SEARCH

SEARCH BY CITATION

Keywords:

  • Escherichia coli;
  • STb toxin;
  • permeabilization;
  • mitochondria;
  • cell alteration

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Previous studies have shown that STb causes microscopic histological alterations in animal intestinal models. Disrupted intestinal epithelium at the villous tips could be the result of an altered physiological cell state induced by the toxin. As a cellular model we used NIH-3T3 cells, a mouse fibroblast cell line, previously shown to be capable of internalizing the STb toxin. Using various probes specific for the cellular physiological state or cell organelles, we investigated STb activity using flow cytometry and confocal microscopy. In NIH-3T3 cells, labelled with propidium iodide and carboxyfluorescein diacetate, STb permeabilized the plasma membrane but the cellular esterases remained active. Confocal microscopy showed that fluorescein isothiocyanate (FITC)-labelled STb toxin molecules were internalized and were found scattered in the cytoplasm. Moreover, important clusters of FITC–STb were observed inside the cells after 6 h and these clusters matched with mitochondria labelling. After cell treatment with STb, using a fluorescent mitochondrial potential sensor, we observed mitochondria hyperpolarization, as an early event of intoxication. This phenomenon increased linearly with the dose of STb. The cell population treated with STb showed histological alterations such as membrane budding, granular cytoplasm and enlarged nucleus. Altogether, these results provide new information, at the cellular level, on the effect of the STb toxin.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

STb is one of the heat-stable enterotoxins produced by enterotoxigenic Escherichia coli. This toxin is responsible for a reversible alteration in intestinal secretion and contributes to diarrhea in various animal species including humans (Dubreuil, 2008). STb is synthesized as a 71-amino acid precursor, which undergoes a 23-residue signal sequence cleavage (Lee et al., 1983; Kupersztoch et al., 1990). A mature STb toxin corresponds to a 48-amino acid peptide with a molecular weight of 5.2 kDa and a basic isoelectric point (pI=9.6) (Handl et al., 1993). The STb tridimensional structure shows two antiparallel α helices (Cys10-Lys22, amphipathic, and Gly38-Ala44, hydrophobic) connected by a glycine-rich loop (Sukumar et al., 1995). Two disulfide bridges (Cys10-Cys48 and Cys21-Cys36) stabilize the structure (Sukumar et al., 1995) and both bridges are necessary for enterotoxicity (Arriaga et al., 1995; Okamoto et al., 1995). Moreover, in vitro STb oligomerization, as hexamers and heptamers, seems to be important for toxicity expression (Labrie et al., 2001b).

In vitro studies on different cell types, including Madin–Darby canine kidney (MDCK), HT-29/C1 human intestinal epithelial cells and primary rat pituitary cells, demonstrated that the intestinal secretion pathway involves the activation of a pertussic toxin-sensitive Gαi3 protein (Dreyfus et al., 1993). This activation results in an increase of the intracellular calcium concentration by a receptor-dependent ligand-gated Ca2+ channel. The high intracellular Ca2+ level is presumably involved in the activation of a calmodulin-dependent protein kinase II (CaMK II), which could open an undetermined ionic channel (Fujii et al., 1997). Furthermore, CaMK II also activates protein kinase C and consequently the cystic fibrosis transmembrane receptor. The elevated intracellular Ca2+ concentration may also induce the activation of phospholipase A2 and C, which catalyze the release of arachidonic acid from the membranar phospholipids. Finally, the production of prostaglandin E2 and serotonin was demonstrated (Hitotsubashi et al., 1992; Peterson & Whipp, 1995; Fujii et al., 1995). These secretion agents are responsible for the release of H2O, HCO3, Na+ and Cl in the intestinal lumen (Dubreuil, 1997, 2008). This activation cascade explains the fluid accumulation observed in the intestinal loop assay (Harel et al., 1991; Hitotsubashi et al., 1992; Labrie et al., 2001a).

Sulfatide, a glycosphingolipid that has been identified as a functional receptor for STb, is present on the intestinal cell membrane (Rousset et al., 1998; Beausoleil & Dubreuil, 2001). This molecule was also found in other pig tissues and epithelia tested. Our laboratory determined a dissociation constant of 2.4±0.6 nM for the STb–sulfatide interaction (Gonçalves et al., 2008). Recently, experiments on pig jejunal brush border membrane vesicles (BBMVs) demonstrated that STb is able to form ionic pores (Gonçalves et al., 2007). Moreover, using trypan blue, STb had been shown to permit adsorption of this stain by cell lines from many animal species and from various organs and epithelia. A good correlation was observed between the trypan blue uptake and the rat loop assay (Labrie et al., 2001a; Beausoleil et al., 2002). STb internalization was observed in rat intestinal loops by transmission electron microscopy (Labrie et al., 2002), revealing that no subcellular compartment seemed to be targeted by the toxin under the conditions tested. Chao & Dreyfus (1997, 1999) had also observed internalization of STb into cultured human and rat intestinal epithelial cells. In both cases a diffuse cytoplasmic distribution of STb was observed. Still, the intracellular target(s) for the STb toxin remains unknown at present.

