PlcR is a pleiotropic regulator of virulence factors in the insect pathogen Bacillus thuringiensis and in the opportunistic human pathogen Bacillus cereus. It activates the transcription of at least 15 genes encoding extracellular proteins, including phospholipases C, proteases and enterotoxins. Expression of the plcR gene is autoregulated and activated at the onset of stationary phase. Here, we used mini-Tn10 transposition to generate a library of B. thuringiensis mutants, with the goal of characterizing genes involved in the expression of the plcR gene. Three mutant strains were identified carrying distinct mini-Tn10 insertions. The mutations impaired plcR expression and caused a deficient haemolytic phenotype, similar to the phenotype of a B. thuringiensis strain in which the plcR gene had been disrupted. The insertion sites of the three mini-Tn10 transposons mapped in a five-gene operon encoding polypeptides homologous to the components of the oligopeptide permease (Opp) system of Bacillus subtilis, and with a similar structural organization. By analogy, the five B. thuringiensis genes were designated oppA, B, C, D and F. In vitro disruption of the B. thuringiensis oppB gene reproduced the effect of the mini-Tn10 insertions (i.e. the loss of haemolytic activity) and reduced the virulence of the strain against insects. These phenotypes are similar to those of a ΔplcR mutant. Opp is required for the import of small peptides into the cell. Therefore, plcR expression might be activated at the onset of stationary phase by the uptake of a signalling peptide acting as a quorum-sensing effector. The opp mutations impaired the sporulation efficiency of B. thuringiensis when the cells were cultured in LB medium. Thus, Opp is on the pathway that ultimately regulates Spo0A phosphorylation, as is the case in B. subtilis. However, analysis of plcR expression in ΔoppB, Δspo0A and ΔoppBΔspo0A mutants indicates that Opp is required for plcR expression via a Spo0A-independent mechanism.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The Gram-positive spore-forming bacterium Bacillus thuringiensis belongs to the diverse Bacillus cereus group. B. cereus is an opportunistic pathogen causing food-borne gastroenteritis resulting from the production of an emetic toxin (Agata et al., 1995) or of enterotoxins, such as Hbl or Nhe (Drobniewski, 1993; Granum and Lund, 1997). In some rare circumstances, B. cereus has been found associated with more severe infections, for example pneumonia and endophthalmitis (Beecher et al., 1995; Miller et al., 1997). B. thuringiensis is well known for its entomopathogenic properties, which are partly caused by the production of a large variety of crystal proteins (Cry toxins) specifically active against insect larvae (Schnepf et al., 1998). However, a contribution of the B. thuringiensis spores to overall virulence has been reported (Li et al., 1987; Dubois and Dean, 1995; Johnson and McGaughey, 1996). It is generally assumed that the toxic effect of the Cry proteins either kills or weakens the susceptible insects, creating favourable conditions for the germination of spores and the development of septicaemia. The opportunistic properties of B. thuringiensis have been demonstrated in mice by infection via nasal instillation of spores (Hernandez et al., 1998; 1999).
PlcR was first identified as a positive regulator that transiently activates transcription of the plcA gene in B. thuringiensis at the onset of stationary phase (Lereclus et al., 1996). plcA encodes a phosphatidylinositol-specific phospholipase C. A genetic screen led to the detection of several B. thuringiensis genes regulated by PlcR, thus indicating that PlcR is a pleiotropic regulator that positively regulates the transcription of various genes encoding extracellular virulence factors, including phospholipases C (PlcA and PlcB), enterotoxins (Hbl and Nhe) and proteases (Agaisse et al., 1999). The plcR gene is also functional in B. cereus, and sequence analysis of DNA regions surrounding plcA and plcR identified three new PlcR-regulated genes in this bacterium (Økstad et al., 1999). These genes encode two proteases and a cell wall hydrolase. PlcR-regulated promoters contain a highly conserved palindromic region (the PlcR box), which is probably the specific recognition site for PlcR activation (Agaisse et al., 1999). Thus, PlcR controls the expression of a large regulon comprising at least 15 genes encoding exported proteins, some of which probably contribute to the overall pathogenicity of B. thuringiensis and B. cereus. Indeed, recent work has shown that the opportunistic pathogenicity of B. cereus and B. thuringiensis is caused, at least in part, by the plcR regulon (Salamitou et al., 2000). The disruption of the plcR gene greatly reduces the haemolytic activity of B. thuringiensis and B. cereus and results in a loss of pathogenicity against both insect larvae and mice. The haemolytic-negative and avirulent phenotypes might result, at least in part, from the non-expression of two PlcR-regulated genes: hblC, the first gene of the hbl operon; and plcB, the first gene of the cereolysin AB operon (Agaisse et al., 1999). The products of these two genes have haemolytic properties (Gilmore et al., 1989; Lindbäck et al., 1999) and are involved in the virulence of B. cereus endophthalmitis (Beecher et al., 2000).
