piggybac-mediated germline transformation of the malaria mosquito Anopheles stephensi using the red fluorescent protein

Transcription

1 JBC Papers in Press. Published on January 22, 2002 as Manuscript C piggybac-mediated germline transformation of the malaria mosquito Anopheles stephensi using the red fluorescent protein dsred as a selectable marker. Tony Nolan, Tom M. Bower, Anthony E. Brown, Andrea Crisanti and Flaminia Catteruccia From the Department of Biological Sciences, SAF Building, Imperial College of Science, Technology and Medicine, Imperial College Road, London SW7 2AZ, UK. To whom correspondence should be addressed: Dept. of Biological Sciences, SAF Building, Imperial College of Science, Technology and Medicine, Imperial College Road, London SW7 2AZ, UK. Tel.: ; Fax: ; f.catteruccia@ic.ac.uk SUMMARY It is estimated that every year malaria infects approximately 300 million people and accounts for the death of 2 million individuals. The Plasmodium parasites that cause malaria in humans are transmitted exclusively by mosquito species belonging to the Anopheles genus. The recent development of a gene transfer technology for Anopheles stephensi mosquitoes, using the Minos transposable element marked with the enhanced green fluorescent protein EGFP(1), provides now a powerful tool to investigate the role of mosquito molecules involved in the interaction with the malaria parasite. Such technology, when further developed with additional markers and transposable elements, will be invaluable for 1 Copyright 2002 by The American Society for Biochemistry and Molecular Biology, Inc.

2 analysing the biology of the vector and for developing malaria resistant mosquitoes to be used as a tool to control malaria transmission in the field. We report here the germline transformation of A. stephensi mosquitoes using a piggybac-based transposon(2) to drive integration of the gene encoding for the red fluorescent protein dsred(3). A. stephensi embryos were injected with transformation vector ppbred, containing the dsred marker cloned within the arms of piggybac. Microscopic analysis of G 1 larvae revealed the presence of 7 fluorescent phenotypes, whose different molecular origins were confirmed by Southern blotting analysis. Sequencing of the insertion sites in two lines demonstrated that integrations had occurred at TTAA nucleotides, in accordance with piggybac-mediated transpositions. INTRODUCTION The recent development of an efficient gene transfer technology for A. stephensi mosquitoes, achieved by using a Minos-based transposon(4) loaded with the EGFP selectable marker(1), has expanded the possibility of studying the genetics of human malaria vectors at the functional level. Although EGFP has proven an invaluable visible marker for identifying transformed individuals in different insect species(1,5-7), the availability of only this selectable marker limits the range of applications of a gene transfer technology in malaria vectors. New molecular tools are now needed in order to exploit fully the potential of germline transformation for Anopheles mosquitoes, with the aim of unravelling the interactions between host molecules and Plasmodium parasites, as well as performing functional studies such as transposon tagging and enhancer trapping. 2

3 Ultimately, this technology could be utilised to develop transgenic mosquitoes with a non-permissive phenotype for parasite development. These mosquitoes could be used in malaria control programs with the aim to replace the permissive wild type vectors. As shown in Drosophila melanogaster, the availability of various molecular and genetic tools to achieve germline transformation has contributed tremendously to our understanding of the fruitfly biology, leading to the identification of hundreds of genes involved in development, immunity, tissue modelling and embryogenesis. The transposable elements P, Minos, piggybac and hermes have, among others, all been successfully utilised to drive integration of exogenous DNA into the fruitfly genome(8-11). At the same time, the employment of an array of visible selectable markers, such as white, rosy, cinnabar, and more recently EGFP(12-15), has greatly facilitated the screening procedure, and allowed complex functional studies of genes and their promoter regions. Germline transformation of A. stephensi mosquitoes is anticipated to open a wide range of applications to explore the genome of these important disease vectors, when sustained by a parallel development of appropriate molecular tools. So far, the availability of a system for germline transformation based on the combination of a single transposable element with a single selectable marker has limited the array of applications of gene transfer technology in Anopheles. In this study, we have achieved germline transformation of A. stephensi mosquitoes by using the piggybac transposable element, initially isolated in the cabbage looper Trichoplusia ni (2), to drive 3

