Cellular Uptake of the Clostridium perfringens Binary Iota-Toxin

Abstract

The binary iota-toxin is produced by Clostridium perfringens type E strains and consists of two separate proteins, the binding component iota b (98 kDa) and an actin-ADP-ribosylating enzyme component iota a (47 kDa). Iota b binds to the cell surface receptor and mediates the translocation of iota a into the cytosol. Here we studied the cellular uptake of iota-toxin into Vero cells. Bafilomycin A1, but not brefeldin A or nocodazole, inhibited the cytotoxic effects of iota-toxin, indicating that toxin is translocated from an endosomal compartment into the cytoplasm. Acidification (pH ≤ 5.0) of the extracellular medium enabled iota a to directly enter the cytosol in the presence of iota b. Activation by chymotrypsin induced oligomerization of iota b in solution. An average mass of 530 ± 28 kDa for oligomers was determined by analytical ultracentrifugation, indicating heptamer formation. The entry of iota-toxin into polarized CaCo-2 cells was studied by measuring the decrease in transepithelial resistance after toxin treatment. Iota-toxin led to a significant decrease in resistance when it was applied to the basolateral surface of the cells but not following application to the apical surface, indicating a polarized localization of the iota-toxin receptor.

Bacterial toxins which act within the cytosol of eucaryotic cells have to be transported across lipid membranes to reach their targets. So far, two mechanisms are known by which the catalytic domain of toxins can be translocated into the cytosol. One group of toxins (e.g., diphtheria toxin, anthrax toxin, and Clostridium botulinum C2 toxin) is trafficked into endosomes after endocytosis, and acidification of this compartment leads to conformational changes in the translocation domain. This domain is then able to insert into the endosomal membrane, eventually resulting in delivery of the catalytic domain into the cytosol (2, 19, 34). A second mechanism employed by some toxins (e.g., cholera toxin) involves transport to the trans-Golgi network (TGN) after endocytosis. In a retrograde manner, these toxins are transported from the TGN to the endoplasmatic reticulum, where the catalytic domain is delivered to the cytosol (24, 32).

Clostridium perfringens iota-toxin belongs to the family of binary toxins, in which the translocation and enzyme domains are located on two individual, nonlinked proteins (37, 38). Other members of the family are Clostridium botulinum C2 toxin (1, 9), Clostridium spiroforme toxin (27), Clostridium difficile ADP-ribosyltransferase (28), the vegetative insecticidal proteins from Bacillus cereus (14), and anthrax toxin from Bacillus anthracis (18). Iota a, the enzyme component of iota-toxin, modifies actin by ADP-ribosylation at arginine-177 (31, 39). The ADP-ribosylation leads to breakdown of the cytoskeleton by inhibiting actin polymerization (1). In contrast to C2 toxin of Clostridium botulinum, iota a ADP-ribosylates both muscle and nonmuscle actin whereas C2I, the enzyme component of C2 toxin, modifies only nonmuscle actin (20, 31). The binding component of iota-toxin, iota b, recognizes an unknown cell surface receptor and mediates cell entry of iota a (9). To obtain biological activity, iota a and iota b have to be activated by protease cleavage (12). Iota b shows significant sequence similarities to the binding components of C2 toxin and anthrax toxin (C2II and protective antigen, respectively) (16, 25). Whereas the structure and function of C2II and protective antigen are well characterized (2, 6, 18), not much is known about iota b. In the present work, we show that after internalization by endocytosis, iota-toxin is delivered to an endosomal compartment by a microtubule-independent mechanism. From this compartment, iota a appears to be translocated to the eucaryotic cytosol by an acidic pH-dependent mechanism comparable to that of C2I. Again similar to C2II, the proteolytic activation of iota b induces heptamer formation in vitro.

MATERIALS AND METHODS

Materials.

