A Novel Secreted Protein Toxin from the Insect Pathogenic Bacterium Xenorhabdus nematophila
2004; Elsevier BV; Volume: 279; Issue: 15 Linguagem: Inglês
10.1074/jbc.m309859200
ISSN1083-351X
AutoresSusan E. Brown, Anh T. Cao, Eric R. Hines, Raymond J. Akhurst, Peter D. East,
Tópico(s)Insect and Pesticide Research
ResumoThe bacterium Xenorhabdus nematophila is an insect pathogen that produces several proteins that enable it to kill insects. Screening of a cosmid library constructed from X. nematophila strain A24 identified a gene that encoded a novel protein that was toxic to insects. The 42-kDa protein encoded by the toxin gene was expressed and purified from a recombinant system, and was shown to kill the larvae of insects such as Galleria mellonella and Helicoverpa armigera when injected at doses of around 30–40 ng/g larvae. Sequencing and bioinformatic analysis suggested that the toxin was a novel protein, and that it was likely to be part of a genomic island involved in pathogenicity. When the native bacteria were grown under laboratory conditions, a soluble form of the 42-kDa toxin was secreted only by bacteria in the phase II state. Preliminary histological analysis of larvae injected with recombinant protein suggested that the toxin primarily acted on the midgut of the insect. Finally, some of the common strategies used by the bacterial pathogens of insects, animals, and plants are discussed. The bacterium Xenorhabdus nematophila is an insect pathogen that produces several proteins that enable it to kill insects. Screening of a cosmid library constructed from X. nematophila strain A24 identified a gene that encoded a novel protein that was toxic to insects. The 42-kDa protein encoded by the toxin gene was expressed and purified from a recombinant system, and was shown to kill the larvae of insects such as Galleria mellonella and Helicoverpa armigera when injected at doses of around 30–40 ng/g larvae. Sequencing and bioinformatic analysis suggested that the toxin was a novel protein, and that it was likely to be part of a genomic island involved in pathogenicity. When the native bacteria were grown under laboratory conditions, a soluble form of the 42-kDa toxin was secreted only by bacteria in the phase II state. Preliminary histological analysis of larvae injected with recombinant protein suggested that the toxin primarily acted on the midgut of the insect. Finally, some of the common strategies used by the bacterial pathogens of insects, animals, and plants are discussed. The availability of more than 100 microbial genomes, including the benign and pathogenic forms of closely related bacteria, is leading to an improved understanding of the common mechanisms used by pathogenic bacteria to invade and survive in hosts. The current evidence suggests that bacterial pathogens of various hosts share many genes that aid virulence and survival (1Keen N. Staskawicz B. Mekalanos J. Ausubel F. Cook R.J. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8752-8753Crossref PubMed Scopus (10) Google Scholar, 2Cao H. Baldini R.L. Rahme L.G. Annu. Rev. Phytopathol. 2001; 39: 259-284Crossref PubMed Scopus (107) Google Scholar). The products of such genes are often called virulence factors and include factors required for host and tissue tropism, cytotoxicity and multiplication within the host (3Shea J.E. Santangelo J.D. Feldman R.G. Curr. Opin. Microbiol. 2000; 3: 451-458Crossref PubMed Scopus (69) Google Scholar). These virulence genes are often present on genomic islands that have been shared among pathogens by horizontal gene transfer (4Hacker J. Kaper J.B. Annu. Rev. Microbiol. 