4 integration of foreign genes. Moreover, we have validated the use of the red fluorescent protein dsred(3) as a visible selectable marker in germline transformation in the same mosquito species. EXPERIMENTAL PROCEDURES Plasmid construction The pmired transformation vector (fig 1a) was derived from plasmid pminot (11) by inserting a DNA fragment containing the dsred gene under the control of the actin5c promoter from D. melanogaster. The helper plasmids phss6hsilmi20 and phsp-pbac, respectively providing the Minos and the piggybac transposase genes, have previously been described (10,16). The ppbred transformation vector (fig 1a) was derived from plasmid pbkoα (17) by cloning a BglII linker, containing an internal NotI site, into the unique BglII site within the piggybac gene. A DNA cassette including the dsred gene under the control of the actin 5C promoter and additional sequences (stuffer) was then cloned as a single NotI fragment into the inserted linker. Embryo microinjection Blood-fed Anopheles stephensi mosquitoes (strain sd 500) were allowed to lay eggs in a solution of 0.1 mm p-nitrophenyl p -guanidinobenzoate (pnpgb) hours after a blood meal. Embryos were then transferred onto glass slides and microinjected in injection buffer essentially as described(1). Embryos were microinjected with a mixture of helper plasmid phss6hsilmi20 (100µgml -1 ) and pmired (400µgml -1 ) in the case of Minos-mediated integrations, and with 4

5 plasmid phsp-pbac (100µgml -1 ) and ppbred (400µgml -1 ) in the case of piggybac-mediated integrations. Hatched larvae were analysed on an inverted microscope using a Texas red filter to detect dsred expression. Southern blot analyses and sequencing of integration sites Genomic DNA from transgenic A. stephensi adults (G 2 generation) and from wildtype mosquitoes was digested with the restriction endonucleases HindII or EcoRI and blotted as described(18). Digested genomic DNA (~4 µg per lane) was separated on a 0.8% agarose gel and transferred onto a nylon membrane. The membranes were hybridised overnight at 65 0 C with different 32 P-labelled probes. In the case of the Minos-mediated integration, genomic DNA was hybridised with the previously described probe M (fig 1a), encompassing both the left and the right arms of Minos(1). In the case of piggybac-mediated integrations, probe P, a polymerase chain reaction (PCR) product encompassing sequences of the left and right arms of the piggybac transposon, was utilised. To sequence the piggybac integration sites, genomic Lambda Zap EcoRI libraries (Stratagene) were constructed from the DNA extracted from G 2 larvae of A. stephensi transgenic lines 10 2 and A 3, and hybridised with probe P (fig 1a). The cloned insertion sites were sequenced using primers annealing internally to the piggybac inverted repeats (ppb-r 5 -CGTACTTACTGTACTTACTGC-3, ppb-l 5 -CATCGCCTTGCAGAAGAGC-3 ). 5

6 RESULTS We tested whether the red fluorescent protein dsred could be utilised as a visible selectable marker for germline transformation of A. stephensi mosquitoes. In a pilot experiment, A. stephensi embryos were microinjected with vector pmired, which contained the dsred gene from Discomona sp. under the transcriptional control of the actin5c promoter from D. melanogaster inserted between the Minos inverted repeats (fig 1a). pmired was microinjected together with plasmid phss6ilmi20, which provided the enzymatic activity necessary for Minos transposition(16). The 23 adults obtained from 102 injected embryos were outcrossed to wild type A. stephensi mosquitoes, and their progeny was analysed for fluorescence using a Texas Red excitation filter (table I). Fluorescent G 1 larvae were detected in the progeny of G 0 female no 7. The segregation of the red allele in following generations was in agreement with the occurrence of a single insertion (data not shown). To assess the nature of the integration event, we performed Southern Blotting analysis of the transgenic population (line MinRED1) on genomic DNA extracted from G 2 larvae and digested with the EcoRI or HindII restriction endonucleases. Two bands could be detected after hybridisation with probe M, which encompassed both arms of Minos, demonstrating that a single insertion of the transposon had occurred (fig 1b)(1). Interestingly, the pattern of expression of fluorescence changed throughout larval development. In newly hatched larvae, a faint fluorescence could be detected in nerve cells around the larval body (fig 2a). The levels of fluorescence increased in second and third instar larvae, which displayed the gut 6

