The epsilon toxin from Clostridium perfringens stimulates calcium-activated chloride channels, generating extracellular vesicles in Xenopus oocytes

Abstract

The epsilon toxin (Etx) from Clostridium perfringens has been identified as a potential trigger of multiple sclerosis, functioning as a pore-forming toxin that selectively targets cells expressing the plasma membrane (PM) myelin and lymphocyte protein (MAL). Previously, we observed that Etx induces the release of intracellular ATP in sensitive cell lines. Here, we aimed to re-examine the mechanism of action of the toxin and investigate the connection between pore formation and ATP release. We examined the impact of Etx on Xenopus laevis oocytes expressing human MAL. Extracellular ATP was assessed using the luciferin-luciferase reaction. Activation of calcium-activated chloride channels (CaCCs) and a decrease in the PM surface were recorded using the two-electrode voltage-clamp technique. To evaluate intracellular Ca²⁺ levels and scramblase activity, fluorescent dyes were employed. Extracellular vesicles were imaged using light and electron microscopy, while toxin oligomers were identified through western blots. Etx triggered intracellular Ca²⁺ mobilization in the Xenopus oocytes expressing hMAL, leading to the activation of CaCCs, ATP release, and a reduction in PM capacitance. The toxin induced the activation of scramblase and, thus, translocated phospholipids from the inner to the outer leaflet of the PM, exposing phosphatidylserine outside in Xenopus oocytes and in an Etx-sensitive cell line. Moreover, Etx caused the formation of extracellular vesicles, not derived from apoptotic bodies, through PM fission. These vesicles carried toxin heptamers and doughnut-like structures in the nanometer size range. In conclusion, ATP release was not directly attributed to the formation of pores in the PM, but to scramblase activity and the formation of extracellular vesicles.

Keywords: TMEM16A; Xenopus laevis oocytes; calcium‐activated chloride channels; epsilon toxin; extracellular vesicles; pore‐forming toxins; scramblase.

FIGURE 1

Etx induced ATP release from individual Xenopus laevis oocytes. (A) ATP release from single oocytes. Etx was added at time 0. The green trace is the mean of the relative increase in light due to the release of ATP from oocytes expressing hMAL. Solution, (FR) with CaCl₂ and Etx 300 nM (N = 10). Black trace, wild‐type (WT) oocytes in the same condition (N = 10). Blue trace, oocytes expressing hMAL treated with 300 nM of the prototoxin, pEtx (N = 5). Vertical bars represent ± SEM at the indicated times. Note that in WT treated with Etx and in hMAL oocytes treated with prototoxin, the negative relative luminescence values were due to the spontaneous decline in the initial light emitted by the luciferin‐luciferase reaction, indicating that no ATP was released. (B) Etx (300 nM) induced currents. Red trace, hMAL expressing oocytes, N = 12. t Black trace, WT oocytes, N = 10). The membrane potential fixed at −40 mV. Traces are the mean of the traces obtained from individual oocytes. Vertical bars represent ± SEM at the indicated times; for clarity, only half of the bars are shown. Violet trace, a known pore‐forming compound (nystatin, 40 μM). A single record showing an increment in the inward current. The solid black box represents the presence of nystatin or Etx in the medium. (C) The light chain of tetanus neurotoxin (TeNT‐LC) was used to test whether ATP release was due to the exocytosis of cortical granules of the oocytes. In the presence of Etx (300 nM). Brown trace, ATP release in oocytes expressing hMAL and injected with TeNT‐LC (brown trace, T‐LC inj, (N = 10 oocytes). Green trace, oocytes expressing hMAL not injected with TeNT‐LC, N = 24. As in (A), Etx was added at time 0. (D) Plot of simultaneous recording of the relative membrane capacitance (red dots) and ATP release/ luminescence (blue trace) of an individual hMAL‐expressing oocyte. Note that while luminescence increases PM capacitance decreases. Etx, 300 nM. Luminescence is expressed in arbitrary units (A.U). As in (A) and (C), Etx was added at time 0. (E) Changes in the relative membrane capacitance. Red line, hMAL oocytes, N = 8. Black line, WT oocytes, N = 11. Membrane potential was clamped at −60 mV. Vertical bars represent ± SEM at the indicated times. The solid black bar indicates the presence of Etx (300 nM) in the medium. Statistical significance (*) p < .05.

