HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 42, 39119-39127, October 18, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/42/39119    most recent M203433200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ouanounou, G. Articles by Molgó, J. Search for Related Content PubMed PubMed Citation Articles by Ouanounou, G. Articles by Molgó, J. Social Bookmarking                      What's this?

Trachynilysin, a Neurosecretory Protein Isolated from Stonefish (Synanceia trachynis) Venom, Forms Nonselective Pores in the Membrane of NG108-15 Cells^*

Gilles Ouanounou^§, Michel Malo^§, Jacques Stinnakre, Arnold S. Kreger^¶, and Jordi Molgó

From the  Laboratoire de Neurobiologie Cellulaire et Moléculaire, UPR 9040 CNRS, 1 avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France and the ^¶ Department of Epidemiology and Preventive Medicine, University of Maryland School of Medicine, Baltimore, Maryland 21201-1595

Received for publication, April 9, 2002, and in revised form, August 9, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Trachynilysin, a protein toxin isolated from the venom of the stonefish Synanceia trachynis, has been reported to elicit massive acetylcholine release from motor nerve endings of isolated neuromuscular preparations and to increase both cytosolic Ca²⁺ and catecholamine release from chromaffin cells. In the present study, we used the patch clamp technique to investigate the effect of trachynilysin on the cytoplasmic membrane of differentiated NG108-15 cells in culture. Trachynilysin increased membrane conductance the most when the negativity of the cell holding membrane potential was reduced. The trachynilysin-induced current was carried by cations and reversed at about 3 mV in standard physiological solutions, which led to strong membrane depolarization and Ca²⁺ influx. La³⁺ blocked the trachynilysin current in a dose-, voltage-, and time-dependent manner, and antibodies raised against the toxin antagonized its effect on the cell membrane. The inside-out configuration of the patch clamp technique allowed the recording of single channel activity from which various multiples of 22 pS elementary conductance were resolved. These results indicate that trachynilysin forms pores in the NG108-15 cell membrane, and they advance our understanding of the toxin's mode of action on motor nerve endings and neurosecretory cells.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Neurotoxins have proved to be valuable tools for understanding molecular mechanisms involved in physiological processes, and their use has been instrumental in the purification and identification of key protein components of excitable membranes and synapses (1-3). Among the neurotoxins that promote neurotransmitter release, the most studied has been -latrotoxin (-LTX)¹ (for recent reviews, see Refs. 4 and 5), a 120 kDa protein purified from the venom of the black widow spider (Lactrodectus mactans tredecimguttatus), which is known to trigger Ca²⁺-dependent and -independent release of a variety of neurotransmitters and hormones (6-11). -LTX forms nonselective cationic pores in planar lipid bilayers (12-14) and in the cytoplasmic membrane of differentiated PC12 (15) and neuroblastoma cells (16). In addition, two families of -LTX receptors that facilitate pore formation in biological membranes (17, 18) have been identified, i.e. neurexins (19, 20) and latrophilins (21-24), and the pores are involved in the toxin's massive release of neurotransmitters (4, 5, 7, 9, 11).

Trachynilysin (TLY), a 159 kDa membrane-perturbing (hemolytic) toxic protein isolated from the venom of the stonefish Synanceia trachynis, also greatly increases quantal acetylcholine release from motor nerve endings (25, 26) and catecholamine secretion from large dense-core vesicles of chromaffin cells (27). However, contrary to -LTX, these two types of release are Ca²⁺-dependent. Interestingly, neither TLY (25) nor -LTX (28) affects the number of large dense-core vesicles containing neuropeptides in motor nerve endings despite the depletion of small clear synaptic vesicles. TLY has also been reported to raise the intracellular Ca²⁺ concentration in cultured mouse hippocampal neurons (29) and adrenal chromaffin cells (27). Simultaneous blockade of L, N, and P/Q voltage-dependent Ca²⁺ channels caused only a minor reduction of TLY-induced catecholamine secretion and little change in Ca²⁺ signals (27), and removal of extracellular Ca²⁺ and addition of EGTA or La³⁺ completely abolished both secretion and Ca²⁺ signals. Moreover, depletion of intracellular Ca²⁺ stores with caffeine inhibited TLY-induced catecholamine secretion by chromaffin cells. These results suggest that transmembrane Ca²⁺ influx and the resulting mobilization of intracellular Ca²⁺ stores are required for TLY-induced secretion (27). To determine the mechanism for the Ca²⁺ influx, we decided to determine whether TLY activates existing ion channels or forms pores in the plasma membrane. Other stonefish toxins; i.e. stonustoxin from Synanceia horrida venom (30) and verrucotoxin from Synanceia verrucosa venom (31), are known to exert their hemolytic activity through pore formation (32).

To the best of our knowledge, voltage clamp studies have not been performed previously with any of the stonefish toxins. In the present studies, the membrane effects and the ability of TLY to form pores were investigated using the patch clamp technique and differentiated NG108-15 hybrid cells (33).

