Introduction

Top Abstract Introduction Methods Results Discussion References

Lubeluzole, the (+)-(S)-enantiomer of a benzothiazole derivative (fig. 1), has neuroprotective properties (De Ryck, 1997). Intravenous post-treatment with lubeluzole rescued sensorimotor function and reduced infarct volume after photochemically induced neocortical infarcts in rats (De Ryck et al., 1996). The (R)-enantiomer (R091154) of lubeluzole was inactive. In the peri-infarct zone surrounding such neocortical infarcts, lubeluzole reduced the infarct-induced rise in extracellular glutamate (Scheller et al., 1995) and normalized paired-pulse, -aminobutyric acid-mediated inhibition (Buchkremer-Ratzmann and Witte, 1997). Again, the (R)-enantiomer was ineffective. Lubeluzole also reduced infarct volume after focal cerebral ischemia induced by middle cerebral artery occlusion (Aronowski et al., 1996) and attenuated delayed neuronal death in a model of global ischemia in rats (Haseldonckx et al., 1996). Experiments on neuronal cultures have shown that lubeluzole inhibits anoxia- and glutamate-induced nitric oxide-related neurotoxicity and that it blocks neurotoxicity induced by nitric oxide donors (Benjamins et al., 1996; Lesage et al., 1996; Maiese et al., 1997). In these in vitro paradigms of neuroprotection, the (R)-enantiomer was either less active or inactive. Therefore, the neuroprotective mechanism of action of lubeluzole may be based on stereospecific down-regulation of the NOS pathway. Lubeluzole is neither an NOS inhibitor nor an NMDA antagonist (Lesage et al., 1996). Although lubeluzole blocks the fast sodium channel and antagonizes veratridine-induced neurotoxicity (Osikowska-Evers et al., 1995; Ashton et al., 1997), these effects are not stereospecific.

View larger version (7K): [in this window] [in a new window]   Fig. 1.   Chemical structure of lubeluzole [(+)-(S)-4-(2-benzothiazolylmethylamino)--[(3,4-difluorophenoxy)methyl]-1-piperidineethanol].

In the present study, we investigated whether lubeluzole and the (R)-enantiomer act on neuronal calcium channels and, if so, whether such effects are stereospecific. Several lines of evidence suggest that intracellular Ca⁺⁺ overload, which leads to activation of proteases, nucleases, phosphatase and other degradative enzymes, can lead to free radical production and neuronal cell death (Choi, 1995; Kristian and Siesjö, 1996). Ischemic Ca⁺⁺ overload can arise by excessive release of excitatory neurotransmitters, such as glutamate, which induces a neuronal influx of Ca⁺⁺ via the NMDA receptor and some AMPA- and kainate-gated ion channels (Lu et al., 1996). Influx of Ca⁺⁺ via inverse Na⁺/Ca⁺⁺ transport after cellular Na⁺ overload can also contribute to Ca⁺⁺ overload (Urenjak and Obrenovitch, 1996). Another direct pathway for intracellular Ca⁺⁺ overload is Ca⁺⁺ influx via voltage-sensitive Ca⁺⁺ channels (VSCC). Ca⁺⁺ influx via presynaptic VSCC triggers neurotransmitter release and can thus also indirectly influence postsynaptic Ca⁺⁺ overload. Partial inhibition of Ca⁺⁺ channels might thus be neuroprotective in ischemic conditions. There are many types of Ca⁺⁺ channels (De Waard et al., 1996; Mori et al., 1996). They can be distinguished partly by their activation voltages. A slight depolarization to 50 mV activates the transient LVA Ca⁺⁺ current (or T current). Depolarization to more positive voltages activates the HVA Ca⁺⁺ current, to which many types of Ca⁺⁺ channels can contribute. To investigate the influence of lubeluzole on Ca⁺⁺ channel currents, we used rat dorsal root ganglion cells, which express both LVA and HVA Ca⁺⁺ channels (Scroggs and Fox, 1992; Mintz et al., 1992; Diochot et al., 1995).