In animal intestinal models, microscopic alterations were observed in the jejunal mucosa following incubation of STb-containing supernatants. The toxin provokes the loss of villous absorptive cells and partial atrophy of villi (Whipp et al., 1986, 1987; Rose et al., 1987). In this study, we were interested in understanding the effect of STb on NIH-3T3 cell membrane permeability as previous studies had shown the formation of transitory pores in pig BBMVs (Gonçalves et al., 2007). NIH-3T3 cells were selected as we showed ready internalization of the toxin in preliminary studies (Dubreuil et al., 2007) and sulfatide, the STb receptor, is present on the surface of these cells (S. Penel & J.D. Dubreuil, unpublished data). This cell line was used recently to study the Bordetella pertussis toxin (Ohnishi et al., 2008). In this study, many cell lines, including NIH-3T3, were compared for morphological changes due to the adenylate cyclase toxin. NIH-3T3 cells were also used to study Clostridium difficile toxin A (Gerhardt et al., 2005).

This study was undertaken to follow the internalization process and determine the route taken by the STb toxin inside cells in culture and also to evaluate the cell's altered state and morphological changes resulting from STb toxin activity.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

NIH-3T3 cell line

The Swiss mouse embryonic fibroblast cell line (NIH-3T3) was kindly supplied by Dr Robert Ivan Nabi (University of British Columbia, Vancouver, BC, Canada). Cells were grown in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% newborn calf serum, nonessential amino acids (1%), vitamins (1%), glutamine (1%) and penicillin/streptomycin antibiotics (1%) (complete medium) (Invitrogen, Burlington, ON, Canada). Cells were grown in 75-cm2 tissue culture flasks at 37 °C, 5% CO2 and routinely trypsinized before 80% confluence was reached.

STb production and purification

The STb toxin was produced according to a previously described method (Gonçalves et al., 2007). STb was quantified spectrophotometrically at 214 nm using aprotinin as the reference protein. The EZ-Label fluorescein isothiocyanate (FITC) protein labelling kit (Pierce Biotechnology, Rockford, IL) was used to prepare the FITC-labelled STb toxin.

Cell treatment conditions

Cells were grown for 2 days in 6-well plates at a density of 50 000 cells per well. Cells were washed with DMEM–100 mM HEPES (pH 7.4) and incubated in 1 mL of the same buffer in the presence or absence of STb (5–20 μg mL−1). After 1 h, the medium was removed and substituted with complete medium and cells were grown for 24 h at 37 °C in a 5% CO2 atmosphere. As an apoptosis control, 10 μM camptothecin (Sigma-Aldrich, Oakville, ON, Canada) was used, in which the cells were placed for 24 h at 37 °C in a 5% CO2 atmosphere.

Membrane integrity evaluation

The combined fluorochromes 5-(6)carboxyfluorescein diacetate (CFDA) and propidium iodide (PI) were used as a stain to evaluate cell membrane permeabilization. Cells were collected using a 0.05% trypsin–EDTA solution (Invitrogen) and 106 cells mL−1 were simultaneously treated with 0.4 μg mL−1 CFDA and 2 μg mL−1 PI. Cells were analyzed using a fluorescence-activated cell sorting (FACS) Vantage SE instrument (Becton Dickinson Biosciences, Mississauga, ON, Canada). The instrument is equipped with an argon ion laser for the excitation of the fluorescent dye at 488 nm. CFDA fluorescent emission was detected in the FL1 channel (530±20 nm), whereas PI was detected in the FL3 channel (630±22 nm). Heat-treated cells (75 °C for 5 min.) were all PI-positive, indicating altered cell membrane integrity and mortality. A sample of untreated cells was stained with PI and the PI-positive cells (red staining) were gated to remove them from the STb-treated cell population analyzed. The control for CFDA consisted of cells not treated with STb, but incubated for the same period of time in the same buffer. In this case, almost all cells were stained with CFDA (green staining), indicating cell viability. Again, gating was performed in order to remove the CFDA-negative cells from the STb-treated cell population analysis. FACS analysis of STb-treated cells was performed using both gates as defined before each experiment with every new cell batch.

Confocal microscopy

An Olympus IX81 microscope, which is part of an FV1000 confocal scanning system (Olympus, Melville, NY) equipped with a krypton–argon laser and the appropriate excitation and emission filters for JC-1, MitoTracker® Red CMXRos (Invitrogen) and FITC, was used. For 4,6-diamidino-2-phenylindole (DAPI, Invitrogen), a laser at 405 nm was used. Sorted cells were centrifuged at 600 g. Each cell pellet was resuspended in 50 μL of phosphate buffer (PB) (10 mM NaHPO4, 10 mM Na2HPO4; pH 7.4), mounted onto glass slides and immediately observed. For labelling with DAPI (a nuclear dye), MitoTracker® Red CMXRos and FITC–STb, cells were grown for 24 h on glass slides at a density of 10 000 cells. The MitoTracker® Red CMXRos probe passively diffuses across the plasma membrane and accumulates in active mitochondria. The MitoTracker® probe contains a thiol-reactive chloromethyl moiety that reacts with thiols on proteins and peptides to form an aldehyde-fixative conjugate, inhibiting the release of the dye from the mitochondria. Fixed cells were washed with DMEM–100 mM HEPES (pH 7.4) and then incubated in 1 mL of DMEM–100 mM HEPES (pH 7.4) with FITC–STb (10 μg). After 1 h, the medium was removed and was substituted with complete medium and slides were incubated at 37 °C in a 5% CO2 atmosphere for 2, 4, 6 or 12 h. Slides were treated with MitoTracker® Red CMXRos at 1 nM in DMEM–100 mM HEPES (pH 7.4) for 15 min at 37 °C. Cells were washed with PB and incubated with 200 μL of DAPI (0.32 μg mL−1) for 5 min at room temperature. After washing twice with PB, the glass slides were mounted in 90% glycerol–PB. Phase-contrast and fluorescence pictures were acquired with built-in fluoview 1000 software (Olympus). Representative cells from three independent experiments were observed.

