Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Tyrosine-specific protein kinase activities have been shown to be involved in the control of cell growth and differentiation. One of the main molecular mechanisms involved is the autophosphorylation of receptor tyrosine kinases in response to ligand binding (Pazin and Williams, 1992; Saltiel and Ohmichi, 1993). Tyrosine kinases seem also to be important in other phenomena such as ligand-induced internalization of antigen receptor in B lymphocytes (Puré and Tardelli, 1992), long-term potentiation in the hippocampus (O'Dell et al., 1991), clustering of acetylcholine receptors (Baker and Peng, 1993) and regulation of N-methyl-D-aspartate receptors (Wang and Salter, 1994). A simple way to check the involvement of these enzymes is to use specific inhibitors. It is thus essential to know if these latter molecules exhibit other pharmacological activities that are not mediated by tyrosine kinase inhibition. Genistein, an isoflavone compound isolated from the fermentation broth of Pseudomonas sp., has been characterized as a tyrosine kinase inhibitor because it strongly inhibits the tyrosine kinase activity of epidermal growth factor receptor, pp60^src, and pp110^gag-fes, whereas it exhibits a considerably weaker effect on protein serine/threonine kinases (Akiyama et al., 1987; Akiyama and Ogawara, 1991). The inhibition was competitive with respect to ATP and was also observed in intact cells.

In this paper, we demonstrate that genistein also blocks voltage-sensitive Na⁺ channels in cultured neurons. Because sodium channel activity in cultured central nervous system neurons is modulated by cAMP-dependent protein kinase and protein kinase C (Li et al., 1993), one could not exclude that genistein-induced channel blockade could be a consequence of the inhibition of the channel phosphorylation by a tyrosine kinase. However, our results clearly suggest that the blockade is caused by direct interaction of genistein with the channel protein.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. [³H]STX (63 Ci/mmol) was from Amersham (Arlington Heights, IL); ²²NaCl and [³H]BTX-B (50 Ci/mmol) were from Dupont New England Nuclear (Boston, MA). Toxin II from the scorpion Androctonus australis Hector (-ScTx) was a generous gift from Prof. H. Rochat (Marseille, France). The pyrethroid RU39568 was from Roussel-Uclaf (Romainville, France). Genistein was from Calbiochem (San Diego, CA) and LC Services Corporation (Woburn, MA); lavendustin A, compound 5 [(2-hydroxylbenzyl)aminobenzoic acid], tyrphostin A₄₇ (RG-50864), methyl 2,5-dihydroxycinnamate (an analog of erbstatin) and daidzein were from LC Services Corporation; veratridine was from Sigma Chemical Co. (St Louis, MO), daidzein from Calbiochem and ouabain from Boehringer (Mannheim, Germany).

Cell culture. Primary cultures of rat fetal brain neurons were prepared essentially as described previously (Jover et al., 1988), except that the culture medium was Dulbecco's modified Eagle medium (GIBCO BRL, Gaithersburg, MD) containing 5% fetal calf serum (Boehringer, Mannheim). Cultures of cerebellar granule cells were obtained as described (Grignon et al., 1993).

Sodium influx. The influx of ²²Na⁺ induced by neurotoxins was measured as described previously (Couraud et al., 1986). Cultured cells were preincubated in the presence of the indicated toxins and drugs in buffer A (5.4 mM KCl, 1.8 mM CaCl₂, 0.8 mM MgSO₄, 10 mM glucose, 25 mM HEPES, 1 mg/ml bovine serum albumin and Tris-base to adjust the pH to 7.4) containing 140 mM choline chloride. After 20 min at 37°C, the preincubation medium was replaced by prewarmed buffer A containing 130 mM choline chloride and 10 mM NaCl to which were added ²²Na⁺ (0.5 µCi/assay), 5 mM ouabain, neurotoxins and drugs at the concentrations specified under "Results." At the end of the incubation time, the medium was aspirated and cells were rinsed three times with 140 mM choline chloride in buffer A at 4°C, dissolved in 0.1 M NaOH and the accumulated radioactivity was measured.

