access to the annotations of mosquito genomes via a genome browser. The DAS protocol allows community researchers to integrate and display their data sets in the genome browser window.
Ensembl http://www.ensembl.org at EMBL-EBI and the Wellcome Trust Sanger Institute, UK aims at developing a system that maintains automatic annotation of large eukaryotic genomes. Presently Ensembl provides access to Ae. aegypti and An. gambiae genomes.
UCSC Genome Bioinformatics http://genome.ucsc.edu at UC Santa Clara, USA contains the reference sequence and working draft assemblies for a large collection of genomes including An. gambiae.

The identification of functional mosquito promoters was not restricted by the absence of whole genome sequences. Cloning and characterization of individual genes combined with transgenesis allowed the discovery of DNA fragments that contain regulatory elements capable of directing restricted temporal and spatial gene expression. The functionality of a number of promoters was demonstrated by determining expression of reporter or anti-pathogen genes in genetically-transformed mosquitoes (Ito et al., 2002; Kokoza et al., 2000). Promoters are available for expressing products that accumulate in the female midgut, hemolymph and salivary glands (Chen et al., 2007b; Moreira et al., 2000; Yoshida and Watanabe, 2006) and promoters were identified that enable stage- and sex-specific expression of genes (Adelman et al., 2007; Catteruccia et al., 2005; Munoz et al., 2004; Smith et al., 2007). However, the availability of three mosquito genomes expands our ability to explore mechanisms involved in regulation of transcription and allows comprehensive searches for genes displaying well-defined spatial and temporal expression patterns.

Transcriptome analyses that utilize high-throughput sequencing of cDNAs (complementary DNA sequences) and determinations of transcription regulation by microarray-based technology are powerful technical approaches to identify genes with suitable expression patterns. More than 200,000 An. gambiae, 300,000 Ae. aegypti, and 200,000 Cx. quinquefasciatus ESTs are available currently at Vectorbase and NCBI (National Center for Biotechnology Information, Table 1), and these provide evidence of transcriptional activity as well as reveal gene structures, particularly at the 5‘ - and 3‘ -ends where nucleotide sequences are often less conserved. Additionally, these large sequence data sets allows the identification of genes involved in mosquito immunity, olfaction, reproduction, insecticide resistance, xenobiotic detoxification, endocrine regulation, and blood-meal acquisition and digestion (Biessmann et al., 2005; Calvo et al., 2007; David et al., 2005; Marinotti et al., 2005; Strode et al., 2008; Waterhouse et al., 2007). These efforts are complemented with modern proteomics approaches to identify gene products (He et al., 2007; Li et al., 2006). Although automated predictions of gene structure are essential to manage large data sets, manual annotation based on sequences of high-quality cDNAs and proteins are important in revealing novel transcriptional products and alternative splicing variants, and providing precise exon boundaries (Ding et al., 2003; Li et al., 2006).

The possibility of conducting genome-wide analyses of gene expression by applications of microarray technologies is another advantage granted by mosquito genome and transcriptome data sets. These applications provide considerable information on distinct transcription profiles of the approximately 15,000 protein-encoding genes of An. gambiae at specific developmental stages (Dana et al., 2006; Dana et al., 2005; Koutsos et al., 2007; Marinotti et al., 2006; Warr et al., 2007). High-density, whole-genome tiling arrays (WGTA) consisting of overlapping oligonucleotides designed to represent the entire (coding and non-coding) non-repetitive genome of a given organism also are current options. Importantly for regulatory DNA searches, comprehensive WGTA-based analyses can provide detailed maps of the transcribed regions (Huber et al., 2006). WGTAs can be used additionally for analyzing DNA–protein interactions (e.g. mapping of transcription factor-binding sites), and the epigenetic status of the genome (Shiu and Borevitz, 2008). As a result of an increasing number of projects directed to mosquito high-throughput DNA sequencing and annotation, and gene expression analyses, specialty databases and exploration tools were developed and are available to mosquito biologists (Table 1).

