Adhesion and cytotoxicity of Bacillus cereus and Bacillus thuringiensis to epithelial cells are FlhA and PlcR dependent, respectively

Microbes and Infection 8 (2006) 1483e1491 www.elsevier.com/locate/micinf

Original article

Adhesion and cytotoxicity of Bacillus cereus and Bacillus thuringiensis to epithelial cells are FlhA and PlcR dependent, respectively Nalini Ramarao*, Didier Lereclus Unite´ Ge´ne´tique Microbienne et Environnement, Institut National de la Recherche Agronomique, La Minie`re, 78285 Guyancourt Cedex, France Received 5 September 2005; accepted 5 January 2006 Available online 3 April 2006

Abstract Some bacteria of the Bacillus cereus group are enteropathogens. The first cells encountered by bacteria following oral contamination of the host are epithelial cells. We studied the capacity of these bacteria to adhere to epithelial cells and the consequences of this interaction. We found that cell adhesion is strain dependent and that a strain mutated in flhA, which encodes a component of flagellum-apparatus formation, is impaired in adhesion, suggesting that flagella are important virulence factors. The bacteria are cytotoxic to epithelial cells and induce substantial cytoplasmic and membrane alterations. However, direct contact between cells and bacteria is not required for cytotoxicity. The determinants of this cytotoxicity are secreted and their expression depends on the pleiotropic regulator PlcR. Adhesion and cytotoxicity of B. cereus to epithelial cells might explain the diarrhea caused by these pathogens. Our findings provide further insight into the pathogenicity of B. cereus group members. Ó 2006 Elsevier SAS. All rights reserved. Keywords: Bacillus cereus; Adhesion; Cytotoxicity; Virulence factors; FlhA; PlcR

1. Introduction The Bacillus cereus group comprises several species of sporulating bacteria, which are found ubiquitously. The species Bacillus anthracis, Bacillus thuringiensis and B. cereus are pathogens [1]. B. anthracis and B. thuringiensis produce specific toxins conferring the ability to colonize hosts as diverse as insects and mammals [2,3]. The genes encoding the insecticidal toxins of B. thuringiensis and the B. anthracis toxins have been extensively studied and are now well characterized. They are plasmid-borne and might have been transferred between species by conjugation. In contrast, the genetic determinants of other aspects of the infectious process (i.e. for adhesion, invasion and colonization) might be common to the three species and located on the chromosome [4]. However, no general and non species-specific virulence factors have been identified. Symptoms associated with B. cereus gastroenteritis and

* Corresponding author. Tel.: þ33 1 30 83 36 36; fax: þ33 1 30 43 80 97. E-mail address: [email protected] (N. Ramarao). 1286-4579/$ - see front matter Ó 2006 Elsevier SAS. All rights reserved. doi:10.1016/j.micinf.2006.01.005

opportunistic infections (endophthalmitis, periodontitis, pneumonia) might be the consequence of the production of various extracellular factors including enterotoxins Hbl and Nhe, and cytotoxin CytK, although no direct link has been demonstrated [5]. The genes encoding these potential virulence factors belong to the PlcR regulon. Their transcription is activated at the onset of the stationary phase of growth by a quorum-sensing mechanism [6,7]. Deletion of plcR from B. cereus and B. thuringiensis reduces, but does not abolish, the virulence of the bacteria in insects, in mice and during endophthalmitis in rabbits [8,9]. Moreover, B. anthracis does not express plcR, and addition of a functional copy does not increase virulence of the bacteria in a murine model [10]. Thus, the PlcR-regulated genes are necessary but not sufficient to account for the pathogenicity of B. cereus in mammals, suggesting that other factors are required for pathogenicity. It has consistently been shown that flagella are important in the early stages of infection [11,12]: they confer motility, and work with a strain mutated in flhA, which encodes a component of flagellumapparatus formation, suggests that they are essential bacterial virulence factors [13].

1484

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

Adhesion of bacteria to eukaryotic cells is often the first key event during infection of susceptible hosts. Microorganisms of the genus Bacillus are known to interact with epithelial cells [14e16], but the bacterial factors responsible for adhesion have not been identified. The role of the adhesion of B. cereus to epithelial cells in virulence is unknown but it may prevent elimination of the bacteria by the cleansing mechanisms of the intestine. Adhesion and cytotoxicity to epithelial cells might be important steps, although not necessarily related, as diarrhea presumably occurs through the destruction of epithelial cells by bacterial toxins produced in the small intestine [17]. Although PlcR may contribute to cytotoxicity, no detailed quantitative assays have been done [8]. Therefore, we studied adhesion of B. cereus and B. thuringiensis to epithelial cells, and invasion, to determine whether they contribute to virulence. Strains were also tested for cytotoxicity to epithelial cells, and bacterial factors involved in adhesion and cytotoxicity to epithelial cells were identified.

