OUP user menu

The enhancin-like metalloprotease from the Bacillus cereus group is regulated by the pleiotropic transcriptional activator PlcR but is not essential for larvicidal activity

Myriam Hajaij-Ellouze, Sinda Fedhila, Didier Lereclus, Christina Nielsen-LeRoux
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00289.x 9-16 First published online: 1 July 2006


Bacillus cereus group bacteria produce virulence factors. Many of these are regulated by the pleiotropic transcriptional activator PlcR, which is implicated in insect virulence. In silico analysis of the B. cereus strain ATCC14579 genome showed an enhancin-like gene preceded by a typical PlcR binding sequence. The gene is predicted to encode a polypeptide showing 23–25% identity with enhancins from several baculoviruses and 31% with that of Yersinia pestis. Viral enhancin acts after oral infection and degrades the peritrophic matrix of various Lepidopteran larvae. To rule out a possible implication of Bacillus enhancin in insect virulence, we sequenced the enhancin gene from the Bacillus thuringiensis 407–crystal minus strain and investigated its gene regulation and larvicidal activity. A typical metalloprotease zinc-binding domain (HEIAH) was detected and the gene was named mpbE (metalloprotease bacillus enhancin). An mpbE′–lacZ transcriptional fusion demonstrated that mpbE belongs to the PlcR regulon. The mpbE mutant was fed to Galleria mellonella larvae, and no significant reduction in virulence was observed. However, this may not exclude MpbE from a role in pathogenesis.

  • Bacillus thuringiensis
  • zinc metalloprotease
  • virulence factor
  • PlcR
  • Bacillus cereus
  • insect


Members of the Bacillus cereus group, such as Bacillus anthracis, B. cereus and Bacillus thuringiensis, are Gram-positive, spore-forming bacteria. Bacillus cereus senso stricto is considered to be an opportunistic human pathogen involved in gastrointestinal diseases (Kotirantaet al,2000; Hoffmasteret al,2004). Bacillus thuringiensis is well known for its entomopathogenic properties, which are mainly due to the production of protein crystals comprising Cry toxins (Schnepfet al,1998; de Maagdet al,2003). However, spores and vegetative cells of B. thuringiensis also play a role in insecticidal activity. Indeed, mutants lacking Cry toxins are still pathogenic when injected into certain lepidopteran larvae (Edlundet al,1976; Heiersonet al,1986; Zhanget al,1993). Vip3A, a vegetative insecticidal protein, may be involved in this process (Donovanet al,2001). Moreover, the addition of B. cereus spores or vegetative cells strongly increases the insecticidal activity of Cry toxins against some weakly susceptible insect species (Liet al,1987; Johnson & McGaughey, 1996; Salamitouet al,2000). Indeed, using Galleria mellonella larvae for infection, it is possible to measure the effect of virulence factors other than Cry toxins following oral infection. Neither bacteria nor toxin alone will induce mortality, whereas their combination will. This effect was defined as resulting from the synergism of the bacteria on the toxicity of the Cry toxins. A large number of B. thuringiensis and B. cereus genes, encoding potential extracellular virulence factors, are positively regulated by the pleiotropic transcriptional activator PlcR (Lerecluset al,1996; Agaisseet al,1999; Goharet al,2002). Alignment of the promoter regions of about 15 PlcR-regulated genes from B. thuringiensis and B. cereus genes revealed the presence of a highly conserved palindromic sequence (TATGNAN4TNCATA), named the PlcR box (Agaisseet al,1999; Okstadet al,1999). The role in virulence of the PlcR regulator (Salamitouet al,2000) and the PlcR-regulated metalloprotease InhA2 (Fedhilaet al,2002) were demonstrated by force-feeding G. melonella larvae. The appearance of bacteria in the haemocoel, when spores are added to the Cry toxins, suggests that some virulence factors may degrade the intestinal barriers to gain access to the larval body. The importance of PlcR for insect virulence is clear, but the role of only a few genes from the regulon has thus far been investigated; hence, it is of interest to identify the role of other PlcR-regulated genes.