Among many advances in the knowledge of mosquito biology that are expected to provide new concepts of vector control, the sequencing of mosquito genomes provides the bases for detailed studies of mechanisms involved in gene regulation. These studies can provide means to define customized promoters as tools for delivering the products of a transgene to specific sites and desired developmental stages within mosquitoes. Cis-acting Regulatory Elements (CREs), such as core and proximal promoters, enhancers, and repressor elements, are specific and discrete stretches of DNA that regulate gene expression via interactions with trans-factors that bind in a sequence-specific manner. The discovery in silico of CREs is an active area of research stimulated most recently by the publication of genomic and associated genome-wide expression data. However, identifying CREs is challenging because they are usually short in length, degenerate in primary structure, and embedded in large amounts of intergenic genomic DNA. Despite these difficulties, a number of computational approaches were developed for CRE discovery and identification of closely-linked CREs, termed modules (Das and Dai, 2007; MacIsaac and Fraenkel, 2006; Hu et al., 2005; Tompa et al., 2005; Gupta and Liu, 2005). Comparative genomics provides a powerful approach to increase the signal-to-noise ratio, based on the concept that selective pressures constrain CRE evolution to a slower rate when compared with the surrounding neutral sequences, and as such, CREs are possible to distinguish from the neutrally-evolving background (“phylogenetic foot-printing”) (Zhang and Gerstein, 2003). This approach was utilized successfully for discovering regulatory motifs in both mammalian and Drosophila species (Stark et al., 2007; Xie et al., 2005), and similar applications in mosquitoes are now possible (Yavatkar et al., 2008).

TRANSGENESIS AS A COMPONENT FOR REGULATING VECTORIAL CAPACITY AND VECTOR COMPETENCE

The hypothesis that genetic transformation of mosquitoes will provide additional alternatives for control has been investigated for just over two decades. However, the prospects for using genetically-engineered mosquitoes to control vector-borne diseases depend on the answers to four questions (Nirmala and James 2003): (1) is it possible to genetically-modulate and introduce new traits such as pathogen refractoriness into mosquitoes; (2) can laboratory achievements be translated into something that can be used in the field; (3) do we know enough about the field to use these strategies safely and effectively; and (4) can the political and social will to introduce genetically-engineered mosquitoes be mustered? The answer to the first question is “yes”, and work is continuing in earnest on the remaining.

Although details in the techniques vary with the species to be genetically transformed, the general procedures and concepts remain the same. Transgenic mosquitoes are created by injecting mosquito embryos with a plasmid containing the desired genetic elements coupled to a marker gene (commonly yielding fluorescent eyes) flanked by the inverted repeat structures of a Class II transposon, along with a helper plasmid that produces the homologous transposase (Table 2). With transposase present, the genetic element is excised from the plasmid and integrated into the mosquito chromosomal DNA. Integrations of transposons into the mosquito genome are thought to be random and in some cases may cause lethal insertions.

Table 2.

Mosquito transgenesis resources.

Procedures for making transgenic mosquitoes:
Mosquitoes (Aedes aegypti including attB Ae. aegypti and several species of Anopheles) are available from MR4, Malaria Research and Reference Reagent Resource Center, ATCC, USA http://www.mr4.org
A video showing how to rear An. gambiae can be found at: http://www.jove.com/index/details.stp?ID=221.
A video showing how to make transgenic Ae. aegypti can be found at: http://www.jove.com/index/details.stp?ID=219.
A video showing how to make transgenic An. stephensi can be found at: http://www.jove.com/index/details.stp?ID=216.
Transgenesis facilities:
Oxitec Ltd, United Kingdom http://www.oxitec.com
The Insect Transgenic Facility at University of Maryland, USA http://www.umbi.umd.edu/itf-cbr

