Photorhabdus Virulence Cassettes Confer Injectable Insecticidal Activity against the Wax Moth
http://www.100md.com
《细菌学杂志》
Center for Molecular Microbiology and Department of Biology and Biochemistry, University of Bath, Bath BA2 7AY, United Kingdom
ABSTRACT
Two recently sequenced genomes of the insect-pathogenic bacterium Photorhabdus and a large Serratia entomophila plasmid, pADAP, have phage-related loci containing putative toxin effector genes, designated the "Photorhabdus virulence cassettes" (PVCs). In S. entomophila, the single plasmid PVC confers antifeeding activity on larvae of a beetle. Here, we show that recombinant Escherichia coli expressing PVC-containing cosmids from Photorhabdus has injectable insecticidal activity against larvae of the wax moth. Electron microscopy showed that the structure of the PVC products is similar to the structure of the antibacterial R-type pyocins. However, unlike these bacteriocins, the PVC products of Photorhabdus have no demonstrable antibacterial activity. Instead, injection of Photorhabdus PVC products destroys insect hemocytes, which undergo dramatic actin cytoskeleton condensation. Comparison of the genomic organizations of several PVCs showed that they have a conserved phage-like structure with a variable number of putative anti-insect effectors encoded at one end. Expression of these putative effectors directly inside cultured cells showed that they are capable of rearranging the actin cytoskeleton. Together, these data show that the PVCs are functional homologs of the S. entomophila antifeeding genes and encode physical structures that resemble bacteriocins. This raises the interesting hypothesis that the PVC products are bacteriocin-like but that they have been modified to attack eukaryotic host cells.
INTRODUCTION
The recent sequencing of the Photorhabdus genome confirmed that this entomopathogenic bacterium produces an astonishing array of putative insecticidal toxins (7). Three classes of toxins have now been characterized. The members of the first class, the toxin complexes, are orally toxic to caterpillar pests (2) and have recently been used to create insect-resistant transgenic plants (14). The members of the second class, the "makes caterpillars floppy" toxins (Mcf1 and Mcf2), are potent toxins that are active upon injection (4, 18). Mcf1 mimics BH3 domain-only proteins, which are proapoptotic proteins found in mitochondria, and promotes apoptosis both in the insect gut and in mammalian tissue culture cells (6). Third and most recently, the "Photorhabdus insect-related" proteins (PirAB) have been shown to be binary toxins with both injectable (17) and oral activities in some insects (7). Given the requirement that Photorhabdus bacteria should kill the insect host into which they are released by their nematode vector and the numerous predicted toxins encoded in the Photorhabdus luminescens TT01 genome (7), it seems likely that other classes of insecticidal toxin remain to be discovered. In this light, here we investigated homologs of an antifeeding locus of the free-living entomopathogenic bacterium Serratia entomophila (11). We speculated that the antifeeding effect of the prophage-like locus in S. entomophila may reflect low antigut toxicity and therefore the possibility that the homologs of this locus in Photorhabdus are also toxic to insects.
In S. entomophila, the 120-kb pADAP plasmid confers "amber disease," a disease associated with "clearing" of midgut of the New Zealand grass grub (12). The pADAP plasmid contains genes encoding both homologs of the insecticidal toxin complex (tc genes) (8), termed sepA, sepB, and sepC (12), which are responsible for insect gut clearance, and also a prophage-like locus responsible for a separate "antifeeding" effect (11). This pADAP prophage-like locus contains 18 putative open reading frames for predicted proteins with high levels of similarity to phage tail and base plate proteins, as well as a putative effector protein which is putatively responsible for the antifeeding activity against the grass grub (11). Similarly, the genomes of two recently sequenced strains of Photorhabdus, P. luminescens strain TT01 (11, 19) and Photorhabdus asymbiotica strain ATCC 43949 (this study), contain numerous copies of these prophage-like loci, each encoding a different putative effector protein. Some of the putative effectors exhibit predicted amino acid similarity to parts of known multidomain toxins, such as the Mcf cytotoxin from P. luminescens (4), toxin A from Clostridium difficile (5), YopT from Yersinia enterocolitica (21), and the active site of cytotoxic necrosis factor (CNF1) from Escherichia coli (13). Others exhibit no predicted similarity to known effectors and so may represent novel effectors with novel modes of action.
In order to determine if the products of Photorhabdus virulence cassettes (PVCs) have insecticidal activity, as suggested by the antifeeding activity of the product of the pADAP PVC locus in S. entomophila, here we investigated their injectable activity against larvae of the wax moth Galleria mellonela. We found not only that E. coli strains with PVC-containing cosmids are toxic to larvae after injection but also that the same E. coli strains produce structures similar to one type of bacteriocins, the R-type pyocins. Unexpectedly, the PVC products have no antibacterial activity but trigger rapid destruction of insect phagocytes. The hemocyte destruction is associated with actin condensation, an effect that can be recapitulated by expressing putative effector proteins from the PVCs directly in tissue culture cells. Transposon mutagenesis of a single PVC showed that injectable toxicity is disrupted by insertion within a central subset of open reading frames. These results support the hypothesis that the PVCs are functional homologs of the antifeeding genes of S. entomophila but leave the precise mechanism of toxicity associated with the bacteriocin-like structure unclear.
MATERIALS AND METHODS
2D electrophoresis and protein identification. Supernatant proteins from E. coli EC100 carrying either the c4DF10 cosmid (which encodes PaPVCpnf) or the control vector pWEB were phenol precipitated, and the protein was resuspended in 150 μl CDU, which contained 9 M urea, 4% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS), 130 mM dithiothreitol (DTT), and 1x HALT protease inhibitor cocktail mixture (Pierce), and was incubated for 2 h at room temperature. Samples were then centrifuged for 30 min at 88,760 x g. For protein quantification, a RediPlate protein quantitation kit (Molecular Probes) was used. For two-dimensional (2D) gel electrophoresis, the Multiphor II system (GE Healthcare) was used for isoelectric focusing and horizontal sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis with Immobiline DryStrip gels and precast 12.5% homogeneous SDS gels (GE Healthcare). Briefly, Immobiline Dry strip gels were rehydrated overnight in rehydration buffer, which consisted of 8 M urea, 2% CHAPS, 0.002% bromophenol blue, 18 mM DTT, and 0.5% IPG buffer, containing 15 μg of the protein sample. Isoelectric focusing was performed at 20°C. For the IPG 3-10 strip the upper limits for current and power were set at 2 mA and 2 W, respectively. In phase 1 the voltage was kept at 300 V for 1 V · h, in phase 2 the voltage was increased to 3,500 V over 2,900 V · h, and in phase 3 the voltage was kept at 3,500 V for 7,100 V · h. The resulting strips were then frozen at –80°C for subsequent analysis. Before SDS gel electrophoresis, strips were equilibrated for 15 min in 50 mM Tris (pH 8.8), 6 M urea, 30% glycerol, 2% SDS, 0.002% bromophenol blue, and 65 mM DTT. In a second equilibration step, the 65 mM DTT was replaced by 135 mM iodoacetamide. The Immobiline DryStrip gel was then transferred onto the SDS polyacrylamide gel and electrophoresed. Before staining, gels were washed overnight in a fixation solution containing 40% ethanol and 10% acetic acid. Silver staining was performed according to the protocol of a PlusOne silver staining kit (GE Healthcare), omitting glutaraldehyde from the sensitizing solution and formaldehyde from the silver nitrate solution. Protein spots from silver-stained gels were cut from the acrylamide and sent to the protein sequencing facility at the University of the West of England. Matrix-assisted laser desorption ionization—time of flight (MALDI-TOF) analysis of trypsin-digested protein samples was used to generate a spectrum, which was compared to all proteins in the database and also to a database of predicted proteins encoded by the P. asymbiotica ATCC 43949 genome.
