391
Journal of Physiology (1992), 447, pp. 391-407 With 6 figures Printed in Great Britain
KINETIC ANALYSIS OF THE GABAB-MEDIATED INHIBITION OF THE HIGH-THRESHOLD Ca2l CURRENT IN CULTURED RAT SENSORY NEURONES BY HIDEHARU TATEBAYASHI AND NOBUKUNI OGATA* From the Department of Pharmacology, Faculty of Medicine, Kyushu University, Fukuoka 812, Japan
(Received 9 May 1991) SUMMARY
1. The action of baclofen on the voltage-gated Ca2+ current (ICa) was studied, using cultured neurones of the newborn rat dorsal root ganglia (DRG). Two major categories of ICa were identified: a small transient current activated positive to -60 mV (low voltage-activated ICa) and a large and slowly inactivating current activated positive to -30 mV (high voltage-activated Ica). 2. Baclofen reversibly blocked the high voltage-activated ICa and slowed the activation phase of the current in a concentration-dependent manner (0 5-50 /IM). The half-maximal effective concentration was about 1f5/kM as measured by a peak of ICa. On the contrary, a high concentration of baclofen (100 /tM) had no detectable effect on the low voltage-activated ICa3. The baclofen-sensitive component of the high voltage-activated ICa was largely inactivated by a depolarized holding potential (Vh) of -40 mV, whereas the baclofenresistant component was not affected by a change in Vh ranging from -110 to -30 mV. 4. The high voltage-activated Ica had two components of current decay: an inactivating component and a quasi-sustained component, with time constants about 420 and 1220 ms, respectively. The time constant of decay for the inactivating component was not affected by replacement of external Ca2+ with Ba2 , whereas that for the quasi-sustained component was markedly prolonged, suggesting that the decay of this component may be due to Ca2+-induced block rather than voltagedependent inactivation. A high concentration of baclofen (50 /gM) selectively blocked the inactivating component. 5. The decay phase of the baclofen-sensitive component of the high voltageactivated ICa was best fitted by a sum of two exponentials, with 29-2 and 481 ms for the fast and slow components, respectively. The time constants of the two components were not affected by an increase in the concentration of baclofen, whereas the amplitudes changed concentration-dependently. 6. The slowed activation of the high voltage-activated Ica by baclofen was partially reversed by a large depolarizing pre-pulse. However, such an acceleration of the current was similarly observed in the control solution. Furthermore, the actual * To whom correspondence should be addressed. MS 9379
392
H. TATEBAYASHI AND N. OGATA
current size increased by the pre-pulse was similar in both the control and baclofencontaining solutions. 7. These results suggest that baclofen selectively blocks the inactivating component of the high voltage-activated ICa which forms a rapid rising phase of this current, thus slowing the activation phase of the total high voltage-activated ICa. INTRODUCTION
Influx of calcium ions through voltage-gated Ca2+ channels is an important signal for a variety of cellular functions, including transmitter release. Recent studies have shown that most excitable membranes contain multiple types of voltage-gated Ca21 channels (Tsien, Lipscombe, Madison, Bley & Fox, 1988; Bean, 1989). For example, three types of Ca2+ channels in chick sensory neurones have been reported (Nowycky, Fox & Tsien, 1985; Fox, Nowycky & Tsien, 1987): a low voltage-activated (LVA) Ttype channel: a high voltage-activated (HVA) inactivating N-type channel, and a HVA sustained L-type channel which is sensitive to dihydropyridines. The N-type Ca21 channel has been implicated in the control of neurotransmitter release (Hirning, Fox, McCleskey, Olivera, Thayer, Miller & Tsien, 1988). However, the property of the N-type Ca2+ channel in cells other than the chick dorsal root ganglion (DRG) has not been fully characterized. The kinetics of the N-type Ca2+ current (ICa) show a considerable variability in different neurones (Fox et al. 1987; Docherty, 1988; Hirning et al. 1988; Gross & MacDonald, 1989). In addition, two components of the HVA N-type ICa have been reported in PC12 cells or SH-SY5Y cells, suggesting an involvement of more than one class of Ca2+ channel (Plummer, Logothetis & Hess, 1989; Seward & Henderson, 1990). The voltage-gated ICa is modulated by a variety of neurotransmitters (Tsien et al 1988; Bean 1989). It is important to clarify which channel type is modulated by various neurotransmitters, because different types of Ca2+ channels may play different physiological