Cell sorting

Cells treated with STb (20 μg mL−1) for 24 h and subsequently with JC-1 were sorted using the FACS Vantage SE instrument. FACS was operated using a 190 mW water-cooled argon ion laser at 488 nm and a 100-μm sort sense flow cell running at 9 psi sheath pressure and a droplet frequency of 20 kHz. Signals were discriminated according to their peak forward scatter signal and were sorted based on a region of interest around their light scatter cluster, together with both red and green, green alone and low green fluorescence. Samples of sorted cells were observed with an Olympus SV1000 confocal microscope.

Evaluation of the mitochondrial membrane potential

The mitochondrial membrane potential was measured using the JC-1 probe (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolocarbocyanine iodide, Invitrogen). JC-1 is a fluorescence mitochondrial potential sensor probe used for detection of membrane potential alterations. JC-1 is a cationic dye exhibiting potential-dependent accumulation in the mitochondria, indicated by a fluorescence emission shift from green (c. 525 nm) to red (c. 590 nm). Membrane depolarization is identified by a change in JC-1 properties from J-aggregates (red) to a monomeric form (green). The ratio between green and red fluorescence depends only on the membrane potential and not on other factors such as mitochondrial size, shape or density, which may influence single-component fluorescence signals (Smiley et al., 1991; Reers et al., 1995; Cossarizza & Salvioli, 2001). Cells were collected as described for membrane integrity evaluation. One milliliter of suspended cells (1 × 106 cells) was treated with the JC-1 probe at 0.25 μg mL−1 for 15 min at 37 °C. Cells were treated with a final concentration of 0.1 μmol mL−1 valinomycin (Sigma-Aldrich) for 15 min or 5 μL of H2O2 for 5 min at room temperature as a control. J-aggregates from JC-1 are formed as a linear function of mitochondrial membrane potential. When excited at 488 nm, these J-aggregates have a maximum emission at 590 nm. As the membrane potential declines, JC-1 reverts to its monomeric form, with a maximum emission at 530 nm. JC-1 fluorescence was measured by FACS (FL-1 channel 530±20 nm and FL-2 channel 585±42 nm).

Data analysis

At least 10 000 cells were acquired for each FACS analysis. Data were collected with the cell quest pro program (Becton Dickinson Biosciences) and further analyzed with winmdi program (version 2.8; Joseph Trotter, Salk Institute for Biological Studies, La Jolla, CA). At least three independent experiments were conducted for each cell treatment. Statistical comparisons were made with the Student t-test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

STb permeabilizes NIH-3T3 cell membranes without affecting cell viability

PI is commonly used as a cell death marker as it is excluded by intact plasma membranes of live cells. The fluorescence conferred by PI is associated with cells that have lost membrane integrity. Moreoverm another marker, CFDA, readily penetrates the cells, where its acetate group is hydrolyzed by nonspecific esterases. The hydrolyzed, negatively charged, product retained by cells presents a green fluorescence. Retention and hydrolysis of CFDA by cells confirms enzyme activity indicative of cell viability.

In this study, we observed a PI–CFDA- and a CFDA-labeled subpopulation for STb-treated cells. A clear separation between the two subpopulations was observed with the various amounts of the STb toxin tested (data not shown). Figure 1 shows the percentage of cells stained by both probes (PI–CFDA) in the presence of various concentrations of STb. The number of PI–CFDA-positive cells is in direct relation to the amount of STb (P<0.05). PI–CFDA-labelled cells indicate that STb permeabilizes the cells and allows PI to penetrate, but the cellular esterases' activity remains intact, corroborating cell viability. Permeabilization observed with NIH-3T3 cells was similar to what had previously been observed on IPEC-J2 cells (Gonçalves et al., 2008). However, permeabilization, with the same quantity of toxin, observed on NIH-3T3 cells was lower than that for BBMVs. We thus pursued the experiments in order to elucidate, besides the fact that the number of receptor molecules (sulfatide) could explain, at last in part, as to why such a difference was observed and also more importantly where the STb toxin could be found in the cell.

image

Figure 1.  Permeabilization of NIH-3T3 cells by the STb toxin. Cells were treated for 1 h with 0–20 μg mL−1 of the STb toxin and then incubated in fresh medium for 24 h at 37°C in a 5% CO2 atmosphere. After addition of PI and CFDA probes, fluorescence was evaluated in 10 000 cells using an FACS. Statistical significance was determined using Student's t-test. Results indicate a dose-dependent relationship between the quantity of STb added and the percentage of permeabilized cells (PI–CFDA positive) (*P<0.05). No statistical difference was observed between treatment with 15 and 20 μg mL−1 of STb.