2. Materials and methods 2.1. Bacterial strains and growth conditions The acrystalliferous B. thuringiensis strain 407 Cry
(Bt 407), the sequenced B. cereus strains ATCC 10987 and ATCC 14579, and the B. cereus diarrheic strains Bc D1, Bc D17 and Bc D23 were used [18e21]. The mutant strains Bt 407 DplcR, Bt 407 DinhA2, Bt 407 DplcA, Bt 407 DflhA and the complemented strain Bt 407 DflhA/flhA have been described previously [8,13,22,23]. The B. subtilis strain 168 and Escherichia coli TG1 were used as controls. Bt 407 DhblC, Bt 407 DcytK and Bt 407 Dtlo were constructed in the laboratory. The cytK and hblC genes were disrupted, through homologous recombination with pRN5101, a derivative of the thermo-sensitive plasmid pE194ts [24]. DNA fragments corresponding to internal regions of the cytK and hblC genes (290 and 284 bp, respectively) were generated from the Bt 407 chromosome by PCR using the primer pairs CytB (50 -CGGGATCCGAGCGCTGTTATCTTGAAGGTT-30 ), CytH (50 -GCCGAAGCTTTCGGGCAAAATGCAAAAACAC-30 ), and HblB (50 -CGGGATCCGTGGCAACTGCGCAAGGCATA-30 ), HblH (50 -GCCGAAGCTTCTCAGCTTCTAGAATAGAGTT-30 ). These DNA fragments were inserted between the BamHI and HindIII sites of pRN5101, and the resulting plasmids were introduced into Bt 407 by electroporation. Transformants resistant to erythromycin were grown at 30  C and transferred to non-permissive temperature (39  C) for about 10 generations. The bacterial cells were then plated onto LB agar plates supplemented with erythromycin and incubated for 24 h at 39  C. Integration of the recombinant plasmid was confirmed by PCR using primers mapping at the ends of pRN5101 and primers external to the insertion site. The insertion mutant strains were designated Bt 407 DhblC and Bt 407 DcytK. The tlo gene encoding thuringiolysin 0 (a cholesterolbinding cytolysin) was disrupted as follows. HindIIIeXbaI (450 bp) and PstIeBamHI (453 bp) DNA fragments were

generated by PCR using Bt 407 chromosomal DNA as the template and primer pairs Tlo7 (5 0 -CCCAAGCTTAT TAGTATAGACTTACCTGGC-3 0 ), Tlo8 (5 0 -CCCTCTA GAATAAACCGTTCTACCATAAGCTAC-3 0 ), and Tlo9 (50 -CCCGAATTCTTTACTGCTGTCGTATTAGGTGGA-30 ), Tlo10 (50 -CCCGGATCCCCATTCCCATGCAAGACCTGTACA30 ), respectively. A KmR cassette carrying the aphA3 gene was purified from pDG783 (laboratory stock) as a 1.5-kb XbaIePstI fragment. The amplified DNA fragments were digested with the appropriate enzymes and inserted between the HindIII and BamHI sites of pRN5101. The resulting plasmid was introduced into Bt 407 by electroporation, and the tlo gene was deleted by a double crossover event as described previously [25]. Chromosomal allele exchange was confirmed by PCR with oligonucleotide primers located upstream from Tlo7 and downstream from Tlo10. All strains were grown in LB medium at 37  C under agitation until the OD600nm reached 2. The growth curves of the various strains were indistinguishable. Bacteria were harvested by centrifugation for 10 min at 3500 rpm. Pellets were washed two times with phosphate-buffered saline (PBS) and resuspended in PBS at 108 bacteria/ml. The culture supernatant was filtered through a 0.2-mm pore-size filter. 2.2. Cell cultures HeLa and CaCo2 epithelial cells were maintained in Dulbeco’s modified Eagle’s minimum essential medium (DMEM, Invitrogen) supplemented with 10% fetal bovine serum (FBS, Invitrogen). Cells were incubated at 37  C under a 5% CO2 atmosphere and saturating humidity. Cells were detached using 0.02% trypsin, counted with a hematocytometer and seeded into multiwell disposable trays containing DMEM þ FBS at a density of 2  105 cells per well. HeLa cells in the trays were infected the day after seeding, and CaCo2 cells 15e20 days after confluence was reached. 2.3. Cell infection assays Cells were covered with fresh DMEM and infected with bacterial suspension at a multiplicity of infection (m.o.i.) of 10 or 50 for 30 min. Non-attached bacteria were removed by carefully washing with PBS. Infected cells were fixed in 3.7% paraformaldehyde (PFA) overnight at 4  C and then washed in PBS. The cells were stained using eosin, or fluorescently immunostained (green) using FITC-phalloidin (Sigma) to stain the actin. Bacteria were either fluorescently labeled (red) before infection using 5-carboxytetramethylrhodamine succinimidyl ester (5-TAMRA, SE) (Molecular Probes) for 20 min at room temperature according to the manufacturer’s instructions or visualized by Gram staining. At least 100 cells in four different experiments were observed under the microscope and the percentage of cells in contact with at least one bacterium was scored as the percentage of infected cells.