roles. GABA (y-aminobutyric acid) is one of the neurotransmitters which block the neuronal voltage-gated Ca2+ channels. This action of GABA is mediated by GABAB receptors (Dunlap, 1981; Deisz & Lux, 1985; Dolphin & Scott, 1986), and involves the pertussis toxin-sensitive G proteins (Holz, Rane & Dunlap, 1986; Scott & Dolphin, 1986; Dolphin & Scott, 1987). The GABAB-induced inhibition of Ca2+ channels appears to be important in the presynaptic control of the transmitter release (Dunlap & Fischbach, 1981). The GABAB receptor-mediated modulation of ICa exhibits remarkable features: the inhibition by a GABAB agonist is more potent in the initial portion of ICa than in the later portion of it, and the activation phase of the current is markedly slowed (Deisz & Lux, 1985; Dolphin & Scott, 1986). The slowing of the activation phase has been explained on the basis of a voltage-dependent recovery from the GABABinduced block (Grassi & Lux, 1989). In this study, we investigated the kinetic effects of baclofen, a specific agonist for GABAB receptors (Bowery, Hill & Hudson, 1980), on different types of ICa in rat DRG neurones. Our results show that baclofen selectively blocks the inactivating component of the HVA Ic. which comprises two distinct components with different sensitivities to baclofen.
INHIBITION OF Ca2+ CURRENTS BY BACLOFEN
393
METHODS
Culture procedures Dissociated cultures of rat DRG neurones were prepared as described by Forda & Kelly (1985). Rats were killed by decapitation under ether anaesthesia. Dorsal root ganglia were dissected from 1- to 3-day-old rats and incubated at 36 °C for 30-40 min in Ca2+- and Mg2+-free saline, containing 0-25% trypsin (Type XI, Sigma, St Louis, MO, USA). The ganglia were then mechanically dissociated with a fire-polished Pasteur pipette. The cells were plated on glass cover-slips coated with poly-L-lysine (Sigma) and maintained in a humidified incubator containing 5 % CO2 in air at 35 °C in Dulbecco's modified Eagle's medium supplemented with 10 % fetal bovine serum (GIBCO), penicillin (40 i.u./ml), and streptomycin (40 ng/ml). After 1-2 days in culture, cytosine fl-D-arabinofuranoside (Sigma) was added to cultures to suppress the growth of non-neuronal cells. Subsequent medium changes were done at 3-4 day intervals. Cells were used after 2-10 days in culture. Electrical Recording The methods for electrical recording used in the present study were similar to those previously described (Ogata, Yoshii & Narahashi, 1990). Membrane currents were recorded with the whole-cell patch-clamp technique (Hamill, Marty, Neher & Sakmann, 1981). The DC resistance of suction electrodes was 15-3 MQ. The pipette solution contained (mM): 120 caesium glutamate, 10 NaCl, 3 Mg-ATP, 2-5 MgCl2, 5 glucose, 5 HEPES, and 5 EGTA. The pH of the pipette solution was adjusted to 7 0 with CsOH. The external solution contained (mM): 120 NaCl, 5 CsCl, 1-8 CaCl2, 1 MgCl2, 5 HEPES, 25 glucose. The pH of the external solution was adjusted to 7-4 with NaOH. Tetrodotoxin (TTX) at a concentration of 0-2 ,UM was added to block TTX-sensitive Na+ current. Membrane currents passing through the pipette were recorded by a current-to-voltage converter designed by M. Yoshii (Ogata et al. 1990). Compensation for the series resistance was performed by adding a part of the output voltage of the current recording to the command pulse. All the experiments were performed with an on-line system which has been developed by M. Yoshii and N. Ogata, using a personal computer (PC-286V, EPSON, Tokyo, Japan). Capacitative and leakage currents were subtracted digitally by the P-P/4 procedure (Ogata et al. 1990). In addition, ICa was subtracted with the current remaining after application of 50 ,tM-Cd2+ unless otherwise specified. Exponential fits were determined by computer using a non-linear sum of the least-squares fitting routine. The liquid junction potential between internal and external solutions was about 11 mV. The data shown here were compensated for this effect by adjusting the zerocurrent potential to the liquid junction potential. Since application of the drug and its wash-out were attained within several seconds, there was usually no detectable 'run-down' of the current. However, to further minimize a possible 'run-down', the 'control' response was recorded after wash-out of the drug solution in all the traces shown in this report. Only cells showing an adequate voltage and space clamp (Ogata & Tatebayashi, 1990) were used. All experiments were done at room temperature (20-25 °C). Results are expressed as means + S.E.M., and n represents the number of cells.