	Methods

Top Abstract Introduction Methods Results Discussion References

Cell preparation. DRG neurons were isolated by use of a modification of the technique of Delree et al. (1989). Briefly, 3-month-old Wistar male rats were decapitated under isoflurane/N₂O anesthesia. DRGs were removed aseptically and freed from connective tissue. They were digested at 37°C in 1 ml of 0.5% collagenase medium for 45 min, to which 1 ml of 0.25% trypsin medium was then added, and digestion was allowed to continue for a further 30 min. The collagenase medium contained 0.5% collagenase (Boehringer-Mannheim, Mannheim, Germany), and NaCl 120 mM, KCl 4.8 mM, KH₂PO₄ 1.2 mM, MgSO₄ 1.2 mM, D-glucose 3 mM, CaCl₂ 0.1 mM, NaHCO₃ 19.7 mM and HEPES 31.4 mM at pH 8. The trypsin medium contained 0.25% trypsin (GIBCO, Belgium), and NaCl 136.7 mM, KCl 2.7 mM, Na₂HPO₄ 16.3 mM and KH₂PO₄ 1.5 mM. The ganglions were washed in DMEM (Flow, Brussels, Belgium) with 10% FCS and then centrifuged. The resuspended cells were further dissociated mechanically by trituration with Pasteur pipettes of decreasing diameters. To remove myelin debris, the cell suspension was layered onto a Percoll solution and centrifuged (800 × g, 12 min). The pellet was washed with DMEM-FCS and centrifuged again (300 × g, 5 min). This purified pellet was resuspended in DMEM-FCS and plated onto a Petri dish coated with FCS for 90 min at 37°C. Nonneuronal cells adhere to this coating and are thereby removed from the cell suspension. The cell suspension was then spun down, and the pellet was resuspended in a 50:50 mixture of DMEM-N1 medium and C6-conditioned DMEM-N1. DMEM-N1 is DMEM with N1 supplements (transferrin 5 µg/ml, putrescine 0.1 mM, insulin 5 µg/ml, progesterone 0.02 µM and Na₂SeO₃ 0.03 µM) to which 6 g/liter D-glucose and 100 units/ml of penicillin/streptomycin are added. C6-conditioned DMEM-N1 was obtained by harvesting DMEM-N1 medium that has been incubated on confluent C6 glioma cultures for 48 hr (Delree et al., 1989; Geerts et al., 1992). Finally, the DRG cells were seeded onto the center of Petri dishes that had previously been coated with 100 µg/ml of poly-L-lysine (1 hr) and then with 10 µg/ml of laminin (overnight). The DRG cells were incubated overnight at 37°C in a humidified atmosphere (95% air/5% CO₂). The next morning, they were stored at room temperature in 95% air/5% CO₂. These cells were used for voltage-clamp experiments on the day of isolation and the 2 following days.

Electrophysiological recording. Whole-cell voltage-clamp (Hamill et al., 1981) was performed at room temperature (18-21°C). A Petri dish containing attached DRG cells was transferred to the stage of a Patch Clamp Tower (Luigs and Neumann, Germany) and examined with an inverted phase-contrast microscope (Diaphot TMD, Nikon). The experimental chamber was continuously perfused with a solution containing NaCl 152 mM, KCl 3 mM, D-glucose 10 mM, HEPES 10 mM, CaCl₂ 2 mM and MgCl₂ 1 mM at pH 7.4. Patch electrodes of borosilicate glass (Jencons, H15/10) were pulled and fire polished by means of a Zeitz DMZ puller (Augsburg, Germany). The electrode resistance ranged from 1.5 to 2.5 M when measured in the bath. After the patch had been broken, the cells were allowed to equilibrate for 5 min with the contents of the electrode.

Solutions. The internal pipette solution contained CsCl 100 mM, EGTA 10 mM, MgCl₂ 1 mM, MgATP 3 mM, TrisGTP 0.3 mM and HEPES 40 mM and was adjusted to pH 7.2 with CsOH. The external solutions contained BaCl₂ 2 mM, tetraethylammonium chloride 135 mM, tetrodotoxin 0.5 µM and HEPES 10 mM and were adjusted to pH 7.4 with tetraethylammonium hydroxide. Because Ba⁺⁺ was used in the extracellular solution instead of Ca⁺⁺, the current through Ca⁺⁺ channels was carried by Ba⁺⁺. This was done because Ba⁺⁺ suppresses residual currents through the delayed rectifier (Adams and Nonner, 1990) and inward rectifier potassium channels (Hagiwara et al., 1978) and because the high-voltage-activated Ca⁺⁺ channel current is then devoid of (Ca⁺⁺)-dependent inactivation. Also, unlike Ca⁺⁺, Ba⁺⁺ does not permeate or barely permeates Na⁺ channels.