Download figure to PowerPoint

The STb toxin is internalized and colocalizes with mitochondria

Cells grown on glass slides and treated with the FITC–STb toxin, for 2, 4, 6 or 12 h, indicate toxin internalization (Fig. 2a). A negative control consisting of free FITC at the same concentration as used for STb labelling was carried out on NIH-3T3 cells and no fluorescence was associated with the cell surface or entered the cell cytoplasm (data not shown). After 2 h, the FITC–STb toxin was present on the plasma membrane and inside the cells (Fig. 2a). Clusters of FITC–STb became visible in the cytoplasm after 6 h, but more clearly after 12 h. Moreover, cells were stained with two specific probes: DAPI (for the cell nucleus) and MitoTracker® Red CMXRos (for mitochondria) (Fig. 2b). The pictures show cells treated with FITC–STb for 1 h and incubated for 6 or 12 h. No colocalization was observed between DAPI and FITC–STb labelling [Fig. 2b(B)]. On the other hand, colocalization between MitoTracker® Red CMXRos and FITC–STb was observed as yellow clusters [Fig. 2b(E), merged images, indicated by arrows] inside the cells, clearly indicating STb colocalization with mitochondria. In addition, mitochondria clusters were seen close to the nucleus [Fig. 2b(e)]. From these data, we focused on mitochondria and more specifically on the study of the physiological state of this organelle.

image

Figure 2.  Internalization of FITC-labelled STb toxin and colocalization with mitochondria. (a) Cells were grown on glass slides and treated with 20mg mL−1 FITC–STb for 2, 4, 6 and 12 h. Images were taken at × 1440. FITC–STb appears on the plasma membrane and inside vesicles after 2 h and can be seen as clusters in the cytoplasm after 6 h,but more clearly after 12 h (arrows). (b) Cells were stained with fluorescent dyes: DAPI and MitoTracker® Red CMXRos. Confocal microscopy images were taken at × 600 (including A–D) and the picture at right at × 1440 (E). Cells were incubated for 6 or 12 h with 20 μg mL−1 FITC–STb. Panel D shows merging of the three probes. Arrows indicate colocalization of FITC–STb with mitochondria. Confocal microscopy pictures are representative of three independent experiments. The bar corresponds to 20 μm.

Download figure to PowerPoint

The STb toxin affects the mitochondrial membrane potential

As colocalization of STb with mitochondria was observed, we characterized NIH-3T3 STb-treated cells using the JC-1 probe. Untreated cells show a strong green fluorescence (Fig. 3a). Cells treated with valinomycin (0.1 μmol L−1) for 10 min, (data not shown, but identical to H2O2) or 5 μL H2O2 for 5 min, two depolarizing compounds (Fig. 3b), allowed discrimination between depolarized (green fluorescence) and hyperpolarized cell populations (red fluorescence).

image

Figure 3.  Effect of STb on the mitochondrial membrane potential. Cells were treated with STb (0–20 μg mL−1) and stained with JC-1. Alteration of the membrane potential was observed as a change in JC-1 properties from a monomeric form (green) to aggregated (red). The Dot plot shows untreated NIH-3T3 cells (a), cells treated with H2O2, a depolarizing agent (b), and with 20 μg mL−1 of STb for 24 h (c). Histogram (d) presents percentages (mean±SD of four independent experiments) of two cell subpopulations separated according to their fluorescence patterns. Significant differences were observed between doses of STb compared with control (P<0.05), but not between 0 and 5 μg mL−1 STb. Graph (e) presents the ratio of red/green fluorescence for the STb-treated cells. STb toxin treatment increased the red to green fluorescence intensity ratio and indicated mitochondria hyperpolarization.

Download figure to PowerPoint

The percentage of cells staining in red after treatment with STb (20 μg mL−1) (Fig. 3d) is attributable to hyperpolarization of the mitochondria. Hyperpolarization of cells treated with STb is a dose-dependent phenomenon as revealed by a linear increase in the red to green ratio (P<0.05) (Fig. 3e). This ratio clearly demonstrates the increased hyperpolarization of mitochondria after STb treatment. Annexin V, an early marker of apoptosis, and DNA fragmentation, a later marker of apoptosis, were performedon our STb-treated cells, but these more ‘classical’ tests did not reveal signs of apoptosis (data not shown).

Cell subpopulations result from STb treatment

After elimination of cells debris using the FACS R1 gating, STb-treated cells, where the JC-1 probe was added, could be sorted into three subpopulations (Fig. 4a). The first subpopulation is stained in red (R2, Fig. 3b); the second subpopulation is strongly stained in green (R3, Fig. 4d); and the third subpopulation is faintly stained in green (R4, Fig. 4c). Samples of sorted cells were observed under phase-contrast microscopy. The R3 subpopulation showed the typical live cell morphology: a round shape without any apparent membrane alteration. R2 (enlarged nucleus and early membrane budding) and R4 (granular cytoplasm and plasma membrane alterations) subpopulations displayed morphological changes.

image

Figure 4.  Cell subpopulations result from STb treatment. Cells were sorted after treatment with 20 μg mL−1 STb toxin. (a) FL1 vs. FL2 dot plot of cells stained with JC-1 after STb treatment. Pictures of sorted cells were obtained using phase-contrast microscopy. Cells in (b), (c) and (d) are from R2, R4 and R3 cluster gates, respectively (a). The R2 subpopulation shows an enlarged nucleus and membrane budding. The R4 subpopulation displays a granular cytoplasm and plasma membrane alterations. The R3 subpopulation shows round-shaped cells without any apparent membrane alteration (normal cells). Images were taken at × 1320. The bars correspond to 20 μm.