7 expression pattern previously described in transgenic A. stephensi transformed with an actin5c-egfp transposon (fig2b)(1). Importantly, fluorescence was more easily detected in the gut of adult mosquitoes than in the case of transgenic lines expressing the EGFP marker (fig 2c). We then tested the ability of the piggybac transposable element to drive integration of exogenous DNA into the germline of A. stephensi mosquitoes. A total of 303 embryos were injected with a mixture of transformation vector ppbred, containing the actin-dsred cassette inserted within the inverted repeats of piggybac, and plasmid phsp-pbac, providing the piggybac transposase gene (fig 1a)(10). The 76 adults surviving injections were divided into three groups of males and one group of females, and each group was outcrossed separately (table 1). G 0 females were group-mated and allowed to lay eggs in isolation, while G 0 males were tested collectively. A minimum of 3 G 0 adults (female no 10 and individuals from male groups A and C, see table I) produced fluorescent progeny, with a minimum integration frequency of 4%, comparable to that obtained using Minos to transform the same mosquito species(1). Due to the testing procedure employed for G 0 males, it was not possible to assess whether the fluorescent G 1 larvae in groups A and C were derived from one or more progenitors. Among the 34 G 0 females, 12 did not lay eggs after multiple feedings, accounting for a total estimated sterility of 35%. Microscopic analysis of G 1 larvae revealed the presence of as many as 7 different fluorescent phenotypes: three from female no 10 ( ), three from male group A (A 1-3 ) and one from male group C (C 1 ), which were reared separately 7

8 (table 1). To determine the nature of the integration events, Southern blotting analysis of the 7 putative transgenic populations was performed (fig 1c). Genomic DNA extracted from each population was digested with EcoRI restriction endonuclease and hybridised with probe P, spanning the two arms of piggybac (fig 1a). With this combination of probe and restriction endonucleases, every insertion of the transposon will be identified by the appearance of two hybridising bands. The Southern blot results confirmed the different molecular origins of the 7 phenotypes. Single insertions were detected in 3 of the 7 lines, while the remaining four showed the occurrence of multiple integrations, a phenomenon typical of piggybac insertions. Legitimate insertions of piggybac elements always occur at a TTAA nucleotide sequence, causing a precise duplication of the site(2). To verify that integrations had occurred according to a precise piggybac mechanism, the sequence of the junctions between the inserted element and the A. stephensi genome was determined for two of the transgenic lines that showed a single insertion of the transposon. Genomic DNA libraries from transgenic lines 10 2 and A 3 were constructed and screened with probe P. Positive clones were sequenced using primers annealing to the piggybac transposon, and in both cases, the insertion sites showed complete piggybac inverted repeats, a TTAA nucleotide and additional sequences of unknown origin, presumably derived from the A. stephensi genome. To confirm these findings, primers annealing to the putative genomic sequences flanking the piggybac inverted repeats were designed to amplify the two empty sites on 8

9 genomic DNA from wild type mosquitoes, and in both cases a band of the expected size was obtained by PCR reactions (not shown). DISCUSSION We have demonstrated the ability of the piggybac transposable element to drive integration of exogenous DNA in A. stephensi mosquitoes. The availability of a second functional transposable element in A. stephensi will introduce important improvements to the transgenic system recently developed in these mosquitoes by using Minos as a delivery vector(1). Minos-transformed mosquito lines can now be further manipulated by introducing additional genes with the use of piggybac transposons. Furthermore, remobilisation studies can be performed, to assess the capability of non-autonomous Minos and piggybac transposons to be remobilised in crosses with transgenic lines producing their specific transposase genes. Since the Minos and piggybac elements do not seem to cross-mobilise each other, they are well suited for such studies. Remobilisation experiments will be fundamental in understanding the feasibility of using autonomous transposable elements to spread malaria-refractory genes throughout a wild population. Our results in A. stephensi mosquitoes, which have been confirmed by the work performed in collaborating laboratories i, testify to the broad host range of piggybac, which makes it a very attractive transposon for germline transformation experiments. An EGFP-loaded piggybac transposon has also been recently validated as DNA delivery vector in A. gambiae mosquitoes, although at frequencies much lower than those obtained in this study(6). 9