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture-- Undifferentiated neuroblastoma × glioma NG108-15 hybrid cells were grown in monolayer cultures using Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum, 100 µM hypoxanthine, 0.4 µM aminopterin, 16 µM thymidine, 2 mM glutamine, and 3 µM glycine, as described previously (34). Three days before performing the experiments, the cells were differentiated by adding 0.5 mM dibutyryl cyclic adenosine-monophosphate (Sigma-Aldrich) to the medium and reducing the serum concentration to 1%. The cultures were maintained at 37 °C in a humidified atmosphere containing 95% air, 5% CO₂. Tissue culture reagents were purchased from Invitrogen.

Trachynilysin-- TLY was purified from stonefish (S. trachynis) venom by sequential anion exchange, fast protein liquid chromatography (FPLC), and size exclusion FPLC (25). Aliquots of the purified toxin (~2 mg of protein/ml) in 10 mM Tris-HCl-buffered saline (pH 7.4) containing 5% glycerol were stored at 60 °C. Before being used in the electrophysiological experiments, TLY was diluted with the standard physiological solution. A fast perfusion system allowed changing the solution around the recorded cell in a fraction of a second. Inside-out patch clamp experiments used TLY diluted with the standard solution filling the patch pipette.

Anti-TLY Antibodies-- A female New Zealand White rabbit was injected subcutaneously with a water-in-oil emulsion composed of equal parts of Hunter's TiterMax adjuvant (Sigma) and a solution of purified TLY. Injections were performed on (i) days 0 and 17 with 32 and 158 µg of toxin protein, respectively, (ii) days 31 and 52 with 317 µg of toxin, and (iii) days 88, 110, and 131 with 1.14 mg of toxin. The rabbit was sedated and exsanguinated on day 139, and the anti-TLY serum was lyophilized and stored at 4 °C. Negative control, normal serum was obtained from a rabbit injected (according to the above protocol) with an emulsion containing only adjuvant and buffer, and the serum was lyophilized and stored at 4 °C. Immunoglobulin G (IgG) antibodies were isolated from reconstituted sera (75 mg/ml deionized water) with an ImmunoPure IgG purification kit (Pierce). The specificity of the anti-TLY IgG was revealed by (i) Western blots probed with the anti-TLY and control IgG preparations and (ii) studies examining the ability of the two preparations to neutralize TLY in vitro hemolytic activity, as described in the below protocols and under "Results" (Fig. 2, A and B).

Extraction of NG108-15 Cell Proteins-- Differentiated NG108-15 cells (2.5 × 10⁶) in culture were extensively washed twice with phosphate-buffered saline, homogenized in 10 mM Tris buffer (pH 7.1), and centrifuged (800 × g, 10 min) to remove the nuclei. The protein content of the supernatant fluids was determined by the Lowry method (35) before aliquots of the fluids were used in the Western-blotting experiments.

Western Blotting-- TLY (0.5 µg of protein/lane) and the protein extract (6 µg/lane) from NG108-15 cells were analyzed under reducing conditions (10% -mercaptoethanol) by SDS-PAGE with a 12% acrylamide gel. After electrotransfer (4 °C, overnight, 0.2 A) of the separated protein bands onto a nitrocellulose membrane, the membrane was blocked (8 h, 4 °C) with Tris-buffered saline containing 0.1% Tween 20, 5% (w/v) skimmed milk powder, and 1% (w/v) bovine serum albumin, and it was probed (12 h, 4 °C) with the anti-TLY or control IgG preparation diluted (1:500) in the blocking buffer. The rabbit IgG-probed membranes were washed with Tris-buffered saline, and bound antibodies were detected after probing (2 h, 4 °C) with a diluted (1:1000) peroxidase-labeled, anti-rabbit IgG preparation using an enhanced chemiluminescence detection system (ECL, Amersham Biosciences) according to the manufacturer's instructions.

Neutralization of TLY Hemolytic Activity-- One hundred and 1000 hemolytic units of TLY, contained in 1 ml volumes of Tris-buffered saline (pH 7.4) supplemented with crystalline bovine albumin (1 mg/ml), were incubated (15 min, 37 °C) with various amounts of anti-TLY IgG and control IgG, and the residual hemolytic activity of the incubated mixtures was determined against rabbit erythrocytes as previously described (36).

Patch Clamp Recordings-- Whole-cell and inside-out patch clamp recordings were performed at room temperature (20-22 °C) as previously described (37). Pipettes were made from borosilicate glass (Clark Electromedical Instrument, Reading, England) and pulled on a BP-7 puller (Narishige, Osaka, Japan). The patch electrodes filled with physiological saline solutions had a resistance of 3-5 M. Membrane currents were recorded with a RK-300 patch clamp amplifier (Biologic, Claix, France), and they were filtered with an 8-pole low pass Bessel filter (Biologic, ibid.) at 8 kHz for whole-cell recordings and at 3 kHz for inside-out patch recordings. The filtered signals were digitized by a 12 bit A/D converter (Labmaster DMA, Scientific Solutions Inc., Mentor, OH) and stored using pCLAMP 5.5 software (Axon Instruments, Union City, CA). Whole-cell and inside-out patch data were acquired at a sample rate of 1 and 10 kHz, respectively. Recordings were analyzed using Origin 5 software (Microcal Software Inc., Northampton, MA), taking advantage of the built-in regression functions and of the Labtalk programming language.

