Journal of Physiology (1992), 457, pp. 559-574 With 9 figures Printed in Great Britain
559
GUANOSINE 3',5'-CYCLIC MONOPHOSPHATE REGULATES CALCIUM CHANNELS IN NEURONES OF RABBIT VESICAL PELVIC GANGLIA
BY T. NISHIMURA, T. AKASU AND J. KRIER* From the Department of Physiology, Kurume University School of Medicine, 67 Asahi-machi, Kurume 830, Japan and the *Department of Physiology, Michigan State University, East Lansing, MI 48824, USA
(Received 27 December 1991) SUMMARY
1. The effects of dibutyryl guanosine 3',5'-cyclic monophosphate (db-cyclic GMP) were studied in vitro on calcium channels of neurones in rabbit vesical parasympathetic ganglia, using intracellular and single-electrode voltage-clamp
recordings. 2. Db-cyclic GMP (100 iM) caused membrane depolarization associated with a decrease in membrane input resistance and an after-hyperpolarization associated with an increase in membrane input resistance. 3. Db-cyclic GMP (0-01-1 mM) caused a concentration-dependent, transient inward current followed by a long-lasting outward current. Membrane conductance was increased and decreased during the inward and outward currents, respectively. 4. The db-cyclic GMP-induced inward current was depressed in nominally calcium-free solutions, by cobalt (1 mM) and nicardipine (10 EM). The mean reversal potentials of the inward current were + 42 and -20 mV in the presence and absence of calcium in the external solution, respectively. 5. The db-cyclic GMP-induced inward current was not altered by lowering the external sodium concentration, raising external potassium concentration or by intracellular injection of caesium. 6. A calcium-insensitive component of the db-cyclic GMP-induced current was increased by lowering the external chloride concentration and blocked by 4acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid, a chloride channel blocker. 7. Voltage-dependent, high-threshold calcium currents were depressed during the db-cyclic GMP-induced inward current and facilitated during the outward current. 8. Cyclic GMP was less potent than db-cyclic GMP in causing both inward and outward currents or modulation of calcium currents. GTP, GDP, GMP, guanosine, 8-bromoadenosine 3',5'-cyclic monophosphate and forskolin did not alter the holding current or voltage-dependent calcium currents. 9. It is concluded that intracellular cyclic GMP causes not only activation of resting calcium and chloride channels but also a transient depression followed by long-lasting facilitation of voltage-dependent calcium currents in neurones of vesical parasympathetic ganglia. MS 8948
560
560T. NISIHIMTRA, T. AKASUr AND J. KRIER INTRODUCTION
Guanine nucleotides are considered second messengers for intracellular signal transduction in various tissues (Greengard, 1976). Guanosine 3',5'-cyclic monophosphate (cyclic GMP) activates light-sensitive conductance in frog and toad retinal rod outer segments (Fesenko, Kolesnikov & Lyubarsky, 1985; Yau & Nakatani, 1985), potassium channels in Xenopus oocytes (Dascal, Lotan & Lass, 1987) and calcium channels in snail neurones (Paupardin-Tritsch, Hammond, Gerschenfeld, Nairn & Greengard, 1986). Cyclic GMP decreases voltage-dependent calcium currents of frog cardiac muscles (Hartzell & Fischmeister, 1986) and pyramidal neurones in guinea-pig hippocampus (Doerner & Alger, 1988). The role of cyclic GMP as an intracellular messenger for neurones in mammalian autonomic ganglia is not precisely understood. For sympathetic neurones the extracellular application of membrane permeable cyclic GMP (McAfee & Greengard, 1972; Dun, Kaibara & Karczmar. 1977, 1978; Hashiguchi, Ushiyama, Kobayashi & Libet, 1978) and intracellular application of cyclic GMP (Gallagher & ShinnickGallagher, 1977) cause a slow depolarizing response. The types of ionic channels regulated by cyclic GMP are also not known. The aim of the present study was to determine if cyclic GMP regulates resting membrane channels and the voltagedependent calcium channels in mammalian parasympathetic neurones. A preliminary account of some of this work has been published (Nishimura, Akasu & Krier, 1990). METHODS
Male white rabbits weighing 2-03-0 kg were anaesthetized with pentobarbitone (40-50 mg/kg i.v.). The methods for isolation of vesical pelvic ganglia were described previously (Nishimura, Tokimasa & Akasu, 1988; Akasu, Nishimura & Tokimasa, 1990a). After removal of the pelvic ganglia, rabbits were killed by intravenous injection of an excess dose of pentobarbitone. Individual ganglia were then pinned onto Sylgard at the bottom of a perfusion chamber (0 5 ml in total volume) and continuously superfused with Krebs solution (3 ml/min) having the following composition (mM): NaCl, 117; KCl, 4 7; CaC12, 2 5; MgCl2 1 2; NaH2PO4, 1-2; NaHCO3, 25; and glucose, 11. Solutions were gassed with 95 % 02, 5 % CO2 and preheated to 35-37 °C at the recording site. Microelectrodes were filled with either 3 M KCl or 2 M CsCl. The tip resistance of microelectrodes was 20-40 MQ. Membrane currents were recorded by the single-electrode voltage-clamp method with an Axoclamp 2A (Axon Instruments) (Akasu et al. 1990a). To block voltage-dependent sodium and potassium currents, the Krebs solution contained tetrodotoxin (TTX, 300 nM) and tetraethylammonium (TEA, 50 mM), where sodium was reduced to 93 2 mm. Caesium ions were ionophoretically injected into pelvic ganglia neurones through the recording microelectrode filled with 2 M CsCl to block potassium channels. Caesium ions (1-2 mM) were also added to the superfusing solution to depress the inward rectification, when current-voltage relationships were obtained (Nishimura et al. 1988). The voltage-dependent calcium currents were evoked by depolarizing step commands to -30 to + 50 mV from a holding potential ranged between -40 and -70 mV. To obtain the voltage-dependent barium current, barium (2-5 mM) was substituted for calcium (Akasu et al. 1990 a). Both calcium and barium inward currents were blocked by application of cobalt (1 mm) or o-conotoxin (500 nM) (Akasu et al. 1990a). Net voltage-dependent barium currents ('Ba) were obtained by subtracting inward currents produced by hyperpolarizing voltage commands of magnitude equal to the depolarizing commands. Subtraction of these currents from the IBa was made by using a Nicolet memory oscilloscope (Nicolet, USA). Calcium- (or barium-) free solution was made by removal of 2-5 mm calcium (or barium) and adding 12 mm magnesium. A low-chloride solution was made by replacement of 67 mM chloride with equimolar isethionate. Forskolin and nicardipine were dissolved in 99.50% ethanol
CYCLIC GUIP AND CXALC.IUMJl/ CHANINS
561 56LS
and diluted in the Krebs solution. The final concentration of ethanol in the Krebs solution was less than 0050% by volume. Other drugs were directly dissolved in Krebs solution and applied by switching the flow with a three-way stopcock. Drugs used were tetrodotoxin from Sankyo; tetraethylammonium chloride from Tokyo Kasei (Japan); 4-acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid disodium 3H20 from Research Organics (USA); w -conotoxin GIVA from Peptide Institute (Japan); nicardipine from Sigma; dibutyryl guanosine 3',5-cyclic monophosphate (db-cyclic GMP) and related compounds (8bromo-cyclic GMP, cyclic (AMP, GTP, GDP, GMP and guanosine) from Sigma; dibutyryl adenosine 3'.5'-cyclic monophosphate, 8-bromoadenosine 3',5'-ccelic monophosphate and forskolin from Sigma 3-isobutyl- 1-methylxanthine from Aldrich. Data are given as means +S.E. of mean. RESULTS
Effects of db- and 8-bromo-cyclic GYIP on the membrane potential Application of db-cyclic GMP (100 aM for 3 min) to neurones of the rabbit pelvic ganglia caused a slow depolarizing response (mean total duration, 5+ 2 min, n = 5) with amplitudes ranging between 10 and 23 mV (mean peak amplitude, 18+7 mV, n = 5) in Krebs solution (Fig. 1A). During the depolarization, membrane input resistance was decreased (Fig. 1B). When db-cyclic GMP was removed from the superfusing solution, the depolarization recovered and was followed by a hyperpolarizing response with a mean amplitude of 10+6 mV, lasting up to 20-30 min (n = 5). The after-hyperpolarization was associated with an increase in membrane input resistance (Fig. 1B). In this particular neurone, a spontaneous hyperpolarizing potential occurred at the resting membrane potential (Nishimura, Akasu & Tokimasa, 1991 b). It was depressed during the depolarization, while it was enhanced during the hyperpolarization produced by db-cyclic GMP. A related compound, 8bromo-cyclic GMP (300 AM), mimicked the effects of db-cyclic GMP in producing a biphasic membrane response in neurones of the pelvic ganglia.