Download figure to PowerPoint

Figures 5a and b show a comparable repartition of cells treated with STb and camptothecin. Figure 5c shows the cell percentages associated with the three subpopulations identified. The R3 (normal cells) subpopulation is more important than the R4 subpopulation except when treated with 5 μg STb (P>0.05). A decrease of R3 subpopulations appears to be related to an increase in the R2 subpopulation. The percentage of R2 cells increases linearly with the STb dose (P<0.05). At 20 μg mL−1 of STb >30% of the cells were in an R2 state. This is comparable to what is observed with 10 μM camptothecin, a positive control for apoptosis.

image

Figure 5.  Effect of STb on the cell subpopulation distribution. Cells treated with STb (0–20 μg mL−1) were stained with JC-1. STb-treated cells (a) compared with camptothecin-treated cells (b). (c) Percentages (mean±SD of four independent experiments) of the three cell subpopulations as identified by their fluorescence-staining patterns. A decrease of the R3 subpopulation is closely associated with an increase of the R2 subpopulation. The R2 cell population increases linearly with the STb dose (P<0.05). Significant differences were observed between control and doses of STb, except for 5 μg mL−1 of STb.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was performed to evaluate the effect of E. coli STb enterotoxin on NIH-3T3 cells. STb was previously described as an enterotoxin permeabilizing several cell lines from pigs and also from other animals (Chao & Dreyfus, 1997, 1999; Beausoleil et al., 2002). In fact, Dreyfus et al. (1993) used MDCK, HT-29/C1 human intestinal epithelial cells and primary rat pituitary cells to study Ca2+ influx mediated by the STb toxin, as these cell types represented models for such studies. In the same way, we used NIH-3T3 as a recognized cell model to study the permeabilization process (Le & Nabi, 2003). Moreover, this toxin, at concentrations similar to that used in the present study, has been shown to permeabilize BBMVs from piglet jejunum in a short period of time (30 s) by forming nonspecific pores (Gonçalves et al., 2007). In the present study, we demonstrated that STb induces, at concentrations comparable to that used in vivo (5 μg per rat loop) (Labrie et al., 2001a), a loss of membrane integrity, allowing PI uptake in NIH-3T3 cells. The relatively small percentage of PI-CFDA–stained cells observed (Fig. 1) could be explained by transient membrane permeability and the R1 gate used to eliminate cell debris. This gating was carried out on a cell sample not treated with STb, but nevertheless, where cell integrity was affected for a small cell population as a result of trypsin treatment (to detach the cells from the culture flasks) and incubation in phosphate-buffered saline. As PI-positive cells were also stained by CFDA, it demonstrated esterase activity, indicating that STb allows PI entry without causing cell mortality. These results corroborated the data obtained by Beausoleil et al. (2002) who, using a trypan blue uptake bioassay, observed no cell mortality of trypan blue-positive cells using a 3-(4,5-dimethylthiazol-2,5-diphenyltetrazolium bromide) (MTT) assay. In this assay, MTT is reduced to formazan in the mitochondria of living cells. Thus, our results confirm the ability of STb, in a short period of time, to permeabilize without killing the cells. The small percentage of cell permeabilized by STb, in our study, could be the result of transient pore opening, and it differs from what was observed in pig jejunal BBMVs, as membrane permeability may have been facilitated because the endocytic machinery of the cell is absent from isolated BBMVs (Gonçalves et al., 2007).