2000; 54: 641-679Crossref PubMed Scopus (938) Google Scholar). The complete genomes available to date have a bias toward vertebrate pathogens, with only a few available for plant pathogens, and one for an invertebrate pathogen, the recently released Photorhabdus luminescens strain TT01 genome (5Duchaud E. Rusniok C. Frangeul L. Buchrieser C. Givaudan A. Taourit S. Bocs S. Boursaux-Eude C. Chandler M. Charles J.F. Dassa E. Derose R. Derzelle S. Freyssinet G. Gaudriault S. Medigue C. Lanois A. Powell K. Siguier P. Vincent R. Wingate V. Zouine M. Glaser P. Boemare N. Danchin A. Kunst F. Nat. Biotechnol. 2003; 21: 1307-1313Crossref PubMed Scopus (472) Google Scholar). The study of insect pathogens such as Xenorhabdus nematophila, P. luminescens, or Bacillus thuringiensis may provide insight into pathogen evolution and may also lead to the development of new insecticides (6Hilder V.A. Boulter D. Crop Prot. 1999; 18: 177-191Crossref Scopus (213) Google Scholar). Xenorhabdus and Photorhabdus are Gram-negative bacteria that live in symbiosis with nematodes (7Forst S. Dowds B. Boemare N. Stackebrandt E. Annu. Rev. Microbiol. 1997; 51: 47-72Crossref PubMed Scopus (492) Google Scholar, 8Burnell A.M. Stock S.P. Nematology. 2000; 2: 31-42Crossref Scopus (165) Google Scholar, 9Akhurst R.J. Dunphy G.B. Beckage N.E. Thompson S.N. Federici B.A. Parasites and Pathogens of Insects. Academic Press, Inc., San Diego, CA1993: 1-23Crossref Scopus (115) Google Scholar). This bacteria-nematode association is highly toxic to many insect species, and in most cases the bacteria alone are highly virulent once they are circulating in the hemocoel of the insect (10Forst S. Nealson K. Microbiol. Rev. 1996; 60: 21-43Crossref PubMed Google Scholar). The bacteria and nematode share a complex life cycle, which includes symbiotic and pathogenic stages. During the symbiotic stage, the bacteria are carried in the gut of the nematode, but after infection of an insect host, the nematodes inject the bacteria into the insect hemocoel. Over several days, the combined actions of the nematode and bacteria kill the insect. Within the hemocoel of the insect carcass, the bacteria grow to stationary phase while the nematodes develop and sexually reproduce. The final stage of development is the re-association of the bacteria and nematodes to form non-feeding infective juveniles, which emerge from the insect carcass to find new hosts. The naturally occurring bacterial symbionts found in the gut of the nematodes are called phase I cells. Variant forms, called phase II, are rarely observed in vivo but are often observed under laboratory conditions (7Forst S. Dowds B. Boemare N. Stackebrandt E. Annu. Rev. Microbiol. 1997; 51: 47-72Crossref PubMed Scopus (492) Google Scholar, 9Akhurst R.J. Dunphy G.B. Beckage N.E. Thompson S.N. Federici B.A. Parasites and Pathogens of Insects. Academic Press, Inc., San Diego, CA1993: 1-23Crossref Scopus (115) Google Scholar, 10Forst S. Nealson K. Microbiol. Rev. 1996; 60: 21-43Crossref PubMed Google Scholar, 11Smigielski A.J. Akhurst R.J. J. Invertebr. Pathol. 1994; 64: 214-220Crossref Scopus (7) Google Scholar) and also in the free living clinical isolates of Photorhabdus (12Peel M.M. Alfredson D.A. Gerrard J.G. Davis J.M. Robson J.M. Mcdougall R.J. Scullie B.L. Akhurst R.J. J. Clin. Microbiol. 1999; 37: 3647-3653Crossref PubMed Google Scholar). The phase II bacteria are still toxic to insects such as Galleria mellonella, despite having many altered properties, including motility and lipase, phospholipase, and protease activities (7Forst S. Dowds B. Boemare N. Stackebrandt E. Annu. Rev. Microbiol. 1997; 51: 47-72Crossref PubMed Scopus (492) Google Scholar, 9Akhurst R.J. Dunphy G.B. Beckage N.E. Thompson S.N. Federici B.A. Parasites and Pathogens of Insects. Academic Press, Inc., San Diego, CA1993: 1-23Crossref Scopus (115) Google Scholar, 10Forst S. Nealson K. Microbiol. Rev. 1996; 60: 21-43Crossref PubMed Google Scholar, 11Smigielski A.J. Akhurst R.J. J. Invertebr. Pathol. 