Cell culture medium was from Biochrom (Berlin, Germany), fetal calf serum was from PAN Systems (Aidenbach, Germany), and cell culture materials were from Falcon (Heidelberg, Germany). The B. anthracis RP31 Ib strain was a gift from M. Mock, Institut Pasteur, Paris, France. The CTR433 anti-Golgi monoclonal antibody was a gift from M. Bornens, Institut Curie, Paris, France. The monoclonal mouse anti-α-tubulin antibody was from Sigma (Deisenhofen, Germany). The Cy3-labeled anti-mouse antibody was from Dianova (Hamburg, Germany). Thrombin and brefeldin A were from Sigma. [³²P]NAD (30 Ci/mmol) was from DuPont NEN (Bad Homburg, Germany) and Na¹²⁵I was from Hartmann (Braunschweig, Germany). Trypsin and trypsin inhibitor were from Boehringer. Bafilomycin A1 and nocodazole were from Calbiochem (Bad Soden, Germany). Vivaspin 4 ML concentrator devices were from Vivascience Ltd. (Binbrook Hill, United Kingdom). Cell culture inserts (pore size, 0.4 μm) were from Falcon.

Purification of C2II and C2I.

Recombinant C2II and C2I were purified as glutathione S-transferase (GST) fusion proteins with the GST gene fusion system from Pharmacia Biotech (Uppsala, Sweden) and cleaved with thrombin as described previously (3, 6). For activation, C2II was incubated for 20 min at 37°C with 0.2 μg of trypsin/μg of C2II and subsequently with 0.4 μg of trypsin inhibitor/μg of C2II for 1 h at 4°C to block the effects of trypsin. Activation and oligomerization of activated C2II were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (3 to 12.5% polyacrylamide) with or without prior boiling of the samples, respectively.

Purification and activation of iota b and iota a.

Recombinant iota b was expressed in B. anthracis RP31 Ib and purified as described previously (35). Briefly, B. anthracis RP31 Ib was grown for 16 h at 37°C under 5% CO₂ in R medium supplemented with sodium bicarbonate. The supernatant was collected by filtration, and proteins were precipitated in 70% ammonium sulfate. The precipitate was dialyzed against 10 mM Tris-HCl (pH 7.5) and concentrated with a Vivaspin 4 ML concentrator. The protein was loaded on a Resource Q column (Pharmacia), eluted with 25 mM NaCl in Tris buffer (pH 7.5), and dialyzed against 10 mM Tris-HCl (pH 7.5). Recombinant iota a was expressed as a GST fusion protein in Escherichia coli BL21 cells. For activation, iota b and iota a were incubated for 20 min at 37°C with 0.2 μg of chymotrypsin/μg of protein and subsequently with 0.4 μg of trypsin inhibitor/μg of protein for 1 h at 4°C to block chymotrypsin effects. Activation of iota b and iota a was analyzed by SDS-PAGE (12.5% polyacrylamide).

Analytical ultracentrifugation.

Molecular mass studies on iota b were carried out using an An-60 Ti rotor in an XL-A type analytical ultracentrifuge (Beckman Instruments, Palo Alto, Calif.) equipped with UV absorbance optics. The sedimentation equilibrium technique was used to directly determine the molecular mass. The experiments were performed by using externally loaded six-channel cells with an optical path length of 12 mm, filled with about 70 μl of solute. This cell type allows the analysis of three solvent-solution pairs. Three of these cells were used to simultaneously analyze different samples in one run. Sedimentation equilibrium was attained after 2 h of overspeed at 10,000 rpm followed by an equilibrium speed at 8,000 rpm at 10°C for about 32 h. The radial absorbances of each compartment were recorded at three different wavelengths between 280 and 310 nm using the molar absorbance coefficients. Molecular mass determination was performed by simultaneously fitting the sets of three radial absorbance distributions described by the equation A_r = A_r,m exp[MF(r² − r_m²)] with F = [(1 − ν̄ρ)ω²]/2RT, using our computer program Polymole (5). In these equations ρ is solvent density, ν̄ is the partial specific volume, ω is the angular velocity, R is the gas constant, and T is the absolute temperature. A_r is the radial absorbance, and A_r,m is the corresponding value at the meniscus position. The partial specific volume was calculated from the amino acid composition and the density increments of the individual amino acids.

Cell culture and cytotoxicity assays.