Statistical analysis. All results are expressed as mean ± S.D., except for the concentration-response curves in figure 3, where mean ± S.E. is used. The difference between the effects of lubeluzole and the (R)-enantiomer on the amplitude of iLVA and iHVA (fig. 3) was evaluated for each concentration by means of the two-sided Student's t test for independent samples. In this procedure, P values were adjusted for the number of different comparisons (five) by applying Bonferroni's inequality. The same statistical method was used to evaluate the difference between two sequences of drug application [from lubeluzole to (R)-enantiomer and from the (R)-enantiomer to lubeluzole] in table 1 and also for the effects on the time constant of the decay of iHVA. In the experiments on the inactivation of iLVA (table 2) and on the voltage dependence of the block of iHVA, all pairwise comparisons of the three treatment groups [control (vehicle), lubeluzole and the (R)-enantiomer] were made using the Tukey-Kramer multiple comparison procedure. Values of P < .05 were considered to indicate statistical significance. Computations were carried out using the SAS system for statistical analysis.

View this table: [in this window] [in a new window]   TABLE 2 Influence of lubeluzole and the (R)-enantiomer on the inactivation of iLVA The inactivation curve was fitted according to the Boltzmann equation (see text), where I is the amplitude of iLVA and V is the potential of the conditioning prepulse, Imax is the maximum amplitude of iLVA (derived from the fitted inactivation curve), Vh is the half-inactivation potential and Sl is the slope factor. The inactivation curve was determined three times at 5-min intervals in each cell. Either all three measurements were performed in the control solution (first row), or the first measurement was carried out in the control solution, the second after a 5-min drug application (second and third row) and the third after a 5-min washout period. The change in Vh (Vh) is the Vh after 5-min application of the control or drug solution (second determination) minus the Vh from the first determination in the control solution. The change in Sl (Sl) was calculated analogously. Ratio Imax is the ratio of the Imax from the second inactivation curve (in control or drug solution) to the Imax from the first inactivation curve in the control solution. N is the number of cells in each group. The values obtained in the drug-treated groups were compared with those in the control group by means of a Tukey-Kramer multiple-comparisons procedure.

	Results

Top Abstract Introduction Methods Results Discussion References

iLVA and iHVA in DRG cells. Some of the DRG cells showed both the transient iLVA (or T current) and the iHVA, whereas others showed only iHVA. The low-voltage-activated current and the high-voltage-activated current could be tested simultaneously by means of the pulse sequence shown in figure 2A. From an HP of 100 mV, first a 155-msec test pulse to 50 mV elicited mainly iLVA (if present). After a 10-msec interval at 100 mV, a 155-msec test pulse to 20 mV elicited iHVA, which was nearly maximal at this potential. The amplitude of iLVA was measured as the transient component of the current at the test pulse of 50 mV, thus as the difference between the peak inward current and the current at the end of this test pulse. The current during the test pulse to 20 mV was contaminated very little by iLVA, because iLVA was still inactivated by the preceding pulse to 50 mV.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Influence of 3 µM lubeluzole and 3 µM (R)-enantiomer on Ca++ channel currents in DRG cells. A, Voltages applied and currents measured in the control solution and after superfusion with 3 µM lubeluzole or the (R)-enantiomer. The numbers in parentheses refer to the time point at which the sweeps were recorded, as shown in B and C. B, Time course of the peak amplitude of the current at the 155-msec test pulse to 50 mV (iLVApeak), the current at the end of the test pulse to 50 mV (iLVAend), the peak current at the 155-msec test pulse to 20 mV (iHVApeak) and the current at the end of the test pulse to 20 mV (iHVAend). The difference between iLVApeak and iLVAend represents the low-voltage-activated current (iLVA). The arrows indicate the time points at which the new solution was applied. C, Current ratios for rLVA, rHVApeak and rHVAend. To obtain these ratios, the time courses of iLVA, iHVApeak and iHVAend in the control period were fitted exponentially and extrapolated to the end of the experiment. Division of each measured current amplitude by the value of the corresponding calculated curve obtained by fitting, at the same time point, yielded the ratios.