Some studies have shown that STb could induce microscopic alterations of the intestinal mucosa due to the loss of villous absorptive cells and partial atrophy of villi (Whipp et al., 1986, 1987; Rose et al., 1987). Internalization of the STb toxin had been observed in rat intestinal loops using anti-STb colloidal gold-labelled antibodies and transmission electron microscopy (Labrie et al., 2002). No clear explanation of STb internalization process in rat intestinal jejunal cells was given at that time. Here, we clearly observe STb internalization into the NIH-3T3 cells. First, the STb toxin appears to be associated with the plasma membrane and after 2 h it enters the cell cytoplasm. Clusters of FITC–STb become visible in the cell cytoplasm and these are seen close to the nucleus [Fig. 2b(E)]. Lyamzaev et al. (2008) reported that during mitoptosis, fragmentation of mitochondria is followed by clustering of mitoptotic bodies in the perinuclear region. Although not proven in our study, this is similar to what was observed by this research team. In spite of the fact that the results obtained with the JC-1 probe indicated a connection between STb and mitochondria, the reason for this colocalization is not obvious. In fact, STb binds to sulfatide, a glycosphingolipid present on the cell membrane, and internalization follows as previously demonstrated in the rat loop assay model (Labrie et al., 2001a). In rat cells, no specific organelle was seen to be particularly targeted by the toxin. JC-1 is frequently used for detection of mitochondrial hyperpolarization (earlier events) and depolarization (later events) occurring during apoptosis (Giovannini et al., 2002; Santos et al., 2003; Lugli et al., 2005). Mitochondria are key organelles for cell survival, their role in programmed cell death is well known and mitochondrial alterations during cell death have been well described (Thress et al., 1999; Gogvadze & Orrenius, 2006; Schwarz et al., 2007). Alteration of the mitochondrial membrane potential can be independent of cytochrome C release, caspase activation and subsequent DNA degradation, which are known classical indicators of apoptosis. A cell treated with STb showed mitochondrial hyperpolarization. In addition, STb treatment induces morphological alterations associated with an apoptotic-like phenomenon (Fig. 4b), such as an enlarged nucleus and membrane budding. It was reported that other toxins can induce apoptosis by a mitochondria-dependent mechanism. For example, Helicobacter pylori VacA toxin is responsible for progressive cell vacuolization as well as gastric epithelium injury. VacA enters HeLa cells and colocalizes with the mitochondria, where it induces, after many hours, a reduction in the mitochondrial membrane potential (Willhite & Blanke, 2004; Blanke, 2005; Cover & Blanke, 2005). Clostridium difficile toxin B also causes apoptosis in epithelial cells by inducing early hyperpolarization of the mitochondria that follows a calcium-associated signalling pathway and precedes the final execution step of apoptosis (depolarization) (Matarrese et al., 2007). Cell subpopulations resulting from STb treatment indicated altered cellular states that could support the inferred cellular death observed in previous studies (Whipp et al., 1986, 1987; Rose et al., 1987). However, an apoptotic-like state was the main effect associated with STb toxin activity as the cell percentage increased linearly with the amount of STb.

Our findings suggest that pore formation could be the basis for STb toxicity (Gonçalves et al., 2007). In fact, the study on BBMVs indicated that only the functional STb toxin displayed a membrane-permeabilizing ability. The pore-forming toxin PorB of Neisseria gonorrhoeae was transported to mitochondria and induced apoptosis in Jurkat T cells (Müller et al., 2000). Mitochondrial membranes represent a target within the cells for many bacterial toxins. Although not demonstrated undoubtedly in our study, a similar phenomenon could be responsible for the altered cell population resulting from STb treatment. A recent study by Braun et al. (2007) established that a pore-forming toxin, pneumolysin, from Streptococcus pneumoniae is the key factor that accounts for cell death in primary neurons. Pnemolysin was shown to colocalize with mitochondrial membranes, altering the membrane potential. The resulting apoptosis was induced without activation of caspase-1, -3 or -8. The toxin activity is essential for the induction of mitochondrial damage and apoptosis.

Although the exact mechanism of mitochondrial targeting of STb is unknown, the effect of STb is probably transitory, explaining the mitochondria hyperpolarization observed in the first hours of intoxication. Oxidized low-density lipoprotein was shown to induce apoptosis in Caco-2 intestinal cells. Mitochondrial hyperpolarization was observed as an early event (up to 24 h), but after 48 h, a decline in mitochondrial potential, typical of apoptosis, was noted (Giovannini et al., 2002). Recently, a relationship between apoptosis, necrosis and mitochondria hyperpolarization was suggested (Skulachev, 2006; Lyamzaev et al., 2008). Mitochondrial alterations can lead to necrosis, apoptosis and mitoptosis. Mitoptosis is used to describe the ‘suicide’ related to mitochondrial disturbances, such as hyperpolarization of the mitochondrial membrane, as was observed in this study. As was concluded by Giovannini et al. (2002), mitochondrial hyperpolarization can now be considered a generalized early feature of apoptotic cell death.

Within a short time period, the pores resulting from STb action could allow leakage of ions and water from the intoxicated cells. However, more importantly, the formation of pores within the plasma membrane or the membrane disturbance resulting from cell permeation could constitute a signalling event triggering the induction of fluid secretion characteristically associated with diarrhea. Targeting of the mitochondrial membrane could represent a key factor in the signal cascade leading to an altered cell state. Structural cellular changes associated with apoptosis could explain the damaged intestinal epithelium, as observed previously in other studies (Whipp et al., 1986, 1987; Rose et al., 1987).