An in silico search for PlcR-regulated genes (based on the genome of B. cereus ATCC 14579) was performed in order to identify factors that might interact with intestinal barriers. Among these was a gene encoding a protein with 23–25% amino acid identity to enhancin-like polypeptides from various baculoviruses (Pophamet al,2001; Liet al,2003). Some enhancins are described as metalloproteases (Leporeet al,1996): the well-characterized enhancin from Tricoplusia ni granular baculovirus (TnGV) degrades a major peritrophic matrix (PM) protein (Wang & Granados, 1997). The PM is composed of a glycoprotein/chitin-rich mucus matrix, which lines the insect midgut and protects the epithelium from abrasion and from direct contact with pathogens (Terra, 2001). Viral enhancin strongly enhances larval baculovirus infection, acting as a synergic factor; it was also found to increase B. thuringiensis larvicidal activities towards lepidopteran larvae (Granadoset al,2001). Until recently, enhancin-like genes were only reported from baculoviruses. Sequencing of bacterial genomes has revealed enhancin family genes from Gram-negative bacteria (Yersinia pestis; Parkhillet al,2001) and from strains of the B. cereus group, such as the B. anthracis Ames strain (Readet al,2003). It is not known whether these bacterial enhancins are involved in host–pathogen interactions and whether they may be important for infection and degradation of intestinal barriers.

Here, we describe the presence of an enhancin gene in the insect pathogen B. thuringiensis 407–crystal minus strain (Bt-407-Cry), and investigations on gene regulation, protein domains and implications of the gene for insect virulence were performed.

Materials and methods

Bacterial strains and growth conditions

The acrystalliferous sporogenic B. thuringiensis (Bt) strain 407 Cry (Lerecluset al,1989) and the Bt-407 CryΔplcR strain carrying a plcR-disrupted gene (Salamitouet al,2000) were used throughout the present study. Escherichia coli K-12 strain TG1 was used as host for the construction of plasmids and cloning experiments and E. coli strain ET 12567 for generating unmethylated plasmid DNA prior to B. thuringiensis transformation. Electroporation was used to transform E. coli (Doweret al,1988) and B. thuringiensis (Lerecluset al,1989). Escherichia coli and B. thuringiensis cells were routinely grown in Luria Bertani (LB) medium with vigorous agitation (175r.p.m.) at 37 and 30°C, respectively. Antibiotic concentrations used for bacterial selection were: ampicillin at 100μgmL−1 (for E. coli) and kanamycin at 200μgmL−1 and erythromycin at 10μgmL−1 (for B. thuringiensis). Bacteria with the Lac+ phenotype were identified on LB plates containing X-Gal (5-bromo-4-chloro-3-indolyl-d-galactopyranoside) at 40μgmL−1. Spores from Bacillus strains were obtained by culturing cells in 40mL of sporulation-specific (HCT) medium (Lecadetet al,1980) at 30°C for 3 days. Spores were harvested by centrifugation (5000g for 10min), washed twice and suspended in 6mL of sterile distilled water. The concentration of the spore preparation was estimated by plating dilutions onto LB agar plates containing appropriate antibiotics.

DNA manipulation

Plasmid DNA was extracted from E. coli and B. thuringiensis by a standard alkaline lysis procedure using QIAprep spin columns (Qiagen), with the following modification in the first step of the lysis procedure for B. thuringiensis: incubation at 37°C for 1h with 5mg of lysozyme (14300Umg−1). Chromosomal DNA was extracted from B. thuringiensis cells harvested in mid-logarithmic phase as described previously (Msadeket al,1990). Restriction enzymes and T4 DNA ligase were used as recommended by the mannufacturer (New England Biolabs). Oligonucleotide primers were synthesized by Genset (Paris, France). PCRs were performed in a Gene Amp PCR system 2400 thermal cycler (Perkin-Elmer) in 100-μL volumes containing 200μM dNTPs, 1.5mM MgSO4, 50pmol of each primer, 0.5μg of DNA template and 0.5units of high-fidelity Pwo DNA polymerase (Roche Boehringer). PCRs were run according to the manufacturer's recommendations and with hybridization at 55°C. Amplified DNA fragments were purified by using the QIAquick PCR purification kit (Qiagen) and separated on 0.7% agarose gels after digestion. Digested DNA fragments were extracted from agarose electrophoresis gels using a centrifugal filter device (QIAquick gel extraction kit, Qiagen).