Cosmid libraries and insect bioassay. Cosmid libraries of P. asymbiotica strain ATCC 43949, P. luminescens strain TT01, and Photorhabdus temperata strain K122 were prepared in E. coli EC100 and arrayed into 96-well microtiter plates by MWG Biotech, Munich, Germany, as described previously (4). Insect bioassays were performed as described elsewhere (17). Briefly, 10-μl portions of a whole bacterial culture or a cell-free supernatant were injected into Galleria larvae that had been cooled at 4°C overnight. The animals were then held at 25°C and scored for "blackening," morbidity, and mortality at 1 and 24 h postinjection. "Heated" bacterial supernatants were heated to 90°C for 20 min to inactivate PVC toxicity.
Recombinant expression of putative effectors. The lopT-like and pnf-like genes from PaPVClopT and PaPVCpnf were PCR amplified from P. asymbiotica strain ATCC 43949 genomic DNA using the following primers: lopTF4 (TAAGAGCTCAATTATAATCTAAAGGTAAT), paLopT_R (ATGAAGCTTTTAGGGATAGTTGTGA), paPNF_F (AATGAATTCAACCATCACGGAGA), and paPNF_R (TGGGCAAGCTTATAACAACCGTT). The PCR products were purified and cloned into the arabinose-inducible expression plasmid pBAD30. The resulting clones were sequenced to confirm their identities, and the optimal expression conditions were determined. Briefly, –80°C glycerol stocks were used to inoculate 5 ml of fresh LB medium supplemented with 0.2% (wt/vol) glucose and the appropriate antibiotic for selection. Bacteria were grown overnight at 30°C with shaking, and 1 ml of the culture was then harvested, resuspended in 100 ml of the same medium, and incubated in an orbital incubator at 37°C until the optical density at 600 nm was 0.7 to 0.9. Cells were then harvested at room temperature by centrifugation at 4,000 rpm for 10 min. The pellet was resuspended in 100 ml of fresh LB medium supplemented with the appropriate antibiotic and the inducer L-arabinose at a concentration of 0.2% (wt/vol). Optimized times for induction were determined experimentally, and cells were then harvested. The bacterial cell pellet was resuspended in 10 ml of 1x phosphate-buffed saline (PBS) and sonicated (four 20-s sonications at 45 mA using a Branson 450 digital Sonifier) fitted with a tapered probe. The freshly sonicated samples were then diluted in 1x PBS for injection into Galleria larvae and for SDS-polyacrylamide gel electrophoresis analysis to confirm expression of the target protein.
Hemocyte counts and actin staining. Injected Galleria larvae were bled onto microscope slides by piercing the larval cuticle. Four animals were bled per treatment, and hemocytes were counted in four random microscopic fields of view on each slide. This provided an estimate of the number of recoverable (circulating) hemocytes. A tetramethyl rhodamine isocyanate (TRITC)-phalloidin conjugate was used to stain the hemocyte cytoskeleton (1). Briefly, preinjected Galleria larvae were bled onto coverslips, and the hemocytes were allowed to attach to the glass surface by incubation at room temperature for 20 min, after which they were fixed and stained with TRITC-phalloidin.
PVC purification and electron microscopy. Overnight cultures of E. coli with PVC-containing cosmids were grown overnight in LB medium at 30°C with aeration. The cultures were centrifuged at 8,000 rpm in a GS3 rotor for 30 min at 4°C. The supernatant was decanted to remove each cell pellet, and the centrifugation procedure was repeated to remove any remaining cells. Each cell-free supernatant was then centrifuged at 150,000 x g for 90 min at 4°C in a Sorvall ultracentrifuge to harvest particulate material. The supernatants were then discarded, and the pellets were washed by gentle resuspension in 1x PBS before a second centrifugation at 150,000 x g for 90 min at 4°C to pellet the particulate material. Each pellet was finally resuspended in 500 μl of ice-cold PBS and stored at 4°C. As a second step in the PVC purification procedure the particulate preparations were further separated by DEAE-Sepharose chromatography. For DEAE-Sepharose chromatography, the particulate material recovered after the first high-speed centrifugation described above was resuspended in 10 ml of ice-cold PBS, and an equivalent volume of DEAE-Sepharose CL-6B anion exchanger (in PBS) was added. This preparation was incubated at room temperature for 15 min. The Sepharose resin was harvested by centrifugation (3,000 x g), and the supernatant was discarded. The resin was resuspended in 40 ml of ice-cold PBS and again harvested by centrifugation. This washing step was repeated another three times, and the resin was finally resuspended in 10 ml of elution buffer (0.5 M NaCl, 50 mM phosphate buffer [pH 7.4]). The resin was removed by centrifugation, and the supernatant containing the PVCs was again centrifuged at 150,000 x g for 90 min at 4°C to pellet the particulate material. The particulate material was finally resuspended in 600 μl of ice-cold PBS and stored at 4°C.
For transmission electron microscopy (TEM) pioloform-covered 300-mesh copper grids that were coated with a fine layer of carbon were used as substrates for the protein fractions. The following four aqueous negative stains were tested with the protein samples: 1% uranyl acetate, 3% ammonium molybdate, 3% methylamine tungstate, and 2% sodium silicotungstate. The preferred stain, 3% methylamine tungstate, produced acceptable contrast and minimum artifacts and was subsequently used for all samples viewed by TEM. The coated grids were exposed to UV light for 16 h immediately prior to use to ensure adequate wetting of the substrate. A 10-μl drop was applied to the TEM grid, and the protein was allowed to settle for 5 min. Liquid was absorbed with filter paper from the edge of the grid and replaced immediately with 10 μl of filtered negative stain. The drop was partially removed with filter paper, and the grids were allowed to air dry thoroughly before they were viewed with a JEOL 1200EX transmission electron microscope (JEOL, Tokyo, Japan) operating at 80 kV.
Effector expression in mammalian cells. The pnf-like and sepC-like genes of PaPVCpnf were PCR amplified from P. asymbiotica genomic DNA using primers (5' to 3') papnfrk5FbamH1 (TTAGGATCCTTAAAATATGCTAATCCT; BamHI site), papnfrk5Rpst1 (TGCCCTCATTCTGCAGAATTATAACA; PstI site), pasepCrk5FbamH1 (TTTAGGATCCCCTAGATATGCTAATTAT; BamHI site), and pasepCrk5Rpst1 (TATTCTGCAGATGACATGAAAATCCA; PstI site), and the plu1690 PlPVCmcf gene was PCR amplified using primers plu1690FbamH1 (AATGGATCCCCAAACAGCAAATATAGT; BamHI site) and plu1690Rpst1 (AATCTGCAGGGTATATATTAATTGT; PstI site) (the underlining indicates mismatched nucleotides incorporated to generate the restriction digestion sites indicated). PCR products were purified using standard techniques and were cloned as myc-tagged fusions into the relevant restriction sites of the mammalian expression plasmid pRK5myc. Clones were sequenced to confirm their identities. These constructs were cotransfected into NIH-3T3 cells along with pEGFPactin as previously described (16). Briefly, the myc-tagged fusion allowed us to detect, visualize, and localize transient expression of the target protein (Pnf, SepC, or Plu1690), while the pEGFPactin revealed the condition of the cytoskeleton in the same cells.