PlcR activates its own expression (Lereclus et al., 1996). However, the burst of plcR expression at the onset of the stationary phase implies that, in addition to its autoregulation, one or several other factors are required to trigger plcR transcription. In B. thuringiensis, plcR is very strongly expressed in a spo0A mutant, and plcR transcription is abolished in cells grown in a sporulation-specific medium (Lereclus et al., 2000). Analysis of the plcR promoter region revealed the presence of two Spo0A binding sites, one on either side of the PlcR box. DNase I protection assays, using the Bacillus subtilis Spo0A∼P protein, showed that the two sequences (TGCCGAA and TGTCGAA) were protected (Lereclus et al., 2000). Therefore, it is likely that Spo0A∼P acts as a repressor of plcR transcription. To detect genes involved in the activation of plcR transcription, we performed a transposon mutagenesis in a B. thuringiensis strain carrying a chromosomal plcA′–lacZ transcriptional fusion as reporter of plcR expression. Mutations in genes of the oligopeptide permease (Opp) system abolished the expression of the plcR regulon and reduced the virulence of B. thuringiensis against insect larvae. Opp and an extracellular signalling peptide might be involved in the regulation of plcR expression.
Isolation of insertion mutations impairing expression of phospholipase and haemolysin
To identify B. thuringiensis genes involved in the regulation of plcR expression, we selected mutants with reduced plcA transcription. Direct selection of mutants with reduced plcR transcription would be difficult because of the low level of plcR expression and its autoactivation (Lereclus et al., 1996). Therefore, a B. thuringiensis strain (407 Cry−[plcA′Z]) carrying a chromosomal transcriptional fusion between the plcA promoter region and the lacZ gene was constructed. The fusion was used as reporter to screen a library of mini-Tn10 insertional mutants (see Experimental procedures). Three independent mini-Tn10 insertion mutants were selected that were Lac− on LB agar plates containing Xgal. These mutants, designated M2, M21 and M35, did not produce detectable amounts of β-galactosidase when grown in LB medium: 1 h after the onset of the stationary phase, the β-galactosidase activity of the mutant strains is < 10 U mg−1 protein versus 600 U mg−1 protein for the parental strain 407 Cry−[plcA′Z].
We checked that the three Tn10 insertions affected plcR expression and did not directly affect the expression of the plcA′–lacZ fusion using agar plates containing sheep erythrocytes. The B. thuringiensis strain 407 Cry−[plcA′Z] grown on sheep blood agar produced intense halos resulting from degradative activity (Fig. 1). In contrast, the haemolytic activity of the mutant colonies, such as that of the 407 Cry−ΔplcR mutant strain, was markedly reduced, although not totally abolished. This is consistent with the mutations affecting genes involved in the expression of plcR or in the activity of its product. Chromosomal DNA from mutant strains M2, M21 and M35 was amplified by polymerase chain reaction (PCR) with primers (PLCR1 and PLCR2) complementary to sequences flanking the plcR gene. The amplification products had a size (1.7 kb) corresponding to the DNA sequence containing the plcR gene of the wild-type B. thuringiensis strain 407 Cry−, thus indicating that the Tn10 had not inserted within the plcR coding sequence or its promoter region.
Characterization of the mini-Tn10 insertion sites
The mini-Tn10 element contains a pUC replicon allowing recovery of the chromosomal DNA sequences flanking the insertion site. Chromosomal DNA from mutants M2, M21 and M35 was digested with HindIII or EcoRI (two restriction enzymes for which sites are not present in the mini-Tn10 element) and treated with T4 DNA ligase. The ligation mixture was used to transform Escherichia coli cells, and transformants were selected for resistance to spectinomycin. The recombinant plasmids isolated from the SpecR transformants were used as templates to determine the nucleotide sequence of the ends of the mini-Tn10 and the chromosomal DNA adjacent to the insertion sites (see Experimental procedures and Fig. 2). The Tn10 insertion site in mutant M2 mapped in a potential ribosome binding site located at an appropriate distance upstream from an ATG starting an open reading frame (ORF). In mutants M21 and M35, the Tn10 insertion sites mapped in ORFs. The amino acid sequences deduced from the three ORFs are similar to components of oligopeptide permease (Opp) systems in both Gram-positive and Gram-negative bacteria (not shown). These preliminary results suggest that the three Tn10 insertions mapped in three different genes of a B. thuringiensis opp operon.
Sequence analysis of the B. thuringiensis opp operon
The DNA regions surrounding the Tn10 insertion sites in mutants M2, M21 and M35 were sequenced directly from the chromosomal DNA of the B. thuringiensis strain 407 Cry−. The nucleotide sequence of a 9364 bp contiguous DNA fragment was thus determined. It contains seven putative ORFs (Fig. 2). Five genes, located in the central part of the DNA fragment, are transcribed in the same direction and are separated by short intercistronic regions. The largest intercistronic region is 130 bp and lies between the first and the second gene; it contains an inverted repeat that can potentially form a stem–loop structure. Two distal ORFs transcribed in the opposite orientation flank this five-gene operon.