Db-cyclic GiVIP-induced inward and outward currents Under voltage-clamp mode, db-cyclic GMP (100 /,M) caused an inward current with amplitude of 2-4 nA in neurones superfused with the Krebs solution (Fig. 2Aa). The inward current tended to return to the initial holding level during the continuous application of db-cyclic GMP. When db-cyclic GMP was removed from the superfusing solution, the inward current was followed by an outward current with an amplitude of 0 5 nA, lasting approximately 20 min in the recovery solution. Figure 2Ab shows the relation between the concentration of db-cyclic GMP and the amplitudes of the inward and the outward currents. At 10 UM, db-cyclic GMP produced inward currents with amplitude of 0 5-0 8 nA. The maximum response (3-4.5 nA) was obtained with 1 mm db-cyclic GMP. The effects of db-cyclic GMP were also compared with other guanosine nucleotides in producing inward and outward currents in the same neurones. Cyclic GMP also produced an inward current response but it was less potent (n = 5) than db-cyclic GMP (Fig. 2Ab). GTP (1 mM), GDP (1 mm), GMP (1 mM), guanosine (1 mM) and db-cyclic AMP (2 mM) had no effect on neurones (n = 6, each). Forskolin (1L1uM), an activator of adenylate cyclase (Seamon & Daly, 1981), caused no current response (n = 4). Pretreatment of neurones with isobutyl methylxanthine (IBMX, 100 AtM), a phosphodiesterase inhibitor, increased to 135 + 9% (n = 4) the amplitude of the inward current produced by db-cyclic GMP.
T. NISHIMURA, T. AKASU AND J. KRIER
562
Conductance change during db-cyclic GMP-induced currents The inward and outward currents produced by db-cyclic GMP (10 ItM-1 mM) were not blocked by bath application of TTX (300 gtM) and TEA (50 mM), where voltagedependent sodium and potassium currents were blocked (Akasu et al. 1990a). DbA
db-cyclic GMP
RP -55 mV C
0.5 nA
2.0 my d
B a
b
c
60 s
d
0 5 nA 200 ms Fig. 1. Depolarization followed by hyperpolarization induced by bath application of dbcyclic GMP (100 1M) in a vesical pelvic neurone. The resting membrane potential was -55 mV. A, period of application is indicated by horizontal bar. Downward deflections in trace indicate hyperpolarizing electrotonic potentials (lower trace) evoked by anodal currents (upper trace) with an amplitude of 0 3 nA for 200 ms. B, expanded records of electrotonic potential (lower traces) and anodal current pulses (upper traces). Records a-d were taken at times marked by respective letters in A.
cyclic GMP (10 gM-1 mM) also caused a biphasic current response in caesium-loaded neurones with amplitudes comparable to those obtained from neurones where sodium and potassium currents were not suppressed (Fig. 2B). To analyse the ionic mechanisms of the db-cyclic GMP-induced responses, the remainder of the experiments were performed on neurones loaded with caesium and superfused with a Krebs solution containing TTX (300 nM), TEA (50 mM) and caesium (2 mM). Membrane conductance, measured by using inward currents evoked by hyperpolarizing voltage jumps from -60 to -85 mV, was increased during the db-cyclic GMP (100 gtM)-induced inward current (Fig. 2B). In contrast, the outward current produced by db-cyclic GMP was associated with a decrease in membrane conductance
(Fig. 2B).
Membrane conductance was calculated by determining the slope of the relationship between membrane potential and current (I-V curve) (Fig. 2C). Vesical pelvic neurones were superfused with a solution containing 2 mm caesium to block inward rectification (Nishimura et at. 1988). Linear I-V curves were obtained between -45 and -100 mV; the chord conductance was 25 nS in the control solution (Fig. 2C). When db-cyclic GMP (100 gtM) was added to the superfusing solution the slope of the I-V curve was increased. The chord conductance increased to 4 nS during the db-
CYCLIC GMP AND CALCIUM CHANNELS 563 cyclic GMP-induced inward current. Extrapolated I-V curves obtained before and during the db-cyclic GMP (100 gtM)-induced inward current met at + 43 mV (Fig. 2 C). For five neurones, the estimated mean reversal potential of the inward current was +42±8mV. A
C
b
db-cyclic GMP
I,,mX
Ax. 0 c)
4
Eo 2
LnA
(
60 s
as
K
10 B
C
db-cyclic GMP a
100 1000 Concentration (uM) +2
mm-a-
mIII
Outward current
1 1 1 1 1 1 1 1 11011
i11
iv
1 nA
-100
0
+50 mV
ITU14Milkfil
b
ii *L-
-60
.1
s ki
itv r-
150 mV iI 200
nA ms
60
s
Control
--2
A' Inward current - nA
-85 mV
Fig. 2. A, biphasic current response produced by bath application of db-cyclic GMP (100 gM) in Krebs solution (a). Horizontal bar indicates the period of application of dbcyclic GMP (100 UM). Membrane potential was held at -65 mV. Graph b shows the relation between the concentration of cyclic GMP analogues and the amplitude of the inward (open symbols) and outward (filled symbols) currents. Responses produced by db-cyclic GMP and cyclic GMP are indicated by circles and triangles, respectively. Each point represents the mean amplitude of current responses obtained from five neurones. Vertical lines indicate S.E. of mean. B, membrane conductance change during db-cyclic GMP (100 gM)-induced currents. Data obtained from a caesium-loaded neurone in the presence of TTX (300 nM) and TEA (50 mM) with sodium reduced to 93 2 mm. The resting membrane conductance was measured from inward current produced by hyperpolarizing step commands with duration of 200 ms (downward deflection). Holding potential was -60 mV. Upper and lower traces represent membrane current and membrane voltage, respectively. Horizontal bar indicates time period of db-cyclic GMP application in the superfusing solution. Note that the size of the current steps are obscured by capacitative transients. Panel b shows expanded records of inward currents (upper traces) and hyperpolarizing step commands (lower traces). Records i-iv were obtained at the times marked by the respective letters in a. Record iv was obtained 15 min after wash-out of dbcyclic GMP. C, current-voltage relations obtained during the inward and outward currents produced by db-cyclic GMP (100 sM). Hyperpolarizing and depolarizing voltage step commands with a duration of 200 ms were applied from a holding potential of -50 mV. Dotted lines indicate extrapolation of I-V curves. Superfusing solution contained CsCl (2 mM).