RESULTS AND DISCUSSION
Genomic organization. To determine the genomic organization of the PVCs in different Photorhabdus species, we identified and compared PVCs from P. luminescens strain TT01 (7), P. temperata strain K122, and P. asymbiotica ATCC 43949. PVC loci were found to be distributed around the genome in both P. asymbiotica ATCC 43949 and P. luminescens. However in the case of P. luminescens, four PVC elements were also clustered in a tandem array between a type IV DNA conjugation pilus operon and a mukB replicon partitioning gene. In P. asymbiotica only one of these PVC elements was present at this locus, while in P. temperata this PVC cassette was replaced by a Mu-like phage, suggesting that there is a common insertion site for mobile elements. A detailed comparison of the different predicted open reading frames in the different PVCs revealed their phage-like structure. Thus, each PVC unit was predicted to encode 15 to 20 proteins, each with predicted similarity to phage tail, phage base plate assembly, fimbrial usher, and putative effector proteins from other pathogenic bacteria (Fig. 1, top panel). We noted that each PVC unit had different putative effector sequences, but they were always in the equivalent position (Fig. 1, top panel), as documented for other prophages (3). The different PVC effectors were also often flanked by transposons, suggesting a possible mechanism for their insertion into the PVC or movement between different PVCs. Finally, we noted that one PVC element was consistently inserted adjacent to the mukB locus, a locus involved in plasmid maintenance and stability (20). This suggests either that mukB may be a communal insertion point for these phage-like elements or that it may regulate the insertion of plasmids containing these elements. Interestingly, a type IV DNA conjugation pilus operon was found immediately 5' of this PVC element insertion site. A similar operon was also found on the pADAP plasmid, where it was responsible for conjugative transfer of the plasmid. It is therefore temping to speculate that this operon in Photorhabdus is involved in horizontal movement of the PVC elements. We selected two model elements from P. asymbiotica, designated PaPVCpnf and PaPVClopT, for further experimental investigation.
Expression and structure. We examined the supernatant of PaPVCpnf cosmid-containing E. coli via 2D gel electrophoresis and subsequent MALDI-TOF analysis of expressed Photorhabdus proteins (Fig. 1, bottom panel). Visual examination of 2D gels of E. coli with and without PVC-carrying cosmids revealed several differences (Fig. 1, bottom panel). Subsequent MALDI-TOF analysis of trypsin digests of proteins present only in the PVC-expressing E. coli and subsequent interrogation of a database of predicted PVC-encoded proteins confirmed that three of the most abundant differentially expressed proteins had a PVC origin and corresponded to the first three predicted open reading frames in the PVC locus (Fig. 1, top panel). To determine the structure of the final exported PVC, we purified preparations of PaPVCpnf-expressing recombinant E. coli by using TEM. Electron microscopy revealed the presence of structures highly reminiscent of R-type pyocins (15). These structures were 30 nm wide and had variable lengths; some examples were more than 800 nm long (Fig. 2). They comprised a rigid contractile outer sheath and a thin hollow inner "needle" structure (approximately 6 nm wide), analogous to R-type pyocins. Examples of both "contracted" and "relaxed" forms were seen (Fig. 2). Although the precise mode of action of R-type pyocins themselves is not known, it has been suggested that their antibacterial activity is achieved by injection of the needle into bacterial cells via contraction of the sheath (15). However, investigation of this potential mode of action for the PVC products was beyond the scope of the current study.
Insecticidal activity. As R-type pyocins are bacteriocins with well-documented antibacterial activity (15), we tested purified PVC products with a range of different bacteria. We were unable to demonstrate any activity against either E. coli or Micrococcus luteus. We therefore also tested insecticidal activity; we injected E. coli K-12 strain EC100 carrying individual PVC-containing cosmids into larvae of the wax moth. Following injection of whole cultures of PVC-expressing E. coli, larvae died within 24 h (Fig. 3A). For E. coli expressing PaPVCpnf, larvae into which cell-free supernatant alone was injected became black and moribund 15 min after injection (Fig. 3B). This injectable activity in the supernatant from PaPVCpnf-expressing E. coli was eliminated by heat treatment prior to injection (Fig. 3C). Surprisingly, a comparison of the mortality data (Fig. 3A) resulting from injection of E. coli carrying PVC products from different Photorhabdus species suggested that PVC products from the human pathogen P. asymbiotica (9) had greater toxicity for insects than PVC products derived from the insect pathogen P. luminescens TT01 (7). At this stage we are uncertain why a PVC product from a pathogen isolated from a human wound should have greater toxicity for insects than a PVC product isolated from an entomopathogenic nematode.
In vivo effect on hemocytes. Following injection of E. coli K-12 strain EC100 not expressing any bacterial virulence factors, the insect immune system normally clears the infecting E. coli via a combination of antibacterial peptide production and the phagocyotosis and nodulation reactions associated with insect hemocytes (10). Therefore, in order to understand why PVC-expressing E. coli is not successfully neutralized by the insect immune system, we bled Galleria larvae 30 min after injection with PaPVCpnf supernatants and examined their hemocytes. Simple counting of recoverable hemocytes revealed a dramatic decrease in the number of intact hemocytes 30 min after injection (Fig. 4A). Further staining of the few remaining hemocytes with TRITC-phalloidin revealed that their actin cytoskeletons were compressed in the residual cell bodies (Fig. 4B). The dramatic decrease in the number of recoverable hemocytes from infected insects as little as 30 min after injection (Fig. 4A) helps explain how PVC-expressing E. coli can persist in the presence of insect phagocytes or hemocytes.
To confirm that the putative effectors within Photorhabdus PVCs do indeed represent novel antieukaryotic effectors, we expressed several effectors directly in transfected mammalian cells and looked for rearrangements of the actin cytoskeleton and changes in nuclear morphology. Putative PVC effectors were transfected into NIH-3T3 cells using the mammalian expression vector pRK5myc (6). This vector expresses each effector as a fusion protein with Myc, which allows detection of it in transfected cells using an anti-Myc antibody (6). To assess the subsequent integrity of the actin cytoskeleton, we cotransfected the same cells with a construct expressing EGFP actin (Fig. 5). Consistent with its dramatic effect on insect hemocytes, the effector that exhibited similarity to Pnf, encoded by PaPVCpnf, dramatically reduced the cell bodies of transfected cells, leaving only a bundle of condensed actin and a disintegrating nucleus (Fig. 5C). Interestingly, a second putative effector also encoded on the PaPVCpnf element, which exhibited similarity to the C-terminal region of the SepC toxin, produced long arms of actin-rich filaments in transfected cells (Fig. 5D). Finally, a third effector, encoded by plu1690, which exhibited similarity to part of the P. luminescens cytotoxin Mcf (4), also reduced the cell bodies of transfected cells, leaving an intact nucleus surrounded by several actin-rich bundles (Fig. 5B). Together, these data suggest that the PVC-encoded effectors are capable of rearranging the actin cytoskeleton of host cells. However, it is not clear if the antihemocytic effects of PVC effector injection are related to this phenotype in mammalian tissue culture cells.
