Photorhabdus luminescens is an enterobacterium that is symbiotic with soil entomopathogenic nematodes and pathogenic to a wide range of insects. P. luminescens promotes its own transmission among susceptible insect populations using its nematode host as vector 1 . Its life cycle comprises a symbiotic stage in the nematode's gut and a virulent stage in the insect larvae, which it kills through toxemia and septicemia. After the nematode attacks a prey insect and P. luminescens is released, the bacterium produces a wide variety of virulence factors ensuring rapid insect killing. Bioconversion of the insect cadaver by exoenzymes produced by the bacteria allows the bacteria to multiply and the nematode to reproduce. During this process P. luminescens produces antibiotics to prevent invasion of the insect cadaver by bacterial or fungal competitors. Finally, elimination of competitors allows P. luminescens and the nematode to reassociate specifically before leaving the insect cadaver 2,3 .To better understand this complex life style, we determined the genome sequence of P. luminescens subspecies laumondii strain TT01 4 , a symbiont of the nematode Heterorhabditis bacteriophora isolated on Trinidad and Tobago. RESULTS General featuresStrain TT01 possesses a single circular chromosome of 5,688,987 bp with an average GC content of 42.8%. No plasmid replicon was found.A total of 4,839 protein-coding genes, including 157 pseudogenes, seven complete sets (23S, 5S and 16S) of ribosomal RNA operons and 85 tRNA genes, were predicted ( Fig. 1; Supplementary Table 1 online). Toxins against insectsMore toxin genes were predicted in the P. luminescens genome than in any other bacterial genome sequenced yet. A large number of these toxins may be involved in the killing of a wide variety of insects. Some may act synergistically or use redundancy for 'overkill' 5 , ensuring a quick death of the host. In addition, some may kill insects by interfering with their development. In the TT01 genome, two paralogs, plu4092 and plu4436, encode proteins similar to juvenile hormone esterases (JHEs) of the insect Leptinotarsa decemlineata 6 . Juvenile hormone maintains the insect in a larval state. Its inactivation by JHE allows metamorphosis to proceed. JHEs may be used to trigger the insect endocrine machinery at an inappropriate time and thus represents a promising approach for insect control 7 . These genes are located downstream of highly related orphan genes (plu4093 and plu4437), suggesting a locus duplication.The toxicity of the proteins encoded by these two loci was verified experimentally. Two Escherichia coli clones, containing the recombinant BAC1A02 and BAC8C11, were shown to be toxic toward insects. BAC1A02, which contains the locus plu4093-plu4092, exhibited substantial oral toxicity toward three mosquito species, Aedes aegypti,

The presence of specific receptors for Bacillus sphaericus binary toxin on brush-border membrane fractions (BBMF) from C u k x pipiens larvae midgut cells was demonstrated by an in vitro binding assay. Both activated and radiolabelled polypeptides from the 51-kDa and 42-kDa binary toxin of B. sphaericus 1593 specifically bound to BBMF. Bacillus sphaericus is a Gram-positive spore-forming bacterium and is one of the most promising agents for the control of mosquitoes. Some strains produce a very high larvicidal activity, especially for mosquito species belonging to the Culex and Anopheles genera, while they are almost inactive on Aedes species. The insecticidal activity is directly related to the synthesis of parasporal protein crystals during sporulation (Davidson, 1983;Payne and Davidson, 1984;Kalfon et al., 1984;Baumann et al., 1985). These crystals contain two major polypeptides of 41.9 kDa and 51.4 kDa, as deduced from DNA sequence data (Hindley and Berry, 1987;Baumann et al., 1988;Arapinis et al., 1988; for complete review see Baumann et al., 1991). Bacillus subtilis or Escherichia coli expressing either of the genes encoding these two proteins are poorly toxic or even non-toxic for mosquito larvae (Baumann et al., 1988;de la Torre et al., 1989;Broadwell et al., 1990). However, transformants expressing both genes are fully toxic, indicating that the protein crystal of B. sphaericus acts as a binary toxin (Broadwell et al., 1990;Baumann and Baumann, 1991).Little is known about the mode of action of these toxins, except that they appear to exert cytopathological effects (Davidson, 1981;Karch and Coz, 1983; Charles, 1987) by binding to susceptible mosquito larval midgut cells (Davidson et al., 1987;Davidson, 1988Davidson, ,1989. In addition, both proteins bind to the gastric caecae and posterior midgut of Culex larvae when administered together (Davidson et al., 1990).In this report, we demonstrate that B. sphaericus binary toxin specifically binds to susceptible mosquito larval midgut membranes, and we measure the affinity of the toxin for its putative membrane receptor. MATERIALS AND METHODS Insects and toxicity assaysLarvae of Culexpipiens ssp. pipiens L. (Montpellier strain) and Aedes aegypti L. (Bora-Bora strain) were reared on cat biscuits at 28 "C. Adult populations were maintained on honey solution, at 2 8 T , in 80% relative humidity with a 16 hi8 h photoperiod. Females were fed on guinea pigs.Toxicity assays were performed on early fourth-instar larvae, in 5-cm diameter Petri dishes containing 10 ml bacterial or crystal suspension, or soluble protein dilutions in demineralized water. Dead larvae were counted after 24 h and 48 h, and lethal concentrations killing 50% of the population (LC,,) evaluated by probit analysis (Finney, 1971). Brush-border membrane isolationBrush-border membrane fractions (BBMF) were prepared from midguts of C. pipiens and A . aegypti larvae following methods modified from those of Biber et al. (1981; described in Wolfersberger et al., 1987) and Houk et al. (1986). Fourthi...