A- ^~ ~ ~
T. NISHIMURA, T. AKASU AND J. KRIER
564
The I-V curves were also obtained during the outward current induced by dbcyclic GMP. The slope of the I-V curve decreased during the outward current; the chord conductance was 12 nS during the outward current (Fig. 2 C). Extrapolation of I-V curves of the control and that obtained during the outward current showed A +8
B
+1
-10~~~~~~~~~~~~~~~ -10 _-150 mV -100 -25
-50
A
-3 nA
-60
C -130 4Ani A mV tN A
14min
I
-100
-50
A
,
0 mV
Fig. 3. Effects of membrane potential on inward and outward currents produced by dbcyclic GMP (100 yM). A, db-cyclic GMP-induced currents recorded at holding potentials ranging between + 8 and - 130 mV. Holding potentials are shown at left of each trace. Horizontal bars indicate the time period of application of db-cyclic (GMP. B, relation between amplitude of the db-cyclic GMIP-induced currents (ordinate) and holding voltage (abscissa). A and A, peak amplitudes of the inward and the outward currents, respectively. Each point was obtained from records in A. C? voltage-dependent properties of inward current decay. The time in which the inward current decays from the peak to 50 % amplitude (half-decay time) (ordinate) is plotted against holding voltage (abscissa). Half-decay time is shown by the horizontal bar with arrow (inset).
an intersection at - 18 mYl. The estimated mean reversal potential for the db-cyclic GMP-induced outward current was -26 + 6 mV (n = 5).
Effects of membrane potential on current responses The inward current increased in amplitude on hyperpolarization, while it decreased at depolarized potentials (Fig. 3A). An almost linear relationship between the amplitude of the inward current and the holding potential was obtained at potentials ranging between - 130 and + 8 m\ (Fig. 3B). There was no reversal of the inward current at potentials ranging between - 130 mV and + 8 mV. The outward current decreased in amplitude on depolarization of the holding potential and eventually reversed its polarity at potentials more positive than -25 m\' in this particular neurone (Fig. 3B). The time to peak of the inward current became faster with membrane hyperpolarization and slower at depolarized membrane potentials (Fig. 3A). Figure 3 C shows that the decay of inward current was also voltage dependent; the half-decay time measured from the peak of the inward current was increased at depolarized holding potentials. The voltage dependence of the half-decay time of the
565
CYCLIC GMP AND CALCIUM CHANNELS
inward current may be in part due to an increase in the amplitude of overlapping outward current at hyperpolarized holding potentials.
Ion dependency of the inward and outward currents Removal of bxtracellular calcium from the superfusing solution containing magnesium (12 mM) caused no detectable membrane currents. The amplitude of the B
A
Control db-cyclic GMP
Ca2` free
+2 0
Ca2 free
-50
-100
0
+50 mV
Control Wash
db-cyclic GMP
LInA 60 s
_2 nA
D
C
Ca2` free db-cyclic GMP
Ca2` free
Ca2` free
Ca2+ free + low Cl-
+
Ia 2+nfree 60 s
SITS Ca2+ free nA 60 s
N~~~~ Fig. 4. A, effects of calcium-free solution (containing zero calcium and 12 mm magnesium) on the db-cyclic GMP (100 #,M)-induced current responses in caesium-loaded neurones. Modified Krebs solution also contained TTX (300 nM) and TEA (50 mM). Periods of bath application of db-cyclic GMP are indicated by horizontal bars. Holding potential was -55 mV. B, steady-state I-V curves obtained by step commands (0-2 Hz, duration 500 ms) from a holding potential of -50 mV in calcium-free solution. The neurone was loaded with caesium ions and superfused with a solution containing TTX (300 nM), TEA (50 mM) and caesium (2 mM). Filled and open symbols were taken in the absence and presence of db-cyclic GMP, respectively. C, effect of lowering chloride concentration on the calcium-insensitive component of the inward current produced by db-cyclic GMP (100 ,SM). Chloride was reduced from 129 1 to 62-1 mm by replacing it with isethionate. D, effect of SITS (500 /M) on the calcium-independent component of the inward current produced by db-cyclic GMP (100 /M). Horizontal bars indicate the time period of dbcyclic GMP application. Holding potentials were -60 and -70 mV in C and D, respectively.
db-cyclic GMP-induced inward current was decreased in the calcium-free solution (Fig. 4A). For six neurones, 69+5% (mean+s.E.M.) depression of the inward current was caused in the calcium-free solution (Table 1). The outward current, which appeared after removal of db-cyclic GMP, was completely inhibited in the calcium-free solution (Fig. 4A). Lowering the sodium concentration in the external solution to 26-2 mm (67 mm sodium chloride was replaced with equimolar choline chloride) produced no change in the holding current or in the amplitudes of db-cyclic GMP-induced inward and outward currents. Similarly, raising the extracellular
566
T. NISHIMURA, T. AKASU AND J. KRIER
potassium concentration in the external solution to 10 mm did not change either the holding current or the amplitudes of db-cyclic GMP-induced inward and outward currents. Statistical data are shown in Table 1. These results suggest that calcium ions are the main charge carrier for the inward current produced by db-cyclic GMP. Effect of calcium channel antagonists Superfusion of the ganglia with cobalt (1 mm) depressed the inward current and completely inhibited the outward current caused by db-cyclic GMP (data not TABLE 1. Calcium-dependent properties of the inward current produced by db-cyclic GMP Amplitude (%) Ionic dependence Calcium-free solution 31+5 (t= 6)* Low-sodium (26-2 mm) solution 95 + 4 (n = 4) High-potassium (10 mM) solution 97 + 3 (n = 3) Calcium antagonists 36 + 8 (n = 4)* Nicardipine (10 ,#M) w-Conotoxin (2 /iM) 93 + 4 (n = 3) * P < 0-001 (Student's t test). Number of experiments is shown in parentheses.
shown). The administration of nicardipine (10 gM), a calcium channel antagonist, depressed the amplitude of the inward current produced by db-cyclic GMP. From four experiments, nicardipine produced 64+ 8% depression of the inward current (Table 1). In contrast, w)-conotoxin (2 gim), a calcium channel antagonist known to block voltage-dependent calcium channels in vesical neurones (Akasu et al. 1990a), did not alter the inward current caused by db-cyclic GMP (Table 1). Chloride component of the inward current In calcium-free solution containing 1 mm cobalt, db-cyclic GMP (100 /LM) produced a 1-2 nA inward shift of membrane current and decreased the slope of the I-V curve. The estimated reversal potential of the calcium-insensitive component of the inward current as measured from the intersection of the I-V curves was -13 mV (-20 + 6 mV, n = 5) (Fig. 4B). The value for the reversal potential of the inward current was near the equilibrium potential for chloride (see Akasu et al. 1990 a). The calcium-insensitive inward current was increased in amplitude in low-chloride solution, where 67 mm sodium chloride was replaced with equimolar sodium isethionate (Fig. 4 C). 4-Acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid (SITS) has been reported to block chloride current (Inoue, 1985; Gray & Ritchie, 1986; Bader, Bertrand & Schlichter, 1987; Akasu et al. 1990a). Addition of SITS (500 gM) to the calcium-free solution completely inhibited the calcium-insensitive component of the inward current (Fig. 4D). These results suggest that a chloride current may be involved in the db-cyclic GMP-induced inward current.
Voltage-dependent calcium channel currents Calcium current (ICa) Voltage-dependent calcium current (ICa) was recorded from caesium-loaded neurones superfused with a Krebs solution containing TTX (300 nM) and TEA (50 mm). Application of depolarizing voltage jumps from -60 to -10 mV evoked an
CYCLIC GMP AND CALCIUM CHANNELS 567 inward calcium current with amplitude of 0-8-3-5 nA followed by an inward tail current (Fig. 5). This tail current is produced by deactivation of voltage-dependent chloride channels (Akasu et al. 1990 a). Db-cyclic GMP (50 /IM) depressed both the ICa and tail current during the inward current (Fig. 5). When db-cyclic GMP was A
db-cyclic GMP
;
t
L
I
J
I
L
l
dI! t L
B
L
il
Il
a m* A
b.
i
7
li
-
_
_
.
!