Influence of lubeluzole and the (R)-enantiomer on iLVA and iHVA. The influence of lubeluzole on Ca⁺⁺ channel currents was investigated and compared with that of the (R)-enantiomer by means of the following protocol: the cells were stimulated every 30 sec with the pulse sequence shown in figure 2A. This was done for 5 min in the control solution, followed by 5 min in the presence of one enantiomer and thereafter by 5 min in the presence of the other enantiomer at the same concentration (fig. 2B).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Concentration-response curves showing the effect of lubeluzole and the (R)-enantiomer on normalized Ca++ channel currents, expressed as rHVApeak, rHVAend and rLVA. The results are given as mean ± S.E. The number of cells tested is given in parentheses, in which the upper and lower numbers refer to lubeluzole and the (R)-enantiomer, respectively. Smooth curves are the best fit to the equation 1/[1 + ([drug]/IC50)n], in which n is the Hill coefficient. The IC50 values (and between brackets the corresponding Hill coefficients) for inhibition of rHVApeak by lubeluzole and the (R)-enantiomer were 2.6 µM (0.7) and 3.5 µM (0.7) respectively; for rHVAend: 0.7 µM (0.9) and 0.9 µM (0.8), respectively; and for rLVA, 1.2 µM (1.3) and 1.2 µM (1.1). The drug effects on rHVApeak and rHVAend were obtained after a 5-min application of the first enantiomer tested, such as time point 2 in figure 2C. Results from the second enantiomer tested in the same cell were not used for calculation of the IC50. Only one drug concentration was tested on each cell so as to avoid errors in the extrapolation of the run-down of iHVA. Only a small fraction of the cells used for the measurement of the concentration-response curves of iHVA (diameter, 32.1 ± 4.1 µm, mean ± S.D., n = 78) expressed iLVA. Therefore, the concentration-response of iLVA was tested in a separate series of experiments on medium-sized DRG cells (35-40 µm), which have a larger iLVA (Scroggs and Fox, 1992). In the latter series, two or three concentrations of a single compound were tested cumulatively in the same cell, because iLVA showed less run-down than iHVA. Drug effects were again measured 5 min after application of each new concentration.

Tonic block of iLVA and iHVA. To test the use-dependency of the block of iLVA and iHVA, the following protocol was used in a separate series of experiments. In the control solution, the cells were stimulated every 30 sec for 5 min by application of the double-pulse protocol shown in figure 4A. Thereafter, stimulation was stopped (fig. 4B). Two minutes later, 10 µM lubeluzole or the (R)-enantiomer was given. Five minute later (i.e., 7 min after stimulation was discontinued), the stimulation was started again. Already on the first stimulation, iLVA and iHVA were greatly reduced by 10 µM lubeluzole or the (R)-enantiomer, and only small further changes in the shape of iLVA and iHVA were observed (fig. 4A). The apparent rate of inactivation of iHVA was very much increased, already from the first stimulation. In addition, no stimulation was needed in the washout period to unblock. This was seen in all medium-sized cells (n = 3 for each enantiomer). These experiments show that lubeluzole and the (R)-enantiomer exert a tonic block of iLVA and iHVA. The decrease in iHVA after resumption of the stimulation in the washout period must be seen as a response of the amplitude of iHVA to the change in frequency of stimulation (1/30 Hz) and is not a normal run-down phenomenon.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Tonic block of iLVA and iHVA by lubeluzole. A, Ca++ channel currents in the control solution, in the presence of 10 µM lubeluzole and after a 5-min washout period. The numbers between parentheses refer to the time points at which the sweeps were recorded, as shown in B. B, Time course of the experiment, in which the cell was stimulated only at the data points shown. No prior stimulation was needed to obtain a block of iLVA and iHVA by lubeluzole or to achieve a relief of block after washout.

Voltage-dependent block of iLVA. To investigate the influence of lubeluzole on the inactivation of iLVA, the 155-msec test pulse to 50 mV was kept constant and the voltage of the 10-sec conditioning prepulse was varied between 60 and 120 mV (fig. 5). The time interval between the test pulses was 15 sec. The curve of the amplitudes (I) of iLVA at the different conditioning potentials (V) was fitted as a function of V according to the Boltzmann equation: I = Imax/{1 + exp[(V  Vh)/Sl]}. This fit yielded the parameters Imax, Vh and Sl; Imax is the maximum amplitude of iLVA in the absence of inactivation, Vh is the conditioning potential at which the inactivation is half-maximum and Sl is the slope factor.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Influence of lubeluzole on the inactivation of the low-voltage-activated Ca++ channel current. The 10-sec conditioning prepulse was varied between 60 and 120 mV, with the 155-msec test pulse maintained at 50 mV. The solutions were applied in the following order: control solution, 1 µM lubeluzole (5 min), washout in control solution (5 min), 3 µM lubeluzole (5 min) and washout 2 (12 min). For each test solution, the curve of the amplitudes of iLVA (I) was fitted as a function of the potential (V) of the conditioning prepulse according to the Boltzmann equation (see text), which yielded the parameters Imax, Vh and Sl, where Imax is the amplitude of iLVA in the absence of inactivation, Vh is the conditioning potential at which the inactivation was half-maximal and Sl is the slope factor. The iLVA values measured and the fitted curves were normalized by division by the individual Imax. The right inset gives the obtained parameter values for the inactivation curves for the different solutions. This shows that lubeluzole reversibly reduces Imax, shifts Vh to a more negative potential and increases the slope factor of iLVA.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Voltage dependence of the block of iLVA by lubeluzole and its time course. A, The 10-sec conditioning prepulse was alternated every 30 sec between 110 and 85 mV, whereas the 155-msec test pulse eliciting iLVA was kept constant at 50 mV. The sweeps show iLVA after the conditioning potentials indicated in the control solution and after application of 1 µM lubeluzole. The numbers between parentheses refer to the time point at which the sweeps were recorded, as shown in B and C. B, Time course of the influence of lubeluzole on iLVA after the conditioning prepulses to 110 (VC 110) and 85 mV (VC 85). C, the curves for iLVA in the control solution were fitted exponentially and extrapolated, and the iLVA values for all time points were divided by the values on the fitted curve at the same time point, yielding the ratios rLVA at the two conditioning potentials. The effect of lubeluzole on iLVA (and rLVA) is larger after a conditioning prepulse to 85 mV than after a conditioning prepulse to 110 mV and is thus voltage dependent.