1994; 64: 214-220Crossref Scopus (7) Google Scholar). Throughout their life cycle, the bacteria and the nematodes produce a variety of metabolites to enable them to colonize and reproduce in the insect host. These metabolites often have overlapping functions, a strategy that is likely to contribute to the success of the nematode-bacteria association against a variety of insect hosts (9Akhurst R.J. Dunphy G.B. Beckage N.E. Thompson S.N. Federici B.A. Parasites and Pathogens of Insects. Academic Press, Inc., San Diego, CA1993: 1-23Crossref Scopus (115) Google Scholar, 13Collmer A. Lindeberg M. Petnicki-Ocwieja T. Schneider D.J. Alfano J.R. Trends Microbiol. 2002; 10: 462-469Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar). The metabolites produced include molecules to help evade the insect immune system, enzymes such as proteases, lipases, and phospholipases to maintain a food supply during reproduction (7Forst S. Dowds B. Boemare N. Stackebrandt E. Annu. Rev. Microbiol. 1997; 51: 47-72Crossref PubMed Scopus (492) Google Scholar, 9Akhurst R.J. Dunphy G.B. Beckage N.E. Thompson S.N. Federici B.A. Parasites and Pathogens of Insects. Academic Press, Inc., San Diego, CA1993: 1-23Crossref Scopus (115) Google Scholar, 14Bowen D. Blackburn M. Rocheleau T. Grutzmacher C. Ffrench-Constant R.H. Insect Biochem. Mol. Biol. 2000; 30: 69-74Crossref PubMed Scopus (35) Google Scholar, 15Thaler J.O. Duvic B. Givaudan A. Boemare N. Appl. Environ. Microbiol. 1998; 64: 2367-2373Crossref PubMed Google Scholar), and antifungal and antibacterial agents to prevent degradation or colonization of the insect carcass while the bacteria and nematodes reproduce (9Akhurst R.J. Dunphy G.B. Beckage N.E. Thompson S.N. Federici B.A. Parasites and Pathogens of Insects. Academic Press, Inc., San Diego, CA1993: 1-23Crossref Scopus (115) Google Scholar, 16Akhurst R. Kaya H. Bedding R. Akhurst R. Nematodes and the Biological Control of Insect Pests. CSIRO Australia, East Melbourne, Australia1993: 127-135Google Scholar). The bacteria and nematodes also produce toxins that are responsible for killing the insect host. Analysis of the genome of P. luminescens identified more predicted toxin genes than any other bacteria sequenced to date, including potential gene products with homology to hemolysin A, chitinase, Rtx (repeats-in-toxin)-like toxin, and δ-endotoxin (5Duchaud E. Rusniok C. Frangeul L. Buchrieser C. Givaudan A. Taourit S. Bocs S. Boursaux-Eude C. Chandler M. Charles J.F. Dassa E. Derose R. Derzelle S. Freyssinet G. Gaudriault S. Medigue C. Lanois A. Powell K. Siguier P. Vincent R. Wingate V. Zouine M. Glaser P. Boemare N. Danchin A. Kunst F. Nat. Biotechnol. 2003; 21: 1307-1313Crossref PubMed Scopus (472) Google Scholar). The only toxins studied in detail are the Tc toxins from P. luminescens strain W14 (17Guo L.N. Fatig R.O. Orr G.L. Schafer B.W. Strickland J.A. Sukhapinda K. Woodsworth A.T. Petell J.K. J. Biol. Chem. 1999; 274: 9836-9842Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar, 18Bowen D.J. Ensign J.C. Appl. Environ. Microbiol. 1998; 64: 3029-3035Crossref PubMed Google Scholar, 19Bowen D. Rocheleau T.A. Blackburn M. Andreev O. Golubeva E. Bhartia R. Ffrench-Constant R.H. Science. 1998; 280: 2129-2132Crossref PubMed Scopus (347) Google Scholar), although a small amount of work has also been done on a 39-kDa toxin from X. nematophila (20Ryu K.G. Bae J.S. Yu Y.S. Park S.H. Biotechnol. Bioprocess Eng. 2000; 5: 141-145Crossref Scopus (9) Google Scholar), the large Xin toxin from X. nematophila strain BJ (21Pan Y.H. Jian H. Zhang J. Liu Z. Chen Z.Y. Yang X.F. Yang H.W. Huang D.F. Prog. Nat. Sci. 2002; 12: 310-312Google Scholar), and the PhlA hemolysin from P. luminescens strain TT01 (22Brillard J. Duchaud E. Boemare N. Kunst F. Givaudan A. J. Bacteriol. 