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

2.4. Adhesion and invasion assays Cells were infected with bacterial suspension at a m.o.i. of 10 for 30 min or 2 h. Non-attached bacteria were removed by washing with PBS. Infected cells were detached by scraping in some cases after incubation with 25 mg/ml of gentamicin for 30 min at room temperature to kill extracellular bacteria. Serial dilutions of gentamicin-treated and non-treated preparations were plated on LB plates to score invasive versus total bacteria, respectively. Results are means of three different experiments done in triplicate. 2.5. Confocal microscopy Intra- or extracellular localization of the bacteria was determined by confocal microscopy. Cells were infected with TAMRA-stained bacteria for 30 min. After PFA fixation, cells were fluorescently stained using FITCephalloidin. Samples were subjected to confocal microscopy and serial pictures were taken to screen cells in three dimensions. At least 100 infected cells were observed for three different experiments and the intracellular versus extracellular location of bacteria was determined. 2.6. Cytotoxicity assays Cells were infected with culture supernatant or bacterial suspension at a m.o.i. (or equivalent for culture supernatants) of 10 or 50 for 15 min to 6 h. After the indicated times, trypan blue dye was added to the cells. Non-permeabilized (viable) cells remained unstained whereas permeabilized (killed) cells allowed the dye to enter the cytoplasm and cells were therefore stained blue. At least 300 cells were counted by eye and cytotoxicity scored as the number of blue cells as a percentage of all cells. Results are mean values of three independent experiments. 2.7. Sample preparation for transmission electron microscopy (TEM) Cells were infected with bacterial suspension for 15 min to 6 h. Non-attached bacteria were removed by washing. The cells were then fixed in 2% gluteraldehyde in cacodylate buffer and post-fixed in 1% osmium tetraoxide. Samples were dehydrated through a series of ethanol solutions and embedded in epon. Samples were cut with a Diatome diamond knife and ultrathin sections were stained with lead citrate. A Philips CM12 electron microscope was used for electron microscopy. 3. Results 3.1. Adhesion of B. cereus group members to epithelial cells is strain dependent We compared the cell-adhesion ability of various strains belonging to the B. cereus group. The infected cells were

1485

visualized by eosin staining or by labeling actin with FITCe phalloidin. The bacteria were stained using Gram staining or TAMRA before infection. The two staining methods (Fig. 1AeD) gave similar results for the percentage of infected cells (not shown). We used fluorescent labeling for the rest of the study. B. thuringiensis Bt 407 adhered to HeLa and CaCo2 cells to a lesser extent than several B. cereus strains. The percentage of HeLa cells infected after 30 min of infection was around 20% for Bt 407, 35% for Bc 10987 and 35e40% for the diarrheic strains Bc D1, Bc D17 and Bc D23 (Fig. 1E). Surprisingly however, the sequenced strain Bc 14579 adhered only poorly to HeLa cells: less than 5% of cells were associated with bacteria (Fig. 1B,D,E). Results were similar for the CaCo2 cells (Fig. 1F) with 20% of cells infected for Bt 407, 40% for Bc 10987 and 1.8% for Bc 14579 (P < 0.002). 3.2. FlhA is required for adhesion The non-motile and non-flagellated Bt 407 DflhA mutant obtained from the strain Bt 407 DplcA [13] was used to study the role of flagella in bacterial adhesion to eukaryotic cells. The ability of the Bt 407 DflhA mutant to adhere to HeLa and CaCo2 cells was substantially lower than that of the strain Bt 407 DplcA: the percentage of cells infected was 31% for Bt 407 DplcA and less than 5% for the flhA deficient mutant in both cells types (P < 0.005) (Fig. 2AeE). The value for the complemented strain Bt 407 DflhA/flhA was restored to a level similar to the strain Bt 407 DplcA with 27% of infected cells, showing that FlhA is involved in the adhesion process. plcRdeficient strains, like the flhA mutant, have impaired motility [9], but the adhesion capacity of the non-motile plcR mutant was not significantly different from that of the wild type. Thus, although requiring FlhA, adhesion is not controlled by PlcR and is not dependent on motility (Fig. 2A). 3.3. B. cereus and B. thuringiensis are poorly or not invasive Microscopic observation of HeLa cells infected with B. cereus and B. thuringiensis indicated that the bacteria were mostly extracellular. To assess the ability of these bacteria to invade epithelial cells, HeLa cells were infected with Bt 407 and Bc 10987, then the total number of bacteria and the number of intracellular bacteria were scored. Five to 6% of initial Bt 407 bacteria adhered to the cells after each 30 min and 2 h (Fig. 3A). More Bc 10987 adhered to HeLa cells: 15% of the bacteria were adhered after 30 min of infection (Fig. 3A). This is consistent with fluorescent observation of infected cells (Fig. 1E). The total number of Bc 10987 bacteria was lower after 2 h of infection due to the removal of detaching cells. For both bacterial strains, the percentage of all bacteria that had invaded the cells never exceeded 2% (P < 0.02, Fig. 3A). Results were similar when CaCo2 cells were used (not shown). Since bacteria mostly located extracellularly, monolayer of CaCo2 cells were infected with Bc 10987 and visualized by confocal microscopy to localize bacteria on the cell surface