Drugs Drugs were applied through a rapid microsuperfusion system (Ogata & Tatebayashi, 1991). This system uses a fine polyethylene needle (20 ,4um tip internal diameter) placed near (about 50 ,um) the cell. Electromagnetic valves are used to apply test solutions. This microsuperfusion system enabled a rapid (latency, less than 0 5 s) and localized application of test solutions and their rapid washout. (-)-Baclofen was a generous gift from Ciba-Geigy (Japan). Baclofen-containing solutions were prepared immediately before the experiment by diluting an appropriate amount of the stock solutions (10 mM) with external solution. RESULTS
The ICa of the rat DRG neurones were classified into two major categories, i.e. a LVA ICa and a HVA ICa. The LVA ICa was elicited by the step depolarization to -60 mV from a holding potential (Vh) of -100 mV and was inactivated completely
394
H TATEBAYASHI AND N. OGATA
within 50-100 ms. The LVA ICa was inactivated when Vh was more positive than -60 mV or blocked in the presence of 50 jtM-Cd2". This type of 'Ca was observed in about 30 % of neurones examined. This was consistent with the findings by other investigators (Fedulova, Kostyuk & Veselovsky, 1985; Regan, Sah & Bean, 1991). These properties are comparable to those of T-type 'Ca in chick DRG neurones (Nowycky et al. 1985; Fox et al. 1987). The HVA ICa was induced by a step depolarization to potentials more positive than -30 mV from Vh of -80 mV. The HVA 'Ca showed a relatively slow inactivation during a 300 ms step depolarization. When evoked from Vh of -40 mV, the HVA ICa showed no or an extremely slow inactivation during the 300 ms step depolarization. In the presence of 50 jtM-Cd2`, only a small outward current was observed. Effects of baclofen on ICa Baclofen in concentration of 0'5-50 /tM markedly reduced the HVA ICa, whereas much higher concentrations of (-)-baclofen (100 /tM) had no effect on the LVA ICa in any of the six cells examined. In subsequent experiments, we selected the cells lacking the LVA ICa to examine the effect of baclofen on the HVA ICa. The initial portion of the HVA ICa evoked from Vh of -80 mV was reduced to a much larger extent than the later portion of the current, when the current was evoked by the 300 ms step depolarization. In addition, in most of the fifty cells examined, baclofen slowed the activation phase of the HVA ICa. The current fully recovered after several seconds of washing. When Vh was depolarized to -40 mV, only the initial portion of the HVA ICa was inhibited by baclofen, and the later portion remained unaffected. Baclofen (50 ItM) had no effect on the outward current evoked in the presence of 50 /tM-Cd2+. The inhibitory effect of baclofen on the HVA ICa was concentration dependent. The concentration-response curves were measured at the peak or at the end of the 300 ms step depolarization. The half-maximal concentrations for the inhibition of the HVA ICa were l1 7 + 0'13 (n = 3) and 2 73 + 0-05 (n = 3) /kM for the peak current and the current persistent during the 300 ms step depolarization, respectively. In the presence of 50 /tM-baclofen, the amplitude of the HVA ICa was reduced to 42'3 + 4*3 % (n = 16) or 68'3 + 4'2 % (n = 16) of the control when measured at the peak or at the end of the step depolarization, respectively. Thus, the initial portion of the HVA Ica was more sensitive to the inhibitory action of baclofen than the later portion.
