Transposable Elements for Insect Transformation Part 2

Transformation Marker Systems

The availability and development of selectable marker systems has played a large part in recent advancements in insect transformation, which have been equal in importance to vector development. The rapid implementation and expansion of P transformation in Drosophila was possible, in large part, due to the availability of several eye-color mutant-rescue systems. These systems depend on the transgenic expression of the dominant-acting wild type gene for an eye-color mutation present in the host strain (see Sarkar and Collins, 2000). Successful transformation of non-drosophilid species was similarly dependent upon the development of analogous systems, with the first transformations in Ceratitis capitata and Aedes aegypti relying on white and cinnabar mutant-rescue systems, respectively. While chemical resistance markers were used initially for non-drosophilid transformation, and can be highly useful for specific applications, their inefficiency and inconsistency when used alone provided ambiguous results for several species (see ffrench-Con-stant and Benedict, 2000). Eye-color marker systems are generally efficient and reliable, and cloned wild type genes from Drosophila often complement orthologous mutant alleles in other insects; however, only a handful of species have stable mutant strains that can serve as suitable hosts for mutant-rescue strategies. The most significant advancement in marker gene development for the wide use of insect transformation has been the development of fluorescent protein markers (see Higgs and Sinkins, 2000; Horn et al., 2002). As dominant-acting neomorphs that do not depend on pre-existing mutations, they are directly useful in almost all host strains. When compared to the white eye-color marker in Drosophila, the enhanced green fluorescent protein (EGFP) gene seemed to be less affected by position effect suppression, and thus has the additional advantage of more reliable detection (Handler and Har-rell, 1999). Certainly for the forseeable future, fluorescent protein markers will continue to be the markers of choice for most insect transformation strategies.

Eye-Color Markers

The first insect transformations used mutant-rescue systems to identify transformant individuals, but in these experiments total genomic DNA was used, rather unreliably, to complement mutations in the respective host strains. The most reliable of these, however, was reversion of the vermilion eye-color mutation in D. melano-gaster (Germeraad et al., 1976). The success of the initial P element transformations in Drosophila also depended on reversion of eye-color mutant strains, but the use of cloned rosy and white genomic DNA within the vector plasmid allowed for much greater efficiency and reliability. The first non-drosophilid transformations in medfly (Loukeris et al, 1995a; Handler et al, 1998, Michel et al, 2001) similarly relied on use of the wild type medfly white gene cDNA that was placed under Drosophila hsp70 regulation (Zwiebel et al., 1995). This gene complemented a mutant allele in a white eye medfly host strain that was isolated more than 20 years earlier. The medfly white gene also complemented the orthologous gene mutation in the oriental fruit fly, yielding, in one line, a nearly complete reversion (Handler and McCombs, 2000). The first transformations of Ae. aegypti used a kynurenine hydroxylase-, white mutant host strain, but for these tests the complementing marker was the wild type form of the D. melanogaster cinnabar gene (Cornel et al., 1997). The D. melanogaster vermilion gene, which encodes tryptophan oxygensase (to), has also been used to complement the orthologous green eye-color mutation in Musca domestica (White et al., 1996), and the Anopheles gambiae tryptophan oxygensase gene complements vermilion in Drosophila (Besansky et al., 1997). The vermilion and cinnabar ortho-logs have also been cloned from Tribolium, and while the white mutation in this species is complemented by tryp-tophan oxygensase, no pre-existing eye-color mutation is complemented by kynurenine hydroxylase (Lorenzen et al., 2002). The use of eye-color mutant-rescue systems has certainly been critical to initial advances in insect transformation, and these markers should have continued utility for those species that have been successfully tested. The use of these markers, however, for development of insect transformation in other species will be limited by the availability of suitable mutant host strains.

