A G Protein Mediates the Inhibition of the Voltage-Dependent Potassium M Current by Muscarine, LHRH,...

European Journal of Neuroscience, Vol. I , pp. 529-542 @European Neuroscience Association 0953-81 &/89 $3.00

A G Protein Mediates the Inhibition of the Voltage- Dependent Potassium M Current by Muscarine, LHRH, Substance P and UTP in Bullfrog Sympathetic Neurons

H. S. Lopez and P. R. Adams Howard Hughes Medical Institute, Department of Neurobiology and Behavior, SUNY at Stony Brook, Stony Brook, NY 11 794, USA

Key words: whole-cell, voltage-clamp, neurotransmitter, modulation

Abstract

The involvement of G proteins in the transduction mechanism of M current (I,) inhibition by extracellular ligands in bullfrog sympathetic neurons was examined using the hydrolysis resistant nucleotide analogues GTPyS and GDPPS. I, was recorded in large (40 - 60 prn) isolated neurons using the patch-clamp technique in the whole-cell configuration, as well as in neurons from the intact ganglion impaled with conventional microelectrodes. In whole-cell recordings I, could be recorded without significant loss for 1 h or more provided ATP was present in the patch pipette. Muscarine, D-Ala'-LHRH, substance P and UTP reversibly inhibited I, in isolated control neurons, with full and rapid recovery of the current following agonist washout. Dialysis of isolated neurons with various concentrations of GTPyS (1 - 100 pM) affected, in a dose-dependent manner, the recovery of I, after its inhibition by brief agonist application. With 50 pM GTPyS, I, inhibition became completely irreversible. Similarly, the reversibility of I, inhibition by muscarine was reduced or abolished by the iontophoretic injection of GTPyS through a second microelectrode into neurons of the intact ganglion. GTPyS by itself caused a slow, agonist-independent suppression of I, in dialysed neurons, thus mimicking agonist action. Dialysis of isolated neurons with GDPpS (100-500 pM) attenuated by half or more the magnitude of I, inhibition by agonist as compared to control neurons. In addition, GDPPS attenuated the response of a given neuron to muscarine and D-Ala'-LHRH, and caused slow increase of I,, as a function of dialysis time. Incubation (2 - 72 h, 4 - 36OC) of isolated neurons or intact ganglions with activated pertussis toxin had no effect on the response to muscarine. Toxin injections to experimental animals were equally ineffective. In contrast to I,,,, the additional inward current with increase in conductance induced by muscarine and D-Ala'-LHRH reversed with agonist washout in GTPyS-dialysed neurons, although more slowly than in control neurons. The results in this study indicate that a G protein, possibly pertussis toxin- insensitive, provides a common coupling step linking muscarinic, substance P, D-Ala'-LHRH and UTP receptors to the inhibition of M current.

Introduction

The excitability of bullfrog ganglion B cells is greatly influenced by chemical signals that modify the properties of membrane ionic conductances (Jones and Adams, 1985; Adams et al., 1986). Inhibition of the voltage-dependent potassium M current (I,) (Adams et al., 1982a) by muscarinic agonists, luteinizing hormone-releasing hormone (LHRH) analogues, tachykinin peptides and nucleotides (Adams and Brown, 1980; Adams et al., 1982b, 1983; Akasu et al., 1983; Jones et al., 1984; Jones, 1985) causes a remarkable increase in membrane excitability to subsequent excitatory input. Thus, while B cells having an intact M current exhibit strong accommodation, rarely firing more than a few spikes during prolonged direct stimulation, inhibition of

I,,, allows sustained repetitive firing during stimulation (Adams et al., 1982b; Jones, 1985). Chemical signals that modulate I, are thus potentially capable of modulating the input-output behaviour of the cell. The physiological significance of the modulation of I, is apparent from the fact that acetylcholine and a peptide structurally similar to teleost fish LHRH-two I, inhibitors-are endogenous neurotransmitters in ganglionic synapses (Blackman et al., 1963; Kuba and Koketsu, 1978; Jan et al., 1979, 1980; Kuffler, 1980; Eiden and Eskay, 1980; Eiden et al., 1982; Jan and Jan, 1982; Sherwood et al., 1983). Indeed, inhibition of I, by synaptically released acetylcholine and the LHRH analogue is the major ionic basis for the slow and late slow

Correspondence to: H. Lopez, as above Received 7 March 1989, accepted 5 May 1989

530 G proteins couple M current inhibition

excitatory postsynaptic potentials (epsps) of B cells (Jan et al., 1979, 1980; Adam and Brown, 1982; Kuffler and Sejnowski, 1983). As predicted from the inhibition of I,, these slow synaptic potentials profoundly affect the response pattern of B cells without eliciting spikes by themselves (Jones, 1985; Adams et al., 1986).