'LI1 I
cd it, __> C -
I
III
I
Il1 nA I
MV I "IbUIfn\1
b 200 ms
-10 ;
Vh -60 mV Fig. 5. A, effects of db-cyclic GMP (50 ,sM) on the ICa evoked by depolarizing voltage jumps from a holding potential of -60 to -10 mV. Modified Krebs solution contained TTX (300 nM), TEA (50 mM) and 2-5 mm calcium. The neurone contained caesium. Upper and lower traces represent membrane current and holding voltage (V,), respectively. B, expanded records of ICas and tail currents. Records a-d were obtained at the times marked by respective letters in A. Records d were taken 20 min after removal of db-cyclic GMP. A and A indicate the ICa and the chloride tail current, respectively. Horizontal lines in current traces indicate current at the holding potential (-60 mV).
removed from the superfusing solution, the ICa and tail currents recovered and then increased in amplitude as long as the outward current occurred (Fig. 5). Barium current (IBa) To block voltage-dependent chloride tail currents, barium ions (2-5 mM) were substituted for calcium ions in the superfusing solution (Akasu et al. 1990 a). Application of depolarizing voltage jumps from -60 to -20 mV evoked an inward current with decay time slower than that of ICa in the presence of barium (2-5 mM) (Fig. 6). It has been reported that the IBa was blocked by cobalt (1 mM) or Wconotoxin (500 nM) (Akasu et al. 1990 a; Akasu, Tsurusaki & Tokimasa, 1990b). However, the IBa was not blocked by application of nicardipine (10 JM) (n = 3). Application of db-cyclic GMP (5 /M-1 mM) in barium-containing solution caused an inward current (amplitude, 0-2-4 nA) associated with an increase in membrane conductance followed by an outward current (0 1-1 8 nA) associated with a decreased membrane conductance. The IBa decreased in amplitude during the db-cyclic GMP (100 puM)-induced inward current (Fig. 6Bb). Partial recovery of the IBa was evident in the continued presence of db-cyclic GMP. In contrast, the IBa was augmented during the occurrence of the outward current (Fig. 6Bc). The potentiation of the IBa lasted for a time period of approximately 20 min in db-cyclic GMP-free solution (Fig. 6 C). At a concentration of 100 /tM, db-cyclic GMP produced 64 + 20°% (n = 8) depression followed by 60 + 10% (n = 8) facilitation of the IBa. 19
PHY 457
568
T. NISHIMURA, T. AKASU AND J. KRIER
Figure 7 shows the effect of db-cyclic GMP on the I-V curve for 'Ba. The neurone was initially held at -50 mV and subjected to voltage jumps between - 100 and +50 mV. IBa appeared near -30 mV and the peak 'Ba occurred at -10 to + 10 mV. Db-cyclic GMP decreased the amplitude of the IBa at potentials from -30 to db-cyclic GMP
A
a
d 12 nA ~~~c
~
~
b
0Q mV a
B
c
b
d
60s 1
Vh-60 mV
___---
2 nA
-300 ms
-
C
-5 -4
-3
db-cyclicGMP
-
.
0 0. @
*°e@*D.............. - ~ ~~~~ ------------2~ ~~~
~~~~~00000 -2~~~~~oo .000 2
-1-
0
5
10
15 20 Time (min)
25
30
35
Fig. 6. Effect of db-cyclic GCMP (100 /M) on the IBa (see Methods) evoked by depolarizing voltage jumps (500 ms duration) from a holding potential of -60 mV to -20 mV. The neurone contained caesium. Modified Krebs solution contained TTX (300 nM) and TEA (50 mM). Calcium (2-5 mM) was replaced with equimolar barium. A, horizontal bar indicates period of application of db-cyclic GMP (100 aM). Record d was taken 20 min after withdrawal of db-cyclic GMP. Downward deflection in lower trace represents hyperpolarizing voltage jump from -60 to -100 mV. B, expanded records of 4BaS (upper traces) and voltage commands (lower traces) taken at the times marked by respective letters in A. C, time course of the inhibition and facilitation of the IBa produced by dbcyclic GMP. The period of application of db-cyclic GMP (100,UM) is indicated by the horizontal bar. Ordinate and abscissa indicate the peak amplitude of the net 'Ba and time, respectively. Each point represents the average of three consecutive 'Bas.
+ 40 mV without causing any change in the threshold or the peak voltage for IBa activation. Figure 7A shows the I-V relation of the IBa obtained by subtracting inward currents produced by hyperpolarizing voltage commands of magnitude equal to the depolarizing commands used to evoke the IBa at various command potentials. Db-cyclic GMP consistently reduced the IBa at command potentials from -30 to + 40 mV during the inward current (Fig. 7A). Figure 7B shows the I-V relation of the IBa for the same neurone during the dbcyclic GMP-induced outward current. The IBa increased in amplitude at command potentials between -30 and + 40 mV. The threshold voltage for the IBa and the voltage for the peak current of the IBa were not changed during the facilitation.
CYCLIC GMP AND CALCIUM CHANNELS A Inward current -50
+50 mV
0 ,.1
db-cyclic GMP
I
B Outward current -50 0 .
A
569
+50 mV
.
/,
/
-
Control -
Control
-4nA
-
-4
0
db-cyclic GMP -\ , -6 nA
Fig. 7. Current-voltage relation (I-V curve) obtained during inward current (A) and outward current (B) produced by db-cyclic GMP (50 gM). The I-V curves were obtained by step commands with duration of 200 ms from a holding potential of -50 mV. The superfusing solution contained caesium (2 mM) to block inward rectification. IBas were obtained from subtraction of inward current produced by hyperpolarizing voltage commands of magnitude equal to depolarizing commands. 0, control I-V curve obtained before application of db-cyclic GMP. A and 0, I-V curves obtained during the db-cyclic GMP-induced inward and outward currents, respectively. A
8-Bromo-cyclic GMP _
2
GMP
Cyclic GMP i
-10
Vh
B
Forskolin
-62 mV
8-Bromo-cyclic AMP
-I 1 0
Vh
50ms
-85 mV
Fig. 8. A, comparison of the effects of 8-bromo-cyclic GMP (100 pm), cyclic GMP (1 mM) and GMP (1 mM) on the IB.. The IBas were evoked by depolarizing voltage jumps to -10 mV from a holding potential of -62 mV. B, effects of 8-bromo-cyclic AMP (1 mM) and forskolin (10 FM) on the IBas evoked by depolarizing voltage jumps to 0 mV from a holding potential of -65 mV. In each panel, record 1 was obtained before application of drugs. Records 2 and 3 were obtained during application of drugs and 5-10 min after withdrawal of drugs, respectively. A and B were obtained from different neurones.