Potential dependence of the HVA ICa Figure 1 shows the effect of pre-pulses on the HVA ICa during and after exposure to 50 /tM-baclofen. Measurements were made using a standard double-pulse protocol. A 5 s conditioning pulse to various potential levels was followed by a 100 ms test pulse to -5 mV. The amplitude of the peak current was plotted against the conditioning potential. Under control conditions, the peak amplitude gradually decreased in the potential range -70 to -30 mV. In the presence of 50 ,um-baclofen, the peak ICawas not dependent on conditioning voltage in the entire range examined. Similar voltage dependencies were observed in all of the four additional measurements. Thus, it is suggested that baclofen preferentially blocks a component
INHIBITION OF Ca2+ CURRENTS BY BACLOFEN
395
which readily inactivates with membrane depolarization. Contrary, a baclofeninsensitive component appears to be relatively resistant to the voltage-dependent inactivation. Repetitive activation of the HVA ICa Figure 2A shows the time course of change in the amplitude of the HVA ICa caused by a train of depolarizing pulses. The peak current amplitude for each pulse in the
A
5s
100Oms
control
lOO1ms L
/test
-80 mV
b
-40 mv -50
Control
-70
7-7 I1 nA Baclofen
(50WpM)
B 01.2 1.0 I
°0-81 0-6 -1 10-100-90 -80 -70 -60 -50 -40 -30
Voltage (mV)
Fig. 1. Effects of baclofen on the voltage dependence of the HVA ICa in cultured neurones of the newborn rat dorsal root ganglia (DRG). A, the control currents (Icontro,) were evoked by a 100 ms step depolarization to -5 mV from a holding potential (Vh) of -80 mV, and then conditioning pulses to various membrane potentials from -110 to -30 mV were delivered for 5 s. Immediately following each conditioning pulse, the test current (Itest) was evoked by the step depolarization identical to that used for Icontroli A family of current traces were recorded in the control solution (A a) or in the presence of 50 /LM-baclofen (A b). B, Itest/Icontrol measured at 7 ms after the onset of the test pulse was plotted against the pre-pulse voltage. 0, control; *, during application of baclofen. In this and subsequent figures: the Ca2+ currents were subtracted with the current remaining after application of 50 ,uM-Cd2+; the 'control' response was recorded after wash-out of the drug solution (see Methods); and downward deflections represent an inward current.
train (In) was normalized to the one for the first pulse (I,), and was plotted against the pulse number (Fig. 2B). Before exposure to baclofen (0), the amplitude of I2 was markedly reduced, whereas the subsequent In with the interpulse interval (AT) ranging from 30 to 480 ms decreased to a smaller extent. In the presence of submaximal concentration of baclofen (1 ,UiM, *), where a considerable amount of the
H. TATEBAYASHI AND N OGATA baclofen-sensitive component remains unaffected, the time course of inhibition was similar to that observed in the control. When a high concentration of baclofen (50 /tM) was introduced (A), the successive reduction of the current amplitude became much smaller throughout the repetitive activation. This attenuation of the 396
A
B 1 *0i
-5 mV -80 mV
100 ms
0.8
WL,AJ,,SJj
In/Il
AT
0.5 0.0 1.0 0.8
AT 30 ms
Baclofen (50 MM) Control
I
In/1l 0.5 0.0 A 1 .0
AT 60 ms
tLtUUW.