Inhibition of I, by extracellular ligands is likely to involve an indirect coupling process since activation of various separate surface receptors inhibit the same conductance. The slow time course of the slow epsp may also reflect a multistep coupling mechanism. In this regard, guanine nucleotide binding proteins (G proteins) have been shown to couple muscarinic receptor-induced cellular responses in a variety of systems (for review see Gilman, 1987), including acetylcholine muscarinic receptors to potassium channels in heart (for review see Brown and Birnbaumer, 1988). In addition, it is becoming increasingly clear that G proteins are involved in the coupling of a variety of transmitter receptors to ion channels in nerve cells (Dunlap et al., 1987; Nicoll, 1988).

Previous results, communicated in abstract form (Lopez et al., 1987). indicated that coupling of muscarinic receptors to the M conductance in bullfrog sympathetic ganglion cells occurs via a G protein. Here we document in full the involvement of G proteins in the transduction mechanism of I, inhibition by muscarine, LHRH, uridine triphosphate (UTP) and substance P (SP) in isolated sympathetic neurons as well as, for muscarinic inhibition of I,,,, in neurons of the intact ganglion. Our data accord with results from studies simultaneously pursued by other laboratories using isolated sympathetic ganglion cells from the grass frog (Rana pipiens) (Pfaffinger, 1988), bullfrog (Jones, 1987) and rat (Marrion, 1987).

Materials and methods

We recorded the M current (I,) in voltage-clamped neurons of both the intact and dissociated sympathetic ganglion of adult bullfrogs (Rana catesbeiana) using either conventional intracellular recordings, or the patch-clamp technique in the whole-cell configuration (Hamill et al., 1981).

Experiments in dissociated cells The lumbar VIIIth, IXth and Xth sympathetic ganglia were dissociated following the procedure of Kuffler and Sejnowski (1983). Briefly, after isolation and cleaning of connective tissue, the ganglia were cut into small pieces and incubated for 30 min at 37°C in Ca2'/Mg2+-free Ringer enzyme solution containing 1 m g / d collagenase (Worthington 4182, CLS 111, 132 and 102 plrng) and 5 m g / d dispase (Boehringer Mannheim), and then mechanically disrupted by repeatedly passing the tissue through the tip of a fire-polished Pasteur pipette. Complete dissociation of the ganglion cells was achieved after a second digestion for 30 min at 37°C in enzyme solution without collagenase, followed by mechanical disruption as indicated above. After washing the cells three times in Ringer solution, the isolated cells were plated in plastic Petri dishes, and stored at 4°C in either Ringer solution (in mM, 120 NaCI, 2.55 KCI, 2 CaC12, 10 MgCI2, 2 Tris; pH 7.2) or L-15 medium (with added 10% fetal calf serum, 0.2% glucose and 1 mM Ca). Cells kept in Ringer solution were used within 4 days, and we found no detectable differences from those maintained in L-15 medium. In about half of the dissociations, minor modifications of the procedure were done. Thus, a single incubation in the collagenase plus dispase enzyme solution with gentle trituration every 10 min produced a better yield of healthy looking cells in a shorter time. In order to maximize the chances of recording from B cells, we selected the largest cells in each