Class II transposons are not the only means of integrating exogenous DNA in mosquito genomes (Venken and Bellen, 2007). Recently, the phage phi C31 was exploited to achieve site-specific insertion of exogenous genes into the Ae. aegypti genome (Nimmo et al., 2006). In its native environment, this bacterial phage uses an integrase to mediate a unidirectional recombination event between an attP site in the phage and the attB site in the Streptomyces genome (Groth et al., 2004). Linking an attP or attB site to a circular DNA fragment can facilitate its stable uptake into a chromosomally-located attB or attP site, respectively (Thyagarajan et al., 2001). This system may mitigate problems posed by random integration, preventing disruption of important or essential genes and associated lethality or imposing fitness loads. To meet these needs, the primary insertion of the docking site must be tested for its impact on fitness. A number of life-table parameters may be used to determine if the insertion is having a net effect on reproduction. Having a pre-established “docking site” for transgenes that has been evaluated for minimal loads and insertion-site influences enables selection of transgenic lines with high fitness. Additionally, the chromosomally-integrated docking sites allow recombination of large transgenesis constructs (10-15 kb) as was demonstrated in other organisms (Venken et al., 2006).

Concepts of transgenic mosquitoes were focused on those carrying anti-pathogen effector genes, and some of these are validated experimentally. Malaria parasites can be incapacitated using effectors such as toxic proteins or peptides, while viruses replication can be suppressed by RNA interference (RNAi) targeting viral structures (Fig. 2). However additional concepts that apply insect transgenesis to control vector-borne diseases led researchers to explore alternative approaches. Vectorial capacity is a numerical index comprising both extrinsic and intrinsic factors that contribute to the successful transmission of a pathogen by a mosquito (Beerntsen et al., 2000; Dye, 1992). These factors include host preference, vector longevity, and frequency of vector feeding. All of these potentially could be modulated genetically as ways of reducing vectorial capacity. Robust expression of exogenous odorant receptors and/or odorant binding proteins in the antenna of anthropophilic mosquitoes could redirect their preferences towards other animals. Alternatively, integration into the genome of mosquitoes of a gene that reduces longevity, under control of a pathogen-regulated promoter, would negatively impact vectorial capacity. These and similar applications envisioned for mosquito control and the control of mosquito-borne disease transmission will benefit greatly from a better understanding of gene regulation mechanisms in these insects.

Figure 2.

Schematic overview of different transgenesis strategies. Strategies result in either population reduction or population replacement. Colored boxes and lines are keyed to specific strategies.

Strategies for insect population reduction are being developed (Alphey et al., 2008). Among them, the release of insects carrying a dominant lethal (RIDL) is a genetic variation of Sterile Insect Technique (SIT), and requires insects to carry a dominant, sex-specific lethal gene whose expression can be repressed during rearing under controlled conditions (Heinrich and Scott, 2000; Thomas et al., 2000). Laboratory experiments with D. melanogaster demonstrated that RIDL has advantages to conventional SITs by having less impact on the survival and ability of adult males to mate (Thomas et al., 2000). Moreover, for public health applications, female mosquitoes are eliminated before release to prevent mating with transgenic males, and most importantly, to avoid release of insects capable of disease transmission. In envisioned RIDL-based programs, transgenic male mosquitoes mate with wild-type females; some or all of the progeny of released individuals die as a consequence of inheriting and expressing one or more dominant lethal genes, and a decline in the population size is anticipated (Atkinson et al., 2007; Phuc et al., 2007). Female-specific lethality can be achieved using sex-specific control DNA sequences or taking advantage of alternative, sex-specific splice variants that occur between males and females (Munoz et al., 2004; Fu et al., 2007). All molecular tools necessary for RIDL implementation are available, namely cis-regulatory regions, effector lethal genes, mass rearing conditions for the population expansion and sexing steps required for the process (Phuc et al., 2007).

Another genetic control approach that combines population replacement with lethality, “death-on-infection”, demands that a population of mosquitoes carry a conditional lethal gene that is activated by the presence of parasites or viruses. This results in the selective death of the infected vectors. Death-on-infection diminishes possible ecological effects associated with populations reduction while reducing transmission rates. Specific pathogen-responsive promoters to drive the expression of toxic proteins, apoptosis effectors, or double-stranded RNA (dsRNA) targeting vital gene transcripts are essential for this approach. Analyses have been conducted of global changes in gene expression in mosquitoes exposed to parasites or viruses (Dong et al., 2006; Meister et al., 2005; Xi et al., 2008) with a goal to identify genes with enhanced level of expression upon infection. Continued studies of mosquito immune responses will contribute the necessary regulatory elements for the development of this approach.