In conclusion, we demonstrated for the first time that STb permeabilizes NIH-3T3 cell membranes without affecting cell viability. The STb toxin targets mitochondria, inducing a hyperpolarization as an early intoxication event. Overall, in NIH-3T3 cells, we observed a progressive alteration of the cellular physiological state and morphology, suggesting the induction of an apoptotic/mitoptotic-like process. This hypothesis will have to be studied in detail by focusing more precisely on this aspect of STb activity.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was support by grants to J.D.D. (#139070-01) from the Natural Sciences and Engineering Research Council of Canada and from the Fonds Québecois de la Recherche sur la Nature et les Technologies (#2007-PR-114426). The authors wish to thank P. Vincent from the Centre de Recherche en Reproduction Animale de l'Université de Montréal for the flow cytometry analysis and his technical assistance with confocal microscopy.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Arriaga YL, Harville BA & Dreyfus LA (1995) Contribution of individual disulfide bonds to biological action of Escherichia coli heat-stable enterotoxin B. Infect Immun 63: 47154720.
  • Beausoleil HE & Dubreuil JD (2001) In vitro binding characteristics and affinity for sulfatide of Escherichia coli STb enterotoxin. Receptor Channels 7: 401411.
  • Beausoleil HE, Labrie V & Dubreuil JD (2002) Trypan blue uptake by chinese hamster ovary cultured epithelial cells: a cellular model to study Escherichia coli STb enterotoxin. Toxicon 40: 185191.
  • Blanke SR (2005) Micro-managing the executioner: pathogen targeting of mitochondria. Trends Microbiol 13: 6471.
  • Braun JS, Hoffmann O, Schickhaus M, Freyer D, Dagand E, Bermpohl D, Mitchell TJ, Bechmann I & Weber JR (2007) Pneumolysin causes neuronal cell death through mitochondrial damage. Infect Immun 75: 42454254.
  • Chao KL & Dreyfus LA (1997) Interaction of Escherichia coli heat-stable enterotoxin B with cultured human intestinal epithelial cells. Infect Immun 65: 32093217.
  • Chao KL & Dreyfus LA (1999) Interaction of Escherichia coli heat-stable enterotoxin B with rat intestinal epithelial cells and membrane lipids. FEMS Microbiol Lett 172: 9197.
  • Cossarizza A & Salvioli S (2001) Analysis of mitochondria during cell death. Method Cell Biol 63: 467486.
  • Cover TL & Blanke SR (2005) Helicobacter pylori VacA, a paradigm for toxin multifunctionality. Nat Rev 3: 320332.
  • Dreyfus LA, Harville B, Howard DE, Shaban R, Beatty DM & Morris SJ (1993) Calcium influx mediated by the Escherichia coli heat-stable enterotoxin B (STB). P Natl Acad Sci USA 90: 32023206.
  • Dubreuil JD (1997) Escherichia coli STb enterotoxin. Microbiology 143: 17831795.
  • Dubreuil JD (2008) Escherichia coli STb toxin and colibacillosis: knowing is half the battle. FEMS Microbiol Lett 278: 137145.
  • Dubreuil JD, Penel S & Gonçalves C (2007) Escherichia coli STb enterotoxin toxicity and internalization investigations: a mini-review. Res J Microbiol 2: 209215.
  • Fujii Y, Kondo Y & Okamoto K (1995) Involvement of prostaglandin E2 synthesis in the intestinal secretory action of Escherichia coli heat-stable enterotoxin II. FEMS Microbiol Lett 130: 259265.
  • Fujii Y, Nomura T, Yamanaka H & Okamoto K (1997) Involvement of Ca(2+)-calmodulin-dependent protein kinase II in the intestinal secretory action of Escherichia coli heat-stable enterotoxin II. Microbiol Immunol 41: 633636.
  • Gerhardt R, Burger S, Tatge H, Genth H, Just I & Hofman F (2005) Comparison of wild type with recombinant Clostridium difficile toxin A. Microb Pathogenesis 38: 7783.
  • Giovannini C, Mataresse P, Scazzocchio B, Sanchez M, Masella R & Malorni W (2002) Mitochondria hyperpolarization is an early event in oxidized low-density lipoprotein-induced apoptosis in Caco-2 intestinal cells. FEBS Lett 523: 200206.
  • Gogvadze V & Orrenius S (2006) Mitochondrial regulation of apoptotic cell death. Chem-Biol Interact 163: 414.
  • Gonçalves C, Vachon V, Schwartz JL & Dubreuil JD (2007) The Escherichia coli enterotoxin STb permeabilizes piglet jejunal brush border membrane vesicles. Infect Immun 75: 22082213.
  • Gonçalves C, Berthiaume F, Mourez M & Dubreuil JD (2008) Escherichia coli STb toxin binding to sulfatide and its inhibition by carragenan. FEMS Microbiol Lett 281: 3035.
  • Handl CE, Harel J, Flock JI & Dubreuil JD (1993) High yield of active STb enterotoxin from a fusion protein (MBP-STb) expressed in Escherichia coli. Protein Expres Purif 4: 275281.
  • Harel J, Lapointe H, Fallara A, Lortie LA, Bigras-Poulin M, Larivière S & Fairbrother JM (1991) Detection of genes for fimbrial antigens and enterotoxins associated with Escherichia coli serogroups isolated from pigs with diarrhea. J Clin Microbiol 29: 745752.
  • Hitotsubashi S, Fujii Y, Yamanaka H & Okamoto K (1992) Some properties of purified Escherichia coli heat-stable enterotoxin II. Infect Immun 60: 44684474.