dish for whole-cell voltage-clamp of the M current (typically 4 0 ~ 60 pm, and sometimes larger). Patch pipettes were pulled from soft glass (Drummond Microcaps, Drummond Scientific Co.) using a PP-83 Narishige patch microelectrode puller. Microelectrode tip resistance was usually 2-3 MQ. A few experiments used pipettes up to 5 MQ. The formation and stability of seals was the same in unpolished and fire-polished pipettes (Microforge from Nanshige Instruments). Pipettes in control cells contained (in mM) 100 KCI, 4 MgCl, 2.5 Na Hepes, 1 EGTA (potassium salt) and 5 Na,ATP. In a different group of cells, 1, 10,20, 50 and 100 pM guanosine 5'-0-3-thiotriphosphate (GTPyS, tetralithium salt, Sigma) or 100 and 500 pM guanosine Zf-0-2-thiodiphosphate (GDPPS, trilithium salt, Sigma) were added to the control solution. After gigaseal formation and attaining the whole- cell configuration for patch recording, the cells were continuously perfused with frog Ringer (in mM, 115 NaCI, 2.5 KCI, 4 CaCI2, 2.5 Tris; pH 7.2). We measured the reduction and recovery of I, following brief bath applications (20 -60 s) of muscarine ( I pM) (Sigma), D-Ala6-LHRH (5 pM) (Peninsula), substance P (50 pM) (Bachem), and UTP (50 or 100 pM) (Sigma) in control cells and in cells dialysed with GTPyS and GDPOS. The effect of agonist on I, and its recovery were quantified off-line by measuring the amplitude of the slow inward relaxations at the end of the hyperpolarizing voltage steps caused by the voltage- and time-dependent closure of a fraction of M channels, previously opened at the holding potential (Adams et al., 1982a). Except for a few experiments using an EPC7 (LIST Medical Electronics) patch amplifier, all recordings were made with an Axoclamp IIA (Axon Instruments, Inc.) amplifier in a discontinuous single-electrode voltage-clamp (dSEVC) protocol (Finkel and Redman. 1984) in order to minimize clamp errors due to series resistance. A cycling rate of 3 - 10 KHz was used. These rates are adequate since the sampling period both greatly exceeds the membrane time constant (10 ms or larger) and provides enough sampling of the slow exponential relaxations (fastest tau at 22°C is about 10 ms; A d a m et al., 1982a). Capacity compensation was continuously monitored on a second oscilloscope and adjusted as necessary to ensure complete head-stage settling. Hyperpolarizing 20-30 mV voltage steps were applied for 1 s from a holding potential of -30 or -40 mV. Current-voltage curves were made by stepping to hyperpolarizing voltages from -30 mV. Voltage and currents records were displayed on an oscilloscope and continuously recorded on chart paper (Could 2400 recorder).

Experiments in the intact ganglion Isolated ganglia from the paravertebral lumbar sympathetic chains were treated with 1 % trypsin (type 111, Sigma) for 8 - 10 min, cleaned of connective tissue, and then pinned in a recording chamber. The isolated ganglia were continuously perfused with either normal frog Ringer solution or an identical solution containing 1 pM muscarine. Large neurons, presumed to be B cells, were impaled, under direct visual control (Wild stereomicroscope), with 3 M KCI-filled, 20-40 MQ, glass microelectrodes (borosilicate glass, 1 mm I.D.) and then voltage- clamped using a dSEVC protocol (see previous section). The effect of brief applications of 1 pM muscarine on control responses and their recovery following washout were recorded. After switching to current clamp mode, the cells were impaled with a second pipette containing 100 mM GTPyS in distilled water. and a negative current ( - 2 to -4 nA) was passed for 2 min. This pipette was then removed and. after switching back to dSEVC, the effect of 1 pM muscarine on I,, was retested. In several control experiments a second 3 M KCI-filled electrode was used to pass negative current. The reduction and recovery

G proteins couple M current inhibition 531

potentials were indistinguishable from those recorded in intact ganglion neurons, except for missing, or very small afterhyperpolarizations. The other conspicuous difference from intact ganglion cells was that at sufficiently high stimulation intensities, long depolarizing pulses evoked sustained repetitive firing in most cells tested. The data presented are from cells with a membrane potential of -45 mV or more negative, and able to fire full size action potentials.

In neurons voltage-clamped at a holding potential of -30 or -40 mV, time-dependent inward and outward M current (I,) relaxations were recorded in response to the onset and offset of long hyperpolarizing voltage steps (Adams et al., 1982a). These hyperpolarizing steps lasted I s and were applied every 10 s. Figure 1A illustrates typical whole-cell records from a dissociated cell patched with an electrode containing the control solution. In this cell, the membrane potential was commanded to -60 mV from a holding potential of -30 mV. A fraction of the M conductance slowly switched off at the command

of I, following the application of rnuscarine before and after the intracellular injection of GTPyS or control currents were quantified off-line as indicated in the previous section.

Results

Experiments were performed with both dissociated cells using the whole-cell recording method (Hamill et al., 1981), and in intact ganglia using microelectrodes (Jones et al., 1984). The data obtained with whole-cell recordings will be described first.