The five central ORFs present similarities to Opp components of various Gram-positive and Gram-negative organisms (not shown), and the greatest similarity was with the proteins encoded by the opp operon of B. subtilis (Perego et al., 1991; Rudner et al., 1991). The general structure of the operon (including the order of the five genes) is identical in the two Bacillus species. The deduced amino acid sequences of the five B. thuringiensis ORFs were similar to the OppA, B, C, D and F proteins of B. subtilis. As for the Opp components of Salmonella typhimurium and B. subtilis (Perego et al., 1991), the least well-conserved protein between B. thuringiensis and B. subtilis is OppA (27% identity), and the best conserved are OppD and OppF (70% identity for both proteins). The conservation for OppB and OppC proteins in the two species is 50% and 49% respectively. OppD and F contain motifs characteristic of nucleotide-binding proteins. The functions of the various components of the Opp system are presumably similar in B. thuringiensis and B. subtilis. OppA is a lipoprotein anchored to the cell membrane and serving as initial receptor for peptide transport. OppB and C are hydrophobic membrane-spanning proteins, and OppD and F are intracellular ATPase subunits associated with the membrane (Perego et al., 1991).
The short truncated ORF upstream from the opp operon encodes the amino-terminal part of a polypeptide with 73% identity with the TrpS protein of B. subtilis (Kunst et al., 1997). The corresponding gene maps to an identical localization in the chromosome of B. subtilis. In contrast, the gene downstream from oppF showed similarity to the B. subtilis gene yisQ, about 60 kb upstream from the B. subtilis opp operon. Nevertheless, the structural organization of the opp genes is highly conserved between the two Bacillus species. Therefore, the same nomenclature was used to designate these genes. The Tn10 mutations selected for their effect on the expression of the plcA′–lacZ transcriptional fusion mapped in the oppB gene (M2), in the oppD gene (M21) and in the oppF gene (M35).
Transcriptional analysis of the B. thuringiensis opp operon
To map the opp promoter(s), the transcriptional activity of various DNA regions of the putative operon was assayed by transcriptional fusion to the lacZ reporter gene on the low-copy-number plasmid pHT304-18′Z (Agaisse and Lereclus, 1994). PCR-amplified DNA fragments containing sequences upstream from oppA, oppC and oppF were inserted upstream from the lacZ gene to give pHToppA′Z, pHToppB′Z and pHToppF′Z respectively (Fig. 2). These plasmids were introduced into B. thuringiensis by electroporation, and β-galactosidase was assayed at various stages of growth in LB medium at 30°C. The β-galactosidase activity of cells harbouring pHToppA′Z was high during the exponential and stationary growth phases (1800 Miller units at t−1 and 5000 Miller units at t1). In contrast, the β-galactosidase activity was very low at all times (< 10 Miller units at t1) in strains harbouring pHToppB′Z and pHToppF′Z. This indicates that the DNA fragments inserted into pHToppB′Z and pHToppF′Z expressed no promoter activity. Presumably, there is a single promoter upstream from oppA, and the five opp genes are transcribed as a single transcription unit. The stem–loop structure between the oppA and oppB genes is probably not a transcription terminator, but might act as a 3′ mRNA stabilizer, as described previously for the stem–loop structure in the intercistronic region of the oppA–oppB genes from S. typhimurium (Hiles et al., 1987).
Primer extension analysis using oligonucleotide Popp and RNA extracted from B. thuringiensis 407 cells grown in LB medium (Fig. 3A) identified the opp transcriptional start site at nucleotide position −83 with reference to the OppA translational start codon (Fig. 3B). The putative −35 and −10 boxes of the opp promoter resemble the consensus promoter recognized by RNA polymerase containing the major sigma A factor (Moran, 1993).
Disruption of the oppB gene results in a haemolytic-negative phenotype
To confirm that the effect of the Tn10 insertions resulted from single mutations in the opp locus, the disruption of the oppB gene was reproduced by homologous recombination. Owing to the low efficiency of transformation and DNA recombination in B. thuringiensis, gene replacement by homologous recombination is not possible with a non-replicative vector and requires the use of a thermosensitive plasmid (pRN5101) as intermediate vector (for the construction, see Experimental procedures). A TetR cassette was used to inactivate oppB and to select the recombinant clones; the inducible xylA promoter and the xylR repressor gene of B. subtilis were used to prevent a potential polar effect of the disruption on the transcription of the downstream opp genes (Fig. 4A). The correct integration of TetR and xylR–pxylA in the recombinant B. thuringiensis 407 Cry−ΔoppB strain was checked by PCR with primers (B5 and B6) external to the oppB fragments manipulated during the construction (not shown).
The effect of the oppB disruption was assayed first on sheep blood agar plates (Fig. 4B). The haemolytic activity of the strain 407 Cry−ΔoppB was very low and similar to that of the Tn10 insertion mutants (Fig. 1). Introduction of pHT304-oppB in the ΔoppB strain restored the haemolytic activity. Therefore, the phenotypes of the mutants M2, M21 and M35 resulted from the single insertions into the opp locus.