The standard external and internal solutions for whole-cell recordings had the following compositions: 154 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES (pH 7.4) (external solution) and 135 mM KCl, 5 mM MgCl₂, 10 mM EGTA, 10 mM HEPES (pH 7.2) (internal solution). The internal solution was filtered through 0.2 µm Millex filters (Millipore, Saint-Quentin en Yvelines, France). The osmotic pressure of the internal and external solutions were 295-305 mosmol and 305-315 mosmol, respectively, as measured with a freezing point osmometer (Knauer, Berlin, Germany). The solutions used to determine the ionic selectivity of the TLY-induced current are shown in Table I. For inside-out patch recordings, the external solution was also used as the internal solution.

Ionic Selectivity Determination-- To minimize the contribution of voltage-gated K⁺ channels to the macroscopic current, some patch clamp experiments were performed at a holding potential of +60 mV. Under this condition, the transient voltage-dependent currents, including K⁺ currents (38, 39), were inactivated by about 95% 2 min after changing the holding potential. Thus, the different voltage-dependent currents were limited without using any pharmacological agent that could interact with the toxin.

To determine the ionic selectivity of the TLY-induced current, its reversal potential was measured in solutions of various ionic composition using ramp-potentials (150 mV/s) from +60 to 60 mV. Derived from the Goldman-Hodgkin-Katz flux equation (40, 41), Equation 1 (42) gives the reversal potential of a cationic current as a function of the concentration and permeability of each ion species.

(Eq. 1)

In this equation, the terms in square brackets are ion concentrations, P_S is the permeability of the S ion species, T is the absolute temperature, R and F are the usual gas and Faraday constants, and i and o are the intra- and extracellular compartments, respectively. Ion activities are used rather than concentrations, as in Equation 2 (43). The activity coefficients used for Na⁺, K⁺, Ca²⁺, and Mg²⁺ were 0.75, 0.75, 0.25, and 0.25, respectively.

(Eq. 2)

Quantification of Macroscopic Current Steps and Low Frequency Fluctuations-- The macroscopic current steps were counted and measured using routines written by G. Ouanounou (this paper) in the Labtalk programming language of Origin software. Current step transitions were detected by computing average currents on a 50-100-ms time window before and after any given point and calculating the difference. A transition occurred when the difference was higher than twice the variance of the "before" window.

To quantify the current variations that did not occur in steps, each detected step was subtracted from the ensuing samples of the recording, which led to a smoothed trace deprived of step transitions. In other words, the new trace exhibited only low frequency components. Derivation of that signal allowed us to dissociate the changes of membrane conductance into increasing and decreasing components according to the sign of the derivative. Separate integration of the two components of the derivative over the entire analyzed period gave the cumulative currents corresponding to each component.

Noise Analysis-- Noise analysis was performed using the fast Fourier transform module included in Origin 5 software after filtering off the lower frequencies. The analyzed periods were divided into sections of 1024 samples, and the resulting power spectra were averaged. The mean spectrum, performed on similar base-line sections, was subtracted before fitting to Lorentz Equation 3,

(Eq. 3)

where S(f) is the power, f is the frequency, and S0 is the constant to which the function tends for low frequencies.

The mean channel lifetime () was computed from the cut-off frequency (f_c).

(Eq. 4)

The elementary conductance was calculated from the variance (²) and the mean current (I_m) of the analyzed signal,

(Eq. 5)

where V_m is the membrane potential, and E_rev is the reversal potential.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Whole-cell Recording of Membrane Currents during TLY Action-- To determine the effect of TLY on ionic permeability, membrane currents of NG108-15 cells were recorded in the standard external solution using the whole-cell configuration of the patch clamp technique at a holding potential of 60 mV. An inward current developed about 1 min after the addition of 12 nM TLY to the external medium. In the continuous presence of TLY, the induced current rapidly became too large to be recorded accurately. Because of the irreversibility of the TLY effect, the toxin was applied for just 1 min. Under these conditions, the TLY-induced current was limited to a recordable intensity range. A 1 min exposure to 12 nM TLY appeared to be just higher than a threshold duration, since shorter applications of that TLY concentration had no effect other than an increase in electrical noise (data not shown). The threshold duration of toxin exposure could be reached by successive applications, alternating TLY-containing medium with TLY-free medium. This suggests that a minimal number of TLY molecules has to bind to the membrane to produce the membrane effect, and this may partly explain the latency of the effect of the toxin. After 1 min of exposure to TLY, the current increased in a stepwise manner to about 500 to 1000 pA and then fluctuated around this range (Fig. 1A).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Whole-cell recordings of the membrane current elicited by exposure of NG108-15 cells to TLY. In this figure as in the others 12 nM TLY was applied for 1 min at a holding potential of 60 mV. A, typical example of the onset of TLY-induced current. The horizontal bar above the trace shows the time of toxin application. B, step-like activities appearing at different time bases. The top trace is from the same cell as that in A. C, time distribution of the membrane current induced by TLY. Three recordings were pooled, including the one shown in A, after the base line was subtracted. Currents were binned in 1 pA classes (see "Results"), and the frequency was normalized for the total recording time. The traces in black represent the Gaussian fits of the different peaks. Inset, plot of the modes of the 15 peaks shown in the histogram versus their ascending rank order. The straight line represents the linear regression across all points.