AT 60 ms
0.8
In/WA AT 120
0.5
ms
i: AT 120 ms
1-04 0.8 AT 240
In//1
ms
0.5 I
r
AT 240 ms I
I
I
1.0 AT 480
In/ll
ms
I|1
nA
0-80-5
Journal of Physiology (1992), 447, pp. 391-407 With 6 figures Printed in Great Britain
KINETIC ANALYSIS OF THE GABAB-MEDIATED INHIBITION OF THE HIGH-THRESHOLD Ca2l CURRENT IN CULTURED RAT SENSORY NEURONES BY HIDEHARU TATEBAYASHI AND NOBUKUNI OGATA* From the Department of Pharmacology, Faculty of Medicine, Kyushu University, Fukuoka 812, Japan
(Received 9 May 1991) SUMMARY
1. The action of baclofen on the voltage-gated Ca2+ current (ICa) was studied, using cultured neurones of the newborn rat dorsal root ganglia (DRG). Two major categories of ICa were identified: a small transient current activated positive to -60 mV (low voltage-activated ICa) and a large and slowly inactivating current activated positive to -30 mV (high voltage-activated Ica). 2. Baclofen reversibly blocked the high voltage-activated ICa and slowed the activation phase of the current in a concentration-dependent manner (0 5-50 /IM). The half-maximal effective concentration was about 1f5/kM as measured by a peak of ICa. On the contrary, a high concentration of baclofen (100 /tM) had no detectable effect on the low voltage-activated ICa3. The baclofen-sensitive component of the high voltage-activated ICa was largely inactivated by a depolarized holding potential (Vh) of -40 mV, whereas the baclofenresistant component was not affected by a change in Vh ranging from -110 to -30 mV. 4. The high voltage-activated Ica had two components of current decay: an inactivating component and a quasi-sustained component, with time constants about 420 and 1220 ms, respectively. The time constant of decay for the inactivating component was not affected by replacement of external Ca2+ with Ba2 , whereas that for the quasi-sustained component was markedly prolonged, suggesting that the decay of this component may be due to Ca2+-induced block rather than voltagedependent inactivation. A high concentration of baclofen (50 /gM) selectively blocked the inactivating component. 5. The decay phase of the baclofen-sensitive component of the high voltageactivated ICa was best fitted by a sum of two exponentials, with 29-2 and 481 ms for the fast and slow components, respectively. The time constants of the two components were not affected by an increase in the concentration of baclofen, whereas the amplitudes changed concentration-dependently. 6. The slowed activation of the high voltage-activated Ica by baclofen was partially reversed by a large depolarizing pre-pulse. However, such an acceleration of the current was similarly observed in the control solution. Furthermore, the actual * To whom correspondence should be addressed. MS 9379
392
H. TATEBAYASHI AND N. OGATA
current size increased by the pre-pulse was similar in both the control and baclofencontaining solutions. 7. These results suggest that baclofen selectively blocks the inactivating component of the high voltage-activated ICa which forms a rapid rising phase of this current, thus slowing the activation phase of the total high voltage-activated ICa. INTRODUCTION
Influx of calcium ions through voltage-gated Ca2+ channels is an important signal for a variety of cellular functions, including transmitter release. Recent studies have shown that most excitable membranes contain multiple types of voltage-gated Ca21 channels (Tsien, Lipscombe, Madison, Bley & Fox, 1988; Bean, 1989). For example, three types of Ca2+ channels in chick sensory neurones have been reported (Nowycky, Fox & Tsien, 1985; Fox, Nowycky & Tsien, 1987): a low voltage-activated (LVA) Ttype channel: a high voltage-activated (HVA) inactivating N-type channel, and a HVA sustained L-type channel which is sensitive to dihydropyridines. The N-type Ca21 channel has been implicated in the control of neurotransmitter release (Hirning, Fox, McCleskey, Olivera, Thayer, Miller & Tsien, 1988). However, the property of the N-type Ca2+ channel in cells other than the chick dorsal root ganglion (DRG) has not been fully characterized. The kinetics of the N-type Ca2+ current (ICa) show a considerable variability in different neurones (Fox et al. 1987; Docherty, 1988; Hirning et al. 1988; Gross & MacDonald, 1989). In addition, two components of the HVA N-type ICa have been reported in PC12 cells or SH-SY5Y cells, suggesting an involvement of more than one class of Ca2+ channel (Plummer, Logothetis & Hess, 1989; Seward & Henderson, 1990). The voltage-gated ICa is modulated by a variety of neurotransmitters (Tsien et al 1988; Bean 1989). It is important to clarify which channel type is modulated by various neurotransmitters, because