FIGURE 2

Intracellular Ca²⁺ is a key element for calcium‐activated chloride channels. (A) ATP release from oocytes injected with hMAL RNA: Green trace, oocytes in FR (1 mM CaCl₂ and 1 mM MgCl₂; N = 10); red trace, oocytes bathed in the MgFR 1 mM EGTA solution (N = 32) with an estimated [Ca²⁺] of 3 nM. Dark blue trace, EGTA (3 μM) injected hMAL oocytes in the FR bathing solution, N = 13. The negative relative luminescence values reflect the spontaneous decline in initial luminescence. Etx (300 nM) was added at time zero. (B) Calbryte™ 520 fluorescence emitted by oocytes. Red trace, hMAL oocytes, N = 5. Black trace, WT oocytes, N = 3. The horizontal solid black bar indicates the presence of Etx (300 nM) in the low Ca²⁺ recording solution (MgFR 1 mM EGTA). Because this fluorescent methodology does not measure the actual intracellular [Ca²⁺], we used a limited number of oocytes to reduce the number of animals utilized (only three). Significance: (*) p < .05; (**) p < .01. (C) Endogenous currents of voltage‐dependent calcium‐activated chloride channels (CaCC) activated by Etx in hMAL oocytes: (a) black traces, before and (b) green traces, 5 min after adding the toxin (300 nM). Notice that slow‐onset currents at positive potentials; (c) black traces, recording before and (d) dark blue traces, 5 min after adding pEtx (300 nM); (e) stimulus protocol (oocyte clamped at −40 mV and square stimuli from −80 mV to +100 mV for 250 ms). (D) Time course of Etx‐induced currents. Upper panel: Red filled circles, hMAL oocytes, N = 11; black filled circles, WT oocytes, N = 14. Each point corresponds to the current increase, at the indicated time, at the end of a pulse of +90 mV for 250 ms. Red empty circles and black empty circles, effect of pEtx on hMAL oocytes (N = 4) and WT oocytes (N = 11) respectively. Lower panel: Currents were stable at −40 mV in all conditions tested, as in the upper panel. Symbols mean same as the upper panel. The horizontal bar represents the addition of Etx or pEtx at 300 nM at the indicated time. In the upper panel, asterisks in gray correspond to differences between hMAL and WT oocytes treated with Etx, while the black asterisks and double asterisks correspond to the comparison between hMAL oocytes treated with Etx or pEtx. Significance: (*) p < .05; (**) p < .01.

FIGURE 3

Etx targets TMEM16A. (A) MAL‐ and Etx‐dependent activation currents. Upper panel, currents at the end of a pulse of +90 mV for 250 ms. Black circles, WT oocytes, data from Figure 2D (N = 14). Sky blue circles represent WT α‐xM oocytes, corresponding to WT oocytes previously injected with the antisense oligonucleotide against endogenous Xenopus MAL (α‐xMAL) (N = 8 oocytes). Lower panel, currents measured at −40 mV. The horizontal bar represents the addition of Etx or pEtx at 300 nM at the indicated time. Significance: (*) p < .05. (B) Effect of MONNA (10 μM), a potent blocker of CaCC, on currents. Upper panel: Currents at +90 mV. Red filled circles, hMAL‐ oocytes (data from Figure 2D, N = 11). Green filled circles, MONNA in hMAL oocytes (gray trace; N = 10, (*) p < .05). Black empty circles, effect of MONNA on WT oocytes (N = 10). For clarity, Etx‐activated currents in WT oocytes (Figure 3A) are not included. Lower panel: MONNA did not alter currents at −40 mV. Symbols are the same as the upper panel. The horizontal bar represents the addition of Etx or pEtx at 300 nM at the indicated time. Significance: (*) p < .05. (C) Etx‐induced ATP release was dependent on MAL and TMEM16 protein expression. The release of ATP is represented by bars corresponding to the accumulated light of individual oocytes during the action of Etx. The ochre bar, labeled ‘hMAL oocytes α‐T16A‐Etx,’ indicates recordings made in hMAL oocytes preinjected with antisense oligonucleotides against endogenous Xenopus oocyte CaCCs (α‐xTMEM16A). The sample sizes were: HMAL oocytes with Etx, N = 19; hMAL oocytes with MONNA and Etx, N = 6; and hMAL oocytes with α‐xT16 and Etx, N = 14. Significance: (*) p < .05. (D) Analysis of Etx‐dependent activation currents in hMAL oocytes injected with α‐xTMEM16A oligonucleotides. Upper panel: +90 mV evoked currents. hMAL oocytes (data from Figure 2D, N = 11). Ochre filled circles, hMAL oocytes injected with oligonucleotide antisense α‐xTMEM16A, labeled as hMAL α‐T16A, N = 14. Black empty circles, WT oocytes injected with α‐xTMEM16A, labeled as WT α‐T16A (N = 10). As in panel 3B, Etx‐activated currents in WT oocytes are not shown. Lower panel: The antisense oligonucleotide did not modify currents at the holding potential of −40 mV. The horizontal bar represents the addition of Etx or pEtx at 300 nM at the indicated time. Significance: (*) p < .05.