2002; 184: 3871-3878Crossref PubMed Scopus (69) Google Scholar). This paper describes the identification, purification, and characterization of a novel insect toxin from X. nematophila strain A24. Isolation of Toxin Gene—High molecular weight genomic DNA was isolated from X. nematophila strain A24 (23Marmur J. J. Mol. Biol. 1961; 3: 208-218Crossref Scopus (8945) Google Scholar), partially digested with Sau3AI to generate fragments in the size range 30–50 kb, dephosphorylated, and ligated to linearized cosmid cloning vector "Supercos" (Stratagene) (24Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1982Google Scholar). The ligated DNA was packaged in vitro using Gigapack II XL Packaging Extract (Stratagene), transfected into the Escherichia coli strain NM554 (Stratagene), and screened by G. mellonella injection bioassay as described below. One clone, designated cos149, was chosen for further analysis. Deletion analysis of this clone (25Henikoff S. Gene. 1984; 28: 351-359Crossref PubMed Scopus (2828) Google Scholar) was performed using the Erase-a-Base kit (Promega) and the enzymes BamHI, ClaI, SphI, SacI, and HindIII (Roche Applied Science). Digested DNA was ligated into pGEM7Z(f)+ (Promega) and transformed by electroporation into E. coli strain DH5α, and the lysates were screened for insecticidal activity using the G. mellonella bioassay. The smallest of the deletion clones that retained insecticidal activity was sequenced on both strands using a combination of vector and gene-specific primers. Similarly, the DNA immediately surrounding the toxin gene and at the 5′ and 3′ ends of the cosmid clones and the intermediate deletion clones was sequenced using a combination of vector and sequence-specific primers. DNA sequencing was performed on an automated sequencer (Applied Biosystems model 377) using ABI Prism™ di-deoxy dye-terminator sequencing mix (Applied Biosystems). DNA was isolated from strains using a standard alkaline lysis procedure and analyzed by agarose gel electrophoresis (24Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1982Google Scholar). Purification of Recombinant Toxin Protein—All protein purification steps were carried out at 4 °C unless otherwise stated. Expression of the full-length protein from X. nematophila strain A24 was performed in the IMPACT system (New England Biolabs), in which the toxin open reading frame was cloned at the 5′ end of a self-splicing intein coding sequence fused to a short DNA sequence encoding a chitin binding domain. Recombinant plasmid containing the X. nematophila strain A24 toxin gene was prepared in the IMPACT vector pCYB3 and transformed into the E. coli strains DH10B or BL21(DE3) by electroporation. Cultures (500 ml) of LB 1The abbreviations used are: LB, Luria-Bertani; BSA, bovine serum albumin; MBP, maltose-binding protein. broth containing 100 μg/ml ampicillin were grown at 37 °C until the A600 reached 0.5–0.6. Protein production was induced with 1 mm isopropyl-β-thiogalactoside, and the cells were grown overnight at 14 °C. The preparation of bacterial cell extracts, affinity isolation of the fusion proteins on chitin columns, on-column dithiothreitol-mediated cleavage of the fusion proteins, and elution of the purified toxin proteins were all performed according to the manufacturer's instructions, except that 1 mm phenylmethylsulfonyl fluoride was added to the lysis and cleavage buffers. E. coli maltose-binding protein (MBP) was prepared from the IMPACT™ system using the same methods and was used as a negative control