Characterization of the maitotoxin-activated cationic current from human skin fibroblasts

Abstract

The maitotoxin (MTX)-induced cationic current (I_mtx) from human skin fibroblasts was characterized using the patch-clamp technique in whole-cell configuration. Under resting conditions (absence of MTX), the main current observed is produced by an outwardly rectifying K⁺ channel which is inhibited by 1 mm TEA. The current reversal potential was −86 mV (n = 12). MTX (500 pm) activated a current with a linear current–voltage relationship and a reversal potential of −10 mV (n = 10). Replacing the extracellular Na⁺ and K⁺ with N-methyl-d-glucamine (NMDG) caused a shift of the reversal potential to a value below −100 mV, indicating that Na⁺ and K⁺, but not NMDG, carry I_mtx. Further ion selectivity experiments showed that Ca²⁺ carries I_mtx also. The resulting permeability sequence obtained with the Goldman–Hodgkin–Katz equation yielded Na⁺ (1) ≈ K⁺ (1) > Ca²⁺ (0.87). The I_mtx activation time course reflected the changes in intracellular Ca²⁺ and Na⁺ measured with the fluorescent indicators fura-2 and SBFI, respectively, suggesting that the activation of I_mtx brings about an increment in intracellular Ca²⁺ and Na⁺. Reducing the extracellular Ca²⁺ concentration below 1.8 mm prevented the activation of I_mtx and the increment in intracellular Na⁺ induced by MTX. Mn²⁺ and Mg²⁺ could not replace Ca²⁺, but Ba²⁺ could replace Ca²⁺. MTX activation of current in 10 mm Ba²⁺ was approximately 50 % of that induced in the presence of 1.8 mm Ca²⁺. When 5 mm of the Ca²⁺ chelator BAPTA was included in the patch pipette, MTX either failed to activate the current or induced a small current (less than 15 % of the control), indicating that intracellular Ca²⁺ is also required for the activation of I_mtx. Intracellular Ba²⁺ can replace Ca²⁺ as an activator of I_mtx. However, in the presence of 10 mm Ba²⁺ the activation by MTX of the current was 50 % less than the activation with nm concentrations of free intracellular Ca²⁺.

Maitotoxin (MTX), a water-soluble polyether obtained from the marine dinoflagellate Gambierdiscus toxicus, is a potent activator of Ca²⁺ influx in a wide variety of cells (Gusovsky & Daly, 1990). This toxin may be one of the molecules involved in the seafood poisoning known as ciguatera (Yasumoto et al. 1976).

Ciguatera is a severe health problem in many countries where seafood is an important part of the diet. There are more than 20 000 cases of this disease each year worldwide. However, it is believed that many more cases are not properly diagnosed or reported (Yasumoto & Satake, 1996). Elucidating the cellular and molecular mechanisms underlying MTX action may help to prevent or treat ciguatera poisoning.

MTX is a powerful tool for studying changes in Ca²⁺ homeostasis. The fact that MTX does not appear to have ionophoretic activity (Takahashi et al. 1983), and that this toxin can activate Ca²⁺ influx in practically all cells tested from a wide variety of organisms, suggests the presence of a highly conserved MTX receptor (Escobar et al. 1998).

Another feature of this toxin is its potency. MTX effects are observed with toxin concentrations ranging from picomolar to nanomolar. Recently, we reported that human skin fibroblasts are very sensitive to MTX. The influx of Ca²⁺ showed a half-activation constant of 450 fm (Gutierrez et al. 1997).

Although the mechanism by which MTX activates the influx of Ca²⁺ remains unsolved, it appears to involve the activation of a current (I_mtx) through Ca²⁺-permeable cationic channels, as indicated by patch-clamp experiments performed on different cells and tissues (reviewed in Escobar et al. 1998). It is not yet clear whether the activation of this channel is direct, or mediated by a second messenger cascade.

The interactions between MTX and its putative receptor appear to be complex, since in some cells the presence of extracellular Ca²⁺ and/or Na⁺ is required for the activation of I_mtx (Escobar et al. 1998). However, no detailed characterization of the cation requirements for the MTX effect has been reported to date.