Chemical Selections

Previous to the development of mutant-rescue marker systems, transformant selections in non-drosophilid insects focused on genes that could confer resistance to particular chemicals or drugs. Importantly, these types of selections could be used for screening transformants en masse by providing the selectable chemical or drug in culture media. Ideally, only transformed individuals would survive the selection, allowing rapid screening of large numbers of G1 insects. For vectors that are inefficient and insects that are difficult to rear, the efficient screening of populations can be essential to identifying transformant individuals. The first drug resistance selection tested used the bacterial neomycin phosphotransferase gene (NPT II or neomycin7), which conferred resistance by inactiva-tion of the neomycin analog G418 (or Geneticin) (Steller and Pirrotta, 1985). This seemed straightforward, since the selection and hsneo marker system (putting NPT II under heat shock regulation) had already been developed and tested in Drosophila for mass transformant screens, and the bacterial resistance gene was thought to be functional in most eukaryotes. The initial P transformations in Drosophila using the pUChsneo vector were generally reliable; however, the marker was not easily transferable to other species. G418 resistance was highly variable, most likely due to species differences in diet, physiology, and symbiotic bacteria, and indeed, variation in resistance in transformed Drosophila has been attributed to strains of yeast used in culture media (see Ashburner, 1989a). Other chemical resistance markers, including organophos-phorus dehydrogenase (opd), conferring resistance to para-oxan (Phillips et al, 1990; Benedict et al., 1995), and the gene for dieldrin resistance (Rdl) (ffrench-Constant et al., 1991), which were initially tested in Drosophila were also problematic when tested in other species. These failures were due in large part to ineffective vector systems, but a common attribute in these studies was the selection of individuals having non-vector related or natural resistance to the respective chemical. While naturally resistant insects could be distinguished from transformed insects by molecular genotype tests, the recurrence of resistant insects in subsequent generations would make use of the transgenic strains highly impractical.

While the problems cited made chemical resistance selections frustrating for several species, and they have not been used for any recent transformation experiments, some successes were reported and the need for mass screening still exists. The initial tests for P transformation in several mosquito species used the pUChsneo vector with G418-resistant transformants being selected, though transformation frequencies were low and all of them resulted from fortuitous recombination events and not P-mediated transposition (Miller et al., 1987; McGrane et al., 1988; Morris et al., 1989). Nonetheless, chemical selections can be very powerful, and, if reliable, they would dramatically improve the efficiency of transformation screens for most insects. It is quite possible that many species will not be amenable to current transformation techniques without markers that allow selection en masse. A potential means of increasing the reliability of chemical resistance screens would be to link a resistance marker to a visible marker within the vector. Initial G1 transformants could be screened en masse for chemical resistance, with surviving individuals verified as transformants and maintained in culture using the visible marker. We have begun to test this type of marking by linking the hsneo construct with a red fluorescent protein marker in the piggyBac vector. Thus far, initial results in Drosophila are highly encouraging (A. Handler and R. Harrell, unpubished).

Of the enzyme systems tested for chemical selection in Drosophila that might be extended to other insects, the alcohol dehydrogenase (Adh) system might have the most promise (Goldberg et al., 1983). An Adh marker gene can complement the adh mutation in Drosophila, eliminating lethal sensitivity to ethanol treatment in mutant hosts. An adh gene has been cloned from the medfly, and a strategy has been developed to use it for genetic sex-ing by male-specific overexpression (Christophides et al., 2001). Conceivably, a similar strategy could be extended to transformant selections, though its use would be limited to medfly and possibly other tephritid species.

Fluorescent Protein Markers

The dramatic advancement of insect transformation in recent years has been due primarily to the development of fluorescent protein markers which are dominant-acting neomorphs that do not depend on pre-existing mutations. The first of these to be tested was the green fluorescent protein (GFP) gene that was isolated from the jellyfish Aequorea victoria (Prasher et al., 1992), and exhibited heterologous functionality in the nematode Caenorhabditis elegans (Chalfie et al., 1994). GFP expression was tested in transformed Drosophila, where it was successfully used as a reporter of gene expression (Plautz et al., 1996; Hazelrigg et al., 1998). GFP was first tested in non-drosophilid insects when GFP-marker Sindbis viruses were successfully used to infect the mosquito Ae. aegypti (Higgs et al., 1996). The dramatic somatic expression of GFP in adults was highly encouraging for the further use of GFP for germ-line transformants.