Vero, EBL, and CaCo-2 cells were cultivated in tissue culture flasks at 37°C under 5% CO₂. EBL cells were grown in minimal essential medium (MEM) supplemented with 15% fetal calf serum. Vero and CaCo-2 cells were grown in Dulbecco's minimal essential medium (DMEM) supplemented with 10% fetal calf serum. All media contained 100 U of penicillin per ml and 100 μg of streptomycin per ml. The cells were routinely trypsinized and reseeded three times a week. For cytotoxicity experiments, the cells were seeded in small dishes and incubated with the respective drug or toxin in serum-free DMEM. For assays with inhibitors of endosomal acidification, microtubule polymerization, and Golgi vesicle trafficking, respectively, the cells were treated with 200 ng of iota b per ml and 100 ng of iota a per ml in the presence or absence of the following drugs: brefeldin A (50 μg/ml), nocodazole (30 mM), and bafilomycin A1 (100 nM). For resistance measurements, CaCo-2 cells were seeded onto cell culture inserts (Falcon). They were grown for 5 to 7 days to form confluent monolayers with high transepithelial resistances (>500 ω/cm²)

ADP-ribosylation assay.

The cells were washed with cold phosphate-buffered saline (PBS), scraped off into 300 μl of cold lysis buffer (2 mM MgCl₂ in 50 mM HEPES [pH 7.4]), and sonicated on ice with 10 strokes each for 5 s at 50% of maximal power (Bandelin Sonopuls HD60). The protein concentration was determined by the method of Bradford (7). ADP-ribosylation was performed as described previously (3). In brief, 50-μg portions of lysate proteins were each incubated with [³²P]NAD (0.5 μM) and 50 ng of iota a for 15 min at 37°C. The reaction was stopped by addition of Laemmli buffer, and the samples were heated for 3 min at 95°C and subjected to SDS-PAGE. [³²P]ADP-ribosylated proteins were detected by autoradiography with a PhosphorImager from Molecular Dynamics (Krefeld, Germany).

Radiolabeling of iota b.

Iota b was labeled with ¹²⁵I by using Iodobeads from Pierce as specified by the manufacturer.

Electrophysiology.

CaCo-2 monolayers confluently growing on cell culture inserts (Falcon) were moved to serum-free DMEM for measurements. Transepithelial resistance was measured in the presence of applied 40-μA currents with an Endohm tissue resistance measurement chamber from World Precision Instruments. Monolayers that did not maintain a resistance of >500 ω/cm² were excluded from the study.

Staining of Golgi and microtubules.

Cells were grown on coverslips and treated with the respective drug or toxin. They were rinsed twice with PBS and fixed with 4% paraformaldehyde for 10 min. They were subsequently treated with ice-cold 50 mM NH₄Cl in PBS for 1 min and with ice-cold PBS plus 1% Triton X-100 for 3 min. After being washed three times with PBS, they were blocked with PBS–0.5% bovine serum albumin (BSA) for 15 min. They were then probed for 30 min with the primary antibody diluted in PBS–0.5% BSA (anti-Golgi, 1:500; anti-tubulin, 1:1,000) and subsequently washed with PBS–0.5% BSA. They were incubated with the secondary antibody (Cy3-labeled anti-mouse antibody, diluted 1:400 in PBS–0.5% BSA) for 30 min. For confocal laser-scanning microscopy, coverslips were embedded in Kaiser's gelatin on glass.

RESULTS

Brefeldin A does not inhibit the uptake of iota-toxin into cells.

The fungal drug brefeldin A inhibits the small GTPase ARF1, which is involved in vesicle formation at the level of the Golgi apparatus (8). This disrupts trafficking between the Golgi and the ER and leads to destruction of the Golgi apparatus. To test whether iota-toxin is transported via the Golgi into the cytosol, we treated Vero cells with 50 μg of brefeldin A per ml prior to iota-toxin addition. As shown in Fig. 1A, brefeldin A did not inhibit iota-toxin-induced cell rounding. Brefeldin A itself showed no morphological effects within the incubation times indicated. Subsequent in vitro ADP-ribosylation assays with [³²P]NAD and iota a confirmed these results. Actin from brefeldin A-pretreated cells was not radioactively labeled, indicating entry and action of iota a (Fig. 1B). Disruption of the Golgi apparatus was monitored by immunofluorescence staining using an antibody directed against a Golgi protein (15) (data not shown). These results suggested that iota-toxin does not translocate into the cytosol via the TGN-Golgi route. We therefore concluded that the iota-toxin-translocating compartment is either the late or the early endosome.