Details of the time course of the TLY-induced current at various time scales are shown in Fig. 1B. The TLY-induced activity exhibited macroscopic steps of various sizes that coexisted with current fluctuations without visible steps. Thus, two types of current variations, each with its own frequency domain, compose the TLY-induced current. Despite a large spectrum of step sizes and low frequency variations, visual inspection suggested that the membrane current presented preferential values. To support this view, the first minutes of three recordings were pooled and analyzed as follows. The current was binned in 1 pA-wide classes, and the relative number of samples at any of these levels was computed and plotted versus the bin size (Fig. 1C). The peaks revealed by this time-distribution showed the most frequent intensities of the TLY-induced current. The modes of these peaks, determined by a Gaussian-fitting regression, appeared to be regularly spaced. The mode values were plotted as a function of their rank when sorted in ascending order, and they were fitted with a linear regression equation (inset, Fig. 1C). The slope of the regression line, 27 pA, represents the mean increment between peaks, which suggests that the TLY current is composed of multiples of a common component.

The presence of steps and low frequency fluctuations precluded a deeper analysis of the TLY-induced channel-like activity. However, another type of analysis was performed, as shown later.

Anti-TLY Antibodies Suppress the Current Induced by TLY-- Before using the anti-TLY IgG preparation in our electrophysiological experiments, its specificity was evaluated by Western blotting and by determining its ability and the ability of the control IgG preparation to neutralize TLY in vitro hemolytic activity. The Western blotting experiments revealed (Fig. 2A) that (i) the anti-TLY IgG bound to TLY - and -subunits but not to any of the proteins isolated from NG108-15 cells, and (ii) labeling of the toxin subunits was not observed when the control IgG preparation was used as a probe. The studies examining the specificity of the anti-TLY IgG preparation ability to neutralize TLY in vitro hemolytic activity revealed (Fig. 2B) that incubating 100 and 1000 hemolytic units of TLY with ~93 and 698 µg, respectively, of anti-TLY IgG eliminated ~90% of the toxin in vitro hemolytic activity. However, incubating 100 hemolytic units of TLY with ~1.34 mg of control IgG did not eliminate any hemolytic activity.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Characterization of anti-TLY antibodies and their ability to inhibit the TLY-induced current. A, Western blot analysis of TLY (0.5 µg of protein/lane) and of proteins (6 µg/lane) isolated from differentiated NG108-15 cells. Protein bands labeled with IgG were detected with an enhanced chemiluminescence detection system after sequential probing with (i) anti-TLY IgG or control IgG and (ii) peroxidase-labeled, anti-rabbit IgG (as described under "Experimental Procedures"). The molecular weight markers are indicated. B, neutralization of TLY in vitro hemolytic activity by anti-TLY IgG (determined as described under "Experimental Procedures"). The data points were fitted by exponential functions. HU, hemolytic units. C, whole-cell recording in the standard solution from a cell to which TLY, control IgG (95 µg/ml, final concentration), and anti-TLY IgG (95 µg/ml, final concentration) were applied. The bars show the duration of the applications. The holding potential was 60 mV.

Our observation that anti-TLY antibodies neutralized TLY hemolytic activity prompted us to determine whether they could also affect the TLY-induced current. TLY was applied (as described under "Experimental Procedures"), and after the macroscopic induced-current developed, anti-TLY IgG (95 µg/ml, final concentration) was added to the external solution. The TLY-induced current progressively decreased in all 4 cells tested (each from a different culture) until it was completely inhibited within 2-4 min (Fig. 2C). However, control, normal IgG (95 µg/ml, final concentration) did not affect the current induced by TLY or the inhibitory action of a subsequent application of anti-TLY IgG (Fig. 2C). In two of the cells, continuous wash-out of the antibodies allowed about 30% recovery of the TLY-induced current before the gigaseal was lost (data not shown). In the other two cells, the seal was lost at the beginning of the wash-out period. In addition, exposure of the cells to anti-TLY IgG followed by wash-out (which had no membrane effect) did not prevent the membrane response to subsequent exposure to TLY (data not shown). These results suggest that the anti-TLY IgG bound to the toxin in a reversible manner, and they confirm that TLY binds irreversibly to the cell membrane.