Related compounds of cyclic GMP The effects of db-cyclic GMP on the IBa were compared with other guanosine derivatives on the same neurone (Fig. 8). Bath application of 8-bromo-cyclic GMP (10-300 pM) caused an inhibition followed by a facilitation of the IBa (Fig. 8A). The efficacy of 8-bromo-cyclic GMP (100 pM) was essentially the same as that of db-cyclic GMP (100 #M) (n = 5). Cyclic GMP (1 mM) was less potent than db-cyclic GMP (100 pM) in producing the inhibition of the IBa (n = 3) (Fig. 8 A). For this neurone, cyclic GMP showed minimal facilitation of the 1Ba (Fig. 8A). GMP (1 mM), GDP (1 mM), GTP (1 mM) and guanosine (1 mM) had no effect on the IBa (n > 2 for each 19-2
12-~ ~3 4n~
T. NISHIMURA, T. AKASU AND J. KRIER
570
agent). IBMX potentiated the effects of db-cyclic GMP on the IBa (not shown). 8Bromo-cyclic adenosine monophosphate (1 mM) and forskolin (10 gM) did not change the IBa (n = 4) (Fig. 8 B). A
Control 1~r
_
Wash 5 min db-cyclic GMP _, 2 t
559
GUANOSINE 3',5'-CYCLIC MONOPHOSPHATE REGULATES CALCIUM CHANNELS IN NEURONES OF RABBIT VESICAL PELVIC GANGLIA
BY T. NISHIMURA, T. AKASU AND J. KRIER* From the Department of Physiology, Kurume University School of Medicine, 67 Asahi-machi, Kurume 830, Japan and the *Department of Physiology, Michigan State University, East Lansing, MI 48824, USA
(Received 27 December 1991) SUMMARY
1. The effects of dibutyryl guanosine 3',5'-cyclic monophosphate (db-cyclic GMP) were studied in vitro on calcium channels of neurones in rabbit vesical parasympathetic ganglia, using intracellular and single-electrode voltage-clamp
recordings. 2. Db-cyclic GMP (100 iM) caused membrane depolarization associated with a decrease in membrane input resistance and an after-hyperpolarization associated with an increase in membrane input resistance. 3. Db-cyclic GMP (0-01-1 mM) caused a concentration-dependent, transient inward current followed by a long-lasting outward current. Membrane conductance was increased and decreased during the inward and outward currents, respectively. 4. The db-cyclic GMP-induced inward current was depressed in nominally calcium-free solutions, by cobalt (1 mM) and nicardipine (10 EM). The mean reversal potentials of the inward current were + 42 and -20 mV in the presence and absence of calcium in the external solution, respectively. 5. The db-cyclic GMP-induced inward current was not altered by lowering the external sodium concentration, raising external potassium concentration or by intracellular injection of caesium. 6. A calcium-insensitive component of the db-cyclic GMP-induced current was increased by lowering the external chloride concentration and blocked by 4acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid, a chloride channel blocker. 7. Voltage-dependent, high-threshold calcium currents were depressed during the db-cyclic GMP-induced inward current and facilitated during the outward current. 8. Cyclic GMP was less potent than db-cyclic GMP in causing both inward and outward currents or modulation of calcium currents. GTP, GDP, GMP, guanosine, 8-bromoadenosine 3',5'-cyclic monophosphate and forskolin did not alter the holding current or voltage-dependent calcium currents. 9. It is concluded that intracellular cyclic GMP causes not only activation of resting calcium and chloride channels but also a transient depression followed by long-lasting facilitation of voltage-dependent calcium currents in neurones of vesical parasympathetic ganglia. MS 8948
560
560T. NISIHIMTRA, T. AKASUr AND J. KRIER INTRODUCTION
Guanine nucleotides are considered second messengers for intracellular signal transduction in various tissues (Greengard, 1976). Guanosine 3',5'-cyclic monophosphate (cyclic GMP) activates light-sensitive conductance in frog and toad retinal rod outer segments (Fesenko, Kolesnikov & Lyubarsky, 1985; Yau & Nakatani, 1985), potassium channels in Xenopus oocytes (Dascal, Lotan & Lass, 1987) and calcium channels in snail neurones (Paupardin-Tritsch, Hammond, Gerschenfeld, Nairn & Greengard, 1986). Cyclic GMP decreases voltage-dependent calcium currents of frog cardiac muscles (Hartzell & Fischmeister, 1986) and pyramidal neurones in guinea-pig hippocampus (Doerner & Alger, 1988). The role of cyclic GMP as an intracellular messenger for neurones in mammalian autonomic ganglia is not precisely understood. For sympathetic neurones the extracellular application of membrane permeable cyclic GMP (McAfee & Greengard, 1972; Dun, Kaibara & Karczmar. 1977, 1978; Hashiguchi, Ushiyama, Kobayashi & Libet, 1978) and intracellular application of cyclic GMP (Gallagher & ShinnickGallagher, 1977) cause a slow depolarizing response. The types of ionic channels regulated by cyclic GMP are also not known. The aim of the present study was to determine if cyclic GMP regulates resting membrane channels and the voltagedependent calcium channels in mammalian parasympathetic neurones. A preliminary account of some of this work has been published (Nishimura, Akasu & Krier, 1990). METHODS
Male white rabbits weighing 2-03-0 kg were anaesthetized with pentobarbitone (40-50 mg/kg i.v.). The methods for isolation of vesical pelvic ganglia were described previously (Nishimura, Tokimasa & Akasu, 1988; Akasu, Nishimura & Tokimasa, 1990a). After removal of the pelvic ganglia, rabbits were killed by intravenous injection of an excess dose of pentobarbitone. Individual ganglia were then pinned onto Sylgard at the bottom of a perfusion chamber (0 5 ml in total volume) and continuously superfused with Krebs solution (3 ml/min) having the following composition (mM): NaCl, 117; KCl, 4 7; CaC12, 2 5; MgCl2 1 2; NaH2PO4, 1-2; NaHCO3, 25; and glucose, 11. Solutions were gassed with 95 % 02, 5 % CO2 and preheated to 35-37 °C at the recording site. Microelectrodes were filled with either 3 M KCl or 2 M CsCl. The tip resistance of microelectrodes was 20-40 MQ. Membrane currents were recorded by the single-electrode voltage-clamp method with an Axoclamp 2A (Axon Instruments) (Akasu et al. 1990a). To block voltage-dependent sodium and potassium currents, the Krebs solution contained tetrodotoxin (TTX, 300 nM) and tetraethylammonium (TEA, 50 mM), where sodium was reduced to 93 2 mm. Caesium ions were ionophoretically injected into pelvic ganglia neurones through the recording microelectrode filled with 2 M CsCl to block potassium channels. Caesium ions (1-2 mM) were also added to the superfusing solution to depress the inward rectification, when current-voltage relationships were obtained (Nishimura et al. 1988). The voltage-dependent calcium currents were evoked by depolarizing step commands to -30 to + 50 mV from a holding potential ranged between -40 and -70 mV. To obtain the voltage-dependent barium current, barium (2-5 mM) was substituted for calcium (Akasu et al. 1990 a). Both calcium and barium inward currents were blocked by application of cobalt (1 mm) or o-conotoxin (500 nM) (Akasu et al. 1990a). Net voltage-dependent barium currents ('Ba) were obtained by subtracting inward currents produced by hyperpolarizing voltage commands of magnitude equal to the depolarizing commands. Subtraction of these currents from the IBa was made by using a Nicolet memory oscilloscope (Nicolet, USA). Calcium- (or barium-) free solution was made by removal of 2-5 mm calcium (or barium) and adding 12 mm magnesium. A low-chloride solution was made by replacement of 67 mM chloride with equimolar isethionate. Forskolin and nicardipine were dissolved in 99.50% ethanol
CYCLIC GUIP AND CXALC.IUMJl/ CHANINS
561 56LS
and diluted in the Krebs solution. The final concentration of ethanol in the Krebs solution was less than 0050% by volume. Other drugs were directly dissolved in Krebs solution and applied by switching the flow with a three-way stopcock. Drugs used were tetrodotoxin from Sankyo; tetraethylammonium chloride from Tokyo Kasei (Japan); 4-acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid disodium 3H20 from Research Organics (USA); w -conotoxin GIVA from Peptide Institute (Japan); nicardipine from Sigma; dibutyryl guanosine 3',5-cyclic monophosphate (db-cyclic GMP) and related compounds (8bromo-cyclic GMP, cyclic (AMP, GTP, GDP, GMP and guanosine) from Sigma; dibutyryl adenosine 3'.5'-cyclic monophosphate, 8-bromoadenosine 3',5'-ccelic monophosphate and forskolin from Sigma 3-isobutyl- 1-methylxanthine from Aldrich. Data are given as means +S.E. of mean. RESULTS
Effects of db- and 8-bromo-cyclic GYIP on the membrane potential Application of db-cyclic GMP (100 aM for 3 min) to neurones of the rabbit pelvic ganglia caused a slow depolarizing response (mean total duration, 5+ 2 min, n = 5) with amplitudes ranging between 10 and 23 mV (mean peak amplitude, 18+7 mV, n = 5) in Krebs solution (Fig. 1A). During the depolarization, membrane input resistance was decreased (Fig. 1B). When db-cyclic GMP was removed from the superfusing solution, the depolarization recovered and was followed by a hyperpolarizing response with a mean amplitude of 10+6 mV, lasting up to 20-30 min (n = 5). The after-hyperpolarization was associated with an increase in membrane input resistance (Fig. 1B). In this particular neurone, a spontaneous hyperpolarizing potential occurred at the resting membrane potential (Nishimura, Akasu & Tokimasa, 1991 b). It was depressed during the depolarization, while it was enhanced during the hyperpolarization produced by db-cyclic GMP. A related compound, 8bromo-cyclic GMP (300 AM), mimicked the effects of db-cyclic GMP in producing a biphasic membrane response in neurones of the pelvic ganglia.