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

1486

A

B

10 µm

10 µm

C

D

10 µm

F

60 50

Infected cells (%)

40 30 20 10

50 40 30 20 10

9 57

7

14

Bc

7

98

40

10

Bc

Bacterial strains

Bt

23 D

17

D 1

D Bc

Bc

57

7

14

98

Bc

40

10

Bt

Bc

9

0 7

0

60

Bc

Infected cells (%)

E

10 µm

Bacterial strains

Fig. 1. Adhesion of B. cereus to epithelial cells is strain dependent. HeLa cells were infected with bacterial suspensions of Bt 407, Bc 10987, Bc 14579, Bc D1, Bc D17 and Bc D23. CaCo2 cells were infected with bacterial suspensions of Bt 407, Bc 10987 and Bc 14579. The cells were fluorescently immunostained (green) using FITCephalloidin to stain the actin (A,B) or stained using eosin (C,D). Bacteria were either fluorescently labeled (red) before infection using TAMRA (A,B) or visualized by Gram staining (C,D). Selected images show HeLa cells infected with Bc 10987 (A,C) and Bc 14579 (B,D). The percentage of HeLa (E) and CaCo2 (F) cells in contact with adherent bacteria (infected cells) was determined.

(Fig. 3B). Pictures were taken sequentially from the basal (top left) to the apical (bottom right) side of the cells. Microvilli were clearly distinguishable as spots inside cells in the last two panels (bottom right). The bacteria were localized extracellularly and most of them were visible in the intercellular space close to the apical side of cells. 3.4. B. cereus and B. thuringiensis are cytotoxic to epithelial cells B. cereus and B. thuringiensis are cytotoxic [8,26]. A trypan blue test was used to assess the effects of the bacteria on the membrane alterations and to quantify the effects. Bt 407

induced cytotoxicity in HeLa cells in a time- and concentration-dependent manner (Fig. 4A). The percentage of cells with permeabilized membranes increased from 10% after 15 min to 60% after 6 h following infection with Bt 407 at a m.o.i. of 10. At a m.o.i. of 50, 1 h of infection was sufficient to kill all the cells. Cytotoxicity was also visualized using TEM. Bacteria first adhered to the cells (Fig. 4B), then, after 30 min of infection, bacteria induced massive cell damage including loss of cell membrane integrity and consequent leakage of the cytoplasm (Fig. 4C). After 2 h of infection, most of the cells completely lost their shape. Bacteria seemed to continue to adhere to the membrane debris (Fig. 4D).

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

A

1487

50

Infected cells (%)

40 30 20 10

Bt pl 40 cR 7

07 Bt 4

flh Bt 4 A/ 07 flh A

Bt flh 40 A 7

Bt pl 407 cA

0

Bacterial strains

B

C

10 µm

D

10 µm

E

10 µm

10 µm

Fig. 2. The bacterial flagella are required for B. thuringiensis adhesion. HeLa cells were infected with bacterial suspensions of Bt 407 DplcA, Bt 407 DflhA, Bt 407 DflhA/flhA, Bt 407 and Bt 407 DplcR. The percentage of infected HeLa cells was determined (A). CaCo2 cells were infected with bacterial suspensions of Bt 407 and Bt 407 DflhA. Selected images show HeLa and CaCo2 cells infected with Bt 407 (B,D, respectively) and with Bt 407 DflhA (C,E, respectively).

3.5. Bacterial-induced cytotoxicity is PlcR-dependent We tested whether the observed cytotoxicity was dependent on direct bacteria-cell contact. Both bacterial suspension and filter culture supernatants of Bt 407 taken in the stationary phase of growth induced cytotoxic effects on both HeLa and CaCo2 cells: For HeLa cells, cytotoxicity was 50% and 80% respectively when used at a m.o.i. of 10, and 100% in both cases when used at a m.o.i. of 50 (Figs. 4A and 5A). Moreover, the poorly adherent strain ATCC 14579 was as cytotoxic as Bt 407 (not shown).

Therefore, cytotoxic components are secreted during bacterial growth. PlcR regulates the expression of secreted proteins, and it was shown, although not quantified, that the cytotoxicity to insect cells of culture supernatant from a plcR-deficient strain is lower than that of the wild type [8]. We tested the cytotoxic activities of bacterial suspensions and culture supernatants from a mutant deficient for plcR. The cytotoxic activity to HeLa cells of the plcR-deficient mutant (bacterial suspension and supernatant) was substantially lower than that of the wild type and less than 10% of cells had permeabilized

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

1488

A 20

Adhesion/invasion (%)

18 16 14 12 Total Intra

10 8 6 4 2 0

7 40 Bt 2h

7 40 Bt min 0 3

7 98 10 in c B 30m

7

89 10 c B 2h

Bacterial strains-Infection time

B

10µm

Fig. 3. B. cereus strains are poorly or not invasive. HeLa cells were infected with bacterial suspensions of Bt 407 and Bc 10987. Serial dilutions of gentamicintreated and non-treated preparations were plated on LB plates to score invasive (intra) versus total (total) bacteria, respectively. Results are means of three independent experiments done in triplicate (A). CaCo2 cells were infected with TAMRA-stained Bc 10987. Localization of the bacteria was determined by confocal microscopy. Fixed successive pictures are shown from the basal to the apical side of the cells (B).