different types of Ca2+ channels may play different physiological roles. GABA (y-aminobutyric acid) is one of the neurotransmitters which block the neuronal voltage-gated Ca2+ channels. This action of GABA is mediated by GABAB receptors (Dunlap, 1981; Deisz & Lux, 1985; Dolphin & Scott, 1986), and involves the pertussis toxin-sensitive G proteins (Holz, Rane & Dunlap, 1986; Scott & Dolphin, 1986; Dolphin & Scott, 1987). The GABAB-induced inhibition of Ca2+ channels appears to be important in the presynaptic control of the transmitter release (Dunlap & Fischbach, 1981). The GABAB receptor-mediated modulation of ICa exhibits remarkable features: the inhibition by a GABAB agonist is more potent in the initial portion of ICa than in the later portion of it, and the activation phase of the current is markedly slowed (Deisz & Lux, 1985; Dolphin & Scott, 1986). The slowing of the activation phase has been explained on the basis of a voltage-dependent recovery from the GABABinduced block (Grassi & Lux, 1989). In this study, we investigated the kinetic effects of baclofen, a specific agonist for GABAB receptors (Bowery, Hill & Hudson, 1980), on different types of ICa in rat DRG neurones. Our results show that baclofen selectively blocks the inactivating component of the HVA Ic. which comprises two distinct components with different sensitivities to baclofen.
INHIBITION OF Ca2+ CURRENTS BY BACLOFEN
393
METHODS
Culture procedures Dissociated cultures of rat DRG neurones were prepared as described by Forda & Kelly (1985). Rats were killed by decapitation under ether anaesthesia. Dorsal root ganglia were dissected from 1- to 3-day-old rats and incubated at 36 °C for 30-40 min in Ca2+- and Mg2+-free saline, containing 0-25% trypsin (Type XI, Sigma, St Louis, MO, USA). The ganglia were then mechanically dissociated with a fire-polished Pasteur pipette. The cells were plated on glass cover-slips coated with poly-L-lysine (Sigma) and maintained in a humidified incubator containing 5 % CO2 in air at 35 °C in Dulbecco's modified Eagle's medium supplemented with 10 % fetal bovine serum (GIBCO), penicillin (40 i.u./ml), and streptomycin (40 ng/ml). After 1-2 days in culture, cytosine fl-D-arabinofuranoside (Sigma) was added to cultures to suppress the growth of non-neuronal cells. Subsequent medium changes were done at 3-4 day intervals. Cells were used after 2-10 days in culture. Electrical Recording The methods for electrical recording used in the present study were similar to those previously described (Ogata, Yoshii & Narahashi, 1990). Membrane currents were recorded with the whole-cell patch-clamp technique (Hamill, Marty, Neher & Sakmann, 1981). The DC resistance of suction electrodes was 15-3 MQ. The pipette solution contained (mM): 120 caesium glutamate, 10 NaCl, 3 Mg-ATP, 2-5 MgCl2, 5 glucose, 5 HEPES, and 5 EGTA. The pH of the pipette solution was adjusted to 7 0 with CsOH. The external solution contained (mM): 120 NaCl, 5 CsCl, 1-8 CaCl2, 1 MgCl2, 5 HEPES, 25 glucose. The pH of the external solution was adjusted to 7-4 with NaOH. Tetrodotoxin (TTX) at a concentration of 0-2 ,UM was added to block TTX-sensitive Na+ current. Membrane currents passing through the pipette were recorded by a current-to-voltage converter designed by M. Yoshii (Ogata et al. 1990). Compensation for the series resistance was performed by adding a part of the output voltage of the current recording to the command pulse. All the experiments were performed with an on-line system which has been developed by M. Yoshii and N. Ogata, using a personal computer (PC-286V, EPSON, Tokyo, Japan). Capacitative and leakage currents were subtracted digitally by the P-P/4 procedure (Ogata et al. 1990). In addition, ICa was subtracted with the current remaining after application of 50 ,tM-Cd2+ unless otherwise specified. Exponential fits were determined by computer using a non-linear sum of the least-squares fitting routine. The liquid junction potential between internal and external solutions was about 11 mV. The data shown here were compensated for this effect by adjusting the zerocurrent potential to the liquid junction potential. Since application of the drug and its wash-out were attained within several seconds, there was usually no detectable 'run-down' of the current. However, to further minimize a possible 'run-down', the 'control' response was recorded after wash-out of the drug solution in all the traces shown in this report. Only cells showing an adequate voltage and space clamp (Ogata & Tatebayashi, 1990) were used. All experiments were done at room temperature (20-25 °C). Results are expressed as means + S.E.M., and n represents the number of cells.