The use of GFP to detect germ-line transformation events was first tested in Drosophila using a construct that placed a modified form of GFP, enhanced GFP (EGFP), with a nuclear localizing sequence, under the regulatory control of the promoter from polyubiquitin (Lee et al., 1988; Handler and Harrell, 1999; see also Davis et al., 1995). The creation and use of thepiggyBac vector pB[Dmw, PUbnlsEGFP] in D. melanogaster allowed for a direct comparison of EGFP expression as a transformation marker to that from the visible mini-white marker. The results from this experiment indicated that not only was the PUbnlsEGFP marker efficient and easily detectable under epifluorescense optics, but also many of the G 1 transformants that expressed GFP did not express a detectable level of white+ (A. Handler and R. Harrell, unpubished). Although the biological basis of this observation is not known, this result provided encouraging evidence for the use of GFP as a marker in non-droso-philids. Several subsequent transformation experiments using EGFP regulated by a variety of promoters in piggy-Bac, Hermes, and Minos vectors confirmed these expectations. Notably, fluorescent protein marker genes allowed germ-line transformation to be tested in several species that otherwise had no visible marker systems, such as the Caribbean fruit fly Anastrepha suspensa, which was transformed with pB[PUbnlsEGFP] (Handler and Harrell, 2000). This vector was subsequently tested in Lucilia cuprina (Heinrich et al., 2002) and An. albimanus (Per-era et al., 2002). Similarly, a Hermes vector marked with EGFP regulated by the Drosophila actin5Cpromoter was first tested in Drosophila (Pinkerton et al., 2000), and was then used to efficiently select transformants in Ae. aegypti (Pinkerton et al., 2000), Stomoxys calcitrans (O’Brochta et al., 2000), and Culex quinquefasciatus (Allen et al., 2001). A Minos vector marked with actin5C-EGFP was used to select An. stephensi transformants (Catteruccia et al. , 2000), and a piggyBac vector marked with EGFP under Bombyx actin 3A promoter regulation was used to transform the lepidopteran species Bombyx mori (Tamura et al., 2000) and Pectinophora gossypiella (Peloquin et al., 2000).

Both the polyubiquitin and actin promoters have activity in all tissues throughout development, making insects marked in this fashion particularly useful for some applications, such as the marking of insects used in biocon-trol release programs (see Handler, 2002a). However, the detection of these markers can occasionally be difficult due to quenching or masking of fluorescence by melanized cuticle or scales. Fluorescent protein expression regulated by strong tissue-specific promoters has proven particularly valuable. Foremost among this type of marker is a series of fluorescent protein open reading frames under the regulatory control of the artificial promoter 3xP3, derived from the Drosophila eyeless gene (Sheng et al., 1997; Horn et al., 2000). Fluorescent protein expressed using 3xP3 is found primarily in the larval nervous system, and the eyes and ocelli of adults. piggyBac, Hermes, and Mos1 vectors containing 3xP3-EGFP were first used to transform D. melanogaster and T. castaneum (piggyBac and Hermes) (Berghammer et al., 1999), and have been widely used in the creation of many species of transgenic insects (see Table 1). The particular strengths and weaknesses for a marker construct such as 3xP3-EGFP are evident from experiments, where it enabled the selection of transgenic Bombyx embryos prior to larval hatching (Thomas et al., 2002), while it is almost undetectable in Ae. aegypti adults having normal eye pigmentation (Kokoza et al., 2001). It must therefore be recognized that the utility of fluorescent protein markers must be considered in the context of the host insect’s structure and physiology during development.

Table 1 Transposon Vectors and Markers Currently used for the Germ-Line Transformation of Various Insect Species

Table 1 Transposon Vectors and Markers Currently used for the Germ-Line Transformation of Various Insect Species

Fluorescent protein genetic markers tend to be more sensitive indicators of genetic transformation than eye-color markers; however, they are subject to qualitative and quantitative variation in their expression. Tissue-specific variation in transgene expression is likely due to local chromatin structure impacting access of promoters to essential transcription factors, while expression of trans-genes in unexpected cells and tissue is likely due to the influence of local enhancer. For example, polyubiquitin-regulated EGFP expression is most intense in the thoracic flight muscles of adult D. melanogaster and tephritid fruit flies. In adult transgenic Caribbean fruit flies containing PUb-EGFP, EGFP was only observed in the thorax, and spectrofluorometric assays revealed as much as five-fold differences in fluorescence among lines with equal copy numbers of transgenes (Handler and Harrell, 2000). In contrast to typical thoracic expression in tephritid flies, PUb-EGFP expression in adult transgenic L. cuprina was limited to female ovaries (Heinrich et al., 2002). PUb-DsRed expression in one transgenic medfly line was most intense in tarsi, while in another it was most intense at the tracheal apertures at the dorsal/ventral midline of the abdomen (A. Handler and R. Krasteva, unpublished). In T. castaneum, 3xP3-EGFP expression is typically in the optic lobes and brain, though several lines have shown atypical muscle-specific expression throughout development (Lorenzen et al, 2003). In various transgenic lines of An. stephensi, the 3xP3-EGFP marker has shown atypical expression in the pylorus and epidermal cells, and in a subset of cells in the rectum (D. O’Brochta, W. Kim, and H. Koo, unpublished).