Mutations in oppB prevent plcR expression
To determine whether the mutations in the opp operon affect the expression of the plcR gene, a plasmid carrying the transcriptional fusion between the plcR promoter and the lacZ gene (pHT304BΩplcR′–lacZ) was introduced into the wild-type 407 Cry− strain and also into the various opp mutants of B. thuringiensis. The transformant strains were grown in LB medium, and β-galactosidase activity was assayed at various stages of growth between t−1 and t3. In the wild-type 407 Cry− strain, plcR transcription is activated at the end of the exponential growth phase, and plcR-directed β-galactosidase expression increases until t1(Fig. 5A) (Lereclus et al., 1996). In the three mutant strains (407 Cry−ΔoppB, M21 and M35), no transcriptional activation was detected, and β-galactosidase activity was very low. In the B. thuringiensis mutant strain 407 Cry−ΔoppB, the downstream opp genes are presumably transcribed from the xylose-induced xylA promoter and, in the M35 mutant, the oppF gene alone should be affected by the Tn10 insertion. Therefore, the oppB and oppF genes are required for the activation of plcR transcription.
The role of Opp in sporulation depends on the growth medium
The opp operon is involved in sporulation in B. subtilis (Perego et al., 1991; Rudner et al., 1991). Mutations in opp genes resulted in a 10- to 100-fold decrease in the sporulation efficiency of B. subtilis cells grown in a sporulation-specific medium. To determine whether a similar effect was observed in B. thuringiensis, the wild-type and mutant strains 407 Cry−ΔoppB, M21 and M35 were grown in HCT, a sporulation-specific medium for B. thuringiensis. Viable cells were counted at the onset of stationary phase, and the rate of sporulation was determined after 3 days (72 h) of growth at 30°C. Surprisingly, none of the opp mutations conferred a sporulation-defective phenotype (Table 1): a large proportion of cells sporulated and gave heat-resistant spores.
Table 1. Sporulation efficiency of opp mutants.
B. thuringiensis strains
Viable counts ml−1 at the onset of stationary phase
a. The percentages were calculated as 100 × the ratio between heat-resistant spores ml−1 and viable cells ml−1. The viable cells were counted at the onset of the stationary phase for the bacteria grown in LB medium and after 3 days for the bacteria grown in HCT medium.
2.8 × 108
5 × 108
3.7 × 108
1.6 × 108
5.8 × 108
2.2 × 108
407 Cry− M2
3.5 × 108
5.3 × 108
4 × 108
407 Cry− M21
2.7 × 108
4.5 × 108
3.8 × 108
407 Cry− M35
4.2 × 108
4.9 × 108
4.3 × 108
3 × 108
3 × 108
2 × 108
5.3 × 108
3.5 × 107
2.6 × 106
407 Cry− M2
4 × 108
3.4 × 107
1.9 × 106
407 Cry− M21
3.6 × 108
4.1 × 107
2.2 × 106
407 Cry− M35
5.3 × 108
6.2 × 107
2.7 × 106
A similar experiment was performed in LB medium in which B. thuringiensis, unlike B. subtilis, is able to sporulate efficiently. In these growth conditions, the sporulation efficiency of each of the three opp mutations was 1% that of the wild type (Table 1). The viability of the opp mutants was also reduced after 3 days of growth in LB medium. This might result from the lysis of a proportion of the cells that did not enter the sporulation process.
Mutation in oppB prevents plcR expression via a Spo0A-independent mechanism
It has been shown previously that Spo0A is involved in the regulation of plcR expression: plcR transcription is very strongly increased in a Δspo0A mutant and is abolished when the cells are grown in HCT medium (Lereclus et al., 2000). Thus, the role of Opp in plcR expression may be a consequence of its role in sporulation (for example, by allowing the uptake of a small peptide affecting the phosphorylation of Spo0A). To determine whether Opp controls plcR expression via a Spo0A-dependent mechanism, plcR expression was assayed in ΔoppB, Δspo0A and ΔoppBΔspo0A mutant strains carrying pHT304BΩplcR′–lacZ(Fig. 5B). β-Galactosidase production was higher in the Δspo0A mutant strain than in the wild-type strain. However, the plcR-directed β-galactosidase expression was abolished in both ΔoppB and ΔoppBΔspo0A mutant strains. Thus, the effect of the ΔoppB mutation on plcR expression is not reversed by the deletion of the spo0A gene. This result indicates that Opp is required for plcR expression via a Spo0A-independent mechanism.
Disruption of the oppB gene affects the virulence of B. thuringiensis against insects
The virulence of B. thuringiensis against insect larvae was determined as the synergistic effect of the spores on the insecticidal activity of the crystal proteins (Salamitou et al., 2000). To assess the pathogenicity of B. thuringiensis strains against insects, spores from wild-type or mutant strains were fed to the larvae in association with the insecticidal toxin Cry1C. The Lepidopteran species Galleria mellonella is a useful model because the larvae are susceptible to the ingestion of B. thuringiensis spore–crystal mixtures, but are only weakly susceptible to the ingestion of crystals alone (Li et al., 1987). The insecticidal toxin Cry1C was fed to the larvae of G. mellonella at a dose (3 µg of toxin per larva) with a low mortality rate. Inclusion of spores from the wild-type 407 Cry− strain resulted in a significant increase in mortality, demonstrating synergy. In contrast, the addition of spores from the opp or plcR mutants had no synergistic or even cumulative effect (Table 2).