FIG. 1.

Influence of brefeldin A on the uptake of iota-toxin into Vero cells. (A) Cells were preincubated with 50 μg of brefeldin A per ml and subjected to a 3-h incubation with iota-toxin (100 ng of iota a per ml and 200 ng of iota b per ml). Phase-contrast micrographs were obtained, and the cells were lysed for a subsequent in vitro ADP-ribosylation assay with iota a. (B) Autoradiography of [³²P]ADP-ribosylated actin from 50 μg of lysate protein each. Lanes: 1, control; 2, iota-toxin; 3, brefeldin A; 4, brefeldin A plus iota-toxin.

Cellular uptake of iota-toxin requires an acidic compartment.

To investigate whether iota-toxin uptake requires an acidic cellular compartment, we applied bafilomycin A1 to cells prior to toxin treatment. Bafilomycin A1 blocks endosomal acidification by specifically inhibiting the vacuolar-type H⁺-ATPase (V-ATPase) (30, 40). As shown in Fig. 2A, cells which were incubated with iota-toxin without bafilomycin A1 were round. By contrast, cells which were treated with iota-toxin in the presence of bafilomycin A1 exhibited normal morphology. Bafilomycin A1 alone did not induce any morphological effects within the incubation times indicated. To confirm this result, the cell lysates were subjected to an in vitro ADP-ribosylation assay with [³²P]NAD and iota a. The autoradiograph of [³²P]ADP-ribosylated actin in Fig. 2B shows that only the actin from cells treated with iota-toxin without bafilomycin A1 was not radiolabeled, indicating entry and action of iota toxin. By contrast, radiolabeling of actin was observed when cells were previously treated with iota-toxin in the presence of bafilomycin A1. These results indicate that bafilomycin A1 prevents iota-toxin uptake and suggest that an acidic compartment is required for translocation of iota-toxin into the cytosol.

FIG. 2.

Bafilomycin A1 inhibits the cytotoxic effects of iota-toxin on Vero cells. Vero cells were pretreated with 100 nM bafilomycin A1 for 30 min at 37°C. Iota-toxin was added (100 ng of iota a per ml and 200 ng of iota b per ml), and the cells were further incubated at 37°C. (A) Micrographs were obtained after 3 h. (B) Cells were lysed, and the lysates were analyzed by an in vitro ADP-ribosylation assay with iota a. Lanes: 1, control; 2, iota-toxin; 3, bafilomycin A1; 4, bafilomycin A1 plus iota-toxin.

Iota-toxin is delivered into cells by extracellular acidification in the presence of bafilomycin A1.

Because iota b mediates the translocation of iota a across the endosomal membrane, we wondered whether iota-toxin is delivered directly into the cytosol after extracellular acidification. To block endosomal toxin uptake, Vero cells were preincubated for 30 min at 37°C with bafilomycin A1. Thereafter, medium at pH 7.5 or 4.5 containing bafilomycin plus iota-toxin was added to the cells. After incubation of the cells for 15 min at 37°C, fresh medium (37°C, pH 7.5) containing bafilomycin A1 was added and cells were further incubated for 3 h at 37°C. As shown in Fig. 3, after bafilomycin treatment only cells which were exposed to pH 4.5 medium exhibited iota-toxin-induced morphology, whereas cells which were not shifted to acidic pH did not round up. Cells did not round up following exposure to pH 4.5 medium containing only iota a and not iota b (data not shown). For a more detailed characterization of the pH dependence of the uptake of iota-toxin, the effects of pH values between 4.5 and 5.5 were studied. Cells which were exposed to pH values between 4.5 and 4.9 in the presence of iota-toxin were completely round. By contrast, cells possessed normal morphology after exposure to pH 5.0 and higher. In this respect, iota-toxin differs from C2 toxin. As shown previously, C2 toxin can be delivered directly into the cytosol in the presence of bafilomycin when the extracellular pH is adjusted to pH 5.4 or below (2).