Homologous recombination techniques for mediating targeted gene knock-outs are available for D. melanogaster (Rong et al., 2002) and are now are being explored in a number of laboratories for the use in mosquitoes. Ablation of genes whose products are essential for pathogen survival could create mosquitoes refractory to parasite or virus infection (atreptic immunity). Gene knock-outs could target mosquito receptors necessary for parasite or virus entrance into host cells. Additionally, the behavior of anthropophilic mosquitoes could be altered from searching for human blood towards taking blood meals from animals that cannot function as pathogen hosts.

MOVING EFFECTOR GENES INTO WILD-MOSQUITO POPULATIONS

While mass release (inundation) is required for those approaches (SIT, RIDL) that do not propagate genes through a target population, implementation of population replacement requires an effective system for gene drive to spread genes into wild populations. Gene drive involves the introduction and establishment of a population replacement effector gene using genetic mechanisms that circumvent Mendelian inheritance (Braig and Yan, 2002; James et al., 2006; Sinkins and Gould, 2006). Standard genetic approaches based on gene segregation and selection require fitness advantages linked tightly to the anti-pathogen gene and are likely to be too protracted in time to be useful. The bases for gene drive systems come from known genetic phenomena, two of which are discussed here. Mobile genetic elements, such as the previously-mentioned Class II transposons, may move rapidly into populations as in the example of D. melanogaster where the P elements spread world-wide in a period of ~50 years (Kidwell, 1983). These mobile genetic elements spread through populations by replicative transposition in the germline, which means that a mating between an animal with an active transposon in its genome and an animal without can result in progeny that all have the element. However, mobilization of naturally-occurring or inserted transposons has yet to be observed in wild populations or transgenic mosquitoes (Sethuraman et al., 2007). This could be a natural consequence of strong selective pressures limiting transposon movement (O'Donnell and Boeke, 2007).

Moving genes into a population without relying on transposon mobilization is possible with Medea (maternal-effect dominant embryonic arrest). Medea refers to a genetic mechanism based on selfish genes that select for their own survival by inducing maternal-effect lethality in all offspring not inheriting the element-bearing chromosome from the maternal and/or paternal genome (Chen et al., 2007a). The Medea system consists of two components, a maternally-expressed toxin and a zygote-expressed antidote. The toxin is regulated by a germline-specific promoter and the antidote is regulated by a zygotic promoter. The toxin is present in all oocytes originating from mothers carrying Medea, but it only induces lethality during embryogenesis. All zygotes with the Medea antidote gene survive, while embryos without die. As demonstrated in cage experiments with D. melanogaster, linking transgenes to a Medea gene drive mechanism would ensure the rapid spread of the desired phenotype throughout wild populations (Chen et al., 2007a). Application of a similar selfish gene-based system to spread a desirable lethal or anti-pathogen effector gene is possible and is the subject of investigation (Chen et al., 2007a; Enserink, 2007). These and other mechanisms for gene drive, including homing endonucleases, underdominance and symbionts (Wolbachia) (Sinkins and Gould, 2006), are essential for further development of population replacement strategies for mosquitoes.

CONCLUSIONS

Considering the advances in both mosquito genomics and transgenesis, we are within reach of creating an optimal genetically-modified mosquito. The theoretical framework is laid out, however, practical hurdles such as difficulties in routine transgenesis (An. gambiae) and the lack of laboratory colonies for some important species (e.g. An. darlingi) need to be overcome. The logistics of the field applications of these technologies remain to be worked out. While experience from SIT approaches can be applied to some of the challenges (mass-rearing and other factory-level developments), others such as mating compatibilities of transgenic mosquitoes, dispersion and risk-assessment represent new areas of investigation and research to find workable solutions. In addition, community desire and political will to try new disease control practices will have to be engaged before transgenic mosquitoes are released as a disease-preventing intervention.