Table 2. Effect of opp and plcR mutations on the virulence of B. thuringiensis in insect larvae (G. mellonella).
Three mini-Tn10 insertion mutants were obtained with affected expression of the plcR gene. We identified the insertion site of the mini-Tn10 transposons, and the nucleotide sequence of the flanking DNA regions was determined. The insertions mapped in a five-gene operon encoding polypeptides homologous to the components of the oligopeptide permease (Opp) system of B. subtilis. The structural organization of the B. thuringiensis opp operon is very similar to that of the B. subtilis and S. typhimurium opp operons. By analogy, the five genes were designated oppA, B, C, D and F. In vitro disruption of the B. thuringiensis oppB gene reproduced the effect of the mini-Tn10 insertions (i.e. the loss of haemolytic activity), and we showed that the different opp mutants did not express a plcR′–lacZ transcriptional fusion in B. thuringiensis. Thus, Opp is required for plcR expression and, consequently, for expression of the plcR regulon.
Opp belongs to the large family of ATP-binding cassette (ABC) transporters and mediates the uptake of oligopeptides of three to six amino acids. In B. subtilis, the Opp system imports extracellular peptides that act as signals to control the development of genetic competence and sporulation (Lazazzera and Grossman, 1998). Presumably, the Opp system of B. thuringiensis allows the uptake of a signalling peptide activating plcR expression. However, it is not possible to infer whether this peptide controls plcR expression directly or indirectly, or at which level (transcriptional or post-transcriptional) the control is exerted. The fact that PlcR activates its own transcription (Lereclus et al., 1996) and that plcR is not expressed in the B. thuringiensis opp mutants at the onset of stationary phase should suggest that PlcR is most probably inactive and requires a molecule imported by Opp in order to be activated.
In B. subtilis, the role of Opp in sporulation results from the uptake of a peptide (PhrA) that inactivates the RapA phosphatase responsible for the dephosphorylation of Spo0F∼P (Perego and Hoch, 1996; Perego, 1997). opp mutations that block the uptake of the PhrA peptide result in a decrease in the concentration of Spo0A∼P, thus preventing sporulation. Here, we show that the role of Opp in sporulation in B. thuringiensis depends on the growth medium: opp mutations significantly reduced the sporulation efficiency in LB medium, but not in a sporulation-specific medium. This differential effect was not observed with B. subtilis cells (Perego et al., 1991), but note that B. subtilis wild-type strains do not sporulate efficiently in LB medium. The B. subtilis Rap phosphatases were compared with the available DNA sequences from Bacillus anthracis, a bacterium closely related to B. thuringiensis (Helgason et al., 2000). Preliminary sequence data for B. anthracis was obtained from The Institute for Genomic Research website at http://www.tigr.org. blast searching reveals the presence of three putative B. anthracis proteins sharing similarity to the Rap proteins (23–27% identity and 46–49% similarity). In addition, ORFs potentially encoding polypeptides of 43–63 amino acids are located downstream from the coding sequences of the three putative proteins. This is reminiscent of the structural organization of the Rap/Phr polypeptides in B. subtilis (Perego et al., 1996). As for the B. subtilis Phr peptides, the three small polypeptides of B. anthracis show a putative signal peptide cleavage site, suggesting that they are exported proteins (Nielsen et al., 1997).
Thus, the regulation of the phosphorelay in B. thuringiensis might be dependent on the uptake of a polypeptide inactivating a phosphatase, as in B. subtilis. The role of Opp in this regulation suggests that the lack of plcR expression in B. thuringiensis opp mutants might be an indirect consequence of the decrease in the concentration of Spo0A∼P in the cell. However, plcR expression is abolished in a sporulation-specific medium (HCT) and is greatly increased in a spo0A mutant of B. thuringiensis, suggesting that plcR transcription is repressed by Spo0A∼P (Lereclus et al., 2000). Therefore, it is unlikely that Opp is involved in plcR expression via the phosphorelay. opp mutations would result in a reduction in the concentration of Spo0A∼P and, consequently, in an increase in plcR expression. This view is reinforced by the fact that the plcR′–lacZ fusion is not expressed in a double Δspo0AΔoppB mutant. Thus, the deletion of the spo0A gene is not able to reverse the effect of the ΔoppB mutation on plcR expression.