10 However, when comparing piggybac to Minos, the fact that the latter element generally creates single integrations may represent an advantage in terms of studying the effects of introducing foreign DNA into the genome of heterologous organisms(1). The presence of piggybac-like transposons in different species seems to indicate the occurrence of horizontal transfer of this element(19). Horizontal transfer has also been postulated for Minos(20). More data are needed on these two transposons before population replacement schemes can be implemented. We have also validated the use of dsred as a selectable marker in germline transformation of A. stephensi mosquitoes. dsred, isolated from Discomona sp.(3), has been utilised as a reporter gene in Saccharomyces cerevisiae and in zebrafish, and very recently as a selectable marker in D. melanogaster(21-23). Previous Minos-mediated integrations in A. stephensi were achieved using the green fluorescent protein EGFP as a selectable marker(1). Interestingly, in A. stephensi mosquitoes, dsred showed slower kinetics of formation than EGFP, as demonstrated by the weak fluorescent phenotype observed in first instar larvae of all dsred transgenic lines. However, dsred is visible in the gut of adult A. stephensi individuals, a feature that would allow the easy identification of transgenic insects in recapture studies after release in the wild. Importantly, the emission profiles of the dsred and EGFP markers do not overlap in the progeny of mixed transgenic populations (data not shown). The availability of an additional selectable marker will extend the range of applications of germline transformation in Anopheles mosquitoes, allowing functional studies such as 10

11 transposon remobilization, enhancer trapping and transposon tagging, which are based on crossing of lines expressing different markers. It will also have an important impact on the manipulation of other insect species for which mutationbased visible markers are not available. In conclusion, the results achieved in this study are anticipated to have an essential impact upon the development of an array of genetic tools for studying the biology of Anopheles mosquitoes. ACKNOWLEDGEMENTS We would like to thank M. Scott and L.S. Alphey for providing us with plasmids pbkoα and phsp-pbac respectively. We thank M. Jacobs-Lorena for discussion of unpublished results. We thank E. Petris for stimulating discussions. F.C. and T.M.B. were supported by the Wellcome Trust, T.N. was supported by a Network grant of the Training and Mobility Program of the European Community. REFERENCES 1. Catteruccia, F., Nolan, T., Loukeris, T. G., Blass, C., Savakis, C., Kafatos, F. C., and Crisanti, A. (2000) Nature 405(6789), Fraser, M. J., Ciszczon, T., Elick, T., and Bauser, C. (1996) Insect Mol Biol 5(2), Matz, M. V., Fradkov, A. F., Labas, Y. A., Savitsky, A. P., Zaraisky, A. G., Markelov, M. L., and Lukyanov, S. A. (1999) Nat Biotechnol 17(10), Franz, G., Loukeris, T. G., Dialektaki, G., Thompson, C. R., and Savakis, C. (1994) PNAS USA 91(11),

12 5. Allen, M. L., O'Brochta, D. A., Atkinson, P. W., and Levesque, C. S. (2001) J Med Entomol 38(5), Grossman, G. L., Rafferty, C. S., Clayton, J. R. T. K., Stevens, T. K., Mukabayire, O., and Benedict, M. Q. (2001) Insect Molecular Biology 10(6), Pinkerton, A. C., Michel, K., O'Brochta, D. A., and Atkinson, P. W. (2000) Insect Mol Biol 9(1), Rubin, G. M., and Spradling, A. C. (1982) Science 218(4570), Sarkar, A., Coates, C. J., Whyard, S., Willhoeft, U., Atkinson, P. W., and O'Brochta, D. A. (1997) Genetica 99(1), Handler, A. M., and Harrell, R. A., 2nd. (1999) Insect Mol Biol 8(4), Loukeris, T. G., Livadaras, I., Arca', B., Zabalou, S., and Savakis, C. (1995b) Science 270, Warren, W. D., Palmer, S., and Howells, A. J. (1996) Genetica 98(3), Rubin, G. M., and Spradling, A. C. (1983) Nucleic Acids Res 11(18), Klemenz, R., Weber, U., and Gehring, W. J. (1987) Nucleic Acids Res 15(10), Zhang, G., Gurtu, V., and Kain, S. R. (1996) Biochem Biophys Res Commun 227(3),