In this study we have characterized I_mtx from human skin fibroblasts utilizing the whole-cell configuration of the patch-clamp technique and fluorescent indicators to measure changes in the intracellular Ca²⁺ and Na⁺ concentrations in response to MTX. The results obtained indicate that MTX activates a cationic current with a linear current–voltage relationship. The time course of activation of I_mtx is correlated with the changes in intracellular Ca²⁺ and Na⁺ as monitored with fluorescent indicators.

We have also characterized the ionic requirements for the MTX effect. We have found that in human skin fibroblasts extracellular Ca²⁺, but not Na⁺, is required for the activation of I_mtx. Ca²⁺ is required only during the initial application of MTX; removal of this cation after the onset of the response did not eliminate I_mtx.

Other divalent cations such as Mg²⁺ or Mn²⁺ cannot replace extracellular Ca²⁺, whereas Ba²⁺ can partially compensate for the absence of free extracellular Ca²⁺. In 10 mm Ba²⁺, MTX activates 50 % of the current obtained with 1.8 mm Ca²⁺.

Intracellular Ca²⁺ is also required for MTX action. When 5 mm BAPTA was included in the patch pipette, MTX failed to activate I_mtx, indicating that nanomolar concentrations of free intracellular Ca²⁺ are also required for the activation of I_mtx. As in the case of extracellular divalent cations, intracellular Mg²⁺ or Mn²⁺ cannot replace Ca²⁺. Only Ba²⁺ at high concentrations can partially compensate for the absence of Ca²⁺.

The results presented here, in combination with those found in the literature, indicate that MTX is a powerful and complex Ca²⁺-mobilizing agent.

METHODS

Solutions and reagents

All salts used are analytical grade (Sigma Chemical Co., St Louis, MO, USA). Ethyleneglycol-bis-(β-aminoethylether)N,N,N ′,N ′-tetraacetic acid (EGTA), 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA), bovine serum albumin (BSA) and trypsin A3 were purchased from the Sigma Chemical Co. Maitotoxin was purchased from Calbiochem. Fura-2 acetoxymethylester (fura-2 AM) was purchased from Molecular Probes. Maitotoxin was dissolved in a 10 mm Mes solution pH 5.6. For patch-clamp experiments, the extracellular solution contained (mm): 140 NaCl, 5 KCl, 1 MgCl₂, 10 Hepes, 5 glucose and 2 CaCl₂; pH 7.4. The CaCl₂ was omitted in low Ca²⁺ solutions and in divalent cation experiments as indicated in the text and figure legends. The intracellular solution contained (mm): 145 potassium aspartate, 5 NaCl, 10 Hepes, 1 MgCl₂ and 2 EGTA; pH 7.2. In experiments performed under low intracellular Ca²⁺ conditions, EGTA was replaced with 5 mm BAPTA. For intracellular Ca²⁺ and Na⁺ measurements, the extracellular solution contained (mm): 120 NaCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 4.75 KCl, 10 glucose, 20 Hepes, 1.8 CaCl₂ and 0.05 % BSA; pH 7.2–7.4. In experiments with low extracellular Na⁺, this cation was replaced with N-methyl-d-glucamine (NMDG). To determine the relative ionic permeabilities for Ca²⁺versus Na⁺ and K⁺ the following extracellular solutions were used (mm): 145 KCl or NaCl, 10 Hepes, 5 glucose and 2 CaCl₂; pH 7.4. When K⁺ was present in the extracellular solution, 2 mm TEA was also added to block K⁺ currents. The intracellular solutions contained (mm): 145 potassium or sodium aspartate, 10 Hepes and 2 EGTA; pH 7.2.

Cell culture

Human skin fibroblast cells were obtained from skin biopsies of healthy donors as previously described (Soto et al. 1996). Cells were given with the informed consent of the donors and all procedures were evaluated and approved by the ethics committee of the Universidad Nacional Autónoma de México and conformed to The Declaration of Helsinki. Cells were maintained in culture at 37 °C (8–13 passages) in a humidity controlled incubator with 5 % in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10 % fetal bovine serum, 2 mm l-glutamine (Gibco), 100 μg ml⁻¹ streptomycin and 25 ng ml⁻¹ amphotericin B (Gibco).