The B. thuringiensis opp genes are transcribed as a single transcription unit during exponential growth and the stationary phase. This is similar to the B. subtilis opp operon (Koide et al., 1999). As opp genes are expressed during exponential growth, the triggering of plcR transcription at the onset of stationary phase cannot result from the appearance of the Opp system in the cell. Activation of plcR transcription might depend on the accumulation of an extracellular peptide acting as a quorum-sensing effector. In B. subtilis, it has been shown that phrA mutants (unable to produce the signalling peptide and thus to sporulate) were complemented by a culture supernatant from an opp mutant, but not by a culture supernatant from a wild-type strain (Perego et al., 1996). This result indicated that the PhrA peptide accumulated in the medium to a sufficient level only in the opp mutant. However, the resuspension of B. thuringiensis wild-type cells (harvested in exponential growth, at t−1) in a conditioned medium (10× stationary phase supernatant from the B. thuringiensis 407 wild-type strain or stationary phase supernatant from the ΔoppB mutant strain, collected at t1) did not advance the onset of plcR transcription (unpublished data). This suggests that the uptake of a signalling peptide is not sufficient to activate plcR expression. To settle this question, the nature and origin of the putative signalling peptide need to be determined. Preliminary results, including those published previously (Lereclus et al., 1996), suggest that a short ORF (designated orf2) located downstream from plcR is involved in the expression of the plcA gene. Studies are in progress to determine whether orf2 plays a role in the regulation of plcR expression, for example via the production of the putative signalling peptide described here.
The plcR regulon is involved in the opportunist infection of insects and mice by B. thuringiensis (Salamitou et al., 2000). In insect larvae, the opportunistic properties of B. thuringiensis are evaluated as the synergistic effect of the spores on the activity of the crystal proteins. The precise cause of this PlcR-dependent effect is not known. However, one or several PlcR-regulated factors might be required, in association with crystal proteins, to cause the death of the larvae or to create favourable conditions for bacterial multiplication, resulting in the death of the larvae from septicaemia (Salamitou et al., 2000). Here, we show that disruption of the oppB gene results in an avirulent phenotype in insects, similar to the ΔplcR mutant strain. In view of the role of the Opp system in plcR expression, this result was expected. However, it shows, indirectly, that the requirement of Opp for plcR expression also applies to in vivo conditions. Our findings are consistent with the uptake of a signalling peptide controlling the virulence of B. thuringiensis (and B. cereus) via the production of the various PlcR-regulated virulence factors.
Bacterial strains and growth conditions
The acrystalliferous strain B. thuringiensis 407 Cry− belonging to serotype 1 (Lereclus et al., 1989) was used throughout this study. The B. thuringiensis 407 Cry−Δspo0A and ΔplcR mutant strains have been described previously (Lereclus et al., 1995; Salamitou et al., 2000). Escherichia coli K-12 strains TG1 [Δ(lac-proAB) supE thi hsdD5 (F ′ traD36 proA+proB+lacIqlacZΔM15)] (Gibson, 1984) and MC1061 [hsdR mcrB araD139 D(araABC-leu)7679 ΔlacX74 galU galK rpsL thi] (Meissner et al., 1987) were used as hosts for the construction of plasmids and cloning experiments. Plasmid DNA used to electrotransform B. thuringiensis was prepared from E. coli strain SCS110 [rpsL (Strr) thr leu endA thi-1 lacY galK galT ara tonA tsx dam dcm supE44Δ(lac-proAB) (F′traD36 proAB lacIqZÆM15)] (Stratagene). E. coli and B. thuringiensis cells were transformed by electroporation as described previously (Dower et al., 1988; Lereclus et al., 1989). E. coli strains were grown at 37°C in Luria broth (LB; 1% tryptone, 0.5% yeast extract, 0.5% NaCl). B. thuringiensis was grown at 30°C in LB or in HCT, a sporulation-specific medium (Lecadet et al., 1980).
Antibiotic concentrations for bacterial selection were as follows: ampicillin, 100 µg ml−1 (for E. coli ); spectinomycin 60 µg ml−1 (for E. coli ) and 250 µg ml−1 (for B. thuringiensis); erythromycin, 2–5 µg ml−1, kanamycin, 200 µg ml−1 and tetracycline, 15 µg ml−1 (for B. thuringiensis). Bacteria with the Lac+ phenotype were identified on LB plates containing Xgal. Columbia medium agar plates (BioMérieux) containing 5% sheep blood were used to evaluate the haemolytic activity of the B. thuringiensis strains. Activation of the xylA promoter in B. thuringiensis was induced by including xylose (20 mM final concentration) in the culture medium.
The sporulation efficiency of the B. thuringiensis strains was determined in LB and HCT medium. The number of viable cells at the onset of the stationary phase was counted as total colony-forming units (cfus) on LB plates. The number of spores was determined as heat-resistant (80°C for 12 min) cfus on LB plates.
Plasmid DNA was extracted from E. coli by a standard alkaline lysis procedure using Qiagen kits. Chromosomal DNA was extracted from B. thuringiensis cells harvested in mid-log exponential phase and purified as described previously (Msadek et al., 1990). Restriction enzymes and T4 DNA ligase were used as recommended by the manufacturers. The oligonucleotide primers used for PCR amplifications (Table 3) were synthesized by Genset. PCRs were performed with a GeneAmp PCR system 2400 thermal cycler (Perkin-Elmer).
Table 3. Primers used in this study.