Frequency-dependent block of iLVA. The frequency dependence of iLVA block was tested by application of a pair of 155-msec test pulses to 50 mV every 30 sec with variation of the interval between the two test pulses. It took longer for iLVA to recover from inactivation in the presence of 1 µM lubeluzole (fig. 7A) than in the control solution. The effect was reversible on return to the control solution. These experiments show that the inhibition of iLVA by lubeluzole is frequency dependent and more pronounced at higher frequencies of stimulation.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Influence of lubeluzole on the recovery from inactivation of iLVA. Every 30 sec, a pair of 155-msec test pulses to 50 mV was given, and the interval between the two test pulses was varied. A, The interval between the two test pulses is represented on the horizontal axis. The ratio of the amplitude of iLVA at the second test pulse to the amplitude of iLVA at the first test pulse is represented on the vertical axis. The results are expressed as mean ± S.D.. The results were obtained in the control solution (n = 4), after a 5-min application of 1 µM lubeluzole (n = 4) and after a 5-min washout period in two of the four cells. Lubeluzole slowed down the recovery from inactivation of iLVA reversibly. B, Ratio of the amplitude of iLVA at the second test pulse to the amplitude at the first test pulse. Every 30 sec, a pair of 155-msec test pulses to 50 mV was given. The interval between the first and second test pulses was 320 msec except in the periods indicated by R and a horizontal bar, in which the recovery from inactivation was tested (see A). This illustrates the time course of the effect of lubeluzole and the (R)-enantiomer on the recovery from inactivation on application and washout of these compounds.

Voltage dependence of block of iHVA. Figure 8 shows the influence of 10 µM lubeluzole on the activation curve in a medium-sized DRG cell expressing iLVA (fig. 8A) and in a small DRG cell not expressing iLVA (fig. 8B). Similar results were obtained with the (R)-enantiomer (not shown). Lubeluzole and the (R)-enantiomer had a stronger inhibitory effect on iLVA than on iHVA. At the concentration of 10 µM, these compounds did not induce a shift in the iHVApeak along the voltage axis. However, they did inhibit iHVAend more strongly at 20 mV and more positive test potentials than at 40 mV. This is partly because the acceleration in the apparent inactivation by these compounds is more pronounced at 20 mV than at 40 mV. This is consistent with a larger open channel block by lubeluzole and the (R)-enantiomer when a larger fraction of the HVA Ca⁺⁺ channels are open. At 0 mV and higher test potentials, the compounds at a concentration of 10 µM blocked iHVAend almost completely.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Influence of 10 µM lubeluzole on the Ca++ channel activation curve. From a 100 mV HP, 155-msec test pulses were given ranging from 80 to +50 mV. The currents were plotted as a function of the test potential. Filled symbols: peak inward current. Empty symbols: current at the end of the 155-msec test pulse. A, Medium-sized DRG cell (40 µm) with a large iLVA, activated already from 70 mV. B, Small DRG cell (25 µm) not expressing iLVA.

60

100 mV

View larger version (11K): [in this window] [in a new window]   Fig. 9.   Voltage dependence of the block of iHVA by lubeluzole. The 50-msec test pulse to 20 mV was preceded by a 10-sec prepulse to 100 mV or 60 mV. For protocol, see text. The block of iHVApeak by lubeluzole was more pronounced when the 10-sec conditioning prepulse was 60 mV than when it was 100 mV. The tail currents on repolarization to the conditioning potential are truncated in the plots.