for the insect bioassays. Purified proteins were exchanged into phosphate-buffered saline, pH 7.4, and concentrated using Ultrafree spin cartridges (Millipore) with a nominal molecular mass cut-off of 5 or 10 kDa. Protein purity was checked by SDS-PAGE, using either the buffer system of Laemmli (26Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (205531) Google Scholar) with a Bio-Rad Mini-Protean apparatus, or the NuPAGE Novex system (Invitrogen). All gels were stained with Fast Stain (Zoion Research). Protein concentrations were determined by the Bradford assay using the Bio-Rad Protein Assay reagent and BSA (Sigma) as a standard. The identity of the protein was confirmed by mass spectroscopy on a Voyager Elite matrix-assisted laser desorption ionization-time of flight-mass spectrometer (Perseptive Biosystems) for both whole and trypsin-digested protein samples. Mass spectroscopy on the whole protein was performed using sinapinic acid as the matrix and BSA as the standard. The in-gel tryptic digest and subsequent mass spectroscopy were carried out essentially as described previously (27Campbell P.M. Harcourt R.L. Crone E.J. Claudianos C. Hammock B.D. Russell R.J. Oakeshott J.G. Insect Biochem. Mol. Biol. 2001; 31: 513-520Crossref PubMed Scopus (51) Google Scholar), except that the gel pieces were not treated with dithiothreitol and iodoacetamide. Theoretical molecular mass and trypsin digest patterns were calculated using Protein Prospector (prospector.ucsf.edu/, version 3.4.1). Bioinformatics—Programs available via BioManager (www.angis.org.au/, release 4.0) were used to analyze the DNA sequences for the presence of open reading frames (Flip). Searches of the non-redundant nucleotide and protein data bases (released Oct. 17, 2003) were performed using the various BLAST programs (version 2.2.6) available via NCBI (www.ncbi.nlm.nih.gov/BLAST/). Motif and fold recognition analysis of the protein sequence was performed using programs such as InterProScan (www.ebi.ac.uk/interpro/, July 22, 2003 (28Zdobnov E.M. Apweiler R. Bioinformatics. 2001; 17: 847-848Crossref PubMed Scopus (2097) Google Scholar)), PANAL (mgd.ahc.umn.edu/panal/ (29Silverstein K.A. Kilian A. Freeman J.L. Johnson J.E. Awad I.A. Retzel E.F. Bioinformatics. 2000; 16: 1157-1158Crossref PubMed Scopus (6) Google Scholar)), Genequiz (jura.ebi.ac.uk:8765/ext-genequiz/ (30Andrade M.A. Brown N.P. Leroy C. Hoersch S. de Daruvar A. Reich C. Franchini A. Tamames J. Valencia A. Ouzounis C. Sander C. Bioinformatics. 1999; 15: 391-412Crossref PubMed Scopus (162) Google Scholar)), and the Structure Prediction Metaserver (bioinfo.pl/Meta/pdb-test.pl/ (31Bujnicki J.M. Elofsson A. Fischer D. Rychlewski L. Bioinformatics. 2001; 17: 750-751Crossref PubMed Scopus (202) Google Scholar)). The secondary structure was predicted using [email protected] (npsa-pbil.ibcp.fr/ (32Combet C. Blanchet C. Geourjon C. Deleage G. Trends Biochem. Sci. 2000; 25: 147-150Abstract Full Text Full Text PDF PubMed Scopus (1392) Google Scholar)). The amino acid composition of the sequence was compared with bacterial sequences available at EBI (www.ebi.ac.uk/proteome/). The sequence was also analyzed using programs available via ExPASy (kr.expasy.org/tools/), including SignalP (version 1.1) and PSORT (version 6.4). Toxin Expression in Native Bacteria—Phase I and phase II bacteria from glycerol stocks of X. nematophila strain A24 were streaked on LB agar plates and grown at 28 °C for 2 days. The phase of the bacteria was confirmed by growth on nutrient agar, bromothymol blue, and tetrazolium chloride agar plates (33Akhurst R.J. J. Gen. Microbiol. 1980; : 303-309Google Scholar) and assays for antibiotic and lecithinase activity (34Boemare N.E. Akhurst R.J. J. Gen. Microbiol. 