Restriction site (underlined)
Subregions of the opp operon were amplified from chromosomal DNA of strain B. thuringiensis 407 Cry− by PCR, and the amplified regions were cloned between the HindIII and BamHI sites of the pHT304-18Z plasmid (Agaisse and Lereclus, 1994) to generate transcriptional fusions with the lacZ gene. The DNA fragments cloned in pHToppA′Z, pHToppB′Z and pHToppF′Z were amplified with the primers opp1 and opp2, opp3 and opp4, opp5 and opp6 respectively (Table 3).
The plasmid pHTS2xyl, carrying the tet gene, the xylR repressor gene and the promoter of the xylA gene, was constructed as follows. The tet gene, conferring resistance to tetracycline, was amplified by PCR from the B. cereus plasmid pBC16 (Polak and Novick, 1982; Tang and Wilson, 1996) using primers Tet1 and Tet2 (Table 3). The primer Tet1 creates a SphI site in addition to the StuI site. The amplified 1.6 kb DNA fragment was digested with StuI and ligated into the SmaI site of the vector pKS− (Stratagene) to give pHTS2. The promoter of the xylA gene and the xylR gene, encoding the transcriptional repressor of xylA, were amplified by PCR from chromosomal DNA of the B. subtilis strain 168, using primers Xyl1 and Xyl2 (Table 3). The amplified 1490 bp DNA fragment was ligated between the SphI and XbaI sites of pHTS2 to give pHTS2xyl.
pHT304BΩplcR′–lacZ, carrying a transcriptional fusion between the promoter region of the plcR gene and the lacZ gene, has been described previously (Lereclus et al., 1996).
A 1.2 kb BamHI–HindIII fragment containing the oppB gene was amplified by PCR from the chromosomal DNA of B. thuringiensis 407 Cry−, using primers B1 and B4. pHT304-18ΩoppB was constructed by inserting the BamHI–HindIII fragment carrying oppB between the BamHI and HindIII sites of pHT304-18. Thus, transcription of the oppB gene is directed from the lacZ promoter region of pHT304-18 (Agaisse and Lereclus, 1994).
Construction of the B. thuringiensis recombinant strains
The plcA gene encoding the phosphatidylinositol-specific phospholipase C of the B. thuringiensis strain 407 Cry− was disrupted by inserting a promoterless lacZ gene into the 5′ part of the coding sequence. A DNA region just upstream from the plcA coding sequence and an internal region of plcA were amplified by PCR using primers PLCA1 and PLCA2, and PLCA3 and PLCA4 respectively (Table 3). The sequences of these primers were selected from the published nucleotide sequence of the B. thuringiensis plcA gene (Lechner et al., 1989). The amplified DNA fragments were digested with appropriate restriction enzymes and inserted separately into pUC18. The promoterless lacZ gene was purified as a 3.2 kb XbaI–EcoRI DNA fragment from pHT304-18Z (Agaisse and Lereclus, 1994). The upstream and internal parts of the plcA gene were purified as HindIII–XbaI and EcoRI–BamHI fragments, respectively, and ligated with the lacZ gene between the HindIII and BamHI sites of the thermosensitive plasmid pRN5101 (Villafane et al., 1987; Lereclus et al., 1992). The ligation mixture was then used to transform E. coli cells to ampicillin resistance, and the plasmid isolated from the transformants was verified by restriction mapping. This recombinant plasmid was introduced into B. thuringiensis strain 407 Cry− by electroporation. Transformants were resistant to erythromycin and had a blue phenotype (Lac+) on LB plates containing Xgal (40 µg ml−1). The chromosomal wild-type copy of plcA was replaced with the disrupted copy by homologous recombination as described previously (Lereclus et al., 1995). In the resulting B. thuringiensis recombinant strain, the lacZ gene was transcribed from the plcA promoter and was thus controlled by the transcriptional activator PlcR (Lereclus et al., 1996): it was designated 407 Cry−[plcA′Z], and was Lac+ and sensitive to erythromycin.
To disrupt the oppB gene, the 5′ and 3′ regions of the gene were amplified by PCR, using primers B1 and B2, and B3 and B4. The 5′ end was purified as a BamHI–EcoRI fragment and the 3′ end as a XbaI–HindIII fragment. A 3.1 kb EcoRI–XbaI fragment containing the tet gene, the xylR repressor gene and the xylA promoter was purified from plasmid pHTS2xyl. This DNA fragment was ligated with the 5′ and 3′ parts of oppB and then inserted between the HindIII and BamHI sites of pRN5101. The resulting construction was verified by restriction mapping and used to transform B. thuringiensis strain 407 Cry− to erythromycin and tetracycline resistance. The chromosomal wild-type copy of oppB was replaced with the disrupted copy by homologous recombination as described previously (Lereclus et al., 1995). The recombinant strain, designated 407 Cry−ΔoppB, was resistant to tetracycline and sensitive to erythromycin. In this strain, the oppC, D and F genes were transcribed from the xylA promoter and thus induced by xylose (20 mM).
The double Δspo0AΔoppB mutant was constructed by disrupting the spo0A gene in the B. thuringiensis Cry−ΔoppB mutant strain. The disruption was introduced by homologous recombination using the thermosensitive plasmid pHT5120 carrying a copy of the spo0A gene disrupted with a gene conferring resistance to kanamycin (Lereclus et al., 1995). The recombinant clones were resistant to kanamycin and tetracycline. The disruption of the spo0A gene in the B. thuringiensisΔspo0AΔoppB mutant strain was verified by PCR using oligonucleotides OA1 and OA2.