membranes after 2 h of infection with Bt 407 DplcR at a m.o.i. of 50 (P < 0.05) (Fig. 5A). This value was similar to those obtained in the control infections with the non-pathogenic species B. subtilis and E. coli. The results with CaCo2 cells were similar to those obtained with HeLa cells and only results with the bacterial supernatants are shown for the CaCo2 cells (Fig. 5B). Similarly, we found that cytotoxic components to CaCo2 cells were secreted and dependent of the PlcR regulon. Hbl and CytK are major PlcR-regulated factors and might be involved in cytotoxicity. However, neither Bt 407 DhblC nor Bt 407 DcytK had impaired cytotoxicity. Mutants lacking plcA, tlo and inhA2 were also tested and their cytotoxicities were similar to the wild type (Fig. 5A,B), implying that other PlcR-regulated components are involved or that the absence of

any single component can be compensated by the presence of other effectors. 4. Discussion A prerequisite for infection is the encounter between the pathogenic bacteria and the target tissue. Epithelial cells comprise the first and major cell type encountered by microorganisms in the mucosa [27] and are the main site of host-pathogen interactions. Here, we show that B. thuringiensis and most B. cereus strains can adhere to epithelial cells. This may be an important virulence mechanism and might contribute to the symptoms associated with B. cereus food poisoning. For instance, the ability of B. cereus to adhere to epithelial cells could prevent elimination of the bacteria by the cleansing

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

1489

A Cytotoxicity (%)

120 100 80 Moi 10 Moi 50

60 40 20 0

15 min

30 min

1h

2h

6h

Infection time

B

C

D

1µm

Fig. 4. B. cereus strains are cytotoxic to epithelial cells. HeLa cells were infected with Bt 407 at a m.o.i. of 10 or 50 for 15 min to 6 h. Non-permeabilized (viable) cells remained unstained whereas permeabilized (killed) cells allowed the dye to enter the cytoplasm and cells were therefore stained blue. At least 300 cells were counted and blue cells scored as a percentage of all cells and this value defined as the percentage of cytotoxicity. Results are mean values of three independent experiments (A). Infected cells were also fixed and subjected to TEM. HeLa cells infected with Bt 407 at a m.o.i. of 10 for 15 min (B), 30 min (C) and 2 h (D) are shown. Note the loss of membrane integrity (arrow), and cytoplasm leakage in C and D.

mechanisms of the intestine. We further show that bacterial binding is dependent on the presence of flagella. Flagella of Bt 407 are required for swimming and swarming motility [13]. Flagella and motility contribute to the virulence of several bacteria by playing a role in the initial adhesion [28,29] and by specifically inducing activation of proinflammatory responses [30]. A Bt 407 mutant lacking flhA is less virulent than the wild type in insects infected orally and in the hemocoel [31]. Although requiring FlhA, adhesion is not controlled by PlcR and thus likely not dependent on motility. The strain ATCC 14579 is motile [9] and possesses a flhA gene [19]. However, this strain adheres only poorly to epithelial cells, and possibly the flagellar apparatus might not be entirely functional. Alternatively, flagella may contribute only to the initial steps of adhesion and a subsequent, stronger interaction might be orchestrated by other bacterial factors present on the strain ATCC 10987 and absent from the strain ATCC 14579. The sequences of these two strains are known, so genomic comparisons may lead to the identification of new adhesins that would be potential virulence factors. The strains we studied are poorly or not invasive. Another study reported that several B. cereus strains were able to partially invade epithelial cells. However, the strains used were different, and the authors showed that invasion was strain dependent, and the strains were treated with chloramphenicol before infection to inhibit bacterial growth and protein synthesis [14]. The extracellular location of the bacteria raises the question as to how bacteria gain access to deeper tissues. Here we show

that B. cereus and B. thuringiensis are highly cytotoxic to both non-polarized as well as polarized epithelial cells and might thus degrade the epithelial cell layer in the host. A mutant lacking plcR is less cytotoxic, and, indeed, PlcR has been consistently implicated in virulence using various infection models [8,9,32]. Possibly, the absence of PlcR impairs the bacterium’s ability to penetrate to deeper tissues and to colonize its host. Cytotoxicity is not contact mediated because the supernatant of Bt 407 cultures (Fig. 5A,B), and the poorly adherent strain ATCC 14579, are as cytotoxic as the wild-type strain Bt 407. Adhesion and cytotoxicity do not appear to correlate, but the two mechanisms are likely to be important for overall bacterial pathogenesis. Indeed, when bacteria encounter the gastrointestinal tract prior to diarrhea, the bacteria first need to resist the cleansing mechanism of the intestine and adhesion might be an important mechanism at this step. The bacteria, attached to the appropriate location can induce cytotoxicity. Indeed, attachment of the microorganisms to the apical face of the enterocyte-like cells might create conditions favorable for bacterial growth and multiplication, thus inducing cytotoxicity through a quorum sensing induced expression of the PlcR regulon providing a microenvironment with high concentration of toxic factors [7]. The bacteria have to be in the stationary phase of growth to induce cytotoxicity, as PlcR is involved and is activated at this stage. Therefore, adhesion and cytotoxicity might both be essential during the early stages of colonization of the gastrointestinal tract prior to diarrhea. PlcR is a pleiotropic regulator that controls the transcription of its