FIG. 3.

Iota-toxin is transferred directly from the plasma membrane to the cytosol by a short exposure to an acidic pH. Vero cells were pretreated with bafilomycin A1 for 30 min at 37°C. Subsequently, the cells were incubated for 15 min with medium at pH 7.5 or 4.5, containing iota-toxin with or without bafilomycin A1. Thereafter the cells were washed and further incubated in serum-free medium. After 3 h, micrographs were obtained. (A) Control; (B) cells incubated with pH 4.5 medium; (C) iota-toxin at pH 7.5; (D) bafilomycin A1 plus iota-toxin at pH 7.5; (E) iota-toxin at pH 4.5; (F) bafilomycin A1 plus iota-toxin at pH 4.5.

Microtubules are not involved in the uptake of iota-toxin.

Transport from the early to the late endosomal compartment depends on microtubules and can be inhibited by the microtubule depolymerizing drug nocodazole (13, 29). To test the influence of the microtubule system on iota-toxin uptake, Vero cells were incubated for 4 h with nocodazole prior to toxin treatment. Disruption of the microtubules was monitored by immunofluorescence staining, using an anti-α-tubulin antibody. Because nocodazole treatment led to significant changes of cell morphology, cellular effects of iota-toxin were determined by subsequent in vitro actin ADP-ribosylation (Fig. 4). Actin from nocodazole-pretreated cells was not radioactively labeled when the cells were subsequently treated with iota a and [³²P]NAD. These data indicate that the microtubule system is not essential for uptake of iota-toxin into eucaryotic cells and suggest that iota a is released from early endosomes into the cytosol.

FIG. 4.

Influence of nocodazole on the effects of iota-toxin on Vero cells. Vero cells were incubated for 4 h with 30 mM nocodazole. Iota-toxin was added (100 ng of iota a per ml, and 200 ng of iota b per ml), and the cells were further incubated for 3 h at 37°C. The cells were lysed, and in vitro ADP-ribosylation with [³²P]NAD and iota a was performed. Control cells were incubated without any drug or toxin. Lanes:1, control; 2, iota-toxin; 3, nocodazole plus iota-toxin.

Iota b forms oligomers after proteolytic activation.

Both C2II and anthrax protective antigen (PA) form heptamers in solution after proteolytic activation (2, 18). Because iota b shows a high degree of homology to both proteins, we investigated the formation of oligomers by iota b. C2II heptamers are SDS stable and can be detected when proteolytically activated C2II is subjected to SDS-PAGE without prior heating (Fig. 5A) (2). When iota b was cleaved by chymotrypsin and heated for 3 min at 95°C, the protein migrated as a 74-kDa protein on SDS-PAGE (Fig. 5A). When iota b was subjected to gradient SDS-PAGE (3 to 12.5% polyacrylamide) without prior heating, two additional bands with molecular masses of approximately 150 to 200 kDa were detected after Coomassie blue staining (Fig. 5A). However, no high-molecular-mass complexes comparable to the C2 heptamers could be detected. The same result was obtained when radiolabeled iota b was used (Fig. 5B, lane 2). For PA, it has been reported (21) that SDS-stable oligomers are formed only at acidic pH (pH 7 or below). Therefore we analyzed radiolabeled iota b after adjusting the pH of the protein solution to 4.5. As shown in Fig. 5B, iota b oligomers were formed after acidification. However, the major part of the activated iota b migrated as a monomer. By contrast, even after acidification no formation of oligomers was observed with inactive iota b (Fig. 5B). These findings indicate that chymotrypsin-cleaved iota b forms oligomers in solution. However, high-molecular-mass complexes comparable to the C2II heptamers were detectable only after acidification of the protein.

FIG. 5.