Dissociated cells

Muscarine and LHRH. Immediately after establishing the whole-cell configuration, the cells were briefly examined in current clamp. The resting membrane potential usually varied between -50 and -70 mV. Brief depolarizing pulses evoked fast spikes, repolarizing to resting potential in 2 ms or less. Spike size was 100 mV or larger. These action

A

mV -30 -60

B I

0

r -- 1111-- - L-_--i 30 sec 1 sec

C mV

-100 -a0 -60 -40

f l 0 O t + + * I 9 O

1 1 50

FIG. I . Whole-cell recording of M current in voltage-clamped dissociated bullfrog ganglion neurons.

(A) Current (top) and voltage (bottom) records from a neuron clamped at -30 mV and stepped to -60 mV during 1 s, every 10 s. The deactivation of I, at the command potential caused the slow inward current relaxation after the initial instantaneous step reflecting the membrane conductance at -30 mV. On repolarization, reactivation of I, caused the slow outward relaxation. The smaller instantaneous step on repolarization reflects the smaller membrane conductance at -60 mV. I, was quantified by measuring the amplitude of the inward relaxation (*).

(B) The average percentage ( f SD) of initial I, in nine cells (except last two points where n=7) was plotted as a function of time in the first 30 min of whole-cell recording. I, could be recorded without significant loss for at least 1 h.

(C) Steady-state I-V curves of I, from a single cell at 5 ( O ) , 10 (0) and 20 (V) min of whole-cell recording, constructed by plotting the displacement current at the end of the command voltage pulse as a function of membrane potential. The outward rectification reflects the voltage-dependent activation of M conductance. No shifts in the voltage dependence of I, were observed. M conductance at -30 mV was 20 nS (reversal potential for I,= -88 mV). Leak conductance was 10 nS as estimated from the slope of the linear part of I-V curve, negative to -70 mV (linear I-V extrapolation yields a leak reversal potential of about - 10 mV).

A m u s 1 p M wash

1 sec

FIG. 2. (A) M current responses to 1 pM muscarine of a control neuron (top records) and a GTPyS-dialysed (50 pM) (middle records) neuron clamped at -30 mV and commanded to -50 mV for I s, every 10 s. In addition, current responses elicited by commanding the membrane to -90 mV from a holding potential of -70 mV (I, is deactivated at these potentials) were used to estimate changes in leakage of the GTPyS-dialysed neuron (bottom records). From left to right, the columns illustrate M current responses before, during a muscarine application lasting 50 s, and after washout. I, inhibition produced an inward shift in holding current, a decrease in the amplitude of the slow current relaxations and a decrease in the size of the instantanous current step at the onset of the command potential (see text). All these changes fully reversed in the control cell. In addition, I, grew bigger than control in the early stage of washout. In contrast, almost no recovery of I, occurred in the GTPyS-dialysed neuron. Rightmost middle and bottom records were taken after 20 min of washout. In these cells, the response to muscarine had n o appreciable leakage component.

(B) M current responses of a control (top records) and a GTPyS-dialysed neuron (50 pM) (bottom records) to a brief application of 5 pM D-Ala6-LHRH (indicated by horizontal bars). The experimental protocol was similar to that in (A), except that the GTPyS-dialysed neuron was commanded to -60 mV. Slow speed records (10 times as slow) show the time course of the change in holding current during agonist application and washout. D-Ala6-LHRH caused I, inhibition in both cells. I, fully recovered in the control cell (with a recovery overshoot), but remained inhibited in the GTPyS-dialysed neuron, even after extensive washout of the agonist (20 min).


potential causing the slow inward relaxation. On repolarization, I, redeveloped causing the slow outward relaxation.