Macroscopic Steps and Low Frequency Current Variations Induced by TLY-- Because the membrane current returned to the initial base line in the presence of anti-TLY antibodies, we assumed that all of the TLY-induced increases in conductance were exactly balanced by conductance decreases, and this situation prompted us to assess the proportion of steps and low frequency fluctuations involved in the conductance changes. Therefore, using the algorithm described under "Experimental Procedures," two recordings (including the one shown in Fig. 2) were decomposed into their step and low frequency components (see Fig. 3, A-B), and the two components were separately analyzed. The number of detected upward events represented only 72% of the downward events. Furthermore, the respective amplitude distributions differed, i.e. the upward events were smaller than the downward events (Fig. 3C). This difference was statistically significant, as demonstrated by the Kolmogorov-Smirnov test performed on the cumulative frequency distribution (p  0.01). On the whole, the sum of the closing (upward) events corresponded to only 57% of the opening (downward) events. This feature resulted in a global downward deflection of the step component, as shown in Fig. 3B (bottom trace). The analysis did not reveal a common substep; however, the frequency distributions (Fig. 3C, insets) revealed that the most frequent step transitions occurred in the 27 pA class, which may explain the preferential levels of the global TLY-induced current already noticed (Fig. 1C) and characterized by a 27 pA increment.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Decomposition and analysis of the macroscopic TLY-induced current. A, expanded fragment of the recording shown in Fig. 2, with gray dots limiting the step transitions detected by the algorithm described under "Experimental Procedures." B, separation of the step (bottom trace) and low frequency (top trace) components of the recording shown in A. The top trace was obtained by subtracting the bottom trace from the initial recording. C, cumulative frequency distributions of downward () and upward () current step transitions extracted from two different recordings, including the one shown in Fig. 2. The insets show the relative frequency-distributions of downward and upward transitions. Bin size was 2 pA. The two distributions are significantly different according to the Kolmogorov-Smirnov test (p  0.01). D, differentiation of the low frequency component of the full trace shown in Fig. 2. The positive and negative parts of the derivative are separately plotted for clarity.

The low frequency component was analyzed by derivation and by separate integration of the positive and negative parts of the derivative (Fig. 3D). This analysis allowed the separation of periods of increasing and decreasing membrane conductance; i.e. the integral of the positive derivative gives the cumulative current, which corresponded to membrane conductance decreases. Conversely, the integral of the negative derivative corresponded to membrane conductance increases. The membrane conductance increases represented 22% of the conductance decreases, which led to a global upward deflection of the low frequency component of the TLY current, as seen in Fig. 3B (top trace). Thus, during TLY action, 85% of the membrane conductance increases occurred by steps, and 15% occurred by low frequency fluctuations, whereas steps and low frequency fluctuations contributed almost equally (49 and 51%, respectively) to the membrane conductance decreases.

Influence of the Membrane Potential on the Conductance Induced by TLY-- All of our previous studies were performed at a membrane potential of 60 mV. Therefore, we decided to examine whether membrane potential influences TLY-induced changes in membrane conductance. At positive potentials, the toxin-induced current was relatively stable and did not exhibit macroscopic steps (only fluctuating at low frequency), in contrast to what was observed at negative potentials (Fig. 4A). Moreover, the TLY membrane conductance increase was greater at positive than at negative holding potentials.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Influence of the holding potential on the membrane conductance changes induced by TLY. A, whole-cell recordings of the toxin-induced current during step potentials from +60 to 60 mV. Four recordings obtained with the same cell are shown after subtraction of the current recorded during control steps (before toxin application). Note that the current steps are only visible at 60 mV. B, current-voltage relationship of TLY-induced current under steady-state conditions. The holding potential was stepped-down by 10 mV every 40 s, from +50 to 100 mV, and the control current (using this protocol) was subtracted. The ordinate represents the mean steady-state current (measured during the last 20 s of each potential change) relative to the mean current recorded at +50 mV. Each point represents the mean ± S.E. from data collected in eight different cells. Note that the membrane conductance increase induced by TLY increased as the negativity of the membrane potential was reduced.

To determine the kinetics of this voltage sensitivity, 60 s step potentials were applied (from the holding potential of +60 mV to the test potential of 60 mV) with different cells previously treated with TLY (Fig. 4A). Transmembrane currents recorded before toxin application were subtracted from currents recorded in the presence of TLY. When the membrane potential was stepped from +60 to 60 mV, the TLY-induced current reversed and then decreased to about 20% of its initial value in ~20 s. The mean current calculated from four recordings (Fig. 4A) was fitted with a single exponential function that had a decay time constant of 5.32 ± 0.05 s (95% confidence interval). Notably, a step from a holding potential of 60 mV to a test potential of +60 mV yielded different kinetics for the membrane conductance changes, i.e. the membrane current increased abruptly about 1 min after the potential change without exhibiting current steps (data not shown).

To determine the current-voltage relationship of the TLY-induced current at the steady state, the potential was changed from +50 mV to 100 mV in 10 mV steps lasting 40 s. The average current during the last 20 s of each step was normalized to the averaged current recorded at +50 mV. As shown by the current-voltage relationship (Fig. 4B), the TLY-induced current reversed at 3 mV in the standard solutions, which suggested that the induced conductance did not have ionic selectivity. The average current was constant from 100 to 40 mV, thus indicating that the membrane conductance increased as the negativity of the membrane potential was reduced. At potentials more positive than the reversal potential, the membrane conductance was almost constant. This corresponds to a so-called "outward rectification."