Formation of oligomers by chymotrypsin-activated iota b. (A) Iota b was incubated with 0.2 μg of chymotrypsin/μg of iota b for 20 min at 37°C and compared with trypsin-activated C2II. The proteins were subjected to SDS-PAGE (3 to 12.5% polyacrylamide) with or without prior heating at 95°C. The proteins were then stained with Coomassie blue. Lanes: 1, protein with prior heating; 2, protein without prior heating. (B) Radiolabeled iota b was activated with chymotrypsin. The pH of native and activated iota b was adjusted to 4.5 or kept at 7.5. The proteins were subjected to SDS-PAGE (3 to 12.5% polyacrylamide) without prior heating. The proteins were visualized by autoradiography. Lanes: 1, iota b at pH 4.5; 2, iota b at pH 7.5; 3, inactive iota b at pH 4.5; 4, inactive iota b at pH 7.5. Iba, chymotrypsin-activated iota b; C2IIa, trypsin activated C2II.

Chymotrypsin-activated iota b forms heptamers in solution.

To determine the size of the iota b oligomers, the molecular mass of the complex was analyzed by analytical ultracentrifugation. Therefore, iota b was activated with chymotrypsin and analyzed at concentrations between 0.1 and 0.5 mg/ml. Since the solutions contained some low-molecular-mass material besides the oligomers, an equilibrium speed of 8,000 rpm was used. This procedure allowed us to consider the low-molecular-mass material to be buried in the baseline with negligible increase of the radial absorbance at sedimentation equilibrium. By using the program Polymole, the data were fitted simultaneously (Fig. 6). An average molecular mass of 530 ± 28 kDa was determined. Similar data were obtained for 0.1 or 0.5 mg of iota b per ml. The molecular mass of about 530 kDa exceeds the value of monomers (75 kDa) about sevenfold. This finding indicates that high-molecular-mass components are on average heptamers. The same result was obtained when the pH of the iota b solution was adjusted to 4.5 prior to the measurements. No heptamers could be detected when native iota b was analyzed without prior activation.

FIG. 6.

Radial absorbance distributions of toxin iota b (loading concentration, 0.2 mg/ml in 10 mM Tris-HCl [pH 7.4]) at sedimentation equilibrium recorded at 295 nm (○), 300 nm (●), and 305 nm (□) at 10°C. From the simultaneous curve fit, an average molecular mass of 530 ± 28 kDa was calculated.

Influence of acidic pH on the activity of iota b.

For PA it has been reported that acidic conditions trigger the conversion of the heptameric prepore to a pore-like form in solution (s-pore) (21). Moreover, transition of the prepore to the s-pore is not reversed when the pH of the PA sample is readjusted to higher pH values. Apparently, the s-pore of PA is inactive, due to its inability to insert into the endosomal membrane and its failure to deliver an enzyme component into the cytosol. Therefore, we tested whether iota b was still active when the pH of a chymotrypsin-cleaved iota b preparation was adjusted to pH 4.5 prior to its addition to cells. Addition of the low-pH preparation of iota b, together with the enzyme component iota a, to cells caused cell rounding, indicating that the acidification had no effect on the biological activity of iota b (Table 1). In a second set of experiments, the pH of the cell culture medium was adjusted to 4.5 prior to the addition of toxin components. Thereafter, cells were incubated for 15 min at 37°C with the toxin components to allow endocytosis. The acidic medium was removed, and the cells were washed and further incubated at 37°C with serum-free medium at pH 7.5. Also under these conditions, the acidic pH had no effect on the cytotoxic activity of iota toxin on Vero cells (Table 1). This result was obtained over a wide iota b concentration range (20 to 500 ng/ml). Therefore, it seems that iota b differs from PA and, as shown in Table 1, also from C2II. Acidification of C2II led to a protein which was no longer able to deliver C2I into the cytosol. Notably, the inactivation of C2II by low pH was reversed when the pH of the protein was readjusted to 7.5 (Table 1).

TABLE 1.

Effect of acidic pH on activity of iota b and C2II

Toxin	pH of:		Cell rounding
	Protein solution	Medium
Iota-toxin	7.5	7.5	+
	7.5	4.5	+
	4.5	7.5	+
	4.5	4.5	+
C2 toxin	7.5	7.5	+
	7.5	4.5	+
	4.5	7.5	+
	4.5	4.5	−

Uptake of iota toxin into polarized CaCo-2 cells.