Patch-clamp experiments

When monolayers reached confluency the cells were mechanically dispersed with a plastic pipette. Cells were placed on 35 mm Petri dishes in extracellular solution and mounted on the stage of a Nikon inverted microscope. The whole-cell configuration of the patch-clamp technique was used to study currents from single fibroblast cells under voltage-clamp (Hamill et al. 1981). Patch pipettes were made of 8161 glass (World Precision Instruments, Inc.) pulled to obtain pipette resistances in the range of 4–6 MΩ when measured in experimental solutions. Liquid junction potentials were compensated in experiments using NMDG-Cl solutions. The holding potential in all experiments was maintained at 0 mV. The cell capacitance varied between 8 and 12 pF. The AxoPatch 200A (Axon Instruments Inc., Foster City, CA, USA) was used as the patch-clamp amplifier. The reference electrode was an Ag–AgCl plug connected to the bath solution via a 100 mm KCl agar bridge. Voltage-clamp data were stored on tape and digitized later for analysis using the Digidata 1200 (Axon Instruments) analogue–digital interface and pCLAMP 6.0.4 software (Axon Instruments). The data were filtered at 10 kHz using a 4-pole Bessel filter (Axon Instruments) and digitized at 100 Hz. The binary data were converted to ASCII and multiple experiments were combined to obtain means ± standard deviation (s.d.) using the SigmaPlot 2.0 software (Jandel Corporation). In all the described experiments, n = number of experiments performed for each condition. All experiments were performed at room temperature (22–25 °C).

Determination of relative permeabilities

Ionic permeability ratios were calculated with a modified version of the Goldman–Hodgkin–Katz (GHK) equation (González-Perrett et al. 2001):

where V is the reversal potential (E_rev), R is the the gas constant, T is the absolute temperature in kelvin, F is the Faraday constant and the ion is K⁺ or Na⁺.

Free divalent cation determinations

Free concentrations of Ca²⁺, Ba²⁺ and Mg²⁺ were determined using the program ‘Bound and Determined’ (BAD 4.35) as previously published (Brooks & Storey, 1992). Briefly, the free ion concentrations reported in this study were calculated by introducing the values of temperature, pH and concentrations of the buffers and all ions present in the experimental solutions into the BAD program.

Intracellular Ca²⁺ measurements

Confluent cells were mechanically dispersed with a plastic pipette after 2 min of incubation with 0.01 % trypsin solution and placed in extracellular solution to a final concentration of 1.5–2 × 10⁶ cells ml⁻¹ and incubated with 5 μm fura-2 AM for 45 min. After this, the cells were rinsed twice and incubated for 30 min in fura-2 AM-free solution. The cells were then rinsed 3–4 times and placed in a cuvette of an Aminco-Bowman series 2 luminescence spectrometer. The excitation wavelength was alternated between 340 and 380 nm every 2 s and the emission was collected at 510 nm. Each experiment was individually calibrated to obtain the maximum fluorescence after disrupting the cells with 0.1 % Triton X-100, and the minimum fluorescence obtained after buffering the Ca²⁺ in the solution with 20 mm EGTA. The values obtained with this procedure were used to calculate the intracellular Ca²⁺ concentration according to previously published equations (Grynkiewicz et al. 1985). Auto-fluorescence of cells not loaded with fura-2 AM was subtracted from each experiment.

Intracellular Na⁺ measurements

Confluent cells were rinsed with Hanks' balanced salt solution (HBSS) and incubated for 2 min in 0.01 % trypsin solution. After this, the cells were rinsed twice with extracellular solution and incubated for 45 min with extracellular solution containing 1 % pluronic acid and 25 μm SBFI AM (Molecular Probes). The cells were rinsed and maintained in SBFI AM-free solution for 15 min, and placed in a cuvette of an Aminco-Bowman series 2 luminescence spectrometer. The excitation wavelength was 358 nm and the emission at 500 nm was collected every 2 s. Na⁺ calibration curves were performed according to previously published procedures for each individual experiment (Harootunian et al. 1989; Zhao et al. 1995).

RESULTS

K⁺ currents present in non-stimulated human skin fibroblasts

First we measured the currents present in human skin fibroblasts under resting conditions (absence of MTX). Figure 1A illustrates the outwardly rectifying current observed in fibroblasts studied with the whole-cell configuration of the patch-clamp technique. This current showed a reversal potential of −86 mV (n = 12).

Figure 1. Basal K⁺ currents from human skin fibroblasts.