1988; 134: 751-761Google Scholar). Single colonies from the LB plates were used to inoculate 5 ml of nutrient broth (13 g/liter, Oxoid), and cultures were grown at 200 rpm for 24 or 48 h at 14, 20, or 28 °C. Cultures (2 ml) were harvested by centrifugation (3000 × g, 10 min), and the supernatant was removed (the soluble secreted material). The cells were resuspended in 1 ml of lysis buffer (20 mm Tris-Cl, 10 mm NaCl, pH 8), and 500 μg of lysozyme was added to 500 μl of this suspension and left on ice for 30 min. After boiling for 2 min, 50 μg of DNase I was added, and the sample was left at 37 °C for 1 h. The resulting suspension was centrifuged (3000 × g, 10 min), and the supernatant was removed (the soluble material from lysis). The final pellet (the insoluble material from lysis) was resuspended in 500 μl of phosphate-buffered saline. The samples were analyzed by SDS-PAGE and Western blot (see below). Antibody Production—Antibody was raised against his6 V16tox, 2S. Brown, A. Cao, P. Dobson, E. Hines, R. Akhurst, P. East, manuscript in preparation. a P. luminescens protein closely related to A24tox, following standard procedures at the Institute for Medical and Veterinary Science (Adelaide, Australia). Approximately 400 μg of purified his6V16tox protein was used to immunize each of two New Zealand White rabbits by subcutaneous injection. Each animal received a primary inoculation and three boosts at 3-week intervals. Following the final boost ∼40 ml of serum was collected from each animal. The antiserum was used without further purification. Western Blots—Western blots were performed using a nitrocellulose membrane (Hybond-C Extra, Amersham Biosciences) on a Novablot semi-dry blotter at 0.8 mA/cm2 in transfer buffer (39 mm glycine, 48 mm Tris-Cl, 0.375% SDS). The membrane was processed at room temperature in Tris-buffered saline containing 0.1% Tween 20 using three 5-min washes between all steps. The steps used were a 1-h or overnight block with 3% BSA, a 1-h incubation with the toxin antibody (1:3000 dilution), and a 1-h incubation with anti-rabbit IgG alkaline phosphatase conjugate (Sigma, 1:30,000 dilution). The blots were visualized using nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate (Promega) in substrate buffer (100 mm Tris-Cl, 5 mm MgCl2, 100 mm NaCl, pH 9.5). Insects—Helicoverpa armigera (Lepidoptera) were reared until fourth instar on an artificial diet (35Akhurst R.J. James W. Bird L.J. Beard C. J. Econ. Entomol. 2003; 96: 1290-1299Crossref PubMed Google Scholar). G. mellonella (Lepidoptera) were reared at 27 °C until fourth instar on an artificial diet containing dried rice cereal (500 g), honey/glycerol mix (154 ml of honey, 179 ml of glycerol, 37 ml of water), brewers yeast (30 g), and methyl 4-hydroxybenzoate (Tegasept, 0.2% w/w). Lucilia cuprina (Diptera) were reared at 27 °C until third instar on an artificial liver and meat meal diet. Bioassays—The activity of crude extracts and purified proteins was determined using an intra-hemocoel injection assay against the insects G. mellonella, H. armigera, and L. cuprina. Larvae were treated with 5–10 μl of either crude or pure sample, which was injected through an intersegmental membrane into the abdominal region. Bioassays were performed on at least 10 larvae that were held at 22 °C, and mortality was recorded daily over 3–6 days. For testing of the cosmid library against G. mellonella, crude extracts were prepared by growing E. coli cultures overnight at 28 °C in LB broth containing 150 μg/ml ampicillin and treating them with 2 mg/ml lysozyme (Amresco) for 15 min. The oral activity of this crude extract was also determined by incorporating 5 ml of crude extract into 95 ml of diet for H. armigera. The