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491

1490

Cytotoxicity (%)

A

120 Moi 10 Moi 50

100 80 60 40 20

pl cA Bt 40 7 Bt tlo 40 7 in hA 2 B. su bt ilis E. co li

cy tK

07 Bt 4

hb lC Bt 4

07

pl cR

07 Bt 4

pl cR

07

40 Bt

SN

Bt 4

SN

7

Bt 4

07 Bt 40 7

0

Bacterial strains

Cytotoxicity (%)

B

120

hbl, plcA, tlo or inhA2 (Fig. 5A,B) and all were as cytotoxic as the wild-type strain (our attempts to construct an nhe mutant were unsuccessful). Thus, other PlcR-regulated factors might be involved in cytotoxicity, or the absence of one cytotoxic component may be compensated by the expression of other PlcR-regulated factors. For example, the cytotoxic strain ATCC 10987 lacks the hbl operon, but synthesizes large amounts of nhe mRNA [26]. Moreover, an hbl mutant of strain ATCC 14579 had a reduced toxicity to Vero cells, but the nhe gene was only very poorly transcribed in both the wild type and the hbl mutant [26]. Detailed analysis of the roles of these components in cytotoxicity may require construction of multiple mutants including mutants lacking each and combinations of hbl, cytK, and nhe.

Moi 10

100

Moi 50

80 60 40 20 B. su bt ilis

in hA 2

7 40

SN

tlo 7 Bt SN

SN

40 Bt

SN

Bt

7

40

pl cA

cy tk 7

hb cR

40

SN

40

Bt

7

7

Bt SN

SN

Bt

40

SN

Bt

40

pl cR

7

0

Bacterial strains Fig. 5. Cytotoxicity to epithelial cells is PlcR dependent. HeLa (A) and CaCo2 (B) cells were infected with bacterial suspension and supernatant of Bt 407, Bt 407 DplcR, Bt 407 DhblC, Bt 407 DcytK, Bt 407 DplcA, Bt 407 Dtlo, Bt 407 DinhA2, B. subtilis and E. coli at a m.o.i. of 10 or 50. After 2 h, at least 300 cells were counted visually and the blue cells scored as a percentage of all cells to determine the percent of cytotoxicity. Results are mean values of three independent experiments.

own gene and of several genes encoding exported proteins putatively involved in virulence [6,33]. About 80% of all extracellular proteins are regulated by PlcR [34], and the following genes (and their products) belong to the PlcR regulon: plcA (PI-PLC), plcB (PC-PLC), cerB (sphingomyelinase), hblCDA (enterotoxic hemolysin BL), nheABC (non-hemolytic enterotoxin), cytK (cytotoxin), inhA2, nprB and nprP2 (metalloproteases), colB (collagenase), sfp (protease), clo and tlo (cereolysin O and thuringiolysin O in B. cereus and B. thuringiensis, respectively). The role of each of these PlcR-regulated genes in pathogenicity has not been determined. However, a B. cereus strain overproducing CytK causes severe gastroenteritis although CytK is not widely found in isolates of B. cereus from diarrheal illness [21,35]. CytK is a pore-forming toxin that, as a purified protein, is cytotoxic to Vero and CaCo2 cells [36]. The metalloprotease InhA2 is necessary although not sufficient for the virulence of B. thuringiensis in insects [22,32]. The hbl mutant has reduced hemolytic activity against sheep erythrocytes but full hemolytic activity against human erythrocytes [26]. Nevertheless, most of these factors present in the bacterial supernatant during the stationary phase of growth [34] are not necessary for cytotoxicity. Indeed, we tested isogenic mutants of the Bt 407 strain lacking cytK,