Representative currents elicited in response to voltage steps from −100 to +100 mV (increment 20 mV) obtained from non-stimulated human skin fibroblasts. A, basal conditions. B, after increasing the extracellular K⁺ concentration from 5 to 25 mm. C, after applying 1 mm TEA. D, I–V curves for the three conditions described: basal (•), after increasing extracellular K⁺ to 25 mm (□) and after addition of 1 mm TEA (▵). Representative traces are shown from one cell (n = 4). Arrows indicate zero current level. Holding potential, V_h = 0 mV.

Increasing the extracellular K⁺ from 5 to 25 mm caused a shift of the reversal potential towards more positive values. Under this condition the reversal potential observed was −40 mV, which is near the predicted potential by the Nernst equation for a K⁺ electrode (−44 mV).

The outwardly rectifying current is completely inhibited by 1 mm TEA as illustrated in Fig. 1C. Figure 1D illustrates the current–voltage (I–V) relationships obtained in resting conditions (•), after increasing the extracellular K⁺ to 25 mm (□) and after the addition of 1 mm TEA (▵).

These results indicate that under resting conditions, the main current present in human skin fibroblasts is produced by an outwardly rectifying K⁺ channel inhibited by TEA.

MTX activates a cationic current carried by K⁺, Na⁺ and Ca²⁺

The effect of 500 pm MTX on human skin fibroblast currents was measured. We have previously shown that at this toxin concentration, the maximal MTX response is achieved (Gutierrez et al. 1997). Interestingly, multiple applications of MTX up to a final concentration of 2 nm did not result in further changes in the amplitude of the MTX-induced current (data not shown). This result supports previously published data indicating that MTX has no ionophoretic activity.

After the addition of MTX, the activation of a linear current was observed when applying voltage ramps during a patch-clamp experiment in the whole-cell configuration (Fig. 2B). The linear current activated by MTX (I_mtx) showed a reversal potential of −10 mV (n = 10, Fig. 2B). Figure 2C compares the control current (before addition of MTX, •) with the current evoked by MTX (Fig. 3C, □). Subtracting I_mtx from the control produced the net current activated by MTX (Fig. 3C, ▴). Due to the large amplitude of I_mtx in comparison with the control outwardly rectifying K⁺ current, the net current obtained after subtraction is very similar to I_mtx but it reverses at a slightly more positive value (−7 mV, Fig. 3C).

Figure 2. MTX activates a non-selective cationic current (I_mtx).

A, I–V curve of the basal K⁺ current present in human skin fibroblasts in the absence of MTX, the current reversal potential is −85.8 ± 4.1 mV (n = 12). B, I–V curve of the current activated by 500 pm MTX, the current reversal potential is −10.10 ± 6.4 mV (n = 10). C, comparison of the I–V curves under control conditions (•), total current after MTX application (control + MTX activated, □) and the net current activated by MTX (total – control, ▴). D, I–V curves for I_mtx obtained with physiological extracellular solution (□) and with NMDG-Cl in the external medium (n = 4, ▪). In the presence of external NMDG the current reversal potential shifts to values more negative than −100 mV. In all the cases means ± s.d. are shown. Where not shown error bars are smaller than symbols. V_h = 0 mV.

Figure 3. Relative cationic permeability of I_mtx.

Representative experiments showing the shift in the reversal potential of I_mtx induced after increasing extracellular calcium from 2 to 5 mm. A, control currents (absence of MTX) in 145 mm symmetrical Na⁺ produced by voltage steps from a holding potential of 0 mV to −100 mV and then to 100 mV in 20 mV steps. B, currents induced by 500 pm MTX in the presence of 2 mm extracellular Ca²⁺. C, same cell as in B after raising extracellular Ca²⁺ to 5 mm. D, current–voltage relationship showing the mean values (n = 4) of the currents in 145 mm Na⁺ and 2 mm Ca²⁺ (○—○) or 5 mm Ca²⁺ (♦—♦). E, control currents (absence of MTX) in 145 mm symmetrical K⁺ and 2 mm TEA produced by voltage steps from a holding potential of 0 mV to −100 mV and then to 100 mV in 20 mV steps. F, currents induced by 500 pm MTX in the presence of 2 mm extracellular Ca²⁺. G, same cell as in F after raising extracellular Ca²⁺ to 5 mm. H, current–voltage relationship showing the mean values (n = 5) of the currents in 145 mm K⁺ and 2 mm Ca²⁺ (□—□) or 5 mm Ca²⁺ (•—•). The voltage scales in D and H show only the data between −30 and +30 mV to evidence the shift in the reversal potential. Horizontal dotted lines indicate the zero current level.