  • Kupersztoch YM, Tachias K, Moomaw CR, Dreyfus LA, Urban R, Slaughter C & Whipp S (1990) Secretion of methanol-insoluble heat-stable enterotoxin (STB): energy- and secA-dependent conversion of pre-STB to an intermediate indistinguishable from the extracellular toxin. J Bacteriol 172: 24272432.
  • Labrie V, Beausoleil HE, Harel J & Dubreuil JD (2001a) Binding to sulfatide and enterotoxicity of various Escherichia coli STb mutants. Microbiology 147: 31413148.
  • Labrie V, Harel J & Dubreuil JD (2001b) Oligomerization of Escherichia coli enterotoxin b through its C-terminal hydrophobic alpha-helix. Biochim Biophys Acta 1535: 128133.
  • Labrie V, Harel J & Dubreuil JD (2002) Escherichia coli heat-stable enterotoxin b (STb) in vivo internalization within rat intestinal epithelial cells. Vet Res 33: 223228.
  • Le PU & Nabi IR (2003) Distinct caveolae-mediated endocytic pathways target the Golgi apparatus and the endoplasmic reticulum. J Cell Sci 116: 10591071.
  • Lee CH, Moseley SL, Moon HW, Whipp SC, Gyles CL & So M (1983) Characterization of the gene encoding heat-stable toxin II and preliminary molecular epidemiological studies of enterotoxigenic Escherichia coli heat-stable toxin II producers. Infect Immun 42: 264268.
  • Lugli E, Troiano L, Ferraresi R, Roat E, Prada N, Nasi M, Pinti M, Cooper EL & Cossarizza A (2005) Characterization of cells with different mitochondrial membrane potential during apoptosis. Cytometry A 68: 2835.
  • Lyamzaev KG, Nepryakhina OK, Saprunova VB, Bakeeva LE, Pletjushkina OY, Chernyak BV & Skulachev VP (2008) Novel mechanism of elimination of malfunctioning mitochondria (mitopsis): formulation of mitoptic bodies and extrusion of mitochondrial material from the cell. Biochim Biophys Acta 1777: 817825.
  • Matarrese P, Falzano L, Fabbri A, Gambardella L, Frank C, Geny B, Popoff MR, Malorni W & Fiorentini C (2007) Clostridium difficile toxin B causes apoptosis in epithelial cells thrilling mitochondria. J Biol Chem 282: 90299041.
  • Müller A, Gunther D, Brinkmann V, Hurwitz R, Meyer TF & Rudel T (2000) Targeting of the pro-apoptotic VDAC-like porin (PorB) of Neisseria gonorrhoeae to mitochondria of infected cells. EMBO J 19: 53325343.
  • Ohnishi H, Miyake M, Kamitani S & Horiguchi Y (2008) The morphological changes in cultured cells caused by Bordetella pertussis adenylate cyclase toxin. FEMS Microbiol Lett 279: 174179.
  • Okamoto K, Baba T, Yamanaka H, Akashi N & Fujii Y (1995) Disulfide bond formation and secretion of Escherichia coli heat-stable enterotoxin II. J Bacteriol 177: 45794586.
  • Peterson JW & Whipp SC (1995) Comparison of the mechanisms of action of cholera toxin and the heat-stable enterotoxins of Escherichia coli. Infect Immun 63: 14521461.
  • Reers M, Smiley ST, Mottola-Hartshorn C, Chen A, Lin M & Chen LB (1995) Mitochondrial membrane potential monitored by JC-1 dye. Method Enzymol 260: 406417.
  • Rose R, Whipp SC & Moon HW (1987) Effects of Escherichia coli heat-stable enterotoxin b on small intestinal villi in pigs, rabbits, and lambs. Vet Pathol 24: 7179.
  • Rousset E, Harel J & Dubreuil JD (1998) Sulfatide from the pig jejunum brush border epithelial cell surface is involved in binding of Escherichia coli enterotoxin b. Infect Immun 66: 56505658.
  • Santos JH, Hunakova L, Chen Y, Bortner C & Van Houten B (2003) Cell sorting experiments link persistent mitochondrial DNA damage with loss of mitochondrial membrane potential and apoptotic cell death. J Biol Chem 278: 17281734.
  • Schwarz M, Andrade-Navarro MA & Gross A (2007) Mitochondrial carriers and pores: key regulators of the mitochondrial apoptotic program? Apoptosis 12: 869876.
  • Skulachev VP (2006) Bioenergetic aspects of apoptosis, necrosis and mitoptosis. Apoptosis 11: 473485.
  • Smiley ST, Reers M, Mottola-Hartshorn C, Lin M, Chen A, Smith TW, Steele Jr GD & Chen LB (1991) Intracellular heterogeneity in mitochondrial membrane potentials revealed by a J-aggregate-forming lipophilic cation JC-1. P Natl Acad Sci USA 88: 36713675.
  • Sukumar M, Rizo J, Wall M, Dreyfus LA, Kupersztoch YM & Gierasch LM (1995) The structure of Escherichia coli heat-stable enterotoxin b by nuclear magnetic resonance and circular dichroism. Protein Sci 4: 17181729.
  • Thress K, Kornbluth S & Smith JJ (1999) Mitochondria at the crossroad of apoptotic cell death. J Bioenerg Biomembr 31: 321326.
  • Whipp SC, Moseley SL & Moon HW (1986) Microscopic alterations in jejunal epithelium of 3-week-old pigs induced by pig-specific, mouse-negative, heat-stable Escherichia coli enterotoxin. Am J Vet Res 47: 615618.
  • Whipp SC, Kokue E, Morgan RW, Rose R & Moon HW (1987) Functional significance of histologic alterations induced by Escherichia coli pig-specific, mouse-negative, heat-stable enterotoxin (STb). Vet Res Commun 11: 4155.
  • Willhite DC & Blanke SR (2004) Helicobacter pylori vacuolating cytotoxin enters cells, localizes to the mitochondria, and induces mitochondrial membrane permeability changes correlated to toxin channel activity. Cell Microbiol 6: 143154.