Frequency-dependent block of iHVA. Lubeluzole and the (R)-enantiomer accelerated the apparent inactivation of iHVA. To study how long this influences the HVA Ca⁺⁺ channels after the return to HP and whether it could induce a frequency dependence of the block of iHVA by these compounds, iHVA was elicited twice every 30 sec, and the interval (at 100 mV) between the first iHVA (iHVA1) and the second iHVA (iHVA2) was varied (fig. 10). The effect was studied in DRG cells not expressing iLVA (fig. 10, A-C) and in DRG cells expressing iLVA (fig. 10, D-F). For the cells expressing iLVA, the test pulses to 20 mV (to elicit iHVA) were preceded by 155-msec pulses to 50 mV to elicit and then inactivate iLVA. Figure 10, A and D, illustrates that lubeluzole blocked iHVApeak to a relatively greater extent after the shorter 100-msec interval at 100 mV (iHVA2) than after the longer 30-sec period at 100 mV (iHVA1). Similarly, iLVA elicited after the 100-msec interval (iLVA2) was blocked to a relatively greater extent by lubeluzole than was iLVA1 (fig. 10D). Similar effects were seen with the (R)-enantiomer. Figure 10, B, C, E and F, shows how long the effects of lubeluzole and the (R)-enantiomer on the apparent inactivation of iHVA lasted. They indicate that the inhibition of iHVA by lubeluzole and the (R)-enantiomer was frequency dependent and more pronounced at shorter intervals between the test pulses to 20 mV. This was tested with lubeluzole (n = 5) and with the (R)-enantiomer (n = 3). The effects were reversible on return to the control solution. Comparison of figure 10 with figure 7 shows that at the concentration of 1 µM, the two compounds had a more pronounced effect on the recovery from inactivation of iLVA than on the recovery from apparent inactivation of iHVA.

View larger version (26K): [in this window] [in a new window]   Fig. 10.   Frequency dependence of the block of iHVA by lubeluzole and the (R)-enantiomer. A-C, DRG cells not expressing iLVA. D, E, F: cells expressing iLVA. A, A 155-msec test pulse to 20 mV was given to elicit iHVA for the first time (iHVA1). After a 100-msec interval at 100 mV, a 20-msec test pulse to 20 mV was given to elicit iHVA again (iHVA2). The corresponding Ba++ currents are shown, in the control solution and after application of 1 µM lubeluzole and 3 µM lubeluzole. B, The recovery from apparent inactivation of iHVA was tested by variation of the interval (at 100 mV) between the test pulses to 20 mV. The horizontal axis shows the interval. The vertical axis shows the ratio of the peak amplitude of iHVA2 to that of iHVA1. This was tested in the control solution, after 5 min of 1 µM lubeluzole and after 5 min of 3 µM lubeluzole in the same cell. C, In another cell, an analogous result was obtained with the (R)-enantiomer. D, In cells expressing iLVA, iLVA was first elicited and inactivated by means of a 155-msec pulse to 50 mV before the test pulse to 20 mV eliciting iHVA. After a 100-msec interval at 100 mV, iLVA and iHVA were elicited again. E, Influence of 1 and 3 µM lubeluzole on the recovery from apparent inactivation of iHVA, tested by variation of the interval at 100 mV (horizontal axis) in the same cell. F, Idem for the (R)-enantiomer in another cell. These experiments show that the decrease in HVApeak induced by lubeluzole and the (R)-enantiomer is larger after a shorter interval and thus at a higher frequency of stimulation.

Block of iHVA in other cell types. Ca⁺⁺ channels of SCG cells are believed to be primarily of the N-type (Boland et al., 1994; Plummer et al., 1989). In rat SCG cells (HP 100 mV), 3 µM lubeluzole decreased iHVApeak and iHVAend by 42 ± 15% and 64 ± 13% (mean ± S.D., n = 9), respectively (fig. 11, A and B). The (R)-enantiomer blocked iHVApeak and iHVAend by 34 ± 15% and 53 ± 19%, respectively (n = 6). In this cell type, too, the time constant of decay of iHVA was smaller in the presence of 3 µM lubeluzole (77.2 ± 23.9 msec, mean ± S.D.) than with 3 µM of the (R)-enantiomer (111.2 ± 41.8 msec) in the same cells (n = 15). The change in time constant after the switch from 3 µM lubeluzole to the (R)-enantiomer was significantly different from the change in time constant after the solutions were given in reverse sequence (P < .001).