[³H]BTX-B binding to rat brain synaptosomes. The synaptosomal crude fraction P2 was prepared as described previously (Jover et al., 1988) and aliquots were stocked at 80°C. The standard binding medium contained 140 mM choline chloride, 5 mM KCl, 1.8 mM CaCl₂, 0.8 mM MgSO₄, 2 mg/ml bovine serum albumin and 20 mM HEPES, adjusted to pH 7.4 with 1 M Tris. Concentrated solutions of veratridine, RU39568, genistein, lavendustin A and tyrphostin A₄₇ were prepared in dimethyl sulfoxide. Unless mentioned, the final concentration of dimethyl sulfoxide was 0.8% (v/v). Concentrated solutions of [³H]BTX-B were prepared in ethanol; in all experiments, the final concentration of ethanol was less than 0.1% (v/v). [³H]BTX-B and RU39568 were diluted in standard binding medium containing 0.04% (v/v) Emulphor EL-620, a nonionic detergent used as an emulsifier as described by Poli et al. (1986).

1

1

Electrophysiological experiments. Cerebellar granule cells in 35-mm dishes (Costar, Cambridge, MA) were used at days 5 and 6 of culture for electrophysiological experiments, which were performed at room temperature (20-22°C) with the single-electrode, whole-cell voltage-clamp technique by use of suction pipettes, ranging from 2 to 4 megohms. The final series resistance of electrodes was 3 to 6 megohms and compensated by 40 to 60%. The Na⁺ gradient was reversed to eliminate variability in the space clamp, allowing recordings of highly reproducible peak currents (Numann et al., 1991; Dargent et al., 1994). The external solution contained 90 mM choline Cl, 5 mM Na acetate, 15 mM tetraethylammonium-Cl, 1 mM MgCl₂, 1.5 mM CaCl₂, 1 mM KCl, 5 mM glucose, 0.2 mM CdCl₂ and 30 mM HEPES (pH adjusted to 7.3 with tetramethylammonium-OH). The internal solution contained 100 mM NaF, 30 mM NaCl, 20 mM CsF, 5 mM HEPES (pH adjusted to 7.3 with CsOH). Tetraethylammonium and Cs were used to ensure minimal K⁺ contribution to the outward Na⁺ channel currents. Currents were recorded by a Biologic (Grenoble, France) RK-300, low pass filtered at 2 kHz with an eight-pole Bessel filter and sampled at 20 kHz with a 12-bit ADC (Labmaster TM 40, Scientific Solution, Foster City, CA). In most experiments, capacitance and leak currents were substracted from active currents with use of a P/4 protocol (Bezanilla and Armstrong, 1977). The total capacitance of cells was 6 to 8 pF. Data acquisition and analysis were controlled by pCLAMP software (Axon Instruments, Foster City, CA), and data were fitted to the following equations with the SigmaPlot nonlinear curve fitter: Na⁺ peak current = a/(1 + exp^(x  V)/k) for the inactivation curve, Na⁺ peak current = a/(1 + exp^(x  V)/k) for the activation curve, in which x, k and a are parameters determined by multiple iterations of the algorithm.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Inhibition of ²²Na⁺ influx by genistein but not by other tyrosine kinase inhibitors in cultured fetal brain neurons. The effects of several tyrosine kinase inhibitors on the influx of ²²Na⁺ induced by neurotoxins were studied in cultured rat brain neurons. The influx through voltage-sensitive Na⁺ channels was revealed by addition of a mixture of -ScTx, which blocks channel inactivation by binding to site 3, and veratridine which alters both activation and inactivation by binding to site 2 (Catterall, 1980). Cells were preincubated for 20 min in a Na⁺-free medium with -ScTx and the different tyrosine kinase inhibitors. ²²Na⁺ influx was then elicited for 30 s in the presence of a mixture of 20 nM -ScTx, 5 µM veratridine and the inhibitors. In these conditions, genistein at 250 µM completely inhibited the toxin-induced ²²Na⁺ influx (fig. 1A). On the contrary, lavendustin A (10 µM), compound 5 (10 µM), tyrphostin A₄₇ (250 µM) and the erbstatin analog (10 µM) had no significant effect on neurotoxin-induced ²²Na⁺ influx. These drugs have been shown to specifically inhibit tyrosine kinase activity in rat hippocampus with IC₅₀ of 18 µM for genistein, 0.5 µM for lavendustin A and compound 5 and 8 µM for tyrphostin A₄₇ (O'Dell et al., 1991), which indicated that at concentrations used in our experiments kinase inhibition was complete. Daidzein, a genistein analog that lacks tyrosine kinase inhibitory activity (Akiyama and Ogawara, 1991), was also able to block ²²Na⁺ uptake. The dose-response curves of genistein and daidzein indicate IC₅₀ values for ²²Na⁺ influx of 60 µM and 195 µM, respectively (fig. 1B). In agreement with the dose-response curve, we observed a 72 ± 2% inhibition of ²²Na⁺ influx at 250 µM daidzein and a complete inhibition at 250 µM genistein (data not shown). To measure the time course of genistein action, we preincubated cultured neurons with 100 µM genistein for different periods of time before a 15-s period of ²²Na⁺ uptake. Figure 1C shows that maximum sodium flux inhibition was obtained within 20 s, which indicated that the time course of genistein interaction with intact cells was very rapid.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Inhibition of 22Na+ uptake by genistein and by daidzein. (A) Cultured neurons (10-13 days in vitro) were preincubated at 37°C for 20 min with 20 nM -ScTx and the indicated drug in a Na+-free medium, then incubated at 37°C for 30 s with 20 nM -ScTx, 5 µM veratridine and the indicated drug in a medium containing 10 mM NaCl, 0.5 µCi/assay 22Na+ and 5 mM ouabain. Cells were washed three times with Na+-free medium at 4°C. Results are expressed as percent of control, i.e., percent of specific 22Na+ uptake in the absence of tyrosine kinase inhibitor. Specific 22Na+ uptake was 22Na+ uptake in the presence of toxins (-ScTx and/or veratridine) minus 22Na+ uptake in the absence of any toxin. Drug concentrations were 250 µM for genistein, 10 µM for lavendustin A, 10 µM for compound 5, 250 µM for tyrphostin A47 and 10 µM for erbstatin analog. (B) Dose-response curves of genistein and daidzein. Cultured neurons were treated as in panel A with increasing concentrations of genistein (open circles) or daidzein (filled circles). (C) Time course of the effect of genistein. Cultured neurons were treated as in panel A except that the preincubation was in the presence of 100 µM genistein during the indicated period of time, and the incubation period was 15 s.