It has been reported previously that iota-toxin presumably binds to a proteinaceous cell surface receptor (36). Here we studied whether this receptor is distributed equally to the apical and basolateral sides of polarized CaCo-2 cells. Iota-toxin (200 ng of iota b per ml, and 100 ng of iota a per ml) was applied to either the apical or the basolateral side of polarized CaCo-2 cells. Cells were incubated at 37°C, and the transepithelial resistance was measured every 30 min. As shown in Fig. 7A, iota-toxin caused a significant decrease of resistance only when it was applied to the basolateral surface of the cells. The decrease of resistance was dependent on the presence of both iota b and iota a (data not shown). By contrast, C2 toxin led to a decrease of resistance when applied to either the apical or the basolateral side of the cells (Fig. 7B). To investigate the reason for the polarity of iota-toxin action on CaCo-2 cells, we examined the hypothesis that the receptor for iota-toxin might be absent from the apical surface of the cells. Radiolabeled, activated iota b was applied to the basolateral or the apical side of polarized CaCo-2 cells. After incubation on ice for 1 h, the membranes were washed and cut from the cell culture inserts. Bound iota b was quantified by scintillation counting. As shown in Fig. 8, the amount of iota b bound to the basolateral surface of the cells was significantly larger than the amount bound to the apical surfaces. These results provide evidence that the iota b receptor is distributed asymmetrically on polarized CaCo-2. In contrast, the C2 toxin receptor is presumably distributed equally to both sides of the cells, indicating that different receptors are involved in the binding and uptake of C2 toxin and iota-toxin.

FIG. 7.

Time course of iota-toxin- and C2 toxin-induced decrease of transepithelial resistance in CaCo-2 cells. (A) Activated iota a (100 ng/ml) and activated iota b (200 ng/ml) were added at time zero to either the apical or the basolateral side of confluent CaCo-2 cells grown in tissue culture inserts. Resistance was measured every 30 min. ⧫, control; ■, iota-toxin apical; □, iota-toxin basolateral (means ± standard deviations [n = 3; for control cells, n = 2]) (B) Activated C2II (200 ng/ml) and C2I (100 ng/ml) were added at time zero to either the apical or the basolateral side of confluent CaCo-2 cells grown in cell culture inserts. Resistance was measured over 200 min. ⧫, control; ■, C2 toxin apical; □, C2 toxin basolateral. The results of one representative experiment are shown.

FIG. 8.

Iota b binds to the basolateral surface of polarized CaCo-2 cells. Confluent monolayers of CaCo-2 cells grown in tissue culture inserts were chilled at 4°C. Radiolabeled iota b was added and the cells were incubated for 1 h on ice. Thereafter, unbound material was washed off with ice-cold PBS. The filters were cut off from the tissue culture inserts, and the amount of bound iota b was determined by scintillation counting.

DISCUSSION

Sequencing of the gene encoding iota b revealed high homology to the binding components of C2 toxin and B. anthracis anthrax toxin (C2II and PA, respectively) (16, 25). Recent studies have shed some light on the molecular mechanisms underlying the cellular uptake of these toxins. To obtain full biological activity, PA and C2II must be proteolytically cleaved (17, 23). Both toxins form heptamers and translocate from endosomes into the cytosol by a pH-dependent step (2, 34). In contrast to C2 and anthrax toxins, it was reported that transport of iota-toxin is inhibited by brefeldin A but not by bafilomycin A1, suggesting that iota-toxin must be transported to the Golgi prior to the release of the A component into the cytosol (26). Additionally, attempts to demonstrate oligomerization of iota b were not successful (26). Therefore, these initial findings suggesting differences among these closely related toxins prompted us to study the uptake of iota-toxin in more detail.

Here we show that brefeldin A does not impair the effects of iota-toxin on Vero cells. The data indicate that the Golgi apparatus is not involved in the transport of iota-toxin to the cytosol. By contrast, the fungal drug bafilomycin A1 completely blocked the effects of iota-toxin on Vero cells, suggesting that an acidic compartment is required for iota-toxin uptake. These results were corroborated by the finding that iota-toxin is efficiently delivered directly from the plasma membrane into the cytosol by extracellular acidification. Taken together, the data indicate that translocation of iota-toxin across a lipid bilayer is triggered by low pH.