Transposon mutagenesis. In order to determine which open reading frames in an individual PVC are required for injectable insecticidal activity, we performed transposon mutagenesis of the PaPVCpnf cosmid and then examined the toxicity of the resulting mutants containing single insertions in single open reading frames. The results of the mutagenesis (Fig. 6) showed that a small central section of open reading frames, including the putative effector gene encoding a protein with similarity to the known effector Lop-T, are required for insecticidal activity.
Conclusions. Here we describe the first functional studies of phage-like PVC structures found as repetitive cassettes in two different genera of entomopathogenic bacteria, Photorhabdus (19) and Serratia (11). These prophage-like loci were described both as an antifeeding locus on the pADAP plasmid of S. entomophila (11) and as repetitive loci in the genomes of P. luminescens TT01 (19) and P. asymbiotica ATCC 43949 (19). We found that these loci encode particles whose structure is similar to the structure of R-type pycocins. However, unlike R-type pyocins, the PVC products have no demonstrable antibiotic activity, but they do have injectable activity against larvae of the wax moth. Together, these data suggest that the PVC products, together with the product of the antifeeding locus on pADAP of S. entomophila (11), may be bacteriocin-like structures modified to have activity against insect phagocytes. However, many central questions related to the biology of both the PVCs and the S. entomophila antifeeding genes remain unanswered. First, what is the relative role of phage-like genes and the putative effectors in PVC toxicity Second, are the hemocytes the primary target of the PVC products, or is their effect more widespread Third, does the mode of action of these novel structures involve the needle-like delivery mechanism inferred for R-type pyocins Finally, what is the role of these prophage-like elements in the ecology of different Photorhabdus species All of these questions about these interesting new structures remain to be investigated. However, the PVCs may well be important in conferring novel virulence either against different types of insects or, in the case of the P. asymbiotica strains isolated from human wounds (9), potentially against humans.
ACKNOWLEDGMENTS
We thank Ursula Potter of the Electron Microscopy facility at Bath for TEM pictures of PVC-expressing bacteria.
REFERENCES
Au, C., P. Dean, S. E. Reynolds, and R. H. ffrench-Constant. 2004. Effect of the insect pathogenic bacterium Photorhabdus on insect phagocytes. Cell. Microbiol. 6:89-95.
Bowen, D., T. A. Rocheleau, M. Blackburn, O. Andreev, E. Golubeva, R. Bhartia, and R. H. ffrench-Constant. 1998. Insecticidal toxins from the bacterium Photorhabdus luminescens. Science 280:2129-2132.
Boyd, E. F., and H. Brussow. 2002. Common themes among bacteriophage-encoded virulence factors and diversity among the bacteriophages involved. Trends Microbiol. 10:521-529.
Daborn, P. J., N. Waterfield, C. P. Silva, C. P. Au, S. Sharma, and R. H. ffrench-Constant. 2002. A single Photorhabdus gene, makes caterpillars floppy (mcf), allows Escherichia coli to persist within and kill insects. Proc. Natl. Acad. Sci. USA 99:10742-10747.
Dove, C. H., S.-Z. Wang, S. B. Price, C. J. Phelps, D. M. Lyerly, T. D. Wilkins, and J. L. Johnson. 1990. Molecular characterization of the Clostridium difficile toxin A gene. Infect. Immun. 58:480-488.
Dowling, A. J., P. J. Daborn, N. R. Waterfield, P. Wang, C. H. Streuli, and R. H. ffrench-Constant. 2004. The insecticidal toxin makes caterpillars floppy (Mcf) promotes apoptosis in mammalian cells. Cell. Microbiol. 6:345-353.
Duchaud, E., C. Rusniok, L. Frangeul, C. Buchrieser, A. Givaudan, S. Taourit, S. Bocs, C. Boursaux-Eude, M. Chandler, J. F. Charles, E. Dassa, R. Derose, S. Derzelle, G. Freyssinet, S. Gaudriault, C. Medigue, A. Lanois, K. Powell, P. Siguier, R. Vincent, V. Wingate, M. Zouine, P. Glaser, N. Boemare, A. Danchin, and F. Kunst. 2003. The genome sequence of the entomopathogenic bacterium Photorhabdus luminescens. Nat. Biotechnol. 21:1307-1313.
ffrench-Constant, R. H., N. Waterfield, and P. Daborn. 2004. Insecticidal toxins from Photorhabdus and Xenorhabdus. Compr. Mol. Insect Sci. 5:239-253.
Gerrard, J., N. Waterfield, R. Vohra, and R. ffrench-Constant. 2004. Human infection with Photorhabdus asymbiotica: an emerging bacterial pathogen. Microbes Infect. 6:229-237.
Gillespie, J. P., M. R. Kanost, and T. Trenczek. 1997. Biological mediators of insect immunity. Annu. Rev. Entomol. 42:611-643.
Hurst, M. R., T. R. Glare, and T. A. Jackson. 2004. Cloning Serratia entomophila antifeeding genes—a putative defective prophage active against the grass grub Costelytra zealandica. J. Bacteriol. 186:5116-5128.
Hurst, M. R., T. R. Glare, T. A. Jackson, and C. W. Ronson. 2000. Plasmid-located pathogenicity determinants of Serratia entomophila, the causal agent of amber disease of grass grub, show similarity to the insecticidal toxins of Photorhabdus luminescens. J. Bacteriol. 182:5127-5138.
Landraud, L., C. Pulcini, P. Gounon, G. Flatau, P. Boquet, and E. Lemichez. 2004. E. coli CNF1 toxin: a two-in-one system for host-cell invasion. Int. J. Med. Microbiol. 293:513-518.
Liu, D., S. Burton, T. Glancy, Z. S. Li, R. Hampton, T. Meade, and D. J. Merlo. 2003. Insect resistance conferred by 283-kDa Photorhabdus luminescens protein TcdA in Arabidopsis thaliana. Nat. Biotechnol. 21:1222-1228.
Michel-Briand, Y., and C. Baysse. 2002. The pyocins of Pseudomonas aeruginosa. Biochimie 84:499-510.
Waterfield, N., M. Hares, A. Dowling, and R. ffrench-Constant. 2005. Potentiation and cellular phenotypes of the insecticidal toxin complexes of Photorhabdus bacteria. Cell. Microbiol. 7:373-382.
Waterfield, N., S. G. Kamita, B. D. Hammock, and R. ffrench-Constant. 2005. The Photorhabdus Pir toxins are similar to a developmentally regulated insect protein but show no juvenile hormone esterase activity. FEMS Microbiol. Lett. 245:47-52.
Waterfield, N. R., P. J. Daborn, A. J. Dowling, G. Yang, M. Hares, and R. H. ffrench-Constant. 2003. The insecticidal toxin makes caterpillars floppy 2 (Mcf2) shows similarity to HrmA, an avirulence protein from a plant pathogen. FEMS Microbiol. Lett. 229:265-270.
Waterfield, N. R., P. J. Daborn, and R. H. ffrench-Constant. 2004. Insect pathogenicity islands in the insect pathogenic bacterium Photorhabdus. Physiol. Entomol. 29:240-250.
Weitao, T., S. Dasgupta, and K. Nordstrom. 2000. Role of the mukB gene in chromosome and plasmid partition in Escherichia coli. Mol. Microbiol. 38:392-400.