Acknowledgements We thank Vincent Sanchis and Christophe Nguyen-The for providing the diarrheic strains and Anne Brit Kolstø for providing the Bc strain ATCC 10987. We thank Christine Longin and Sophie Chat for TEM. We thank all members of the GME laboratory and Philippe Sansonetti for helpful discussion. We thank Myriam Gominet, Denis Robichon and Cosette Grandvalet for construction of the mutant strains Bt 407DhblC, Bt 407DcytK and Bt 407Dtlo. We thank Elisabeth Guillemet and Seav-ly Tran for their help with the CaCo2 cells experiments. This work was supported by the Institut National de la Recherche Agronomique (AIP Microbiologie, no. P00244). References [1] E. Helgason, O.A. Økstad, D.A. Caugant, H.A. Johansen, A. Fouet, M. Mock, I. Hegna, A.B. Kolstø, Bacillus anthracis, Bacillus cereus, and Bacillus thuringiensis-One species on the basis of genetic evidence, Appl. Environ. Microbiol. 66 (2000) 2627e2630. [2] M. Mock, A. Fouet, Anthrax, Annu. Rev. Microbiol. 55 (2001) 647e671. [3] E. Schnepf, N. Crickmore, J. Van Rie, D. Lereclus, J. Baum, J. Feitelson, D.R. Zeigler, D.H. Dean, Bacillus thuringiensis and its pesticidal crystal proteins, Microbiol. Mol. Biol. Rev. 62 (1998) 775e806. [4] A.B. Kolstø, D. Lereclus, M. Mock, Genome structure and evolution of the Bacillus cereus group, Curr. Top. Microbiol. Immunol. 264 (2002) 95e108. [5] A. Kotiranta, K. Lounatmaa, M. Haapasalo, Epidemiology and pathogenesis of Bacillus cereus infections, Microb. Infect. 2 (2000) 189e198. [6] H. Agaisse, M. Gominet, O.A. Økstad, A.B. Kolstø, D. Lereclus, PlcR is a pleiotropic regulator of extracellular virulence factor gene expression in Bacillus thuringiensis, Mol. Microbiol. 32 (1999) 1043e1053. [7] L. Slamti, D. Lereclus, A cell-cell signaling peptide activates the PlcR virulence regulon in bacteria of the Bacillus cereus group, EMBO J 21 (2002) 4550e4559. [8] S. Salamitou, F. Ramisse, M. Brehe´lin, D. Bourguet, N. Gilois, M. Gominet, E. Hernandez, D. Lereclus, The plcR regulon is involved in the opportunistic properties of Bacillus thuringiensis and Bacillus cereus in mice and insects, Microbiology 146 (2000) 2825e2832. [9] M.C. Callegan, S.T. Kane, D.C. Cochran, M.S. Gilmore, M. Gominet, D. Lereclus, Relationship of PlcR-regulated factors to Bacillus endophthalmitis virulence, Infect. Immun. 71 (2003) 3116e3124. [10] T. Mignot, M. Mock, D. Robichon, A. Landier, D. Lereclus, A. Fouet, The incompatibility between the PlcR- and AtxA-controlled regulons may have selected a nonsense mutation in Bacillus anthracis, Mol. Microbiol. 42 (2001) 1189e1198.

N. Ramarao, D. Lereclus / Microbes and Infection 8 (2006) 1483e1491 [11] M.-Y. Zhang, A. Lo¨vgren, M.G. Low, R. Lande´n, Characterization of an avirulent pleiotropic mutant of the insect pathogen Bacillus thuringiensis: reduced expression of flagellin and phospholipases, Infect. Immun. 61 (1993) 4947e4954. [12] A. Lo¨vgren, M.-Y. Zhang, A. Engstro¨m, R. Lande´n, Identification of two expressed flagellin genes in the insect pathogen Bacillus thuringiensis subsp. alesti, J. Gen. Microbiol. 139 (1993) 21e30. [13] E. Ghelardi, F. Celandroni, S. Salvetti, D.J. Beecher, M. Gominet, D. Lereclus, A.C. Wong, S. Senesi, Requirement of flhA for swarming differentiation, flagellin export, and secretion of virulenceassociated proteins in Bacillus thuringiensis, J. Bacteriol. 184 (2002) 6424e6433. [14] J. Minnaard, V. Lievin-Le Moal, M.H. Coconnier, A.L. Servin, P.F. Pe´rez, Disassembly of F-actin cytoskeleton after interaction of Bacillus cereus with fully differentiated human intestinal Caco-2 cells, Infect. Immun. 72 (2004) 3106e3112. [15] N.J. Rowan, K. Deans, J.G. Anderson, C.G. Gemmell, I.S. Hunter, T. Chaithong, Putative virulence factor expression by clinical and food isolates of Bacillus spp. after growth in reconstituted infant milk formulae, Appl. Environ. Microbiol. 67 (2001) 3873e3881. [16] A. Andersson, P.E. Granum, U. Ronner, The adhesion of Bacillus cereus spores to epithelial cells might be an additional virulence mechanism, Int. J. Food Microbiol. 39 (1998) 93e99. [17] P.E. Granum, T. Lund, Bacillus cereus and its food poisoning toxins, FEMS Microbiol. Lett. 157 (1997) 223e228. [18] D. Lereclus, O. Arantes, J. Chaufaux, M.-M. Lecadet, Transformation and expression of a cloned v-endotoxin gene in Bacillus thuringiensis, FEMS Microbiol. Lett. 60 (1989) 211e218. [19] N. Ivanova, A. Sorokin, I. Anderson, N. Galleron, B. Candelon, V. Kapatral, A. Bhattacharyya, G. Reznik, N. Mikhailova, A. Lapidus, L. Chu, M. Mazur, E. Goltsman, N. Larsen, M. D’Souza, T. Walunas, Y. Grechkin, G. Pusch, R. Haselkorn, M. Fonstein, S.D. Ehrlich, R. Overbeek, N. Kyrpides, Genome sequence of Bacillus cereus and comparative analysis with Bacillus anthracis, Nature 423 (2003) 87e91. [20] D.A. Rasko, I.J. Rave, O.A. Okstad, E. Helgason, R.Z. Cer, L. Jiang, K.A. Shores, D.E. Fouts, N. Tourasse, S.V. Angiuoli, J.F. Kolonay, W.C. Nelson, A.B. Kolsto, C.M. Fraser, T.D. Read, The genome sequence of Bacillus cereus ATCC 10987 reveals metabolic adaptations and a large plasmid related to Bacillus anthracis pXO1, Nucleic Acids Res. 32 (2004) 977e988. [21] M.H. Guinebretie`re, V. Broussolle, C. Nguyen-The, Enterotoxigenic profiles of food-poisoning and food-borne Bacillus cereus strains, J. Clin. Microbiol. 40 (2002) 3053e3056. [22] S. Fedhila, P. Nel, D. Lereclus, The InhA2 metalloprotease of Bacillus thuringiensis strain 407 is required for pathogenicity in insects infected via the oral route, J. Bacteriol. 184 (2002) 3296e3304.