Determination of the mpbE nucleotide sequence from Bt-407 Cry

Four pairs of oligonucleotides [M1 (EnS1–EnS2), M2 (EnS3–EnS4), M3 (EnS5–EnS6), M4 (EnS7–EnS8)] (Table 1) were designed from the gene AAP10320.1 from the available nucleotide sequence of B. cereus ATCC 14579 (contig 1543). These primers were used to amplify four fragments of 657, 678, 614 and 684bp, respectively, using Bt-407 Cry chromosomal DNA as a template. The four fragments are overlapping and cover the entire mpbE region from position −153 to +78 with respect to the ATG start and TAA terminal codons of the mpbE coding sequence, respectively. PCR products were purified using the QIA quick PCR purification kit (Qiagen) and then eluted from agarose gel following electrophoresis and sequenced by Genome Express (Paris, France).

View this table:
Table 1

Primer sequences used in this study

  • Restriction sites are underlined.

Mapping of mRNA start site by primer extension

Total RNA was extracted from wild-type B. thuringiensis (407 Cry) cells grown in LB at 30°C with shaking (175r.p.m.). The mpbE transcription start site was determined by primer extension as described by Agaisse & Lereclus (1996), using a synthetic 32-mer oligonucleotide EnExt (Table 1), which is complementary to the DNA sequence at positions +4 to +36 with respect to the translational start site of the mpbE gene. DNA sequencing was performed by the dideoxy chain termination method with the primer EnExt and using the double-stranded pHT304mpbE′Z as template.

Plasmid and mutant strain constructions

The mpbE gene in Bt-407 Cry was disrupted as follows. A 880-bp BamHI–EcoRI and a 920-bp SmaI–NcoI DNA fragment, corresponding to the chromosomal DNA regions upstream and downstream of the mpbE gene, respectively, were generated by PCR using the oligonucleotide pairs En1–En2 and En3–En4 (determined from the Bc14579 sequence) (Table 1). A kanamycin resistance (KmR) cassette (1.6kb) was purified from plasmid pDG783 (Guérout-Fleuryet al,1995). The amplified DNA fragments and the KmR cassette were digested and cloned between the BamHI and the NcoI sites of the thermosensitive plasmid pMAD (Arnaudet al,2004). The resulting plasmid was verified by restriction mapping. Transformed Bt-407Cry strains that were resistant to kanamycin, LacZ (white on X-Gal) and sensitive to erythromycin arose through a double crossover event resulting in deletion of mpbE and introduction of the KmR cassette (Lerecluset al,1995). The chromosomal allele exchange was checked by PCR using appropriate oligonucleotide primers. The corresponding mutant strain was designated Bt-407 CryΔmpbE.

Construction of the mpbE–lacZ transcriptional fusion

The mpbE–lacZ transcriptional fusion was constructed by cloning a 546-pb PstI/BamHI DNA fragment harbouring the mpbE promoter between the PstI and BamHI sites of pHT304-18′Z (Agaisse et al, 1994). The DNA fragment was generated by PCR amplification performed on Bt-407 Cry chromosomal DNA with the primers FwLacEnh and RvLacEnh (Table 1). The recombinant plasmid, designated pHT304mpbE′-Z, was introduced into B. thuringiensis wild-type and plcR mutant strains by electroporation. The transformants were named Bt-407 Cry (pHT304mpbE′-Z) and Bt-407 CryplcR (pHT304mpbE′-Z), respectively, and were resistant to erythromycin (10μgmL−1).