Zumbihl, R., M. Aepfelbacher, A. Andor, C. A. Jacobi, K. Ruckdeschel, B. Rouot, and J. Heesemann. 1999. The cytotoxin YopT of Yersinia enterocolitica induces modification and cellular redistribution of the small GTP-binding protein RhoA. J. Biol. Chem. 274:29289-29293.(G. Yang, A. J. Dowling, U)
ABSTRACT
Two recently sequenced genomes of the insect-pathogenic bacterium Photorhabdus and a large Serratia entomophila plasmid, pADAP, have phage-related loci containing putative toxin effector genes, designated the "Photorhabdus virulence cassettes" (PVCs). In S. entomophila, the single plasmid PVC confers antifeeding activity on larvae of a beetle. Here, we show that recombinant Escherichia coli expressing PVC-containing cosmids from Photorhabdus has injectable insecticidal activity against larvae of the wax moth. Electron microscopy showed that the structure of the PVC products is similar to the structure of the antibacterial R-type pyocins. However, unlike these bacteriocins, the PVC products of Photorhabdus have no demonstrable antibacterial activity. Instead, injection of Photorhabdus PVC products destroys insect hemocytes, which undergo dramatic actin cytoskeleton condensation. Comparison of the genomic organizations of several PVCs showed that they have a conserved phage-like structure with a variable number of putative anti-insect effectors encoded at one end. Expression of these putative effectors directly inside cultured cells showed that they are capable of rearranging the actin cytoskeleton. Together, these data show that the PVCs are functional homologs of the S. entomophila antifeeding genes and encode physical structures that resemble bacteriocins. This raises the interesting hypothesis that the PVC products are bacteriocin-like but that they have been modified to attack eukaryotic host cells.
INTRODUCTION
The recent sequencing of the Photorhabdus genome confirmed that this entomopathogenic bacterium produces an astonishing array of putative insecticidal toxins (7). Three classes of toxins have now been characterized. The members of the first class, the toxin complexes, are orally toxic to caterpillar pests (2) and have recently been used to create insect-resistant transgenic plants (14). The members of the second class, the "makes caterpillars floppy" toxins (Mcf1 and Mcf2), are potent toxins that are active upon injection (4, 18). Mcf1 mimics BH3 domain-only proteins, which are proapoptotic proteins found in mitochondria, and promotes apoptosis both in the insect gut and in mammalian tissue culture cells (6). Third and most recently, the "Photorhabdus insect-related" proteins (PirAB) have been shown to be binary toxins with both injectable (17) and oral activities in some insects (7). Given the requirement that Photorhabdus bacteria should kill the insect host into which they are released by their nematode vector and the numerous predicted toxins encoded in the Photorhabdus luminescens TT01 genome (7), it seems likely that other classes of insecticidal toxin remain to be discovered. In this light, here we investigated homologs of an antifeeding locus of the free-living entomopathogenic bacterium Serratia entomophila (11). We speculated that the antifeeding effect of the prophage-like locus in S. entomophila may reflect low antigut toxicity and therefore the possibility that the homologs of this locus in Photorhabdus are also toxic to insects.
In S. entomophila, the 120-kb pADAP plasmid confers "amber disease," a disease associated with "clearing" of midgut of the New Zealand grass grub (12). The pADAP plasmid contains genes encoding both homologs of the insecticidal toxin complex (tc genes) (8), termed sepA, sepB, and sepC (12), which are responsible for insect gut clearance, and also a prophage-like locus responsible for a separate "antifeeding" effect (11). This pADAP prophage-like locus contains 18 putative open reading frames for predicted proteins with high levels of similarity to phage tail and base plate proteins, as well as a putative effector protein which is putatively responsible for the antifeeding activity against the grass grub (11). Similarly, the genomes of two recently sequenced strains of Photorhabdus, P. luminescens strain TT01 (11, 19) and Photorhabdus asymbiotica strain ATCC 43949 (this study), contain numerous copies of these prophage-like loci, each encoding a different putative effector protein. Some of the putative effectors exhibit predicted amino acid similarity to parts of known multidomain toxins, such as the Mcf cytotoxin from P. luminescens (4), toxin A from Clostridium difficile (5), YopT from Yersinia enterocolitica (21), and the active site of cytotoxic necrosis factor (CNF1) from Escherichia coli (13). Others exhibit no predicted similarity to known effectors and so may represent novel effectors with novel modes of action.
In order to determine if the products of Photorhabdus virulence cassettes (PVCs) have insecticidal activity, as suggested by the antifeeding activity of the product of the pADAP PVC locus in S. entomophila, here we investigated their injectable activity against larvae of the wax moth Galleria mellonela. We found not only that E. coli strains with PVC-containing cosmids are toxic to larvae after injection but also that the same E. coli strains produce structures similar to one type of bacteriocins, the R-type pyocins. Unexpectedly, the PVC products have no antibacterial activity but trigger rapid destruction of insect phagocytes. The hemocyte destruction is associated with actin condensation, an effect that can be recapitulated by expressing putative effector proteins from the PVCs directly in tissue culture cells. Transposon mutagenesis of a single PVC showed that injectable toxicity is disrupted by insertion within a central subset of open reading frames. These results support the hypothesis that the PVCs are functional homologs of the antifeeding genes of S. entomophila but leave the precise mechanism of toxicity associated with the bacteriocin-like structure unclear.
MATERIALS AND METHODS
2D electrophoresis and protein identification. Supernatant proteins from E. coli EC100 carrying either the c4DF10 cosmid (which encodes PaPVCpnf) or the control vector pWEB were phenol precipitated, and the protein was resuspended in 150 μl CDU, which contained 9 M urea, 4% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS), 130 mM dithiothreitol (DTT), and 1x HALT protease inhibitor cocktail mixture (Pierce), and was incubated for 2 h at room temperature. Samples were then centrifuged for 30 min at 88,760 x g. For protein quantification, a RediPlate protein quantitation kit (Molecular Probes) was used. For two-dimensional (2D) gel electrophoresis, the Multiphor II system (GE Healthcare) was used for isoelectric focusing and horizontal sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis with Immobiline DryStrip gels and precast 12.5% homogeneous SDS gels (GE Healthcare). Briefly, Immobiline Dry strip gels were rehydrated overnight in rehydration buffer, which consisted of 8 M urea, 2% CHAPS, 0.002% bromophenol blue, 18 mM DTT, and 0.5% IPG buffer, containing 15 μg of the protein sample. Isoelectric focusing was performed at 20°C. For the IPG 3-10 strip the upper limits for current and power were set at 2 mA and 2 W, respectively. In phase 1 the voltage was kept at 300 V for 1 V · h, in phase 2 the voltage was increased to 3,500 V over 2,900 V · h, and in phase 3 the voltage was kept at 3,500 V for 7,100 V · h. The resulting strips were then frozen at –80°C for subsequent analysis. Before SDS gel electrophoresis, strips were equilibrated for 15 min in 50 mM Tris (pH 8.8), 6 M urea, 30% glycerol, 2% SDS, 0.002% bromophenol blue, and 65 mM DTT. In a second equilibration step, the 65 mM DTT was replaced by 135 mM iodoacetamide. The Immobiline DryStrip gel was then transferred onto the SDS polyacrylamide gel and electrophoresed. Before staining, gels were washed overnight in a fixation solution containing 40% ethanol and 10% acetic acid. Silver staining was performed according to the protocol of a PlusOne silver staining kit (GE Healthcare), omitting glutaraldehyde from the sensitizing solution and formaldehyde from the silver nitrate solution. Protein spots from silver-stained gels were cut from the acrylamide and sent to the protein sequencing facility at the University of the West of England. Matrix-assisted laser desorption ionization—time of flight (MALDI-TOF) analysis of trypsin-digested protein samples was used to generate a spectrum, which was compared to all proteins in the database and also to a database of predicted proteins encoded by the P. asymbiotica ATCC 43949 genome.