Iota-toxin activity in Vero cells was not influenced by the microtubule-depolymerizing drug nocodazole. Likewise, the cytotoxic action of C2 toxin is not blocked by nocodazole (2). These findings indicate that both toxins enter the cytosol from an early endosomal compartment. However, whereas for C2 toxin an extracellular pH of 5.4 was sufficient for the translocation of C2I into the cytosol (2), iota-toxin required a pH below 5.0 for translocation. The reason for this discrepancy is not known. However, the different pH requirements for toxin uptake may indicate spatial and temporal differences of the membrane translocation process of iota-toxin and C2 toxin.

A prerequisite for the delivery of the enzyme components of C2 and anthrax toxins into the cytosol is the activation of the binding components by partial proteolysis (17, 23). After activation, C2II and PA form heptamers in solution (2, 22). It has been reported previously that iota b also must be activated to obtain biological activity (12). Here we show by analytical ultracentrifugation that, similar to C2II and PA, iota b forms heptamers after activation with chymotrypsin. The C2II heptamers are SDS stable (2), and the PA heptamers reach a similar stability after acidification of the protein solution (21). By contrast, iota b heptamers appear to be more labile than PA and C2II heptamers. Even after the pH of the protein solution was adjusted to 4.5, the majority of iota b migrated in SDS gels as monomers. However, the migration behavior of iota b at pH 4.5 clearly differed from that at pH 7.5. Therefore, we assume that conformational changes occur after acidification. For PA it has been reported that low pH triggers the conversion of a prepore-like conformation to a pore-like conformation (s-pore) (21). The formation of the s-pore is reportedly irreversible and prevents membrane insertion. Also, C2II undergoes conformational changes and/or membrane insertion at low pH (2). By comparing the influence of acidic pH on the activity of iota b and C2II, we observed additional important differences between the two toxins. Whereas low pH had no influence on the biological activity of iota b, C2II was inactive at low pH but regained activity after readjustment of the pH. Thus, conformational changes of iota-toxin induced by low pH did not inhibit receptor binding, binding of the enzyme component iota a, or transport of iota a into the cytosol. By contrast, structural changes of free C2II induced by low pH appear to inhibit subsequent binding and/or interaction with the enzyme component and its translocation.

At present, the role of iota b oligomerization in the intoxication process remains unclear. However, sequence homologies to PA and C2II and the similar uptake mechanisms of all three toxins suggest that iota b forms a heptameric pore which is also involved in translocation of iota a across the endosomal membrane.

Iota-toxin decreased the transepithelial resistance of CaCo-2 cells only when applied to the basolateral cell surfaces. This is most probably due to polarized localization of the toxin receptor. In this respect, iota-toxin resembles edema toxin of B. anthracis, which elicits a response only when applied to basolateral surfaces of polarized T84 cells (4). By contrast, C2 toxin is able to enter polarized CaCo-2 cells from the apical and the basolateral surfaces. This finding is in line with the notion that iota-toxin and C2 toxin bind to different cellular receptors (11). Whereas C2II binds to asparagine-linked complex and/or hybrid carbohydrates (10), it has been suggested that iota b binds to a proteinaceous receptor (36). The two receptors also differ with respect to cell surface expression. The C2 receptor is ubiquitously expressed, and therefore all known cell lines are sensitive to C2 toxin (reference 33 and unpublished results). By contrast, several cell lines are known which are highly resistant to iota toxin, e.g., HeLa, NIH 3T3, and MRC-5 cells (reference 36 and unpublished results).

Taken together, our data show that iota b has features in common with C2II and PA with respect to activation, oligomerization into a heptamer, and cellular uptake mechanism. This is in agreement with the similar primary structures for domains I to III of these toxins (18). However, besides binding to different cell membrane receptors, important differences exist between the toxins with respect to the stability of the oligomers, pH dependence of toxin translocation, and reversibility of pH-induced structural changes.