Mini-Tn10 insertional mutagenesis
The thermosensitive plasmid pIC333 (Steinmetz and Richter, 1994) was used to deliver the mini-Tn10 and to produce a library of insertional mutants in B. thuringiensis. The mini-Tn10 in pIC333 is a 2.5 kb element composed of the two ends of Tn10 flanking a gene conferring resistance to spectinomycin and the replication region of pBR322. The thermosensitive replicon (pE194ts), the erm gene and the gene encoding the transposase of Tn10 are outside the mini-Tn10 element. The general strategy used to select and characterize the mutations was as follows. About 1 µg of pIC333 was used to transform B. thuringiensis 407 Cry−[plcA′Z] by electroporation. Transformants were selected at 28°C on LB plates containing erythromycin (2 µg ml−1) and spectinomycin (250 µg ml−1). All the transformants (about 50 000) were pooled and cultured at 28°C in LB medium (100 ml) containing erythromycin (2 µg ml−1) for 3 h. Cells were diluted (1:100) and grown in LB medium (without antibiotics) at 40°C (non-permissive temperature for the replication of the plasmid). This step was repeated three times to multiply the B. thuringiensis cells for about 15 generations. Appropriate dilutions of the cells were then plated on LB agar containing spectinomycin (250 µg ml−1) and Xgal (80 µg ml−1) and incubated at 40°C for 24 h. White colonies were selected (M2, M21 and M35) and analysed as follows: chromosomal DNA was cut with EcoRI or HindIII (two restriction sites that are not present in the mini-Tn10) and ligated. The ligation mixture was used to transform E. coli for resistance to spectinomycin (60 µg ml−1). Plasmid DNA was prepared from the E. coli transformants, and the restriction map of the plasmid was determined to verify the presence of the mini-Tn10 (a 2.2 kb BamHI fragment is characteristic of the mini-Tn10). The junction fragments between the ends of Tn10 and the chromosomal DNA were sequenced from plasmid DNA using the dideoxy chain termination kit from Pharmacia and primers E1 and E3 (Table 3), matching the ends of Tn10.
Sequencing of the opp operon
The DNA regions surrounding the sites of mini-Tn10 insertion in M2, M21 and M35 were sequenced directly by inverse PCR from B. thuringiensis chromosomal DNA. The purification and sequencing of the amplified DNA fragments and sequence assembly were performed as described previously (Økstad et al., 1999). The nucleotide sequence of the 9364 bp DNA region including the B. thuringiensis opp operon has been deposited in the GenBank database under accession no. AF305387.
RNA extraction and primer extension
B. thuringiensis strain 407 Cry− was grown with shaking in LB at 30°C. RNA extraction and primer extension were performed as described previously (Agaisse and Lereclus, 1996). To detect the transcription start site of the opp operon, primer extension was performed using the synthetic oligonucleotide Poop (Table 3) complementary to the 5′ end of the oppA gene (Fig. 3).
The B. thuringiensis cells containing lacZ transcriptional fusions were cultured in LB medium at 30°C, and β-galactosidase assays were performed as described previously (Msadek et al., 1990). The specific activities are expressed in units of β-galactosidase mg−1 protein (Miller units).
Preparation of Cry toxins and bioassay of insecticidal activity
The Cry1C toxins were prepared from the asporogenic strain 407 ΔsigK (Bravo et al., 1996) transformed with pHTIC (Sanchis et al., 1996). The cells were grown in HCT medium (100 ml) at 30°C for 2 days. The culture was centrifuged at 10 000 g for 20 min, and the pellet was washed twice in sterile water and then resuspended in 10 ml of sterile water. The Cry1C crystal preparation was sonicated briefly before use to prevent aggregation. The concentration of crystal protein was determined using the Bio-Rad protein assay.
Galleria mellonella eggs were hatched at 30°C, and the larvae were reared on beeswax and pollen (Naturalim France). Last instar larvae (150–350 mg) were force fed using 0.5 × 25 mm needles (Burkard Manufacturing) and a microinjector (automatic microapplicator; Burkard Manufacturing) with spore–crystal mixtures in sterile water (10 µl larva−1). The larvae were kept individually in boxes containing beeswax and pollen at 25°C. They were checked daily, and casualties were recorded for 7 days.
We thank Georges Rapoport, in whose laboratory part of this work was conducted. We are grateful to Hervé Agaisse and Tarek Msadek for helpful discussions, to Véronique Dartois for her help with the mini-Tn10 mutagenesis, and to Alex Edelman for English corrections. We thank the referees of this paper for their valuable comments and suggestions. This work was supported by the Institut Pasteur, the Institut National de la Recherche Agronomique (AIP Microbiologie) and the Centre National de la Recherche Scientifique. Funds for sequencing were supplied by the European Community (EC contract BIO4-CT96-0655).