Db-cyclic GiVIP-induced inward and outward currents Under voltage-clamp mode, db-cyclic GMP (100 /,M) caused an inward current with amplitude of 2-4 nA in neurones superfused with the Krebs solution (Fig. 2Aa). The inward current tended to return to the initial holding level during the continuous application of db-cyclic GMP. When db-cyclic GMP was removed from the superfusing solution, the inward current was followed by an outward current with an amplitude of 0 5 nA, lasting approximately 20 min in the recovery solution. Figure 2Ab shows the relation between the concentration of db-cyclic GMP and the amplitudes of the inward and the outward currents. At 10 UM, db-cyclic GMP produced inward currents with amplitude of 0 5-0 8 nA. The maximum response (3-4.5 nA) was obtained with 1 mm db-cyclic GMP. The effects of db-cyclic GMP were also compared with other guanosine nucleotides in producing inward and outward currents in the same neurones. Cyclic GMP also produced an inward current response but it was less potent (n = 5) than db-cyclic GMP (Fig. 2Ab). GTP (1 mM), GDP (1 mm), GMP (1 mM), guanosine (1 mM) and db-cyclic AMP (2 mM) had no effect on neurones (n = 6, each). Forskolin (1L1uM), an activator of adenylate cyclase (Seamon & Daly, 1981), caused no current response (n = 4). Pretreatment of neurones with isobutyl methylxanthine (IBMX, 100 AtM), a phosphodiesterase inhibitor, increased to 135 + 9% (n = 4) the amplitude of the inward current produced by db-cyclic GMP.
T. NISHIMURA, T. AKASU AND J. KRIER
562
Conductance change during db-cyclic GMP-induced currents The inward and outward currents produced by db-cyclic GMP (10 ItM-1 mM) were not blocked by bath application of TTX (300 gtM) and TEA (50 mM), where voltagedependent sodium and potassium currents were blocked (Akasu et al. 1990a). DbA
db-cyclic GMP
RP -55 mV C
0.5 nA
2.0 my d
B a
b
c
60 s
d
0 5 nA 200 ms Fig. 1. Depolarization followed by hyperpolarization induced by bath application of dbcyclic GMP (100 1M) in a vesical pelvic neurone. The resting membrane potential was -55 mV. A, period of application is indicated by horizontal bar. Downward deflections in trace indicate hyperpolarizing electrotonic potentials (lower trace) evoked by anodal currents (upper trace) with an amplitude of 0 3 nA for 200 ms. B, expanded records of electrotonic potential (lower traces) and anodal current pulses (upper traces). Records a-d were taken at times marked by respective letters in A.
cyclic GMP (10 gM-1 mM) also caused a biphasic current response in caesium-loaded neurones with amplitudes comparable to those obtained from neurones where sodium and potassium currents were not suppressed (Fig. 2B). To analyse the ionic mechanisms of the db-cyclic GMP-induced responses, the remainder of the experiments were performed on neurones loaded with caesium and superfused with a Krebs solution containing TTX (300 nM), TEA (50 mM) and caesium (2 mM). Membrane conductance, measured by using inward currents evoked by hyperpolarizing voltage jumps from -60 to -85 mV, was increased during the db-cyclic GMP (100 gtM)-induced inward current (Fig. 2B). In contrast, the outward current produced by db-cyclic GMP was associated with a decrease in membrane conductance
(Fig. 2B).
Membrane conductance was calculated by determining the slope of the relationship between membrane potential and current (I-V curve) (Fig. 2C). Vesical pelvic neurones were superfused with a solution containing 2 mm caesium to block inward rectification (Nishimura et at. 1988). Linear I-V curves were obtained between -45 and -100 mV; the chord conductance was 25 nS in the control solution (Fig. 2C). When db-cyclic GMP (100 gtM) was added to the superfusing solution the slope of the I-V curve was increased. The chord conductance increased to 4 nS during the db-
CYCLIC GMP AND CALCIUM CHANNELS 563 cyclic GMP-induced inward current. Extrapolated I-V curves obtained before and during the db-cyclic GMP (100 gtM)-induced inward current met at + 43 mV (Fig. 2 C). For five neurones, the estimated mean reversal potential of the inward current was +42±8mV. A
C
b
db-cyclic GMP
I,,mX
Ax. 0 c)
4
Eo 2
LnA
(
60 s
as
K
10 B
C
db-cyclic GMP a
100 1000 Concentration (uM) +2
mm-a-
mIII
Outward current
1 1 1 1 1 1 1 1 11011
i11
iv
1 nA
-100
0
+50 mV
ITU14Milkfil
b
ii *L-
-60
.1
s ki
itv r-
150 mV iI 200
nA ms
60
s
Control
--2
A' Inward current - nA
-85 mV
Fig. 2. A, biphasic current response produced by bath application of db-cyclic GMP (100 gM) in Krebs solution (a). Horizontal bar indicates the period of application of dbcyclic GMP (100 UM). Membrane potential was held at -65 mV. Graph b shows the relation between the concentration of cyclic GMP analogues and the amplitude of the inward (open symbols) and outward (filled symbols) currents. Responses produced by db-cyclic GMP and cyclic GMP are indicated by circles and triangles, respectively. Each point represents the mean amplitude of current responses obtained from five neurones. Vertical lines indicate S.E. of mean. B, membrane conductance change during db-cyclic GMP (100 gM)-induced currents. Data obtained from a caesium-loaded neurone in the presence of TTX (300 nM) and TEA (50 mM) with sodium reduced to 93 2 mm. The resting membrane conductance was measured from inward current produced by hyperpolarizing step commands with duration of 200 ms (downward deflection). Holding potential was -60 mV. Upper and lower traces represent membrane current and membrane voltage, respectively. Horizontal bar indicates time period of db-cyclic GMP application in the superfusing solution. Note that the size of the current steps are obscured by capacitative transients. Panel b shows expanded records of inward currents (upper traces) and hyperpolarizing step commands (lower traces). Records i-iv were obtained at the times marked by the respective letters in a. Record iv was obtained 15 min after wash-out of dbcyclic GMP. C, current-voltage relations obtained during the inward and outward currents produced by db-cyclic GMP (100 sM). Hyperpolarizing and depolarizing voltage step commands with a duration of 200 ms were applied from a holding potential of -50 mV. Dotted lines indicate extrapolation of I-V curves. Superfusing solution contained CsCl (2 mM).