Inhibition of sodium current by genistein in cultured cerebellar granule cells. Voltage-sensitive sodium currents were measured in cultured cerebellar granule cells by the patch-clamp technique in the whole-cell configuration. The Na⁺ gradient was reversed to eliminate space-clamp variability (Numann et al., 1991). Addition of 100 µM genistein in the extracellular medium induced a progressive decrease in the amplitude of whole-cell outward Na⁺ current. The inhibition was partial (70% decrease), and a plateau was reached after 5 min (fig. 2). At higher concentrations block was complete (data not shown). Figure 3A shows Na⁺ peak current amplitude evoked by 8-mV depolarization steps from 60 mV to +60 mV. In the conditions of reversed Na⁺ gradient, the value calculated for the reversal potential for Na⁺ was 82 mV, and Na⁺ currents were only observed outward. Application of 20 µM genistein induced a significant reduction of the Na⁺ peak current amplitude (n = 6 cells). Figure 3B indicates that changes in the voltage dependency of both activation and inactivation could be detected. After treatment with 20 µM genistein, a shift to the left of the voltage-inactivation curve was observed, whereas the voltage-activation curve shifted about 20 mV toward more positive potentials. This shift was mainly caused by a change in the slope of the curve which makes the interpretation difficult. The changes were complete 5 min after genistein was added to the cell bath medium. In the same experimental conditions, lavendustin A (10 µM) was ineffective and tyrphostin A₄₇ (100 µM) showed no significant effect (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Whole-cell outward Na+ currents recorded from a granule cerebellar cell before (upper), 2 min (middle) and 5 min (bottom) after application of 100 µM genistein. Superimposed current traces evoked by depolarization from 60 mV to +60 mV in 8-mV steps. The holding potential was 90 mV. Test pulses were applied at 0.25 Hz with a 45 ms duration. No additional change occurred after a 5-min application.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   (A) Current/voltage relationships for Na+ peak current obtained before (open circles) and 10 min after (filled circles) 20 µM genistein application. Test pulses were applied as indicated in figure 2. Reduction of Na+ peak current amplitude by genistein was observed. The data points show the mean ± S.E.M. for six granule cerebellar cells. (B) Voltage dependence of inactivation and activation in cells before (open symbols) and 10 min after (closed symbols) application of 20 µM genistein. Currents during the test pulse were normalized to the largest outward current and plotted versus prepulse potential (inactivation) or test pulse (activation). Steady-state inactivation curves were determined with a 200-ms prepulse from 110 mV to 0 mV in 10-mV steps, followed by a test pulse to +40 mV. Current-voltage plots of the peak currents derived from values shown in panel A. The data points are mean ± S.E.M. for six cells.