Cosmid libraries and insect bioassay. Cosmid libraries of P. asymbiotica strain ATCC 43949, P. luminescens strain TT01, and Photorhabdus temperata strain K122 were prepared in E. coli EC100 and arrayed into 96-well microtiter plates by MWG Biotech, Munich, Germany, as described previously (4). Insect bioassays were performed as described elsewhere (17). Briefly, 10-μl portions of a whole bacterial culture or a cell-free supernatant were injected into Galleria larvae that had been cooled at 4°C overnight. The animals were then held at 25°C and scored for "blackening," morbidity, and mortality at 1 and 24 h postinjection. "Heated" bacterial supernatants were heated to 90°C for 20 min to inactivate PVC toxicity.
Recombinant expression of putative effectors. The lopT-like and pnf-like genes from PaPVClopT and PaPVCpnf were PCR amplified from P. asymbiotica strain ATCC 43949 genomic DNA using the following primers: lopTF4 (TAAGAGCTCAATTATAATCTAAAGGTAAT), paLopT_R (ATGAAGCTTTTAGGGATAGTTGTGA), paPNF_F (AATGAATTCAACCATCACGGAGA), and paPNF_R (TGGGCAAGCTTATAACAACCGTT). The PCR products were purified and cloned into the arabinose-inducible expression plasmid pBAD30. The resulting clones were sequenced to confirm their identities, and the optimal expression conditions were determined. Briefly, –80°C glycerol stocks were used to inoculate 5 ml of fresh LB medium supplemented with 0.2% (wt/vol) glucose and the appropriate antibiotic for selection. Bacteria were grown overnight at 30°C with shaking, and 1 ml of the culture was then harvested, resuspended in 100 ml of the same medium, and incubated in an orbital incubator at 37°C until the optical density at 600 nm was 0.7 to 0.9. Cells were then harvested at room temperature by centrifugation at 4,000 rpm for 10 min. The pellet was resuspended in 100 ml of fresh LB medium supplemented with the appropriate antibiotic and the inducer L-arabinose at a concentration of 0.2% (wt/vol). Optimized times for induction were determined experimentally, and cells were then harvested. The bacterial cell pellet was resuspended in 10 ml of 1x phosphate-buffed saline (PBS) and sonicated (four 20-s sonications at 45 mA using a Branson 450 digital Sonifier) fitted with a tapered probe. The freshly sonicated samples were then diluted in 1x PBS for injection into Galleria larvae and for SDS-polyacrylamide gel electrophoresis analysis to confirm expression of the target protein.
Hemocyte counts and actin staining. Injected Galleria larvae were bled onto microscope slides by piercing the larval cuticle. Four animals were bled per treatment, and hemocytes were counted in four random microscopic fields of view on each slide. This provided an estimate of the number of recoverable (circulating) hemocytes. A tetramethyl rhodamine isocyanate (TRITC)-phalloidin conjugate was used to stain the hemocyte cytoskeleton (1). Briefly, preinjected Galleria larvae were bled onto coverslips, and the hemocytes were allowed to attach to the glass surface by incubation at room temperature for 20 min, after which they were fixed and stained with TRITC-phalloidin.
PVC purification and electron microscopy. Overnight cultures of E. coli with PVC-containing cosmids were grown overnight in LB medium at 30°C with aeration. The cultures were centrifuged at 8,000 rpm in a GS3 rotor for 30 min at 4°C. The supernatant was decanted to remove each cell pellet, and the centrifugation procedure was repeated to remove any remaining cells. Each cell-free supernatant was then centrifuged at 150,000 x g for 90 min at 4°C in a Sorvall ultracentrifuge to harvest particulate material. The supernatants were then discarded, and the pellets were washed by gentle resuspension in 1x PBS before a second centrifugation at 150,000 x g for 90 min at 4°C to pellet the particulate material. Each pellet was finally resuspended in 500 μl of ice-cold PBS and stored at 4°C. As a second step in the PVC purification procedure the particulate preparations were further separated by DEAE-Sepharose chromatography. For DEAE-Sepharose chromatography, the particulate material recovered after the first high-speed centrifugation described above was resuspended in 10 ml of ice-cold PBS, and an equivalent volume of DEAE-Sepharose CL-6B anion exchanger (in PBS) was added. This preparation was incubated at room temperature for 15 min. The Sepharose resin was harvested by centrifugation (3,000 x g), and the supernatant was discarded. The resin was resuspended in 40 ml of ice-cold PBS and again harvested by centrifugation. This washing step was repeated another three times, and the resin was finally resuspended in 10 ml of elution buffer (0.5 M NaCl, 50 mM phosphate buffer [pH 7.4]). The resin was removed by centrifugation, and the supernatant containing the PVCs was again centrifuged at 150,000 x g for 90 min at 4°C to pellet the particulate material. The particulate material was finally resuspended in 600 μl of ice-cold PBS and stored at 4°C.
For transmission electron microscopy (TEM) pioloform-covered 300-mesh copper grids that were coated with a fine layer of carbon were used as substrates for the protein fractions. The following four aqueous negative stains were tested with the protein samples: 1% uranyl acetate, 3% ammonium molybdate, 3% methylamine tungstate, and 2% sodium silicotungstate. The preferred stain, 3% methylamine tungstate, produced acceptable contrast and minimum artifacts and was subsequently used for all samples viewed by TEM. The coated grids were exposed to UV light for 16 h immediately prior to use to ensure adequate wetting of the substrate. A 10-μl drop was applied to the TEM grid, and the protein was allowed to settle for 5 min. Liquid was absorbed with filter paper from the edge of the grid and replaced immediately with 10 μl of filtered negative stain. The drop was partially removed with filter paper, and the grids were allowed to air dry thoroughly before they were viewed with a JEOL 1200EX transmission electron microscope (JEOL, Tokyo, Japan) operating at 80 kV.
Effector expression in mammalian cells. The pnf-like and sepC-like genes of PaPVCpnf were PCR amplified from P. asymbiotica genomic DNA using primers (5' to 3') papnfrk5FbamH1 (TTAGGATCCTTAAAATATGCTAATCCT; BamHI site), papnfrk5Rpst1 (TGCCCTCATTCTGCAGAATTATAACA; PstI site), pasepCrk5FbamH1 (TTTAGGATCCCCTAGATATGCTAATTAT; BamHI site), and pasepCrk5Rpst1 (TATTCTGCAGATGACATGAAAATCCA; PstI site), and the plu1690 PlPVCmcf gene was PCR amplified using primers plu1690FbamH1 (AATGGATCCCCAAACAGCAAATATAGT; BamHI site) and plu1690Rpst1 (AATCTGCAGGGTATATATTAATTGT; PstI site) (the underlining indicates mismatched nucleotides incorporated to generate the restriction digestion sites indicated). PCR products were purified using standard techniques and were cloned as myc-tagged fusions into the relevant restriction sites of the mammalian expression plasmid pRK5myc. Clones were sequenced to confirm their identities. These constructs were cotransfected into NIH-3T3 cells along with pEGFPactin as previously described (16). Briefly, the myc-tagged fusion allowed us to detect, visualize, and localize transient expression of the target protein (Pnf, SepC, or Plu1690), while the pEGFPactin revealed the condition of the cytoskeleton in the same cells.