To determine the ionic selectivity of I_mtx, the cations from the extracellular bath solution were replaced with NMDG. Replacing the extracellular solution with NMDG-Cl resulted in a shift of the reversal potential of I_mtx to a value more negative than −100 mV, indicating that I_mtx carries cations but not NMDG (Fig. 2D). Under these experimental conditions, the calculated reversal potential for an anionic current according to the Nernst equation is near 0 mV.

Further ionic substitution experiments were performed to determine the relative permeability of Ca²⁺ with respect to K⁺ and Na⁺ (P_Ca/P_K and P_Ca/P_Na). I_mtx was activated with [Ca²⁺]_o = 2 mm and symmetrical intra- and extracellular concentrations of K⁺ or Na⁺ (145 mm). When K⁺ was present as the main cation, 2 mm TEA was added to the extracellular bath to inhibit the basal K⁺ current present in fibroblasts. When I_mtx was fully activated, [Ca²⁺]_o was raised to 5 mm and the change in the reversal potential (E_rev) was monitored (Fig. 3). Addition of 5 mm Ca²⁺ with Na⁺ as the main cation present produced a shift in the E_rev of −4.9 mV (Fig. 3D, n = 4). In a similar set of experiments with K⁺ as the main cation present, the shift in the E_rev was −4.8 mV (Fig. 3H, n = 5). Calculating the relative permeabilities for each condition using a modified form of the GHK equation (Methods) resulted in P_Ca/P_Na and P_Ca/P_K = 0.87.

To evaluate the contribution of MTX to the changes in intracellular Ca²⁺ and Na⁺, cells were loaded with fluorescent indicators to measure the time course of changes in fluorescence. Whole-cell patch-clamp experiments were performed to follow the time course of I_mtx activation.

Figure 4 illustrates representative experiments of the time course of activation of I_mtx in response to 500 pm MTX (Fig. 4A) compared to the changes in intracellular Ca²⁺ (Fig. 4B) and Na⁺ (Fig. 4C) as monitored by the fluorescent indicators fura-2 and SBFI, respectively. As illustrated in this figure, the time course of I_mtx followed very closely the increment in intracellular Ca²⁺ and Na⁺ induced by MTX. This result along with the selectivity experiments previously shown strongly suggests that the increments in intracellular Ca²⁺ and Na⁺ are the result of the activation of I_mtx. As we have previously shown, MTX does not induce release of Ca²⁺ from intracellular storage compartments in these cells (Gutierrez et al. 1997). Therefore, the changes in intracellular Ca²⁺ observed after MTX must result from Ca²⁺ influx.

Figure 4. Time course of I_mtx and the increments in intracellular Ca²⁺ and Na⁺ concentrations.

Concomitant measurements of current using patch clamp and intracellular Na⁺ and Ca²⁺ with fluorescent indicators. A, a typical experiment showing the time course of I_mtx activation. The current was evoked with continuous ramps from −100 to +100 mV. The ramps were interrupted to perform voltage steps from −140 to +120 mV in increments of 20 mV (V_h = 0 mV) to obtain I-V curves. Voltage protocols shown in the inset. Vertical arrows indicate the time of addition of 500 pm MTX. B, time course of the changes in [Ca²⁺]_i in response to MTX monitored with the fluorescent indicator fura-2. C, time course of the changes in [Na⁺]_i monitored with the fluorescent indicator SBFI. Time scale is the same in all panels (shown at the bottom of the figure). Representative traces are shown from eight experiments.

Extracellular Ca²⁺ is required for the activation of I_mtx

To investigate the dependence of extracellular Ca²⁺ on I_mtx, we performed experiments in which the concentration of Ca²⁺ in the extracellular medium was modified.

Figure 5A illustrates the effect of 500 pm MTX on a cell bathed in the low Ca²⁺ extracellular solution. As shown in the figure, under these conditions MTX failed to activate I_mtx. However, addition of 2 mm Ca²⁺ in the presence of MTX caused activation of I_mtx (Fig. 5A, n = 5). To demonstrate that the current induced after Ca²⁺ application was not the result of Ca²⁺ re-addition, we performed experiments in which the extracellular Ca²⁺ was removed and later re-added in the absence of toxin (Fig. 5B). Under these conditions no current was observed until MTX was added.