Beta-galactosidase assays

Bacteria harbouring the mpbE–lacZ transcriptional fusion were cultured in LB medium at 30°C with shaking at 175r.p.m. Beta-galactosidase-specific activities were measured as described by Bouillaut (2005).

Insects and in vivo experiments

Galleria mellonella eggs were hatched at 25°C, and the larvae were reared on bees' wax and pollen (Naturalim, 39330 Mouchard, France). Trypsin-activated Cry1C toxin was prepared from the asporogenic Bt-407 ΔsigK (Bravoet al,1996) transformed with pHT1C (Sanchiset al,1996) as described by Bouillaut (2005). For the infection experiments, groups of 20–30 last-instar G. mellonella larvae, weighing around 200 mg, were force-fed with 10μL of a mixture containing 106 spores from Bt-407 Cry and Bt-407 CryΔmpbE and 2μg purified and activated Cry1C toxin or with 10μL toxin or spores alone, using a thin needle and a microinjector (Burkard Manufacturing). The larvae were kept in individual boxes at 37°C. A control group was fed with NaPi buffer. Mortality was recorded after 24 and 48h. Experiments were repeated at least three times.

Results and discussion

The MpbE metalloprotease of Bt-407 Cry

A DNA fragment carrying the enhancin-like gene was amplified by PCR using Bt-407 Cry chromosomal DNA as template and primers based on the genome sequence of B. cereus ATCC14579. The PCR fragment was sequenced on both strands. The sequence revealed an ORF of 2229bp potentially encoding a polypeptide of 743 amino acid residues with a calculated molecular mass of 85.4kDa. This ORF was named mpbE (metalloprotease Bacillus Enhancin) (GeneBank accession no. DQ151839). Comparisons of enhancin-like proteins from various origins (Bacillus, Yersinia and Baculovirus) were made in order to determine a relationship between conserved domains, function and phylogeny (Fig. 1). BLAST protein alignment indicated about 97% similarity (E-value 0.0) with enhancin from other members of the B. cereus group and 50% (E-value 10−89) with enhancin from Yersinia pestis. The enhancin gene is present in three out of four B. cereus sequenced genomes (Bc ATCC14579 (AAP10320.1) and BcE33L (YP_084679.1) but shows only low homology (20% amino acid similarity) with a gene in strain Bc ATCC 10987. It is also found in the recently sequenced B. thuringiensis serovar konkukion strain 97-27 (YP_037494) and reported from B. cereus strain NCTC 7464 (Gallowayet al,2005). Thus, this gene is well conserved within the B. cereus group (Ivanovaet al,2003). MpbE showed 23–25% similarity (E-value 10−32 to 10−22) when compared with Baculovirus enhancins. Interestingly, only 20% amino acid identity was found between two Baculoviruses, TnGV and MacoNPV. The common sequences were mainly located in the N-terminal region containing the specific zinc-binding motifs (Popham et al, 2003), indicating a strong motif conservation among various pathogens.

Figure 1

 Partial alignment of MpbE protein from Bacillus thuringiensis strain 407 Cry (Bt-407) with enhancin family proteins from B. thuringiensis serovar konkukion (Bt kon), Bacillus cereus (Bc ATCC14579), Bacillus anthracis Ames (Ba), Yersinia pesti (Y. pestis), Mamestra configurata NPV baculovirus (MacoNPV) and Tricoplusia ni GV baculovirus (Tni GV). The first block shows N-terminal sequences; the arrow indicates the putative peptide signal sequence of 24 amino acids and the arrowhead the cleavage site. The second block shows amino acids in the region 230–260. The highly conserved metalloprotease zinc-binding domain (HEIAH) is shown boxed with a single line whereas the conserved glutamic acid (E) is shown boxed with a double line.