Apparent competition between genistein and veratridine on ²²Na⁺ influx in cultured fetal brain neurons. We have investigated the effect of veratridine on the dose-response curves of genistein on ²²Na⁺-specific uptake by rat fetal brain neurons in culture. Figure 4 shows that, in the absence of veratridine in the preincubation medium, the IC₅₀ for genistein was 70 µM, whereas, when 5 µM veratridine was present in the preincubation medium, the dose-response curve was shifted toward higher concentrations of genistein, giving an IC₅₀ of 150 µM. At a high concentration of veratridine, i.e., 50 µM, 500 µM genistein was unable to inhibit ²²Na⁺ influx into the cells. Higher concentrations of genistein could not be checked because of the insolubility of the drug. This result suggests that competition between veratridine and genistein may occur.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   The effect of veratridine on the inhibition of 22Na+ uptake by genistein was measured as in figure 1, except that cultured neurons were preincubated in the absence (open circles) or the presence of 5 µM (closed triangles) or 50 µM (closed diamonds) veratridine.

Inhibition of [³H]BTX-B binding to rat brain synaptosomes by genistein. To analyze the molecular mechanism of the apparent competition between genistein and veratridine, we have looked at the effect of genistein on the binding of [³H]BTX-B to rat brain synaptosomes. Because the level of specific binding of the latter toxin was low, experiments were done in the presence of the pyrethroid RU39568 (10 µM) that was shown to increase the affinity of [³H]BTX-B to site 2 (Lombet et al., 1988; Trainer et al., 1993). In these conditions, binding equilibrium of [³H]BTX-B to synaptosomes was obtained after 16 h at 26°C as shown in figure 5A. Dissociation experiments (fig. 5B) allowed the calculation of a dissociation rate constant k₁ of 5.0 × 10⁵ s¹ and with data from the association kinetics the association rate constant k₁ was calculated at 5.5 × 10³ s¹ M¹, which gives an equilibrium dissociation constant K_d = k₁/k₁ of 9 nM. A value of 15 nM has been obtained in similar conditions, i.e., in the presence of 10 µM RU39568 on synaptosomal membranes by Lombet et al. (1988), whereas a higher affinity (K_d = 1.5 nM) has been measured on solubilized and purified sodium channel (Trainer et al., 1993). The difference could be a consequence of the voltage dependence of pyrethroid interaction with sodium channels, synaptosomal membranes and frozen P2 fractions probably being depolarized compared with purified and reconstituted channels.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   (A) Time course of formation of the [3H]BTX-B/receptor complex in the absence or presence of genistein. Synaptosomes were incubated at 26°C with 1 nM [3H]BTX-B and 10 µM RU39568 in the presence (closed circles) and absence (open circles) of 500 µM genistein. At the indicated times, [3H]BTX-B bound was determined as described under "Materials and Methods." (Insert) linearized plot, B represents the [3H]BTX-B bound at each time and Beq is the bound ligand at equilibrium. (B) Dissociation of the [3H]BTX-B/receptor complex in the absence or presence of genistein. Synaptosomes were incubated as described under "Materials and Methods." Dissociation was initiated by adding 600 µM veratridine, in the presence (closed circles) or absence (open circles) of 500 µM genistein and at the indicated times, bound [3H]BTX-B was determined. The results presented show data of a single experiment performed in triplicate. (Insert) linearized plot, B represents the [3H]BTX-B bound at each time and Bo is the bound ligand at t = 0. (C) Dose-response curve of genistein effect on the binding of [3H]BTX-B to synaptosomes. [3H]BTX-B was incubated for 16 h at 26°C in the presence of increasing concentrations of genistein as described under "Materials and Methods." The results represent data of a single experiment performed in quadruplicate.