Figure 5. Extracellular cation requirement for the activation of I_mtx.

A, addition of MTX to the extracellular medium (in the absence of Ca²⁺) did not induce the activation of I_mtx. The subsequent addition of Ca²⁺ (2 mm) resulted in a typical activation of I_mtx. B, when 2 mm Ca²⁺ is added before MTX current activation is produced only after toxin application. C, the effect of extracellular Ca²⁺ appears to be an irreversible process, the repeated addition of EGTA (5 mm each occasion) after the activation of I_mtx does not have any effect on the current. D, the rise in [Na⁺]_i is not observed when extracellular Ca²⁺ is chelated with 5 mm EGTA prior to the addition of MTX. With 2 mm Ca²⁺ in the bath solution (Control), MTX induced typical increments in [Na⁺]_i. The voltage ramp used in A is the same as that used in Fig. 3A. Representative traces are shown from five experiments for each condition.

Figure 5C shows that after I_mtx is fully activated by the toxin, removal of extracellular Ca²⁺ has no effect on the current. This result indicates that Ca²⁺ is required to induce the current, but not to maintain I_mtx, suggesting that Ca²⁺ is required for the association of MTX with its putative receptor.

Figure 5D shows that Ca²⁺ is required also for the increment in intracellular Na⁺ induced by MTX and measured with SBFI. This result further supports the hypothesis that the increments in intracellular Na⁺ are the result of I_mtx activation.

Since Ca²⁺ is only required for the activation of I_mtx, the relative permeability ratios for K⁺ and Na⁺ (P_K/P_Na) can be calculated once I_mtx is fully activated by chelating extracellular Ca²⁺ with the addition of 5 mm EGTA to the bath solution. From these experiments the P_K/P_Na obtained was ∼1. The permeability ratios P_Ca/P_Na and P_Ca/P_K previously calculated (Fig. 3), give a selectivity of I_mtx of Na⁺ (1) ≈ K⁺ > Ca²⁺ (0.87).

Extracellular Ca²⁺ cannot be replaced by other divalent cations for I_mtx activation

Since extracellular free Ca²⁺ is required for the activation of I_mtx, it was important to determine if other divalent cations could replace Ca²⁺ for the activation of I_mtx. MTX failed to activate I_mtx when Mg²⁺ (up to 10 mm) replaced Ca²⁺ in the extracellular solution (Fig. 6A). Similar results were obtained with Mn²⁺ (data not shown). MTX in the presence of 5 mm extracellular Ba²⁺ induced about 25 % of the current amplitude compared to 2 mm Ca²⁺(Fig. 6B).

Figure 6. Extracellular divalent cations required for I_mtx activation.

A, Mg²⁺ does not replace the action of Ca²⁺ in the activation of I_mtx Only after the addition of Ca²⁺ was I_mtx induced. B, Ba²⁺ was partially effective in replacing extracellular Ca²⁺ in the activation of I_mtx. C, two extracellular concentrations of Ca²⁺ (1 and 2 mm). The activation of I_mtx is observed only with 2 mm Ca²⁺. Bars above the current recordings show the cation present in the bath. The patch-clamp stimulation protocol shown here is the same as in Fig. 3A. D, data obtained from the experiments in (A–C) and at other concentrations given in the text. The maximum current activated in the presence of 2 mm [Ca²⁺]_o is considered 100 % of the response. The symbols indicate mean and s.d. of n number of experiments for: Ca²⁺ (n = 25), Ba²⁺ (n = 17) and Mg²⁺ (n = 10). Where not shown error bars are smaller than symbols.

The percentage of current induced in the presence of Ba²⁺, Mg²⁺ and Ca²⁺ is compared in Fig. 6D. In 10 mm Ba²⁺, MTX induced about half the current amplitude activated in 2 nm Ca²⁺. The Ca²⁺ requirement reached a saturation point around 1.8 mm, and further additions of Ca²⁺ did not result in I_mtx amplitude increments (Fig. 6D).