Ionic Selectivity of TLY-induced Current-- To determine the ion(s) taking part in the TLY-induced current, the reversal potential was measured in solutions of various ionic composition using ramp potentials (from +60 to 60 mV) in a whole-cell configuration. Ramp potentials performed before TLY application were used as controls (Fig. 5, middle trace). The TLY-induced current reversed at about 3 mV (n = 17) in standard solutions (Fig. 5, bottom traces), as was observed previously at the steady state. The reversal potentials of the TLY-induced current recorded in external solutions of various ionic composition are shown in Table I. For each cell, the reversal potential was determined in the external standard solution and in one of the test solutions. Methanesulfonate substitution for Cl did not affect the reversal potential (n = 5), whereas substituting half of the NaCl content with sucrose significantly shifted the reversal potential to a more negative value (n = 3). Thus, neither methanesulfonate nor Cl contribute to the TLY current. In contrast, TLY caused all of the physiological cations to permeate through the membrane. The relative permeability values, for which calculated reversal potentials using Equation 2 fit best with experimental values, were P_K/P_Na = 1.3, P_Ca/P_Na = 1.4, and P_Mg/P_Na = 1.4. 

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Determination of reversal potential of the TLY-induced current. Upper trace, ramp potentials (150 mV/s, +60 to 60 mV) applied to determine the reversal potential. Middle trace, control membrane current recorded before TLY application. Lower traces, recordings of the TLY-induced current from one cell obtained at various times after TLY application (after subtraction of the control current). The current amplitude scale applies to the middle and bottom recordings.

La³⁺ Blocks the Current Induced by TLY-- The La³⁺ ion is known to inhibit TLY-induced release of catecholamines from chromaffin cells (27) and the -LTX-induced current (11, 17) and to perturb the structure of -LTX pores (11, 44). Therefore, it was of interest to determine whether the current induced by TLY in NG108-15 cells was affected by this cation. At 60 mV, 1 mM La³⁺ added to the external medium completely blocked the current induced by TLY. In the presence of 100 µM La³⁺, the current returned to the base line, suppressing all low frequency current fluctuations and leaving only fast current transients (Fig. 6, A-B). Lower La³⁺ concentrations partially affected the TLY-induced current. Because of the current fluctuations already described, the La³⁺ effect was quantified by rationing the mean current during La³⁺ application to the control current. A Boltzmann equation fit of the data points, obtained with various La³⁺ concentrations, gave an EC₅₀ = 512 ± 56 nM (95% confidence limits, Fig. 6C). At all concentrations used, the blockade of the TLY-induced current by La³⁺ was rapidly reversed by washing with the standard medium.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Blockade of TLY-induced current by La3+. A and B, whole cell recording of the TLY-induced current in the presence of 100 µM La3+ in the external solution (upper bar) at 60 mV. Note the persistence of current transients (also visible in B) on expanded scales. C, dose-response curve for La3+ blockade of the TLY-induced current (holding potential, 60 mV). The data points represent the means ± S.E. of the relative currents calculated from 20 s period recordings made before and after the application of the La3+ concentrations (data were obtained from 4 different cells). The data points were fitted by the Boltzmann equation, and they gave an EC50 = 512 ± 56 nM (95% confidence limits).

A closer examination of the La³⁺ blockade at 100 µM showed that the transient currents described above exhibited a fast on-current followed by an exponential decay (Fig. 6B). This suggests that the transients correspond to the current steps described above, which are blocked in a time-dependent manner by La³⁺ ions. In addition, the use of the ramp-potential protocol revealed that, in the presence of La³⁺ (100 µM), the current exhibited an outward rectification (data not shown), thus indicating that the La³⁺ effect is voltage-dependent.

Elementary TLY-induced Current-- To determine whether the observed macroscopic changes in membrane conductance were due to channel-like activity and whether this activity could explain the time course of the macroscopic TLY-induced current, elementary currents were recorded from membrane patches using the "inside-out" configuration. A drop of buffered TLY solution was deposited at the back of the pipette previously filled with standard external solution. The time necessary for TLY to diffuse to the pipette tip allowed the sealing of the inside-out patch clamp configuration and the confirmation of the absence of endogenous ionic conductances in the membrane patch.

At a pipette potential of 60 mV (i.e. a positive membrane potential), only one channel conductance (22 pS) was detected; i.e. several minutes after inside-out patch achievement, a single channel activity developed (Fig. 7, A-A'), and the number of channels (as well as the opening/closing rate) then increased rapidly. At a pipette potential of +60 mV (i.e. negative membrane potential), several channel conductances were detected. The induction of channel activities began with a 22 pS elementary conductance (Fig. 7, B-B'). After relative stabilization in the open state of 4-8 channels of 22 pS, bursts of 72 and 86 pS channel activity appeared from this current level as base line (Fig. 7, C-D, C'-D'). As time progressed, channel activity became complex and showed elementary conductance multiples of ~22 pS and an increased opening/closing rate. However, periods of single channel activity alternated with periods during which the current continuously changed without evident channel activity but with an increased noise level (Fig. 7E). During these periods, the current evolved between two levels corresponding to the open and closed states of 72 and 86 pS channels, without summation with these channel activities. The frequency spectrum of these periods (provided by fast Fourier transform noise analysis) could be fitted by a Lorentz equation, which indicated that the noise resulted from summation of discrete events (see Ref. 45). These events had a mean open time of about 3.5 ms and an elementary conductance of ~22 pS. It should be emphasized that this analysis corresponds to the early stages of TLY activity, because the patch current rapidly became too complex to analyze.