1491

[23] M. Gominet, L. Slamti, N. Gilois, M. Rose, D. Lereclus, Oligopeptide permease is required for expression of the Bacillus thuringiensis PlcR regulon and for virulence, Mol. Microbiol. 40 (2001) 963e975. [24] R. Villafane, D.H. Bechhofer, C.S. Narayanan, D. Dubnau, Replication control genes of plasmid pE194, J. Bacteriol. 169 (1987) 4822e4829. [25] D. Lereclus, H. Agaisse, M. Gominet, J. Chaufaux, Overproduction of encapsulated insecticidal crystal proteins in a Bacillus thuringiensis spo0A mutant, Bio/Technology 13 (1995) 67e71. [26] T. Lindba¨ck, O.A. Økstad, A.L. Rishovd, A.B. Kolstø, Insertional inactivation of hblC encoding the L2 component of Bacillus cereus ATCC 14579 haemolysin BL strongly reduces enterotoxigenic activity, but not the haemolytic activity against human erythrocytes, Microbiology 145 (1999) 3139e3146. [27] H.C. Ramos, M. Rumbo, J.C. Sirard, Bacterial flagellins: mediators of pathogenicity and host immune responses in mucosa, Trends Microbiol. 12 (2004) 509e517. [28] K. Richardson, Roles of motility and flagellar structure in pathogenicity of Vibrio cholerae: analysis of motility mutants in three animal models, Infect. Immun. 59 (1991) 2727e2736. [29] L. Dons, E. Eriksson, Y. Jin, M.E. Rottenberg, K. Kristensson, C.N. Larsen, J. Bresciani, J.E. Olsen, Role of flagellin and the twocomponent CheA/CheY system of Listeria monocytogenes in host cell invasion and virulence, Infect. Immun. 72 (2004) 3237e3244. [30] H.A. Zeng, Q. Carlson, Y. Guo, Y.L. Yu, S.J. Collier-Hyams, L.A. Madara, T.A. Gewirtz, S. Neish, Flagellin is the major proinflammatory determinant of enteropathogenic Salmonella, J. Immunol. 171 (2003) 3668e3674. [31] L. Bouillaut, N. Ramarao, C. Buisson, N. Gilois, M. Gohar, D. Lereclus, C. Nilesen-LeRoux, FlhA influences Bacillus thuringiensis PlcRregulated gene transcription, protein production, and virulence, Appl. Environ. Microbiol. 71 (2005) 8903e8910. [32] S. Fedhila, M. Gohar, L. Slamti, P. Nel, D. Lereclus, The Bacillus thuringiensis PlcR-regulated gene inhA2 is necessary, but not sufficient, for virulence, J. Bacteriol. 185 (2003) 2820e2825. [33] D. Lereclus, H. Agaisse, M. Gominet, S. Salamitou, V. Sanchis, Identification of a Bacillus thuringiensis gene that positively regulates transcription of the phosphatidylinositol-specific phospholipase C gene at the onset of the stationary phase, J Bacteriol. 178 (1996) 2749e2756. [34] M. Gohar, O.A. Økstad, N. Gilois, V. Sanchis, A.-B. Kolstø, D. Lereclus, Two-dimensional electrophoresis analysis of the extracellular proteome of Bacillus cereus reveals the importance of the PlcR regulon, Proteomics 2 (2002) 784e791. [35] T. Lund, M.-L. DeBuyser, P.E. Granum, A new cytotoxin from Bacillus cereus that may cause necrotic enteritis, Mol. Microbiol. 38 (2000) 254e261. [36] S.P. Hardy, T. Lund, P.E. Granum, CytK toxin of Bacillus cereus forms pores in planar lipid bilayers and is cytotoxic to intestinal epithelia, FEMS Microbiol. Lett. 197 (2001) 47e51.