RESULTS AND DISCUSSION
Genomic organization. To determine the genomic organization of the PVCs in different Photorhabdus species, we identified and compared PVCs from P. luminescens strain TT01 (7), P. temperata strain K122, and P. asymbiotica ATCC 43949. PVC loci were found to be distributed around the genome in both P. asymbiotica ATCC 43949 and P. luminescens. However in the case of P. luminescens, four PVC elements were also clustered in a tandem array between a type IV DNA conjugation pilus operon and a mukB replicon partitioning gene. In P. asymbiotica only one of these PVC elements was present at this locus, while in P. temperata this PVC cassette was replaced by a Mu-like phage, suggesting that there is a common insertion site for mobile elements. A detailed comparison of the different predicted open reading frames in the different PVCs revealed their phage-like structure. Thus, each PVC unit was predicted to encode 15 to 20 proteins, each with predicted similarity to phage tail, phage base plate assembly, fimbrial usher, and putative effector proteins from other pathogenic bacteria (Fig. 1, top panel). We noted that each PVC unit had different putative effector sequences, but they were always in the equivalent position (Fig. 1, top panel), as documented for other prophages (3). The different PVC effectors were also often flanked by transposons, suggesting a possible mechanism for their insertion into the PVC or movement between different PVCs. Finally, we noted that one PVC element was consistently inserted adjacent to the mukB locus, a locus involved in plasmid maintenance and stability (20). This suggests either that mukB may be a communal insertion point for these phage-like elements or that it may regulate the insertion of plasmids containing these elements. Interestingly, a type IV DNA conjugation pilus operon was found immediately 5' of this PVC element insertion site. A similar operon was also found on the pADAP plasmid, where it was responsible for conjugative transfer of the plasmid. It is therefore temping to speculate that this operon in Photorhabdus is involved in horizontal movement of the PVC elements. We selected two model elements from P. asymbiotica, designated PaPVCpnf and PaPVClopT, for further experimental investigation.
Expression and structure. We examined the supernatant of PaPVCpnf cosmid-containing E. coli via 2D gel electrophoresis and subsequent MALDI-TOF analysis of expressed Photorhabdus proteins (Fig. 1, bottom panel). Visual examination of 2D gels of E. coli with and without PVC-carrying cosmids revealed several differences (Fig. 1, bottom panel). Subsequent MALDI-TOF analysis of trypsin digests of proteins present only in the PVC-expressing E. coli and subsequent interrogation of a database of predicted PVC-encoded proteins confirmed that three of the most abundant differentially expressed proteins had a PVC origin and corresponded to the first three predicted open reading frames in the PVC locus (Fig. 1, top panel). To determine the structure of the final exported PVC, we purified preparations of PaPVCpnf-expressing recombinant E. coli by using TEM. Electron microscopy revealed the presence of structures highly reminiscent of R-type pyocins (15). These structures were 30 nm wide and had variable lengths; some examples were more than 800 nm long (Fig. 2). They comprised a rigid contractile outer sheath and a thin hollow inner "needle" structure (approximately 6 nm wide), analogous to R-type pyocins. Examples of both "contracted" and "relaxed" forms were seen (Fig. 2). Although the precise mode of action of R-type pyocins themselves is not known, it has been suggested that their antibacterial activity is achieved by injection of the needle into bacterial cells via contraction of the sheath (15). However, investigation of this potential mode of action for the PVC products was beyond the scope of the current study.
Insecticidal activity. As R-type pyocins are bacteriocins with well-documented antibacterial activity (15), we tested purified PVC products with a range of different bacteria. We were unable to demonstrate any activity against either E. coli or Micrococcus luteus. We therefore also tested insecticidal activity; we injected E. coli K-12 strain EC100 carrying individual PVC-containing cosmids into larvae of the wax moth. Following injection of whole cultures of PVC-expressing E. coli, larvae died within 24 h (Fig. 3A). For E. coli expressing PaPVCpnf, larvae into which cell-free supernatant alone was injected became black and moribund 15 min after injection (Fig. 3B). This injectable activity in the supernatant from PaPVCpnf-expressing E. coli was eliminated by heat treatment prior to injection (Fig. 3C). Surprisingly, a comparison of the mortality data (Fig. 3A) resulting from injection of E. coli carrying PVC products from different Photorhabdus species suggested that PVC products from the human pathogen P. asymbiotica (9) had greater toxicity for insects than PVC products derived from the insect pathogen P. luminescens TT01 (7). At this stage we are uncertain why a PVC product from a pathogen isolated from a human wound should have greater toxicity for insects than a PVC product isolated from an entomopathogenic nematode.
In vivo effect on hemocytes. Following injection of E. coli K-12 strain EC100 not expressing any bacterial virulence factors, the insect immune system normally clears the infecting E. coli via a combination of antibacterial peptide production and the phagocyotosis and nodulation reactions associated with insect hemocytes (10). Therefore, in order to understand why PVC-expressing E. coli is not successfully neutralized by the insect immune system, we bled Galleria larvae 30 min after injection with PaPVCpnf supernatants and examined their hemocytes. Simple counting of recoverable hemocytes revealed a dramatic decrease in the number of intact hemocytes 30 min after injection (Fig. 4A). Further staining of the few remaining hemocytes with TRITC-phalloidin revealed that their actin cytoskeletons were compressed in the residual cell bodies (Fig. 4B). The dramatic decrease in the number of recoverable hemocytes from infected insects as little as 30 min after injection (Fig. 4A) helps explain how PVC-expressing E. coli can persist in the presence of insect phagocytes or hemocytes.
To confirm that the putative effectors within Photorhabdus PVCs do indeed represent novel antieukaryotic effectors, we expressed several effectors directly in transfected mammalian cells and looked for rearrangements of the actin cytoskeleton and changes in nuclear morphology. Putative PVC effectors were transfected into NIH-3T3 cells using the mammalian expression vector pRK5myc (6). This vector expresses each effector as a fusion protein with Myc, which allows detection of it in transfected cells using an anti-Myc antibody (6). To assess the subsequent integrity of the actin cytoskeleton, we cotransfected the same cells with a construct expressing EGFP actin (Fig. 5). Consistent with its dramatic effect on insect hemocytes, the effector that exhibited similarity to Pnf, encoded by PaPVCpnf, dramatically reduced the cell bodies of transfected cells, leaving only a bundle of condensed actin and a disintegrating nucleus (Fig. 5C). Interestingly, a second putative effector also encoded on the PaPVCpnf element, which exhibited similarity to the C-terminal region of the SepC toxin, produced long arms of actin-rich filaments in transfected cells (Fig. 5D). Finally, a third effector, encoded by plu1690, which exhibited similarity to part of the P. luminescens cytotoxin Mcf (4), also reduced the cell bodies of transfected cells, leaving an intact nucleus surrounded by several actin-rich bundles (Fig. 5B). Together, these data suggest that the PVC-encoded effectors are capable of rearranging the actin cytoskeleton of host cells. However, it is not clear if the antihemocytic effects of PVC effector injection are related to this phenotype in mammalian tissue culture cells.
Transposon mutagenesis. In order to determine which open reading frames in an individual PVC are required for injectable insecticidal activity, we performed transposon mutagenesis of the PaPVCpnf cosmid and then examined the toxicity of the resulting mutants containing single insertions in single open reading frames. The results of the mutagenesis (Fig. 6) showed that a small central section of open reading frames, including the putative effector gene encoding a protein with similarity to the known effector Lop-T, are required for insecticidal activity.