With ATP added to the patch pipette, I, was stable over time (Fig. lB), and could generally be recorded without significant loss for as long as the neurons conserved a healthy morphological appearance. While most of our experiments were completed within 30-40 min, the stability of I, allowed for reliable records extending for I h or more. The onset of significant spontaneous loss of I, invariably coincided with signs of rapid cell deterioration, such as severe swelling or extensive vacuolation. A few (n =4) healthy-looking cells showed a moderate decrease of I, (about 20% on average) within 30 min of recording. GTP in the patch pipette solution was not required for I, stability. In several cells, the steady-state current -voltage (I -V) curve was examined at different times during the recording session. These I-V curves, of which examples are shown in Figure lC, were constructed by plotting the steady-state inward shift of the holding current at the end of hyperpolarizing voltage commands (up to 100 mV), from a holding potential of -30 mV. The outward-going rectification from about -70 mV towards more depolarized potentials reflects the voltage-dependent opening of M channels in this range. As illustrated by the I-V curves of the neuron in Figure lC, made at 5, 10 and 20 min from the establishment of the whole-cell configuration, no voltage shifts or any other detectable changes were observed. Occasionally, a slight increase of leakage over time resulted in a small offset of the whole I-V curve in the negative direction of the current axis.

A brief bath application of 1 pM muscarine (Fig. 2A) or 5 pM D-Ala6-LHRH (Fig. 2B) inhibited I,, so producing three characteristic effects (Adams et al., 1982b): (i) an inward shift in the holding current resulting from a reduction of a steady K current flowing through M channels at the holding potential; (ii) a reduction in the amplitude of the inward and outward current relaxations following the onset and offset of the command voltage steps; and (iii) a reduction in the size of the instantaneous inward current step at the onset of the hyperpolarizing pulse due to the reduced membrane chord conductance. The amplitude of the inward current relaxations, which we measured to quantify I,, was on average reduced 48% ( f 14 SD, n=5) by 1 pM muscarine, and 83% (*7 SD, n=10) by 5 pM D-Ala6-LHRH. All these effects readily and fully reversed soon after washout of the agonists with frog Ringer control solution (Fig. 2; see average data in Fig. 5). The sensitivity of I, to these concentrations of both agonists was similar to that found in the intact ganglion with conventional microelectrode recording (Adams et al., 1982b; Jones, 1985). In the majority of the cells, I, transiently grew bigger than control in the earlier stage of washout, and this was accompanied by a net outward current at the holding potential (Figs 2A and B, and 9B). This phenomenon is not exclusive to dialysed cells. Less frequently and prominently, it is also observed in neurons of the intact ganglion recorded with microelectrodes (not shown; see an example in Fig. 5 of Jones, 1985). The overshoot of 1, during recovery was rarely maintained, relaxing back to preagonist levels within about 2 min, and occasionally past control, resembling slow oscillations. It was also observed in cells dialysed with the lowest GTPyS concentration (Fig. 4), and with GDPOS (not shown). The reported average data on I, recovery are from records taken once the transient overshoot appeared to decay to a steady-state value.

The response to agonist was as stable over time as I, itself. Thus, repeated sessions of inhibition and subsequent recovery of I, could be readily accomplished by repeated brief bath applications of 1 pM muscarine or 5 pM D-Ala6-LHRH, separated by periods of washout

% 100

P r e a g Im 50 0 n i

t 0 S

mus

1

L

D - A I a:- L H R H

I r! FIG. 3.lnhibition and recovery of I, following several applications of I pM muscarine (left) or 5 p M D-Ala6-LHRH (right). The bars indicate the average amplitude of I,, ( *SD). expressed as percentage of preagonist 1,. The extent of I, inhibition was similar in subsequent applications of each agonist. Also, I,, fully recovered after every single agonist application. Last agonist application was done 20-45 rnin (muscarine) and 20-60 rnin (D-Ala6-LHRH) after the initiation of recording.

(Fig. 3). There was essentially a similar extent of inhibition as well as full recovery of I, following the termination of every single application of either muscarine or D-Ala6-LHRH. Even in the few neurons showing the small and slowly progressing decrease of I, previously mentioned, I, was equally inhibited and recovered in full after each brief application of agonist. Thus, no essential components for either inhibition or recovery of I, were lost in our dialysing conditions. The extent of I, inhibition by 5 pM D-Ala6-LHRH was consistently larger than that by 1 pM muscarine (Fig. 3).

In addition to the inward current due to the inhibition of the M conductance, about half of the control cells responded to both 1 pM muscarine and 5 pM D-Ala6-LHRH with a second component of inward current of variable extent, usually reversing in full after washout of the agonists. The characteristics of this additional inward current response corresponded to those described in other reports (Adams et al., 1982b; Katayama and Nishi, 1982; Jones et al., 1984; Jones, 1985: Kuffler and Sejnowksi, 1983). Thus, unlike I, inhibition, it involved an increase in conductance, and instantaneously appeared on hyperpolarization at the command voltage with no detectable time-dependent relaxations. Also, it developed and disappeared with a slower time course than that of I, inhibition. The additional inward current was also observed in cells dialysed with nucleotide analogues (Fig. 7B).