temperature dependence of the insecticidal activity was determined using crude extracts of cultures of A24tox in pGEM7Z(f)+/DH5α that were grown overnight at 37 °C, treated with lysozyme (1 mg/ml), and left at room temperature for 30 min. These extracts were injected into G. mellonella larvae that were kept at 20 and 25 °C. The activity of purified recombinant A24tox was tested against all three insects, using a dose range of 1–1000 ng of protein per larva for G. mellonella and H. armigera and 10–100 ng of protein per larva for L. cuprina. For G. mellonella and H. armigera, data were fitted to the probit curve using POLO-PC (LeOra software, 1987), a program that corrects for natural mortality. Histology—Fourth instar H. armigera larvae were injected with 10 μl of purified A24tox (100 ng) or MBP (100 ng). At 18 h post-injection, fixative (4% formaldehyde in 100 mm sodium phosphate, pH ∼7.2) was injected into three larvae; holes were punctured into the cuticle, and the larvae were transferred to fixative overnight at 4 °C. The larvae were embedded in paraffin wax, sectioned at 6 μm, and the sections stained with hematoxylin and eosin. Images were captured with a ProgRes 3012 digital camera on a Leica Diaplan microscope. Isolation of Toxin Gene—Xenorhabdus and Photorhabdus bacteria are known to be toxic to insects, but to date the only toxins from the bacteria that have been studied in detail are the Tc toxins. In order to identify other insect toxins from the bacteria, crude extracts from a cosmid library constructed from X. nematophila strain A24 were tested for insecticidal activity using a G. mellonella injection bioassay. The library screen identified two clones with insecticidal activity, whereas control lysates prepared from E. coli NM554 cells containing nonrecombinant Supercos vector showed no insecticidal activity. Both cosmid clones appeared to contain the same region of ∼35 kb of X. nematophila genomic DNA. Deletion analysis of this cosmid clone, further injection bioassays, and DNA sequencing identified an open reading frame of 1104 bp (368 amino acids) that was required for toxicity (Fig. 1). The DNA sequence has been deposited in GenBank™ (accession number AX029373). An analysis of the 5′ region of the gene identified possible imperfect –10 and –35 RNA polymerase recognition sites and an imperfect Shine-Dalgarno sequence, suggesting that the open reading frame was likely to be transcribed. Note that it is not possible to say whether the 1.1-kb toxin gene is part of the genome or a megaplasmid, as the cosmid clone had a size of ∼35 kb, and X. nematophila strain A24 is known to contain megaplasmids of size 70 and 120 kb (11Smigielski A.J. Akhurst R.J. J. Invertebr. Pathol. 1994; 64: 214-220Crossref Scopus (7) Google Scholar). Purification and Characterization of Toxin Protein—Biochemical characterization of the toxin encoded by the cloned gene from X. nematophila strain A24 required recombinant expression of the full-length protein. Initial attempts to overproduce the protein from commonly used E. coli vectors and strains, such as pGEM7Zf(+) in DH10B, resulted in the expression of very small amounts of protein, most of which was insoluble. Similarly, the toxin was mostly insoluble when expressed as a glutathione S-transferase fusion protein using the vector pGEX (Amersham Biosciences). Eventually, a useful amount of soluble protein without N- or C-terminal modifications was obtained by expressing the toxin as a fusion protein in the IMPACT system. After expression and purification according to the manufacturer's instructions, the A24tox construct produced a major soluble protein product of the expected size of ∼42 kDa, as determined by SDS-PAGE (Fig. 2A) and mass spectroscopy (see below). The final preparations contained 90% or more of the toxin protein, along with contaminants that varied slightly depending on the cleavage and wash conditions used during the purification process. Western blotting with an antibody known to be highly specific for the toxin indicated that some of these contaminants were related to A24tox, in particular the uncleaved fusion protein that runs at about 90 kDa on SDS-PAGE (Fig. 2B). E. coli GroEL (∼57–60 kDa) can also co-purify in small amounts with the target protein (www.neb.com/neb/faqs/impact.html). After purification and concentration, the yield of A24tox was ∼0.4 mg per liter of E. coli culture. The identity of the protein was confirmed by mass spectroscopy, where an analysis of the tryptic digest of A24tox identified peptides that covered 77% of the predicted open reading frame, including peptides from both the N and C termini (Fig. 1). The measured molecular weight (41,485) was within instrumental error (0.05%) of the calculated molecular weight (41,502). Both the whole protein and tryptic digest spectra suggested that the N-terminal methionine residue of the recombinant protein had not been processed. Biological Activity of the Toxin—The biological activity of the recombinant purified protein was tested by an intra-hemocoel injection bioassay, a method that delivers the bacteria directly into the hemolymph and thus mimics the release of the bacteria into the hemolymph that occurs soon after a nematode infects the insect host (7Forst S. Dowds B. Boemare N. Stackebrandt E. Annu. Rev. Microbiol. 1997; 51: 47-72Crossref PubMed Scopus (492) Google Scholar, 8Burnell A.M. Stock S.P. Nematology. 2000; 2: 31-42Crossref Scopus (165) Google Scholar). The bioassay results indicated that purified A24tox killed a high percentage of G. mellonella, H. armigera, or L. cuprina larvae after a single injection of at least 10 ng of toxin per larva. MBP purified from the IMPACT system was used as a control for these experiments and did not cause significant larval mortality (0–5%) at any of the quantities tested. This confirmed that A24tox was responsible for the observed toxicity and not any of the minor contaminants present in the toxin purification. With H. armigera, the larvae ceased feeding almost immediately and were dead after 2–3 days, and for G. mellonella, the larvae were unable to spin cocoon silk and were dead after 6 days. Mortality was dependent on the concentration of the toxin, as shown in Fig. 3A for G. mellonella and H. armigera. The LD50 values were calculated by probit analysis (Fig. 3A) and are listed with the 95% confidence limits shown in parentheses. G. mellonella was killed effectively by A24tox (LD50 = 26 ng (11–48 ng)), as was H. armigera (LD50 = 5.2 ng (1.4–9.0 ng)). Because the average larval mass at the time of injection was ∼140 mg for H. armigera and ∼800 mg for G. mellonella, this indicated that the LD50 values were in the range 30–40 ng/g. The A24tox protein was also active by injection against L. cuprina larvae, with an observed mortality of 60, 100, and 100% at doses of 10, 50, and 100 ng, respectively, compared with a MBP control mortality of 5–10%. Crude extracts of A24tox from E. coli strains, which were known to be active by injection, did not have any oral activity when incorporated into the H. armigera diet (data not shown). The activity of an E. coli crude extract against G. mellonella larvae was found to be temperature-dependent (Fig. 3B), with the toxin causing 100% mortality in approximately half the time at 25 °C as compared with 20 °C. It is clear that a major target of the toxin is the caterpillar midgut (Fig. 4). Control animals injected with MBP showed the expected histology for hea
Referência(s)