View larger version (19K): [in this window] [in a new window]   Fig. 11.   Influence of 3 µM lubeluzole and the (R)-enantiomer on Ca++ channel currents in a superior cervical ganglion cell (A and B) and cerebellar Purkinje neuron (C and D). A and C, Ba++ currents elicited by the pulse sequence shown. The numbers between parentheses refer to the time point at which the sweeps were recorded (see B and D). B and D, Time course of the drug effects on the Ba++ currents measured.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The results show that lubeluzole and the (R)-enantiomer block iLVA and iHVA in a concentration-, voltage- and frequency-dependent manner.

When lubeluzole or the (R)-enantiomer was applied, rHVA decreased in two phases: a first phase, which was nearly completed in 5 min, and thereafter a phase of slow progressive decrease. The latter slow phase of decline in Ca⁺⁺ channel currents might imply that the effect of prolonged application of lower concentrations of these compounds might be larger than the effect detected after 5 min. The small surface-to-volume ratio of the dorsal root ganglion cells prolongs the intracellular equilibration and thus also the influx into the cell and delays equilibration of lubeluzole or the (R)-enantiomer within and close to the cell membrane. This may be reflected in the time course of the block of Ca⁺⁺ channels by these compounds. The slow onset of block of iLVA and iHVA, elicited every 30 sec, seems not to be due to a use-dependency of the block because application of these compounds over 5 min in the absence of stimulation produced nearly the steady state block from the first stimulation after the 5-min rest period. Consequently, the block by lubeluzole or the (R)-enantiomer has a large tonic component.

The observation that lubeluzole and the (R)-enantiomer do not accelerate the inactivation of iLVA suggests that these compounds do not exert an open channel block on iLVA at the 50 mV test potential used.

These compounds not only decrease the amplitude of iLVA at its maximal availability (conditioning potential 120 mV) but also produce a shift of the inactivation curve in the negative direction. This negative shift causes an additional decrease in iLVA at less negative membrane potentials and means that the percentage decrease in iLVA caused by these compounds is larger when the cell is more depolarized. The influence on iLVA is thus voltage dependent. In addition, these compounds also clearly slow down the recovery from inactivation of iLVA. Consequently, they inhibit iLVA more potently as the frequency of stimulation increases. The negative shift of the inactivation curve of iLVA and the slowing down of the recovery from inactivation caused by lubeluzole and its (R)-enantiomer show that they make the inactivated state of the T channel more probable, possibly by binding preferably to the inactivated state (Sanguinetti and Kass, 1984).

There also was some voltage dependence of the effect of lubeluzole and the (R)-enantiomer on iHVA, in that both compounds inhibited iHVA more when the conditioning potential was 60 mV than when it was 100 mV. The voltage dependence was not tested at a more depolarized conditioning voltage because iHVA is elicited above 50 mV, which would result in a continuous influx of Ba⁺⁺ into the cell during the preceding conditioning pulse.

Lubeluzole and the (R)-enantiomer accelerated the apparent inactivation of iHVA (decay of iHVA during a test pulse), resulting in a stronger inhibition of iHVAend than of iHVApeak. This reduced the peak amplitude of iHVA elicited by a second test pulse when the interval between the two test pulses was short, in the presence of these compounds. After a longer interval, the peak iHVA with the second test pulse approached the peak iHVA obtained with the first test pulse. It took a few seconds (after repolarization to HP) for the amplitude to recover from the apparent inactivation of iHVA induced by these compounds. This behavior therefore induced some frequency dependence of the drug effect on iHVA, the drug effect having been greater at higher frequencies of stimulation. The acceleration of the apparent inactivation of iHVA by these compounds is probably due to an open channel block of HVA Ca⁺⁺ channels because this acceleration was more pronounced at test potentials at which iHVA was more activated.

At the highest concentrations of 3 and 10 µM, lubeluzole accelerated the decay of iHVA significantly more than the (R)-enantiomer. This was the only clear difference in effect observed with these compounds. We found no difference in effect on iHVA at the lower concentrations of 0.1 and 0.3 µM, probably because then the acceleration of iHVA by these compounds was small.

It has been reported that iHVA of small DRG cells consists of ~30% N-type Ca⁺⁺ channel current, ~53% L current and ~18% another type and that iHVA of medium-sized DRG-cells consists of ~36% N-current, ~7% L current and ~58% of another Ca⁺⁺ current (Scroggs and Fox, 1992), including the P current (Mintz et al., 1992; Diochot et al., 1995). In DRG cells of embryonic mice, iHVA is composed of L-, N-, P-, Q- and possibly also R-type Ca⁺⁺ channel currents (Diochot et al., 1995). The fact that 10 µM lubeluzole blocked iHVAend nearly completely at 20 mV and more positive test potentials, in small and medium-sized DRG cells, suggests that it blocks all or most high-voltage-activated Ca⁺⁺ channel current types contributing to iHVA in DRG cells. However, the fact that these compounds block iLVA more potently than the total iHVA shows that these compounds display some selectivity in their inhibition of Ca⁺⁺ channels.