GTPyS. When the dissociated neurons were dialysed with GTPyS, by adding the nucleotide to the patch pipette solution before making the seal, the recovery of I, following its inhibition by a brief application of either muscarine or D-Ala6-LHRH was reduced or abolished altogether. Thus, in sharp contrast to control cells in which I, rapidly and fully recovered following washout, brief bath application of the same concentration of muscarine (Fig. 2A) and D-Ala6-LHRH (Fig. 2B) to neurons patched with pipettes containing 50 pM GTPyS caused an almost irreversible inhibition of I,. The M current relaxations now remained inhibited after the cessation of muscarine or peptidergic receptor stimulation, and did not recover even after prolonged washout (up to 40 min). Consistently, the inward shift in holding current induced by both agonists did not reverse either since it resulted from the

D-Ala*-LHRH 10 p M GTP-y-S

- 1 min 11 n A

- - 20 P M GTP-Y-S -

~

"1 111-

Tfh,, , , , , . .,, , , , , , - , , , , , , m , r , , , , , , , , , , , , , , , , , , ,

11 nA 50 p M GTP-y-S -

m 1 0 . 5 nA

FIG. 4 .Concentration-dependency of the effect of GTPyS on the recovery of I,, after inhibition by D-Alah-LHRH. Membrane potential was held at -30 mV and commanded to -50 mV for I s, every 10 s. Only slow records are shown. Several applications of 5 pM D-Ala6-LHRH (horizontal bars) were done in cells dialysed with 10 (top), 20 (middle) and 50 pM (bottom) GTPyS. Since the responses had no leakage component, the inward shift in holding current due to the reduction of a steady K current flowing through M channels at the holding potential accurately reflects I, inhibition (Adams et al.. 1982; see also text). Ten micromolar had no effect on the peptide response, except for a suppression of the overshoot of I,, immediately following washout in the second application of the agonist. I, recovery was partially and totally arrested by 20 and 50 FM GTPyS, respectively. Notice that after partial recovery I,, is normally inhibited by agonist.

inhibition of a steady K current flowing through M channels at the holding potential. This point is particularly clear in the examples of Figure 2, where the leakage contribution to the muscarinic and peptidergic response is negligible.

The effect of GTPyS on the reversibility of muscarinic and peptidergic inhibition of I, was dose-dependent. This dependency was examined by testing muscarinic response in neurons dialysed with 1, 10, 20, 50 and 100 pM GTPyS, and D-Ala6-LHRH responses in neurons dialysed with 10,20 and 50 pM GTPyS. Figure 4 shows actual responses from cells dialysed with different concentrations of GTPyS, while the bar plots of Figure 5 summarise the results. With 10 pM GTPyS in the patch pipette, there was full reversion of I, inhibition by either agonist. Indeed, full recovery of I, occurred after each of several agonist applications, as Figure 4 illustrates for D-Ala6-LHRH responses. Increasing the GTPyS concentration in the patch pipette to 20 pM resulted in about half-maximal inhibition of the reversibility of both responses, which otherwise looked normal. As before, multiple brief applications of agonist repeatedly inhibited I, but now, as exemplified by peptidergic responses in the slow records of Figure 4, I, only partially recovered after each agonist washout. This concentration of GTPyS was ineffective in two cells (not included in the average). As illustrated in the records of Figure 2, the effect of muscarine became nearly totally irreversible with 50 pM GTPyS (or 100 pM, not shown), with an average percentage of recovery of 5 %. With 50 pM GTPyS in the pipette, the percentage of I, inhibition by D-Ala6-LHRH was rather variable but always powerful (Figs 2 and 4), ranging from 84% to loo%, and very poorly reversing, with an overall percentage of recovery of 14%. In contrast to the results with lower GTPyS concentrations, no significant I, recovered even after a single application of agonist, as illustrated by the slow records of Figure 4 for D-Ala6-LHRH. In this figure, the inward shift in holding current accurately reflects the magnitude of I, inhibition (Adams et al., 1982b) since the leakage component is negligible. Comparison with the changes in the inward relaxation from the corresponding fast records showed that this was so. For example, assuming a reversal potential