Intracellular Ca²⁺ is also required for the activation of I_mtx

The requirement of intracellular Ca²⁺ for the activation of I_mtx was also evaluated (Fig. 7). When 5 mm BAPTA was included in the pipette solution to chelate the Ca²⁺ in the intracellular medium, MTX failed to activate I_mtx (Fig. 7B). These results were obtained in the presence of 2 mm extracellular Ca²⁺, indicating that free Ca²⁺ is required on both sides of the cell membrane for the activation of I_mtx. Figure 7C shows the I_mtx current density obtained with 1 mm free intracellular Ca²⁺ and in experiments with 5 mm BAPTA (no Ca²⁺ added). With BAPTA in the pipette, MTX induced 4.6 % of the current activated without BAPTA (control 1.05 ± 0.3 nA pF⁻¹, BAPTA 0.05 ± 0.01 nA pF⁻¹, n = 5).

Figure 7. Intracellular Ca²⁺ is required also for activation of I_mtx.

A, control experiment showing the activation of I_mtx in response to 500 pm MTX with [Ca²⁺]_o = 2 mm and [Ca²⁺]_i = 1 μm. B, activation of I_mtx when [Ca²⁺]_o = 2 mm and [Ca²⁺]_i is near zero with 5 mm BAPTA in the pipette solution. The voltage ramp used here is the same as in Fig. 3A. C, mean ± s.d. from experiments with or without BAPTA in the pipette solution. In BAPTA loaded cells only 4.57 % of the current obtained in control conditions was observed (n = 5). The values of the I_mtx current were obtained by subtracting the basal fibroblast K⁺ current (I₁, ▾) from the total current activated by MTX (I₂, ▿, n = 10). The graph shows the values of the current corrected by capacitance (I₂ – I₁/C). The capacitance of the cells varied between 8 and 12 pF.

Intracellular Ca²⁺ can be replaced by Ba²⁺ but not by Mg²⁺

The ability of divalent cations other than intracellular Ca²⁺ as activators of I_mtx was evaluated by including BAPTA (5 mm) in the pipette solution along with a fixed concentration of Mg²⁺ or Ba²⁺. The extracellular solution contained physiological concentrations of Ca²⁺ as described in Methods. The free intracellular concentration of Ca²⁺, Mg²⁺ and Ba²⁺ obtained under these conditions was calculated as described in Methods.

Only Ba²⁺ was capable of replacing Ca²⁺. However, about 10 times more Ba²⁺ was required. Even at concentrations as high as 8 mm intracellular Mg²⁺, MTX failed to induce I_mtx (n = 5, data not shown).

DISCUSSION

Since the original finding showing that MTX may be one of the entities responsible for the seafood poisoning known as ciguatera, many attempts have been made to elucidate the molecular mechanism by which this toxin induces its effects. Evidence has accumulated in recent years indicating that MTX is a very powerful activator of Ca²⁺ influx in practically all cells tested. Due to the fact that MTX can activate Ca²⁺ influx in a wide variety of cells from different organisms, it was originally proposed that this toxin was acting as an ionophore (Gusovsky & Daly, 1990). However, several lines of evidence argue against the ionophoretic activity of MTX. Firstly, MTX has no effect on liposomes or mitochondria (Takahashi et al. 1983). Secondly, we have previously shown that trypsin treatment of human skin fibroblasts eliminates the effects of MTX on Ca²⁺ influx, suggesting that a trypsin-sensitive protein from the plasma membrane is mediating the effects of MTX on Ca²⁺ (Gutierrez et al. 1997). Thirdly, MTX cannot activate channels in cell patches in the inside-out configuration obtained from Madin-Darby canine kidney (MDCK) cells. However, activation of cationic channels can be achieved in patches in the outside-out configuration (Dietl & Völkl, 1994), showing that MTX can activate the channel only when applied to the extracellular surface of the membrane. Furthermore, MTX induced only one or two single channels over a wide range of toxin concentrations in patches obtained from guinea-pig single ventricular myocytes (Nishio et al. 1996). In this report we have found that the rise in intracellular Ca²⁺ and Na⁺ concentrations and the activation of I_mtx by MTX saturate and occur at very low toxin concentrations. This type of behaviour is atypical for an ionophore. All these results strongly suggest that MTX is inducing Ca²⁺ and Na⁺ influx via activation of non-selective cationic channels such as the one shown in this study. These observations suggest that MTX selectively recognizes a target in the cell membrane, which opens the possibility that MTX could be acting through a widely distributed, highly conserved receptor or channel.