That lubeluzole and the (R)-enantiomer also block N- and P-type HVA Ca⁺⁺ channels is supported by our observation that 3 µM of these compounds clearly blocked iHVA in rat superior cervical ganglion cells, in which mainly the N-channel contributes to iHVA (Boland et al., 1994; Plummer et al., 1989), and in rat cerebellar Purkinje cells, expressing mainly the P-channel (Mintz et al., 1992).

Lubeluzole and the (R)-enantiomer also inhibit the TTX-sensitive Na⁺ current in isolated hippocampal cells with IC₅₀ values of 3.1 and 1.6 µM, respectively (Osikowska-Evers et al., 1995) and protect against veratridine-induced Na⁺ overload and neurotoxicity in hippocampal slices (IC₅₀ = 0.54 and 0.69 µM, Ashton et al., 1997). Inhibition of Na⁺ currents can have neuroprotective effects (Taylor and Meldrum, 1995; Urenjak and Obrenovitch, 1996). However, the effects of lubeluzole on Na⁺ channels were not stereospecific, in contrast to the neuroprotective effect in the photochemical stroke model (Scheller et al., 1995; De Ryck et al., 1996; Buchkremer-Ratzmann and Witte, 1997).

Partial block of VSCC could also be neuroprotective after ischemia. It has been suggested that block of iLVA could reduce neuronal firing frequency (Akaike, 1991; Huguenard, 1996). Block of several types of HVA Ca⁺⁺ channels, such as the N-, P- and Q-types, can reduce neurotransmission (Gaur et al., 1994, Reuter, 1996) and glutamate release. This can lead to a reduced activation of NMDA, AMPA and kainate receptors and hereby diminish Ca⁺⁺ influx and intracellular Ca⁺⁺ overload, which is thought to play an important role in ischemic neuronal damage (Choi, 1995; Kristian and Siesjö, 1996). A temporary reduction in neurotransmission can also reduce the need for restorative ion pumping and thus be energy saving in a cerebral region at risk as a consequence of a stroke. Hence, the ability of lubeluzole to block VSCC might in principle contribute to its protective effect in stroke.

Although the concentrations at which lubeluzole blocked Ca⁺⁺ channels in these voltage-clamp experiments were rather high compared with the effective plasma concentrations in the rat stroke model (0.23 µM, of which 1% is unbound) (De Ryck et al., 1996), lubeluzole could have a more potent effect on VSCC than might be expected from the IC₅₀ values for inhibition of iLVA and iHVA. Indeed, its effect appeared not to be complete within 5 min of application. The IC₅₀ values were also determined at a 100 mV HP and at a very low frequency of 0.033 Hz (to slow down the run-down of the Ca⁺⁺ channel currents). The compounds inhibited iLVA and iHVA more potently at a more depolarized conditioning potential, which may be relevant for pathological situations in which the cells are more depolarized. The observed frequency dependence of the effect of lubeluzole and the (R)-enantiomer on iLVA and iHVA indicates that the inhibition is also more potent at higher frequencies of stimulation. The fact that lubeluzole accelerated the apparent inactivation of iHVA is also interesting because it means that lubeluzole blocks iHVA especially during long depolarizations, which may be the situation in cells at risk after ischemia.

In contrast to the photochemical stroke model (Scheller et al., 1995; De Ryck et al., 1996; Buchkremer-Ratzmann and Witte, 1997), in which lubeluzole is protective and the (R)-enantiomer is virtually inactive, we found a much smaller difference in effect on VSCC between the two enantiomers, mainly in the acceleration of the decay of iHVA, and only at the highest concentrations of 3 and 10 µM. On the other hand, stereospecificity has been found in some in vitro models of neuroprotection (Benjamins et al., 1996; Lesage et al., 1996; Maiese et al., 1997).

In conclusion, inhibition by lubeluzole of VSCC may, in principle, contribute to the neuroprotective effect via a reduction in intracellular Ca⁺⁺ overload. However, the small difference in effect on VSCC between the two enantiomers in DRG cells does not explain the clear stereospecificity of the neuroprotection in the photochemical stroke model.