The amino acid sequence of MpbE and homologues revealed a putative signal peptide cleavage site between positions 24 and 25 (Fig. 1), suggesting that MpbE is exported. However, the prediction analysis (SignalP 3.0, http://www.cbs.dtu.dk) indicates a low score for MpbE due to the histidine residue in position 16 (when changed for a leucine the score increased from 0.053 to 0.881). This amino acid difference may be due to an error during sequencing. The apparent lack of the signal peptide in enhancin from Bc14579 is due to a mistake in genome annotation. Several other bacterial zinc metalloproteases have been shown to be extracellular proteins that need signal and leader sequences to aid transport across the bacterial cell membrane (Miyoshi & Shinoda, 2000). No signal peptide was found in enhancin from Yersinia and Baculovius, suggesting that these molecules may use another translocation system. The MpbE amino acid sequence contains the zinc-binding consensus motif HEXXH (Jongeneelet al,1989). In this motif, the two histidine residues bind the Zn2+ ions and the glutamic acid residue about 25 residues downstream is involved in nucleophilic attack (Bodeet al,1993). In fact, a glutamic acid was found 21 amino acids downstream of the first histidine in all aligned enhancin proteins (Fig. 1). MpbE belongs to the zincin superfamily of zinc-requiring metalloproteases (Miyoshi & Shinoda, 2000); however, examination of the sequence failed to find any relationship with the previously reported metalloprotease families, including InhA metalloproteases, secreted by B. thuringiensis (Fedhilaet al,2002). This suggests that MpbE may form a new family. Interestingly, an identical HEIAH motif was found for all strains from the B. cereus group, Yersinia and the MacoNVP virus enhancins, but not from TnGV (Fig. 1). This strong similarity was also indicated by Galloway (2005), in accordance with the results of Li (2003).

MpbE expression is PlcR dependent

A typical PlcR DNA binding sequence (TATGCAAATTACATA) was located 138bp upstream from the ATG (presumed start codon), at a distance also found for other PlcR-regulated genes (Agaisseet al,1999). The mpbE transcriptional start was mapped by primer extension analysis using a synthetic oligonucleotide EnExt (Table 1) and total RNA extracted from Bt-407 Cry cells harvested at different stages of growth in LB medium [T−0.5, T0, T+0.5 and T+1h with respect to the onset (T0) of stationary phase]. Transcripts were only detected after entry to stationary phase at T0. The 5′ end of the transcript was found 91bp upstream from the presumed mpbE start codon (Fig. 2b). The putative −10 (TATAAT) and −35 (TTGATA) boxes of the mpbE promoter (Fig. 2a) are similar to the −35 and −10 regions of the Sigma A promoter of Bacillus subtilis and are located 10 and 34bp downstream from the PlcR box. To assess the transcriptional activity of the mpbE promoter, and to analyse whether PlcR controls mpbE expression, we constructed a transcriptional fusion between the 546-bp DNA region extending upstream from the mpbE start codon and the lacZ gene in pHT304-18′Z. The recombinant plasmid (pHT304mpbE′–Z) was introduced into B. thuringiensis wild-type and plcR mutant strains. Bt-407 Cry carrying pHT304mpbE′–Z was cultured in LB medium and in a sporulation-specific medium (HCT) at 30°C. Beta-galactosidase production was measured at different stages of growth between T1 and T4 (Fig. 3). In LB medium, the level of mpbE-directed ß-galactosidase synthesis was very low during the exponential growth phase. It increased at the onset of the stationary phase and reached a maximum specific activity of 530 Miller units at T+2. A level of activity in the mid range was found for PlcR-regulated factors (plcA, hblc, inhA2) measured in LB medium (Agaisseet al,1999; Fedhilaet al,2002; Bouillautet al,2005.) In HCT medium, no ß-galactosidase was produced (<10 Miller units). This suggests that the mpbE gene is preferentially expressed when the cells are grown in a relatively rich medium (i.e. LB medium). In contrast with the wild-type strain, the mpbE–lacZ fusion was not expressed (ß-galactosidase level of <10 Miller units) in a plcR mutant strain grown in LB medium. (Fig. 2). These results indicate that PlcR positively controls mpbE expression during bacterial growth.