FIGURE 4

Generation of EVs in Etx‐treated hMAL oocytes. (A) Interference light microscopy image of the oocyte surface before adding the toxin. The dark zone corresponds to the oocyte as a non‐transparent body. The interface between the oocyte surface and the medium was diaphanous. (B) Image of this interface after 30 min exposure to Etx. In the area delimited with a white discontinuous line, the focal plane was adjusted to reveal vesicles after the addition of toxin. Vesicle‐like structures accumulated around the oocyte with the appearance of bright or dark spheres of different diameters (scale bar for I and J: 100 μm, Movie S1). These observations were made in oocytes from six independent experiments. (C) Image sequence (C_a–C_e) of the formation and excision of an EV taken from Movie S2. The black surface corresponds to a delimited area of an oocyte. The black arrows point to the vesicle being excised from the oocyte. Images were taken every 30 s. In (C_a), the vesicle appears fuzzy and out of focus, but in contact with the PM; in (C_b), (C_c) and (C_d), it is moving towards the plane of focus; in (C_d) and (C_e) appears clearly separated from the PM; in (C_e), it is nearly in the plane of focus. Arrowheads in (C_a) point to an attached EV and the shadow of another vesicle (scale bar: 100 μm). (D–G) Effect of pEtx. The interface between the oocyte's surface and the surrounding medium is transparent. Notably, remnants of the follicular layer are visible on the oocyte's surface, indicated by the arrow. The focal plane was periodically adjusted to capture these remnants. Our observations from four distinct experiments reveal that the presence of EVs remained minimal within a 30‐min period when pEtx was administered at a concentration of 300 nM. Magnification is the same as that of (B).

FIGURE 5

Ultrastructural changes of the PM induced by Etx. (A) Electron micrograph of a WT oocyte segment containing cytoplasmic cortical granules (GC) and PM. Note the highly undulated PM with numerous microvilli (mv) covered with a glycocalyx (gc). Endoplasmic reticulum cisternae (ER) were abundant near the PM (scale bar: 1 μm). (B–D) correspond to higher magnifications of the square's insights from (A). The arrow shows the ER in close contact with the PM (i.e., ER–PM junctions). Scale bar for A = 0.2 μm; for (B–D), scale bar = 0.1 μm). (E) The surface and the PM of a defolliculated WT oocyte. Note the microvilli (mv) of the PM, the CG close to the PM, the smaller dark granules containing melanin (MG) and the yolk granules (YG) in the deepest part of the micrograph. The black shadow corresponds to the copper grid. (F–H) Etx‐induced EVs generation in hMAL oocytes. (F) Effect of 30 min' exposure of hMAL oocyte to Etx (300 nM). While granules showed no apparent change, the PM showed sophisticated changes, especially the presence of spherical structures (arrows) (scale bar for E and F: 10 μm). (G) shows the content of the inset from (F), where a spherical structure is delimited by a double PM separated by a very thin portion of cytoplasm containing internal membranes (arrowheads). (H) Another example of a spherical structure (arrow) from a different hMAL oocyte with slightly different organization, also showing internal membranous structures (arrowheads) (scale bar for G and H: 1 μm). Images were obtained from two independent experiments.

FIGURE 6

Testing the apoptotic effect of Etx. MOLT‐4 cells were treated with pEtx, (P) or Etx (T) for 5, 10, and 20 min, as well as overnight (O/N). The shorter times (5, 10, and 20 min) correspond to periods preceding the formation of membrane blebs. To determine whether EVs induced by Etx are apoptotic bodies due to early apoptosis activation, the cleavage of caspase‐3 was assayed. In the upper panel, Etx is detected both as a monomer and as oligomers, with oligomer mobility indicated by a black arrow. Oligomers, which become more evident after 20 min of incubation, serve as a fingerprint of Etx action. In the middle panel, no cleaved caspase‐3 bands were observed except in cells treated overnight (O/N) with staurosporine (ST) at two concentrations used. The gray arrow indicates the cleaved caspase‐3 band. In contrast, no active caspase‐3 was detected in cells incubated overnight with Etx. Note that the vehicle in which staurosporine was diluted (dimethyl sulfoxide, DMSO) did not activate caspase‐3 either. In the lower panel, tubulin was detected as a loading control in the electrophoretic analysis. The concentration of Etx and pEtx was 100 nM. This figure is representative of the three independent experiments performed.

FIGURE 7

Etx oligomers in EVs. (A) Immunoblotting of EVs. Anti‐Etx antibody was used to detect toxin molecules present in EV, which were collected after centrifuging the solution that contained hMAL oocytes exposed to Etx. Exposure to pEtx did not yield detectable amounts of EV. The first and second wash lanes correspond to the supernatant collected from the first and second centrifuges, before resuspension of the EV. Etx monomers were detected in the first supernatant but not in the second. After the second centrifugation Etx was detected in the sediment (EV lane). The arrow shows monomeric Etx and the arrowhead shows Etx oligomers. This experiment was repeated in four frogs and the most representative image is shown here. (B–G) In five experiments, we isolated EV from the bath medium after 30 min of exposure of hMAL oocytes to Etx (300 nM), and negative staining was performed. (B) An example of negative staining of an EV. Only part of the vesicle with a diameter larger than 600 nm is displayed. Black arrowheads indicate the limit of the membrane. Inside of this delimited region, diverse circular light structures, with a doughnut‐like shape, are seen and some of them are labeled with white arrowheads (scale bar: 200 nm). (C) Upper panel shows the profile of the gray value (in arbitrary units (AU) of one of the doughnut‐like selected structures. We differentiate between the pore diameter (sky blue) and the inner barrel diameter (brown), which is represented in the lower panel. (D) Mean distance of the pore and inner barrel diameter (N = 12 selected structures). (E–G) Close‐up of the distinctive structures (scale bar: 10 nm).