%

I m

i n h i b i t i 0 n

%

Irn

r e C 0 V e r Y

Ioo1


(51 (101 (5) (6) (6) (6) (7) (6)

10 20 50

[GTP-')'-S] i n p a t c h p i p e t t e (1M)

100

80

60

40

20

10 20 50

[GTP-Y-S] i n p a t c h p i p e t t e (UM) FIG 5 . Concentration dependency of the effect of GTPyS on muscarine and D-Alah-LHRH. Upper plot: average percentage ( kSD) of I, inhibition by 1 pM muscarine (clear bars) and 5 pM D-Ala6-LHRH (hatched bars) as a function of GTPyS concentration. The number of cells is shown on the top of the bars. Bottom plot: percentage of recovery (*SD) of I, after washout of muscarine (clear bars) and D-Ala6-LHRH (hatched bars) as a function of GTPyS concentration. Percentage of recovery is defined as the absolute value of (I,,, increase after washout/I, decrease during agonist) x 100. The number of cells is indicated in the upper plot.

of -90 mV, the changes in conductance during the three applications of agonist illustrated in the middle records agreed very well, in nS, (-36,-36), (-23,-21) and(-7,-7), thefirstnumberineachpair estimated from the inward shift in holding current, and the second number from the change in the inward relaxation. It is worth noting that the recoverable fraction of I, following agonist washout in GTPyS-dialysed cells remained fully sensitive to subsequent agonist inhibition, though with only partial recovery after each test (middle record in Fig. 4).

The extent of I, inhibition by muscarine and D-Ala6-LHRH was not appreciably changed by dialysis with GTPyS in the range 1-20 pM (Fig. 5). However, some of the cells dialysed with 50 pM GTPyS responded to the agonists with a slightly enhanced inhibitory response,


50 )IM GTP-Y-S 20 min D-Ala6-LHRH 5 clM

A - Fa 2 mM

* * 1 min * * *

-30 7 -50 JmV 1 sec

FIG. 6. Spontaneous decrease in I, in a neuron dialysed with 50 pM GTPyS. Slow (top) and fast (bottom) current records in a neuron held at -30 mV and stepped to -50 mV for 1 s, every 10 s. I, gradually decreased over time in the absence of agonist application. At 20 min of whole-cell recording little I, remained. Dashed lines above the fast records, obtained at the times marked with asterisks, indicate the level of holding current at the beginning of the experiment. No increase in leak was observed in this cell. This neuron measured 32 X47 pm, and access resistance was 3.3 MQ.

m *

which was reflected in a stronger average response as compared to control (Fig. 5) . The GTPyS-dialysed neuron of Figure 2 provides an example. In that cell, 1 pM muscarine inhibited 73% of I,, in contrast to an inhibition of about 50% in control cells as well as in cells dialysed with lower GTPyS concentrations. These results hint at a shift to the left of the dose-response curve for I, inhibition, although the effects of varying the muscarine concentration were not examined.

GTPyS by itself was capable of gradually switching off I,, independently of agonist application. Figure 6 shows an example. The inclusion of 50 pM in the patch pipette produced a slow but steady diminution of I, over time (n=5), with very little current remaining after 20-40 min. The I-V relationship for I, was examined after the reduction of I, seemed complete by commanding to negative and positive potentials from a holding potential of either -30 or -60 mV. No shift in the voltage-dependence of I, was found by this procedure.

In several GTPyS-dialysed cells tested with muscarine or D-Ala6-LHRH, steady-state I -V curves were constructed at different stages of the experiment. The shape of these curves in GTPyS- dialysed neurons (10-50 pM) did not differ from that in control neurons (see Fig. 1) before or during the application of muscarinic or peptidergic agonists (not shown). The inhibition of 1, by muscarine or D-Ala6-LHRH caused a virtually identical diminution of the outward-going rectification positive to -70 mV both in control and GTPyS-dialysed neurons. The recovery of the outward rectification in the latter was a function of GTPyS concentration. Thus, the persistent decrease of I, following the agonist-induced inhibition in GTPyS- dialysed cells was not due to shifts in voltage dependence or any other change in the I-V relationship.