No abstract

The mosquitocidal activity of Bacillus sphaericus is because of a binary toxin (Bin), which binds to Culex pipiens maltase 1 (Cpm1), an ␣-glucosidase present in the midgut of Culex pipiens larvae. In this work, we studied the molecular basis of the resistance to Bin developed by a strain (GEO) of C. pipiens. Immunohistochemical and in situ hybridization experiments showed that Cpm1 was undetectable in the midgut of GEO larvae, although the gene was correctly transcribed. The sequence of the cpm1GEO cDNA differs from the sequence we previously reported for a susceptible strain (cpm1IP) by seven mutations: six missense mutations and a mutation leading to the premature termination of translation. When produced in insect cells, Cpm1IP was attached to the membrane by a glycosylphosphatidylinositol (GPI). In contrast, the premature termination of translation of Cpm1GEO resulted in the targeting of the protein to the extracellular compartment because of truncation of the GPI-anchoring site. The interaction between Bin and Cpm1GEO and the enzyme activity of the receptor were not affected. Thus, Bin is not toxic to GEO larvae because it cannot interact with the midgut cell membrane, even though its receptor site is unaffected. This mechanism contrasts with other known resistance mechanisms in which point mutations decrease the affinity of binding between the receptor and the toxin. E nvironmentally safe toxins produced by Bacillus thuringiensis and͞or Bacillus sphaericus have been integrated in management programs to control crop pests such as Heliothis virescens and Plutella xylostella, and disease vectors such as the mosquitoes Anopheles gambiae and Culex pipiens (1, 2). However, the potential benefits of these biopesticides may be rapidly lost because of the proliferation of highly resistant insect populations (3-6). Control strategies to delay or prevent the development of resistance have been developed, based on several assumptions. The most important of these assumptions are that the resistance gene is recessive and that the rate of mutation to generate resistance alleles is low. Currently, it is difficult to evaluate the success of these strategies because we lack adequate methods for monitoring resistance alleles because of our very restricted knowledge of the mechanisms of resistance to bioinsecticides. Bacillus sphaericus is toxic to mosquitoes, mainly because it produces a binary toxin (Bin) in crystals during sporulation. Following the ingestion and solubilization of crystals by larvae, the released toxin is activated and interacts with the brushborder membrane of the midgut epithelium. In a previous study, we reported the partial purification of a Bin-binding protein from IP, a susceptible strain of C. pipiens. This receptor displayed sequence similarity to ␣-glucosidases and other maltase-like proteins, and was thus named Cpm1, for Culex pipiens maltase 1 (7). We recently isolated the cDNA encoding Cpm1 from IP larvae (cpm1 IP ) and showed that Cpm1 has ␣-glucosidase activity when produced in bacteria (8). ...