13 16. Klinakis, A. G., Loukeris, T. G., Pavlopoulos, A., and Savakis, C. (2000) Insect Mol Biol 9(3), Thibault, S. T., Luu, H. T., Vann, N., and Miller, T. A. (1999) Insect Mol Biol 8(1), Sambrook, Fritsch, and Maniatis. (1989) Molecular Cloning. A laboratory manual, Second edition Ed. (Press, C. S. H. L., Ed.) 19. Handler, A. M., and McCombs, S. D. (2000) Insect Mol Biol 9(6), Arca, B., and Savakis, C. (2000) Genetica 108(3), Rodrigues, F., van Hemert, M., Steensma, H. Y., Corte-Real, M., and Leao, C. (2001) J Bacteriol 183(12), Finley, K. R., Davidson, A. E., and Ekker, S. C. (2001) Biotechniques 31(1), 66-70, Handler, A. M., and Harrell, R. A., 2nd. (2001) Biotechniques 31(4), 820, i Crisanti and Jacobs-Lorena, manuscript in preparation FIGURE LEGENDS Figure 1 The Minos and piggybac transposable elements integrate into the A. stephensi genome. a, Maps of transformation vectors pminred and ppbred. Actin5C, D. melanogaster Actin5C promoter; HspT, D. melanogaster Hsp70 terminator sequence. The dsred gene is indicate by a black arrow. ML; Minos left arm; MR, Minos right arm; pbacl, piggybac left arm; pbacr, piggybac right arm. H, HincII; E, EcoRI. Black bars represent the probes (M and P) used in the 13

14 Southern Blot analyses. b, Southern Blot analysis of genomic DNA from transgenic line MinRED 1, digested with either HincII or EcoRI and hybridised with probe M (fig 1a). c, Southern Blot analysis of genomic DNA from piggybactransgenic lines , A 1-3 and C 1, or from control wild type mosquitoes (WT) digested with EcoRI and hybridised with probe P (fig 1a). Figure 2 A. stephensi mosquitoes expressing dsred. a and b, confocal fluorescence microphotographs of transgenic first and third instar larvae, showing differential dsred expression. c, adult female mosquito showing strong dsred expression in the gut. 14

15 Table I Outcome of injection experiments Injection experiments were performed on A. stephensi embryos using two different transformation plasmids. The total numbers of injected embryos, hatching larvae and surviving adults are shown. The percentages of embryos that hatched to larvae and survived to adults are indicated in brackets. Adults were outcrossed to wild type (wt) mosquitoes. Where asterisks are present, this indicates that more than one phenotype was isolated in the G 1 progeny. Plasmid Embryos injected Larvae Adults wt outcross Group/ founder Fluorescent/ total G1s pmired (31%) 23 (22%) 13 males 0/7, females no 7 3/1,561 ppbred (39%) 76 (25%) 15 males A 19/2,340 *** 20 males B 0/4,793 7 males C 25/214 * 34 females no 10 38/2,797 *** 14

16 Table 2 Sequence of integration sites Putative piggybac insertions were isolated and sequenced after using probe P to screen Lambda Zap genomic libraries from transgenic families 10 2 and A 3. Both sides (left and right) of the insertions were sequenced, confirming that duplication of the TTAA nucleotide had occurred. The piggybac left and right arms are shown in lower case, the TTAA nucleotide sequence of the insertion site is shown in bold, while the putative A. stephensi genomic sequences are shown in capitals. Family Flanking sequence piggybac arm A 3 A 3 CTCCTCTCTTAAACACCTTAAccctagaaagat atctttctagggttaaagcagtggagtggagg CCGCGCATCGATCGCATTAAccctagaaagat atctttctagggttaaggagggcgatctcgtga left right left right 15

17

18