A- ^~ ~ ~
T. NISHIMURA, T. AKASU AND J. KRIER
564
The I-V curves were also obtained during the outward current induced by dbcyclic GMP. The slope of the I-V curve decreased during the outward current; the chord conductance was 12 nS during the outward current (Fig. 2 C). Extrapolation of I-V curves of the control and that obtained during the outward current showed A +8
B
+1
-10~~~~~~~~~~~~~~~ -10 _-150 mV -100 -25
-50
A
-3 nA
-60
C -130 4Ani A mV tN A
14min
I
-100
-50
A
,
0 mV
Fig. 3. Effects of membrane potential on inward and outward currents produced by dbcyclic GMP (100 yM). A, db-cyclic GMP-induced currents recorded at holding potentials ranging between + 8 and - 130 mV. Holding potentials are shown at left of each trace. Horizontal bars indicate the time period of application of db-cyclic (GMP. B, relation between amplitude of the db-cyclic GMIP-induced currents (ordinate) and holding voltage (abscissa). A and A, peak amplitudes of the inward and the outward currents, respectively. Each point was obtained from records in A. C? voltage-dependent properties of inward current decay. The time in which the inward current decays from the peak to 50 % amplitude (half-decay time) (ordinate) is plotted against holding voltage (abscissa). Half-decay time is shown by the horizontal bar with arrow (inset).
an intersection at - 18 mYl. The estimated mean reversal potential for the db-cyclic GMP-induced outward current was -26 + 6 mV (n = 5).
Effects of membrane potential on current responses The inward current increased in amplitude on hyperpolarization, while it decreased at depolarized potentials (Fig. 3A). An almost linear relationship between the amplitude of the inward current and the holding potential was obtained at potentials ranging between - 130 and + 8 m\ (Fig. 3B). There was no reversal of the inward current at potentials ranging between - 130 mV and + 8 mV. The outward current decreased in amplitude on depolarization of the holding potential and eventually reversed its polarity at potentials more positive than -25 m\' in this particular neurone (Fig. 3B). The time to peak of the inward current became faster with membrane hyperpolarization and slower at depolarized membrane potentials (Fig. 3A). Figure 3 C shows that the decay of inward current was also voltage dependent; the half-decay time measured from the peak of the inward current was increased at depolarized holding potentials. The voltage dependence of the half-decay time of the
565
CYCLIC GMP AND CALCIUM CHANNELS
inward current may be in part due to an increase in the amplitude of overlapping outward current at hyperpolarized holding potentials.
Ion dependency of the inward and outward currents Removal of bxtracellular calcium from the superfusing solution containing magnesium (12 mM) caused no detectable membrane currents. The amplitude of the B
A
Control db-cyclic GMP
Ca2` free
+2 0
Ca2 free
-50
-100
0
+50 mV
Control Wash
db-cyclic GMP
LInA 60 s
_2 nA
D
C
Ca2` free db-cyclic GMP
Ca2` free
Ca2` free
Ca2+ free + low Cl-
+
Ia 2+nfree 60 s
SITS Ca2+ free nA 60 s
N~~~~ Fig. 4. A, effects of calcium-free solution (containing zero calcium and 12 mm magnesium) on the db-cyclic GMP (100 #,M)-induced current responses in caesium-loaded neurones. Modified Krebs solution also contained TTX (300 nM) and TEA (50 mM). Periods of bath application of db-cyclic GMP are indicated by horizontal bars. Holding potential was -55 mV. B, steady-state I-V curves obtained by step commands (0-2 Hz, duration 500 ms) from a holding potential of -50 mV in calcium-free solution. The neurone was loaded with caesium ions and superfused with a solution containing TTX (300 nM), TEA (50 mM) and caesium (2 mM). Filled and open symbols were taken in the absence and presence of db-cyclic GMP, respectively. C, effect of lowering chloride concentration on the calcium-insensitive component of the inward current produced by db-cyclic GMP (100 ,SM). Chloride was reduced from 129 1 to 62-1 mm by replacing it with isethionate. D, effect of SITS (500 /M) on the calcium-independent component of the inward current produced by db-cyclic GMP (100 /M). Horizontal bars indicate the time period of dbcyclic GMP application. Holding potentials were -60 and -70 mV in C and D, respectively.
db-cyclic GMP-induced inward current was decreased in the calcium-free solution (Fig. 4A). For six neurones, 69+5% (mean+s.E.M.) depression of the inward current was caused in the calcium-free solution (Table 1). The outward current, which appeared after removal of db-cyclic GMP, was completely inhibited in the calcium-free solution (Fig. 4A). Lowering the sodium concentration in the external solution to 26-2 mm (67 mm sodium chloride was replaced with equimolar choline chloride) produced no change in the holding current or in the amplitudes of db-cyclic GMP-induced inward and outward currents. Similarly, raising the extracellular
566
T. NISHIMURA, T. AKASU AND J. KRIER
potassium concentration in the external solution to 10 mm did not change either the holding current or the amplitudes of db-cyclic GMP-induced inward and outward currents. Statistical data are shown in Table 1. These results suggest that calcium ions are the main charge carrier for the inward current produced by db-cyclic GMP. Effect of calcium channel antagonists Superfusion of the ganglia with cobalt (1 mm) depressed the inward current and completely inhibited the outward current caused by db-cyclic GMP (data not TABLE 1. Calcium-dependent properties of the inward current produced by db-cyclic GMP Amplitude (%) Ionic dependence Calcium-free solution 31+5 (t= 6)* Low-sodium (26-2 mm) solution 95 + 4 (n = 4) High-potassium (10 mM) solution 97 + 3 (n = 3) Calcium antagonists 36 + 8 (n = 4)* Nicardipine (10 ,#M) w-Conotoxin (2 /iM) 93 + 4 (n = 3) * P < 0-001 (Student's t test). Number of experiments is shown in parentheses.
shown). The administration of nicardipine (10 gM), a calcium channel antagonist, depressed the amplitude of the inward current produced by db-cyclic GMP. From four experiments, nicardipine produced 64+ 8% depression of the inward current (Table 1). In contrast, w)-conotoxin (2 gim), a calcium channel antagonist known to block voltage-dependent calcium channels in vesical neurones (Akasu et al. 1990a), did not alter the inward current caused by db-cyclic GMP (Table 1). Chloride component of the inward current In calcium-free solution containing 1 mm cobalt, db-cyclic GMP (100 /LM) produced a 1-2 nA inward shift of membrane current and decreased the slope of the I-V curve. The estimated reversal potential of the calcium-insensitive component of the inward current as measured from the intersection of the I-V curves was -13 mV (-20 + 6 mV, n = 5) (Fig. 4B). The value for the reversal potential of the inward current was near the equilibrium potential for chloride (see Akasu et al. 1990 a). The calcium-insensitive inward current was increased in amplitude in low-chloride solution, where 67 mm sodium chloride was replaced with equimolar sodium isethionate (Fig. 4 C). 4-Acetamido-4'-isothiocyanostilbene-2,2'-disulphonic acid (SITS) has been reported to block chloride current (Inoue, 1985; Gray & Ritchie, 1986; Bader, Bertrand & Schlichter, 1987; Akasu et al. 1990a). Addition of SITS (500 gM) to the calcium-free solution completely inhibited the calcium-insensitive component of the inward current (Fig. 4D). These results suggest that a chloride current may be involved in the db-cyclic GMP-induced inward current.
Voltage-dependent calcium channel currents Calcium current (ICa) Voltage-dependent calcium current (ICa) was recorded from caesium-loaded neurones superfused with a Krebs solution containing TTX (300 nM) and TEA (50 mm). Application of depolarizing voltage jumps from -60 to -10 mV evoked an
CYCLIC GMP AND CALCIUM CHANNELS 567 inward calcium current with amplitude of 0-8-3-5 nA followed by an inward tail current (Fig. 5). This tail current is produced by deactivation of voltage-dependent chloride channels (Akasu et al. 1990 a). Db-cyclic GMP (50 /IM) depressed both the ICa and tail current during the inward current (Fig. 5). When db-cyclic GMP was A
db-cyclic GMP
;
t
L
I
J
I
L
l
dI! t L
B
L
il
Il
a m* A
b.
i
7
li
-
_
_
.
!