MTX effects on human skin fibroblasts

Human skin fibroblasts provide an excellent model to study the mechanism of action of MTX. These cells are highly sensitive to this toxin; the half-activation constant for Ca²⁺ mobilization is in the femtomolar range (Gutierrez et al. 1997). The results presented here show that MTX activates a non-selective cationic current (I_mtx) in these cells. The activation of I_mtx is correlated with the increments in intracellular Ca²⁺ and Na⁺. The activation of I_mtx and the increments in intracellular Na⁺ induced by MTX require the presence of intracellular and extracellular free Ca²⁺.

Previous studies have also reported macroscopic and single-channel currents activated by MTX in different cellular types (Escobar et al. 1998; Schilling et al. 1999). In all of them MTX (1 pm to 30 nm) induced large voltage-independent currents similar to the one reported here. The relative permeability calculated in the present study for I_mtx (Na⁺= K⁺ > Ca²⁺) is in agreement with previous reports showing the activation of non-selective cationic currents by MTX (reviewed in Escobar et al. 1998).

In a recent study using the SK45 fibroblast cell line, Schilling et al. (1999) described a dual effect of MTX. At low toxin concentrations (<2 nm), MTX activates non-selective cation channels (NSCCs); at higher toxin concentrations (>2 nm) a cytolytic/oncotic pore (COP) is activated. This COP becomes evident after 5 min of exposure to 2 nm MTX and permeates large molecules (<900 Da) such as fura-2 and NMDG. The MTX-induced COP resembles the large pore formed by stimulation of purinergic P2Z/P2X₇ receptors (Schilling et al. 1999). The present report focuses on the effects of low MTX concentrations on ionic conductances and the changes in intracellular Ca²⁺ and Na⁺. Our results indicate that I_mtx does not permeate fura-2, as intracellular Ca²⁺ increments reach a plateau phase after MTX stimulation (Fig. 4A and Gutierrez et al. 1997). If MTX promoted fura-2 leakage, a plateau phase would not be observed. Furthermore, we have previously shown that in human skin fibroblasts the increments in fluorescence promoted by increasing MTX concentrations saturate, arguing against the possibility that MTX is promoting fura-2 leakage from the cell. Finally, the shift that occurs in the reversal potential when the extracellular cations are replaced with NMDG indicates that this cation does not permeate the MTX-activated channel.

The activation of I_mtx in human skin fibroblasts requires the presence of free Ca²⁺ in the intracellular and extracellular solutions. Extracellular Ca²⁺ is only required for the activation of I_mtx, since removal of Ca²⁺ after I_mtx activation does not affect the amplitude of the ionic current.

Ba²⁺ is an inefficient replacement for Ca²⁺ on both sides of the membrane, around 10 times the concentration being required to achieve the same activation of current. Other divalent cations do not replace Ca²⁺. In other cell types, Ca²⁺ is also required for I_mtx activation but no dose–response curve for Ca²⁺ or any evidence about the role of other divalent cations has been presented to date (Gusovsky & Daly, 1990; Escobar et al. 1998; Weber et al. 2000). In guinea-pig single ventricular myocytes, I_mtx is activated in the absence of free Ca²⁺ on both sides of the membrane (Nishio et al. 1996). This is one of few reports found in the literature where Ca²⁺ is not required for MTX activation.

There are several possible reasons for the requirement of free extracellular Ca²⁺ for I_mtx activation. MTX may require extracellular Ca²⁺ to interact with its putative receptor in the plasma membrane. The hydrophobic region of MTX resembles that of gangliosides and it has been shown that the ganglioside GM1 inhibits some MTX effects (Konoki et al. 1999). Gangliosides aggregate in membranes, forming stable clusters (Thompson & Tillack, 1985). Extracellular Ca²⁺ may be required for the insertion of MTX into the plasma membrane.

Another possibility is that the NSCC activated by MTX is Ca²⁺ dependent. Several NSCCs are activated by intracellular Ca²⁺ (Hofmann et al. 2000). More complex mechanisms could be envisioned involving calcium– calmodulin interactions or calcium-dependent phosphorylation processes.