Figure 2

 Analysis of the mpbE promoter region. (a) Sequence of mpbE promoter region. The putative PlcR DNA binding sequence and −10 and −35 regions of the mpbE promoter are shown boxed. The potential ribosome binding site (RBS) and the presumed start codon (AGT) are indicated. (b) Mapping of the mpbE transcriptional start site region by primer extension. Total RNA was isolated from Bacillus thuringiensis strain 407 Cry 0.5h before, at onset (0), and at 0.5 and 1h after the onset of stationary phase. RNA was subjected to primer extension using the oligonucleotide EnExt (Table 1). Lanes A, C, G and T show the sequence of the promoter region of the mpbE gene. The nucleotide marked with an asterisk (A*) corresponds to the transcriptional start site at position −91 with respect to the ATG start codon.

Figure 3

 PlcR positively regulates the expression of mpbE. Specific ß-galactosidase activity was measured by analysis of transcriptional lacZ fusions to the mpbE promoter in Bacillus thuringiensis (Bt) 407 Cry [pHT 304-18ZpmpbE] (filled squares) and Bt-407 CryΔplcR [pHT 304-18ZpmpbE] (filled circles) grown in LB medium and Bt-407 Cry [pHT 304-18ZpmpbE] (open diamonds) grown in HCT medium. Bacteria were grown at 30°C.

Pathogencity in insects

To assess the pathogenicity of B. thuringiensis against insects, spores were fed to larvae either alone or in association with the insecticidal crystal toxin Cry1C. Galleria mellonella larvae are not susceptible to the ingestion of B. thuringiensis vegetative bacteria, spores or Cry toxins alone. However, by mixing bacteria and Cry toxins G. mellonella can be used to identify chromosomal virulence factors. Indeed, the importance of the PlcR regulon (Salamitouet al,2000) and the metalloprotease InhA2 (Fedhilaet al,2002) were demonstrated with this ‘synergy-model’.

We compared the larval mortality obtained with the parental Bt-407 Cry and the ΔmpbE mutant, following force feedings. Two micrograms of Cry1C toxin alone induced 12±9% mortality; 106 spores alone resulted in 7±2.2% and 3.5±2.3% mortality for the parental Bt-407-Cry and ΔmpbE mutant strains, respectively. The association of spores and Cry toxin indicated a clear synergy for both strains; an average of 42.8±14.3% mortality was found for the parental strain and 33.5±12.8% for the ΔmpbE mutant. No larval mortality was found for larvae fed with NaPi buffer. Although some difference in survival was observed (Fig. 4) no significant reduction in mortality was found with the mutant strain, indicating that MpbE is not essential for virulence towards G. mellonella larvae under the conditions used here. However, it cannot be excluded that MpbE has a role during the infection process, which cannot be identified while scoring mortality with a single high dose only. MpbE may have a cytotoxic effect or even have a role in degradation of the peritrophic matrix, but this was not recorded by Galloway (2005). To gain further insight into the mode of action of Bacillus enhancin, purification is needed to identify host target molecules, and in vivo expression approaches could provide information concerning the localization and level of gene expression during the infection process. We have described here a new potential virulence gene that is a part of the PlcR virulence regulon, and which is conserved within various invertebrate pathogens, which interact with host intestinal barriers. Enhancin, although not essential, might be involved with the overall virulence of the B. cereus group bacteria, including the interaction with intestinal mucus of vertebrates.

Figure 4

 Effect of mpbE mutation on virulence against the insect Galleria mellonella by force-feeding assays. Last instar larvae were force-fed with 2μg Cry1C toxin per larva, 106 spores alone per larva or with a mixture of Cry1C toxin and spores. Results shown are the mean of three independent experiments. Vertical bars indicate standard errors of the means. Control larvae were force-fed with NaPi buffer alone.


We are grateful to Myriam Gominet for help with gene sequencing and to Christophe Buisson for performing bioassays. M.H.-E. was supported by a grant from AUF (Agence universitaire de la Francophonie). This study was supported by research funds from Institut National de la Recherche Agronomique (INRA): AIP Microbiolgie No. 2003/P00244 and project No. 0071-2001-02: Colonisation du biotope insecte par des bactéries pathogènes.


View Abstract