The use of GFP will certainly continue to be a useful and popular insect transformation marker, but there is also a need for a variety of distinguishable fluorescent protein markers to permit the detection of multiple independent transgenes, and, when used in concert, for conditional gene expression systems and gene discovery methods, such as enhancer traps (Bellen et al., 1989; Wilson et al., 1989; Brand et al., 1994). After testing 3xP3-EGFP, the 3xP3 promoter was linked to the GFP red-shifted variants that emit blue (BFP), cyan (CFP), and yellow (YFP) fluorescence, which were tested in Drosophila, and have proven useful individually as reporters and for identifying transformants (Horn and Wimmer, 2000). BFP and GFP have distinct enough emission spectra to be used together, though BFP photobleaches quickly and is not useful for many applications. While use of EGFP with ECFP is also problematic, ECFP and EYFP can be distinguished when using appropriate filter sets. For details on appropriate filter sets for particular applications, see Horn et al. (2002), and the website for Chroma Technology Corp. (Brattle-boro, VT; www.chroma.com), which manufactures filters for most of the stereozoom fluorescence microscopes used for insect studies.

The most spectrally distinct fluorescent protein from GFP and its variants is a red fluorescent protein (RFP), known as DsRed, isolated from the Indo-Pacific sea coral Discosoma striata (Matz et al. 1999). It was first tested in insects by linking it to the polyubiquitin promoter in a pig-gyBac vector (pB[PUb-DsRed1]) and tested in Drosophila, where it exhibited highly intense expression (Handler and Harrell, 2001). Importantly, DsRed expression was completely distinguishable from EGFP when the two transgenic lines were interbred, and when co-expressed as an hsp70-Gal4/UAS-DsRed reporter in lines having vectors marked with EGFP. DsRed and its variants have since been incorporated into several mosquito and fruit fly species (Nolan et al., 2002; A. Handler, unpublished). Some of the original RFP variants include those found in a mutagenesis screen for rapid maturation and increased solubility (DsRed.T1/T3/T4), though their relative brightness is less intense than the wild type form (Bevis and Glick, 2002). The DsRed.T4 variant is available as DsRed-Express (Clontech), along with further variant forms which include monomeric RFPs that are preferred for fusion protein labeling (Strack et al., 2008). Both EGFP and DsRed are highly stable and generally resistant to photobleaching, and could be detected in teph-ritid flies several weeks after death, though DsRed and its variants are the relatively more stable of the two. Notably, PUb-DsRed.T3-marked transformant Caribbean fruit flies were unambiguously distinguished from unmarked wild type flies after being kept in liquid traps (torula yeast borax and propylene glycol) in field conditions for up to 3 weeks (Nirmala et al., 2010). This is highly advantageous for the use of these genes as markers for released insects that might only be retrieved several weeks after death in traps. A drawback for fluorescent proteins, and DsRed in particular, is that they require oligomerization and slow maturation that can take up to 48 hours, resulting in low intensity in early development. However, variants of DsRed with shorter maturation times (Campbell et al., 2002), and new fluorescent proteins with enhanced properties for specific applications, are becoming available on a consistent basis (see Matz et al., 2002; Chudakov et al., 2010).

In addition to providing new markers that are more easily identifiable, additional distinguishable markers will be invaluable to new methods of vector manipulation (see section 4.5.6). These currently include vectors for post-integration stabilization of transposon vectors, requiring either two or three markers (Handler et al., 2004; Dafa’alla et al., 2006), and repeatable targeting of genomic insertion sites requiring independent markers for each transgene insertion (Horn and Handler, 2005; Nimmo et al., 2006; Schetelig et al., 2010). These new and variant fluorescent proteins, many of which are available from the Clontech Living Colors® collection, include proteins isolated from various reef corals and sea ane-nomes, and rapidly maturing monomeric forms of the previously discovered fluorescent proteins. The AmCyan and ZsGreen FP markers were used for the first time in a study that showed that the D. melanogaster scs/scs’ and gypsy insulators, and the chicken fi-globin HS4 insulator, are effective means to minimize genomic position effect suppression of transgene expression in piggyBac vectors (Sarkar et al., 2006). ZsGreen, isolated from an Anthozoa reef coral, has since been placed under L. cuprina hsp83 promoter regulation to more efficiently select L. cuprina transformants compared to use of PUb-nls-EGFP (Concha et al., 2010). This argues for enhanced transformant selection with new fluorescent proteins under conspecific promoter regulation, which has been further supported by use of the same marker to identify the first L. sericata transformants (Concha et al., 2010).