1

5

1)

View larger version (41K): [in this window] [in a new window]   Fig. 6.   Effect of the tyrosine kinase inhibitors, lavendustin A and tyrphostin A47, on the binding of [3H]BTX-B to rat brain synaptosomes. Synaptosomes were preincubated for 30 min at room temperature, either in absence (left bars) or in presence (right bars) of 5 µM lavendustin A (lav A) and 100 µM tyrphostin A47 (tyr A47), before adding 1 nM [3H]BTX-B, with (filled bars) or without (empty bars) 250 µM genistein and incubation was carried out for 8 h at 26°C. Specifically bound [3H]BTX-B was measured as described under "Materials and Methods."

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In this paper, we have analyzed the effects of genistein, a specific tyrosine kinase inhibitor, on voltage-sensitive sodium channels. We have shown that this drug induced a blockage of ²²Na⁺ influx through neurotoxin-activated sodium channels in cultured brain neurons and a decrease of Na⁺ current in cultured cerebellar granule cells. At this point, the question was whether the inhibitory effect was mediated by the inhibition of a tyrosine kinase activity. Several data argue against this hypothesis: 1) only genistein but not the other tested tyrosine kinase inhibitors was active on ²²Na⁺ influx or Na⁺ current; 2) the effect of genistein was mimicked by daidzein which is described as a genistein analog inactive on tyrosine kinase activity; 3) the time course of genistein activity on intact neuronal cells was very rapid because the maximum effect was obtained in less than 20 s, which seems incompatible with an effect caused by a dephosphorylation revealed by the inhibition of a tyrosine kinase; 4) a direct interaction between genistein and sodium channels was visualized in rat brain synaptosomal fractions by the inhibition of [³H]BTX-B specific binding induced by genistein and not by other tyrosine kinase inhibitors. The two latter arguments also allow the elimination of a possible effect through the inhibition of another protein kinase, the protein histidine kinase, which has been shown to be sensitive to genistein with an IC₅₀ of 110 µM (Huang et al., 1992).

Regarding the site of action of genistein, it is clear that it has no effect on binding of STX to neurotoxin receptor site 1. In contrast, genistein has strong effects on veratridine and batrachotoxin action and binding at receptor site 2. This could be caused either by direct competition at this binding site or by an indirect allosteric interaction similar to what has been observed with local anesthetics and some antiarrhythmic and anticonvulsant drugs (Catterall, 1987). These drugs have been shown to accelerate the dissociation of the preformed batrachotoxin-receptor complex (Postma and Catterall, 1984) whereas, on the contrary, genistein induced a small decrease in the dissociation kinetics, which is not in agreement with negative cooperativity. However, like local anesthetics, genistein induced a shift in the voltage dependence of inactivation to the more negative potentials (Catterall, 1987; Ragsdale et al., 1991). Although genistein alters the voltage dependence of the rat brain Na⁺ channel, it could not be ignored that it may also decrease the channel conductance.

An alternative explanation for the activity of genistein is that the inhibition of [³H]BTX-B binding was caused by competition between genistein and RU39568 for the same binding site, inducing a decrease in the level of bound pyrethroid and thus a decrease in its stimulatory action on site 2. This hypothesis can be excluded because the effect of genistein on [³H]BTX-B binding was measured at two concentrations of RU39568, 10 µM and 50 µM, and no change in the apparent affinity of genistein was detected (data not shown).

Thus, it seems that genistein competes with BTX for the same binding site on rat brain Na⁺ channels, but we cannot exclude an allosteric effect that does not induce an increase in the dissociation kinetics of [³H]BTX-B from the preformed complex.

In conclusion, this paper mainly demonstrates that genistein, a drug very often used as a specific tyrosine kinase inhibitor, is also able to block neuronal voltage-sensitive sodium channels in a direct manner and not through the inhibition of a tyrosine kinase activity.