Conclusions. Here we describe the first functional studies of phage-like PVC structures found as repetitive cassettes in two different genera of entomopathogenic bacteria, Photorhabdus (19) and Serratia (11). These prophage-like loci were described both as an antifeeding locus on the pADAP plasmid of S. entomophila (11) and as repetitive loci in the genomes of P. luminescens TT01 (19) and P. asymbiotica ATCC 43949 (19). We found that these loci encode particles whose structure is similar to the structure of R-type pycocins. However, unlike R-type pyocins, the PVC products have no demonstrable antibiotic activity, but they do have injectable activity against larvae of the wax moth. Together, these data suggest that the PVC products, together with the product of the antifeeding locus on pADAP of S. entomophila (11), may be bacteriocin-like structures modified to have activity against insect phagocytes. However, many central questions related to the biology of both the PVCs and the S. entomophila antifeeding genes remain unanswered. First, what is the relative role of phage-like genes and the putative effectors in PVC toxicity Second, are the hemocytes the primary target of the PVC products, or is their effect more widespread Third, does the mode of action of these novel structures involve the needle-like delivery mechanism inferred for R-type pyocins Finally, what is the role of these prophage-like elements in the ecology of different Photorhabdus species All of these questions about these interesting new structures remain to be investigated. However, the PVCs may well be important in conferring novel virulence either against different types of insects or, in the case of the P. asymbiotica strains isolated from human wounds (9), potentially against humans.
ACKNOWLEDGMENTS
We thank Ursula Potter of the Electron Microscopy facility at Bath for TEM pictures of PVC-expressing bacteria.
REFERENCES
Au, C., P. Dean, S. E. Reynolds, and R. H. ffrench-Constant. 2004. Effect of the insect pathogenic bacterium Photorhabdus on insect phagocytes. Cell. Microbiol. 6:89-95.
Bowen, D., T. A. Rocheleau, M. Blackburn, O. Andreev, E. Golubeva, R. Bhartia, and R. H. ffrench-Constant. 1998. Insecticidal toxins from the bacterium Photorhabdus luminescens. Science 280:2129-2132.
Boyd, E. F., and H. Brussow. 2002. Common themes among bacteriophage-encoded virulence factors and diversity among the bacteriophages involved. Trends Microbiol. 10:521-529.
Daborn, P. J., N. Waterfield, C. P. Silva, C. P. Au, S. Sharma, and R. H. ffrench-Constant. 2002. A single Photorhabdus gene, makes caterpillars floppy (mcf), allows Escherichia coli to persist within and kill insects. Proc. Natl. Acad. Sci. USA 99:10742-10747.
Dove, C. H., S.-Z. Wang, S. B. Price, C. J. Phelps, D. M. Lyerly, T. D. Wilkins, and J. L. Johnson. 1990. Molecular characterization of the Clostridium difficile toxin A gene. Infect. Immun. 58:480-488.
Dowling, A. J., P. J. Daborn, N. R. Waterfield, P. Wang, C. H. Streuli, and R. H. ffrench-Constant. 2004. The insecticidal toxin makes caterpillars floppy (Mcf) promotes apoptosis in mammalian cells. Cell. Microbiol. 6:345-353.
Duchaud, E., C. Rusniok, L. Frangeul, C. Buchrieser, A. Givaudan, S. Taourit, S. Bocs, C. Boursaux-Eude, M. Chandler, J. F. Charles, E. Dassa, R. Derose, S. Derzelle, G. Freyssinet, S. Gaudriault, C. Medigue, A. Lanois, K. Powell, P. Siguier, R. Vincent, V. Wingate, M. Zouine, P. Glaser, N. Boemare, A. Danchin, and F. Kunst. 2003. The genome sequence of the entomopathogenic bacterium Photorhabdus luminescens. Nat. Biotechnol. 21:1307-1313.
ffrench-Constant, R. H., N. Waterfield, and P. Daborn. 2004. Insecticidal toxins from Photorhabdus and Xenorhabdus. Compr. Mol. Insect Sci. 5:239-253.
Gerrard, J., N. Waterfield, R. Vohra, and R. ffrench-Constant. 2004. Human infection with Photorhabdus asymbiotica: an emerging bacterial pathogen. Microbes Infect. 6:229-237.
Gillespie, J. P., M. R. Kanost, and T. Trenczek. 1997. Biological mediators of insect immunity. Annu. Rev. Entomol. 42:611-643.
Hurst, M. R., T. R. Glare, and T. A. Jackson. 2004. Cloning Serratia entomophila antifeeding genes—a putative defective prophage active against the grass grub Costelytra zealandica. J. Bacteriol. 186:5116-5128.
Hurst, M. R., T. R. Glare, T. A. Jackson, and C. W. Ronson. 2000. Plasmid-located pathogenicity determinants of Serratia entomophila, the causal agent of amber disease of grass grub, show similarity to the insecticidal toxins of Photorhabdus luminescens. J. Bacteriol. 182:5127-5138.
Landraud, L., C. Pulcini, P. Gounon, G. Flatau, P. Boquet, and E. Lemichez. 2004. E. coli CNF1 toxin: a two-in-one system for host-cell invasion. Int. J. Med. Microbiol. 293:513-518.
Liu, D., S. Burton, T. Glancy, Z. S. Li, R. Hampton, T. Meade, and D. J. Merlo. 2003. Insect resistance conferred by 283-kDa Photorhabdus luminescens protein TcdA in Arabidopsis thaliana. Nat. Biotechnol. 21:1222-1228.
Michel-Briand, Y., and C. Baysse. 2002. The pyocins of Pseudomonas aeruginosa. Biochimie 84:499-510.
Waterfield, N., M. Hares, A. Dowling, and R. ffrench-Constant. 2005. Potentiation and cellular phenotypes of the insecticidal toxin complexes of Photorhabdus bacteria. Cell. Microbiol. 7:373-382.
Waterfield, N., S. G. Kamita, B. D. Hammock, and R. ffrench-Constant. 2005. The Photorhabdus Pir toxins are similar to a developmentally regulated insect protein but show no juvenile hormone esterase activity. FEMS Microbiol. Lett. 245:47-52.
Waterfield, N. R., P. J. Daborn, A. J. Dowling, G. Yang, M. Hares, and R. H. ffrench-Constant. 2003. The insecticidal toxin makes caterpillars floppy 2 (Mcf2) shows similarity to HrmA, an avirulence protein from a plant pathogen. FEMS Microbiol. Lett. 229:265-270.
Waterfield, N. R., P. J. Daborn, and R. H. ffrench-Constant. 2004. Insect pathogenicity islands in the insect pathogenic bacterium Photorhabdus. Physiol. Entomol. 29:240-250.
Weitao, T., S. Dasgupta, and K. Nordstrom. 2000. Role of the mukB gene in chromosome and plasmid partition in Escherichia coli. Mol. Microbiol. 38:392-400.
Zumbihl, R., M. Aepfelbacher, A. Andor, C. A. Jacobi, K. Ruckdeschel, B. Rouot, and J. Heesemann. 1999. The cytotoxin YopT of Yersinia enterocolitica induces modification and cellular redistribution of the small GTP-binding protein RhoA. J. Biol. Chem. 274:29289-29293.(G. Yang, A. J. Dowling, U)