'LI1 I
cd it, __> C -
I
III
I
Il1 nA I
MV I "IbUIfn\1
b 200 ms
-10 ;
Vh -60 mV Fig. 5. A, effects of db-cyclic GMP (50 ,sM) on the ICa evoked by depolarizing voltage jumps from a holding potential of -60 to -10 mV. Modified Krebs solution contained TTX (300 nM), TEA (50 mM) and 2-5 mm calcium. The neurone contained caesium. Upper and lower traces represent membrane current and holding voltage (V,), respectively. B, expanded records of ICas and tail currents. Records a-d were obtained at the times marked by respective letters in A. Records d were taken 20 min after removal of db-cyclic GMP. A and A indicate the ICa and the chloride tail current, respectively. Horizontal lines in current traces indicate current at the holding potential (-60 mV).
removed from the superfusing solution, the ICa and tail currents recovered and then increased in amplitude as long as the outward current occurred (Fig. 5). Barium current (IBa) To block voltage-dependent chloride tail currents, barium ions (2-5 mM) were substituted for calcium ions in the superfusing solution (Akasu et al. 1990 a). Application of depolarizing voltage jumps from -60 to -20 mV evoked an inward current with decay time slower than that of ICa in the presence of barium (2-5 mM) (Fig. 6). It has been reported that the IBa was blocked by cobalt (1 mM) or Wconotoxin (500 nM) (Akasu et al. 1990 a; Akasu, Tsurusaki & Tokimasa, 1990b). However, the IBa was not blocked by application of nicardipine (10 JM) (n = 3). Application of db-cyclic GMP (5 /M-1 mM) in barium-containing solution caused an inward current (amplitude, 0-2-4 nA) associated with an increase in membrane conductance followed by an outward current (0 1-1 8 nA) associated with a decreased membrane conductance. The IBa decreased in amplitude during the db-cyclic GMP (100 puM)-induced inward current (Fig. 6Bb). Partial recovery of the IBa was evident in the continued presence of db-cyclic GMP. In contrast, the IBa was augmented during the occurrence of the outward current (Fig. 6Bc). The potentiation of the IBa lasted for a time period of approximately 20 min in db-cyclic GMP-free solution (Fig. 6 C). At a concentration of 100 /tM, db-cyclic GMP produced 64 + 20°% (n = 8) depression followed by 60 + 10% (n = 8) facilitation of the IBa. 19
PHY 457
568
T. NISHIMURA, T. AKASU AND J. KRIER
Figure 7 shows the effect of db-cyclic GMP on the I-V curve for 'Ba. The neurone was initially held at -50 mV and subjected to voltage jumps between - 100 and +50 mV. IBa appeared near -30 mV and the peak 'Ba occurred at -10 to + 10 mV. Db-cyclic GMP decreased the amplitude of the IBa at potentials from -30 to db-cyclic GMP
A
a
d 12 nA ~~~c
~
~
b
0Q mV a
B
c
b
d
60s 1
Vh-60 mV
___---
2 nA
-300 ms
-
C
-5 -4
-3
db-cyclicGMP
-
.
0 0. @
*°e@*D.............. - ~ ~~~~ ------------2~ ~~~
~~~~~00000 -2~~~~~oo .000 2
-1-
0
5
10
15 20 Time (min)
25
30
35
Fig. 6. Effect of db-cyclic GCMP (100 /M) on the IBa (see Methods) evoked by depolarizing voltage jumps (500 ms duration) from a holding potential of -60 mV to -20 mV. The neurone contained caesium. Modified Krebs solution contained TTX (300 nM) and TEA (50 mM). Calcium (2-5 mM) was replaced with equimolar barium. A, horizontal bar indicates period of application of db-cyclic GMP (100 aM). Record d was taken 20 min after withdrawal of db-cyclic GMP. Downward deflection in lower trace represents hyperpolarizing voltage jump from -60 to -100 mV. B, expanded records of 4BaS (upper traces) and voltage commands (lower traces) taken at the times marked by respective letters in A. C, time course of the inhibition and facilitation of the IBa produced by dbcyclic GMP. The period of application of db-cyclic GMP (100,UM) is indicated by the horizontal bar. Ordinate and abscissa indicate the peak amplitude of the net 'Ba and time, respectively. Each point represents the average of three consecutive 'Bas.
+ 40 mV without causing any change in the threshold or the peak voltage for IBa activation. Figure 7A shows the I-V relation of the IBa obtained by subtracting inward currents produced by hyperpolarizing voltage commands of magnitude equal to the depolarizing commands used to evoke the IBa at various command potentials. Db-cyclic GMP consistently reduced the IBa at command potentials from -30 to + 40 mV during the inward current (Fig. 7A). Figure 7B shows the I-V relation of the IBa for the same neurone during the dbcyclic GMP-induced outward current. The IBa increased in amplitude at command potentials between -30 and + 40 mV. The threshold voltage for the IBa and the voltage for the peak current of the IBa were not changed during the facilitation.
CYCLIC GMP AND CALCIUM CHANNELS A Inward current -50
+50 mV
0 ,.1
db-cyclic GMP
I
B Outward current -50 0 .
A
569
+50 mV
.
/,
/
-
Control -
Control
-4nA
-
-4
0
db-cyclic GMP -\ , -6 nA
Fig. 7. Current-voltage relation (I-V curve) obtained during inward current (A) and outward current (B) produced by db-cyclic GMP (50 gM). The I-V curves were obtained by step commands with duration of 200 ms from a holding potential of -50 mV. The superfusing solution contained caesium (2 mM) to block inward rectification. IBas were obtained from subtraction of inward current produced by hyperpolarizing voltage commands of magnitude equal to depolarizing commands. 0, control I-V curve obtained before application of db-cyclic GMP. A and 0, I-V curves obtained during the db-cyclic GMP-induced inward and outward currents, respectively. A
8-Bromo-cyclic GMP _
2
GMP
Cyclic GMP i
-10
Vh
B
Forskolin
-62 mV
8-Bromo-cyclic AMP
-I 1 0
Vh
50ms
-85 mV
Fig. 8. A, comparison of the effects of 8-bromo-cyclic GMP (100 pm), cyclic GMP (1 mM) and GMP (1 mM) on the IB.. The IBas were evoked by depolarizing voltage jumps to -10 mV from a holding potential of -62 mV. B, effects of 8-bromo-cyclic AMP (1 mM) and forskolin (10 FM) on the IBas evoked by depolarizing voltage jumps to 0 mV from a holding potential of -65 mV. In each panel, record 1 was obtained before application of drugs. Records 2 and 3 were obtained during application of drugs and 5-10 min after withdrawal of drugs, respectively. A and B were obtained from different neurones.
Related compounds of cyclic GMP The effects of db-cyclic GMP on the IBa were compared with other guanosine derivatives on the same neurone (Fig. 8). Bath application of 8-bromo-cyclic GMP (10-300 pM) caused an inhibition followed by a facilitation of the IBa (Fig. 8A). The efficacy of 8-bromo-cyclic GMP (100 pM) was essentially the same as that of db-cyclic GMP (100 #M) (n = 5). Cyclic GMP (1 mM) was less potent than db-cyclic GMP (100 pM) in producing the inhibition of the IBa (n = 3) (Fig. 8 A). For this neurone, cyclic GMP showed minimal facilitation of the 1Ba (Fig. 8A). GMP (1 mM), GDP (1 mM), GTP (1 mM) and guanosine (1 mM) had no effect on the IBa (n > 2 for each 19-2
12-~ ~3 4n~
T. NISHIMURA, T. AKASU AND J. KRIER
570
agent). IBMX potentiated the effects of db-cyclic GMP on the IBa (not shown). 8Bromo-cyclic adenosine monophosphate (1 mM) and forskolin (10 gM) did not change the IBa (n = 4) (Fig. 8 B). A
Control 1~r
_
Wash 5 min db-cyclic GMP _, 2 t