Drugs Drugs were applied through a rapid microsuperfusion system (Ogata & Tatebayashi, 1991). This system uses a fine polyethylene needle (20 ,4um tip internal diameter) placed near (about 50 ,um) the cell. Electromagnetic valves are used to apply test solutions. This microsuperfusion system enabled a rapid (latency, less than 0 5 s) and localized application of test solutions and their rapid washout. (-)-Baclofen was a generous gift from Ciba-Geigy (Japan). Baclofen-containing solutions were prepared immediately before the experiment by diluting an appropriate amount of the stock solutions (10 mM) with external solution. RESULTS
The ICa of the rat DRG neurones were classified into two major categories, i.e. a LVA ICa and a HVA ICa. The LVA ICa was elicited by the step depolarization to -60 mV from a holding potential (Vh) of -100 mV and was inactivated completely
394
H TATEBAYASHI AND N. OGATA
within 50-100 ms. The LVA ICa was inactivated when Vh was more positive than -60 mV or blocked in the presence of 50 jtM-Cd2". This type of 'Ca was observed in about 30 % of neurones examined. This was consistent with the findings by other investigators (Fedulova, Kostyuk & Veselovsky, 1985; Regan, Sah & Bean, 1991). These properties are comparable to those of T-type 'Ca in chick DRG neurones (Nowycky et al. 1985; Fox et al. 1987). The HVA ICa was induced by a step depolarization to potentials more positive than -30 mV from Vh of -80 mV. The HVA 'Ca showed a relatively slow inactivation during a 300 ms step depolarization. When evoked from Vh of -40 mV, the HVA ICa showed no or an extremely slow inactivation during the 300 ms step depolarization. In the presence of 50 jtM-Cd2`, only a small outward current was observed. Effects of baclofen on ICa Baclofen in concentration of 0'5-50 /tM markedly reduced the HVA ICa, whereas much higher concentrations of (-)-baclofen (100 /tM) had no effect on the LVA ICa in any of the six cells examined. In subsequent experiments, we selected the cells lacking the LVA ICa to examine the effect of baclofen on the HVA ICa. The initial portion of the HVA ICa evoked from Vh of -80 mV was reduced to a much larger extent than the later portion of the current, when the current was evoked by the 300 ms step depolarization. In addition, in most of the fifty cells examined, baclofen slowed the activation phase of the HVA ICa. The current fully recovered after several seconds of washing. When Vh was depolarized to -40 mV, only the initial portion of the HVA ICa was inhibited by baclofen, and the later portion remained unaffected. Baclofen (50 ItM) had no effect on the outward current evoked in the presence of 50 /tM-Cd2+. The inhibitory effect of baclofen on the HVA ICa was concentration dependent. The concentration-response curves were measured at the peak or at the end of the 300 ms step depolarization. The half-maximal concentrations for the inhibition of the HVA ICa were l1 7 + 0'13 (n = 3) and 2 73 + 0-05 (n = 3) /kM for the peak current and the current persistent during the 300 ms step depolarization, respectively. In the presence of 50 /tM-baclofen, the amplitude of the HVA ICa was reduced to 42'3 + 4*3 % (n = 16) or 68'3 + 4'2 % (n = 16) of the control when measured at the peak or at the end of the step depolarization, respectively. Thus, the initial portion of the HVA Ica was more sensitive to the inhibitory action of baclofen than the later portion.