View larger version (38K): [in this window] [in a new window]   Fig. 7.   Elementary TLY-induced currents. Elementary currents were recorded with the inside-out patch clamp configuration. The pipettes were filled with the standard external solution, and a drop of TLY was deposited at the back of the pipette before initiating the sealing procedure. A, induction of channel activities just after inside-out patch achievement at a pipette potential of 60 mV. A', probability distribution (0.1 pA bin size) of the current shown in A ( = 22 pS). B, as in A, but at a pipette potential of +60 mV in another cell. B', probability distribution (0.1 pA bin size) of the current shown in B ( = 22 pS). C-D, the same patch and same potential as in B; new channel activities are superimposed with the current due to the constant opening of a few 22 pS channels. C'-D', probability distribution (0.1 pA bin size) of the current shown in C-D ( = 86, 72 pS, respectively). E, the same cell as in B; shows periods of unresolved channel activity (between arrows) alternating with periods similar to those shown in C-D. E', frequency spectrum obtained by fast Fourier transform of several recordings similar to that shown between the arrows in E. The data points are fitted to a Lorentz equation, with S(0) = 1.58 105 pA2·s, and fc = 45 Hz.

This study shows that the TLY-induced macroscopic current is due at least in part to ion channel-like activity. Interestingly, at the elementary level as at the macroscopic one TLY activity appears to be constituted of various components at negative membrane potentials, whereas only a single component is detectable at positive potentials.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The present study demonstrates that brief exposure to TLY induces after a short latent period a potent and irreversible increase in the membrane conductance of NG108-15 cells. Under voltage clamp conditions, the TLY-induced increase in membrane conductance leads to a transmembrane cationic current which, when measured in standard media, reverses at a membrane potential of 3 mV and exhibits an outward rectification. A 1 min exposure to 12 nM TLY appeared to be the time threshold before which time TLY had no apparent effect, and this minimum duration could be reached by successive exposures. This observation suggests that TLY binds irreversibly to the membrane and that a threshold amount of TLY-bound molecules presumably is needed to alter membrane permeability.

Two hypotheses may be proposed to explain the above-mentioned results; either TLY activates one or several types of endogenous ion channels or TLY forms ionic pores by insertion into the cell membrane. Our experimental results revealed that (i) the TLY-induced macroscopic current is made of slow fluctuations and current steps of up to 160 pA (at 60 mV), making it particularly unstable, (ii) TLY elicited channel activity in membrane patches devoid of endogenous channels, and (iii) the observed TLY effect required a threshold amount of bound toxin. These features are inconsistent with the idea that TLY forms pores by activating preexisting channels. Indeed, the observed current steps correspond to a conductance change of about 2.8 nS, which is much higher than that of any endogenous single channel conductance described to date. Furthermore, considering the irreversibility of the TLY effect, it is difficult to understand or explain how preexisting channels activated by TLY would fluctuate as observed. Also, features similar to those described above have been described for exogenous polypeptides, such as equinatoxin II (46) and -LTX (12, 15-17), and for endogenous polypeptides such as amylin (47), all of which are known to form pores by insertion into the cell membrane. Thus, our data support the idea that the TLY-elicited pore activity we observed is caused by insertion of TLY into the cell membrane rather than by TLY-induced activation of preexisting silent channels.

Experiments using various holding potentials revealed a large difference in the time course of the TLY-induced current depending on the polarity of the membrane potential. At positive membrane potentials, the macroscopic toxin-induced current fluctuated at low frequency, and inside-out patch recordings exhibited only one elementary conductance of 22 pS. Thus, at positive membrane potentials, the macroscopic TLY current is due to the summation of 22 pS channel-like activities. However, at negative membrane potentials, the macroscopic TLY current was complex and consisted of low and high frequency components. The high frequency component was composed of downward and upward macroscopic current steps, the former being more frequent and exhibiting higher amplitudes than the latter. Thus, step activity leads to a global downward deflection of the membrane current, which is compensated for by an opposite behavior of the low frequency component. This complex behavior was also observed at the channel level, which revealed several conductance values and progressive transitions combined with channel-like activity. Taken together, these observations suggest that the macroscopic downward current steps result from the opening of a large conductance that can close gradually or by small steps. Another possibility is that elementary channels behave independently at positive membrane potentials but that they cooperate at negative potentials, with a higher cooperativity coefficient for opening than for closing. The noise analysis of single channel recordings (Fig. 7, E-E'), which revealed the existence of a discrete event, together with the observation of various conductances (all multiples of the discrete event) are in favor of the second hypothesis.

The influence of the membrane potential on the TLY-induced current at the steady state revealed a strong outward rectification. In contrast, fast ramp currents varied linearly with the membrane potential, which indicates that the elementary conductance did not change with the potential. Notably, current steps still occurred during ramp potentials but only at negative potentials, and they led to abrupt changes in the slope of the current (Fig. 5, bottom). Thus, the outward rectification (Fig. 4B) is due to an increased number of opened TLY-induced conductances overcompensating for the smaller elementary conductance at positive potentials.