Potential dependence of the HVA ICa Figure 1 shows the effect of pre-pulses on the HVA ICa during and after exposure to 50 /tM-baclofen. Measurements were made using a standard double-pulse protocol. A 5 s conditioning pulse to various potential levels was followed by a 100 ms test pulse to -5 mV. The amplitude of the peak current was plotted against the conditioning potential. Under control conditions, the peak amplitude gradually decreased in the potential range -70 to -30 mV. In the presence of 50 ,um-baclofen, the peak ICawas not dependent on conditioning voltage in the entire range examined. Similar voltage dependencies were observed in all of the four additional measurements. Thus, it is suggested that baclofen preferentially blocks a component
INHIBITION OF Ca2+ CURRENTS BY BACLOFEN
395
which readily inactivates with membrane depolarization. Contrary, a baclofeninsensitive component appears to be relatively resistant to the voltage-dependent inactivation. Repetitive activation of the HVA ICa Figure 2A shows the time course of change in the amplitude of the HVA ICa caused by a train of depolarizing pulses. The peak current amplitude for each pulse in the
A
5s
100Oms
control
lOO1ms L
/test
-80 mV
b
-40 mv -50
Control
-70
7-7 I1 nA Baclofen
(50WpM)
B 01.2 1.0 I
°0-81 0-6 -1 10-100-90 -80 -70 -60 -50 -40 -30
Voltage (mV)
Fig. 1. Effects of baclofen on the voltage dependence of the HVA ICa in cultured neurones of the newborn rat dorsal root ganglia (DRG). A, the control currents (Icontro,) were evoked by a 100 ms step depolarization to -5 mV from a holding potential (Vh) of -80 mV, and then conditioning pulses to various membrane potentials from -110 to -30 mV were delivered for 5 s. Immediately following each conditioning pulse, the test current (Itest) was evoked by the step depolarization identical to that used for Icontroli A family of current traces were recorded in the control solution (A a) or in the presence of 50 /LM-baclofen (A b). B, Itest/Icontrol measured at 7 ms after the onset of the test pulse was plotted against the pre-pulse voltage. 0, control; *, during application of baclofen. In this and subsequent figures: the Ca2+ currents were subtracted with the current remaining after application of 50 ,uM-Cd2+; the 'control' response was recorded after wash-out of the drug solution (see Methods); and downward deflections represent an inward current.
train (In) was normalized to the one for the first pulse (I,), and was plotted against the pulse number (Fig. 2B). Before exposure to baclofen (0), the amplitude of I2 was markedly reduced, whereas the subsequent In with the interpulse interval (AT) ranging from 30 to 480 ms decreased to a smaller extent. In the presence of submaximal concentration of baclofen (1 ,UiM, *), where a considerable amount of the
H. TATEBAYASHI AND N OGATA baclofen-sensitive component remains unaffected, the time course of inhibition was similar to that observed in the control. When a high concentration of baclofen (50 /tM) was introduced (A), the successive reduction of the current amplitude became much smaller throughout the repetitive activation. This attenuation of the 396
A
B 1 *0i
-5 mV -80 mV
100 ms
0.8
WL,AJ,,SJj
In/Il
AT
0.5 0.0 1.0 0.8
AT 30 ms
Baclofen (50 MM) Control
I
In/1l 0.5 0.0 A 1 .0
AT 60 ms
tLtUUW.
AT 60 ms
0.8
In/WA AT 120
0.5
ms
i: AT 120 ms
1-04 0.8 AT 240
In//1
ms
0.5 I
r
AT 240 ms I
I
I
1.0 AT 480
In/ll
ms
I|1
nA
0-80-5