La³⁺ was found to block the TLY-induced current. However its EC₅₀ was about 5-fold higher for the TLY current than has been reported (48) for voltage-sensitive Ca²⁺ channels in the same cell line. Moreover, the voltage dependence of the La³⁺ effect indicates that it acts via binding to a site located in the pore formed by TLY. In addition, La³⁺ blockade was time- and concentration-dependent. Thus, at 100 µM the kinetics of the blockade of TLY action by La³⁺ were slow enough to allow the TLY-induced conductance opening to produce macroscopic downward current transitions, but they were sufficiently fast to block the TLY pore before its closing and to hide the low frequency fluctuations.

The single-channel activity induced by TLY may result from the formation of TLY channels with intrinsic gates oscillating between the open and closed state or from oligomerization-deoligomerization of the TLY channels. Our results do not allow us to distinguish between these two possibilities. The complete amino acid sequence of TLY is not yet known, but its reported (25) N-terminal amino acid sequence is similar to that of stonustoxin, for which an amphiphilic -helix structure was predicted (49) for each subunit. To explain the pore-forming property of stonustoxin, the so-called "carpet-like" model (for review, see Ref. 50) has been proposed (32). This model predicts the existence of a threshold amount of bound toxin for membrane permeation and an instability of the pore structure. Our observations with TLY are in agreement with this model.

The ability of TLY to form cationic pores in the cytoplasmic membrane of NG108-15 cells leads to membrane depolarization, which in turn increases membrane permeation due to outward rectification. Under normal conditions in which the membrane potential is not controlled this mechanism of amplification leads to a rapid and complete depolarization (51) that should induce activation of endogenous voltage-gated channels. However, previous experiments in our laboratory found that low and high voltage-activated Ca²⁺ currents in NG108-15 cells were not modified by TLY.² In addition, the slight participation of voltage-gated L, N, and P/Q Ca²⁺ channels for TLY-induced catecholamine release from chromaffin cells (27) may result from the depolarizing effect of the toxin. Considering the high Ca²⁺ permeability of TLY pores and the potent increase in membrane conductance induced by the toxin, the Ca²⁺ influx through the pores would be expected to be large and to be higher than Ca²⁺ influx through voltage-gated Ca²⁺ channels in NG108-15 cells. Thus, the Ca²⁺ permeability of the TLY-induced pores is sufficient to account for the external Ca²⁺ dependence of the TLY-induced release of neurotransmitters from motor nerve terminals and chromaffin cells (25-27). Also, the elicited increase in Ca²⁺ permeability may be responsible for mobilizing intracellular Ca²⁺ stores in chromaffin cells (27). Moreover, the latent period between TLY application and its increasing intracellular Ca²⁺ and stimulating catecholamine release in chromaffin cells is similar to that observed in the present studies of pore formation by TLY. Finally, La³⁺, which inhibits NG108-15 cell membrane permeation by TLY, also blocks the Ca²⁺ increase and catecholamine secretion induced in chromaffin cells by TLY (27). Although the Ca²⁺ permeability of the TLY pores may account for most of the biological effects of TLY, it does not explain why TLY does not trigger the release of large dense core vesicles containing neuropeptides from motor nerve endings. Thus, further studies are needed to explain this seeming paradox.

	ACKNOWLEDGEMENTS

We thank Dr. B. Rouzaire-Dubois for providing the NG108-15 cells used in this study, L. Prado de Carvalho for assistance during preliminary experiments, L. Lane-Guermonprez for help with Western blotting, Dr. J.-M. Dubois and the late Dr. R. Kado for critical discussions and comments on the manuscript, and David McColm for assistance in helping to obtain the S. trachynis venom used as the source of trachynilysin needed in our study.

	FOOTNOTES

^* This work was supported in part by the CNRS, by a grant from the Direction des Systèmes de Forces et de la Prospective (to J. M.), and by National Institutes of Health Public Health Service Grant GM-43728 (to A. S. K.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Supported by fellowships from the Ministère de l'Education Nationale de la Recherche et de la Technologie and from the Association Française contre les Myopathies.

To whom correspondence should be addressed. Tel.: 33-169-82-36-42; Fax: 33-169-82-94-66; E-mail: Jordi.Molgo@nbcm.cnrs-gif.fr.

Published, JBC Papers in Press, August 12, 2002, DOI 10.1074/jbc.M203433200

² G. Ouanounou, unpublished data.

	ABBREVIATIONS

The abbreviations used are: -LTX, -latrotoxin; TLY, trachynilysin; EC₅₀, effective concentration (i.e. 50% of maximum).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This Article Abstract Full Text (PDF) All Versions of this Article: 277/42/39119    most recent M203433200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ouanounou, G. Articles by Molgó, J. Search for Related Content PubMed PubMed Citation Articles by Ouanounou, G. Articles by Molgó, J. Social Bookmarking                      What's this?