In contrast to the receptor-coupled inhibition response of I,, the presumed direct blockade of the M channels with 2-4 mM Ba (Constanti et al., 1981) was in no way affected by GTPyS (up to 50 pM). Figure 7A shows an example of the response of a single cell to Ba and then to D-Ala6-LHRH, when voltage-clamped with a patch pipette containing 50 pM GTPyS. I, fully recovered following washout of Ba, while it was irretrievably inhibited by D-Ala6-LHRH. Similar results were obtained for muscarine (not shown).

In addition to its effect on I,, GTPyS sometimes enhanced the non-M, additional inward current component of the response to

f 1 #

FIG. 7. (A) Differential efect of GTPyS on the inhibition of I,, by Ba or receptor-coupled agonists. Slow (top) and fast (bottom) current records illustrate the inhibition and recovery of I, in a single neuron dialysed with 50 pM GTPyS. Fast records were obtained at the times marked by asterisks. The inhibition of I, by 2 mM Ba, which directly blocks M channels (Constanti et al., 1981). readily reversed following washout. In contrast. I, was irreversibly inhibited by the subsequent application of D-Alah-LHRH. The last fast record corresponds to 20 min after D-Ala6-LHRH application, and shows a small increase in leakage. Similar differential effects of GTPyS were obtained for Ba and muscarine (not shown).

(B) GTPyS affected the reversibility of the M current component, but not of the additional inward current component of the response to D-Ala6-LHRH and muscarine. Records from a neuron dialysed with 50 pM GTPyS illustrate responses to 5 pM D-Ala6-LHRH consisting of 1, inhibition plus an additional inward current component. The latter is clearly visible as an inward current instantaneously appearing at -50 mV (compare instantaneous current levels at the onset of the hyperpolarizing pulse before and after agonist application). While the additional inward current component reversed about 10 min after agonist washout, I, was irreversibly inhibited (records at later times not shown).

muscarine and D-Ala6-LHRH. Firstly, the frequency of neurons responding with the additional inward current component slightly increased. Secondly, and more noticeably, several cells responded to the application of agonist with an unusually large non-M inward current, especially when dialysed with the highest GTPyS concentration (50 pM). Figure 7B shows an example of such additional inward current response in a cell dialysed with 50 pM GTPyS when briefly challenged with 5 pM D-Ala6-LHRH. Although this enhanced additional inward current still had a delayed onset relative to the inhibition of I,, it showed a faster, as well as larger, increase in leakage, and made a larger contribution to the inward shift of holding current following agonist application. Unlike inhibition of I,, however, this enhanced leakage current response induced by muscarine and D-Ala6-LHRH in GTPyS-dialysed neurons was reversed by agonist washout, although


of I, by D-Ala6-LHRH in the control cell (upper record in A) amounted to 77%, whereas the GDPPS-dialysed neuron responded very poorly to a similar concentration of this agonist, with only a 12% inhibition. The average inhibition of I, in GDPPS-dialysed neurons was only 19% (k7 SD, n=6) for muscarine, and 27% (+ 16 SD, n=5) for D-Ala6-LHRH, about half and one-third of that obtained in control neurons, respectively.

Secondly, some GDPPS-dialysed cells initially showing strong inhibitory responses to muscarine or D-Ala6-LHRH became increasingly insensitive to these agonists over time. Because this effect developed with a slow time course, it provided a situation in which each cell in a way served as its own control. Figure 8B illustrates this observation for muscarine. At 15 min of whole-cell recording 1 pM muscarine inhibited I, as much as in control cells. The inhibition, however, was much poorer after 35 min of whole-cell recording, as reflected by the smaller change in amplitude of the slow current relaxations, and the inward shift of holding current following application of the agonist. Notice that in control neurons (Fig. 3), the extent of I, inhibition by either agonist remained unchanged when tested at different times after the establishment of the whole-cell configuration, including those times at which I, inhibition was clearly depressed in GDPPS-dialysed neurons.

The recordings in Figure 8B illustrate an observation which is relatively common (5/12) in cells in which GDPPS reduced the agonist-induced inhibition of I,, but which is never seen in control or GTPyS-treated cells. I, spontaneously and steadily grew over tens of minutes, as indicated by both larger slow relaxations and larger holding current at the holding potential of -30 mV (compare pre-muscarine records at 15 and 35 min). This finding seems analogous to the increase in the transient component of the calcium channel current reported by Dolphin and Scott (1987) in rat sensory neurons dialysed with GDPDS. This effect of GDPPS developed after, and has a much slower time course than an early increase of I, often seen in the first 2 min or so after establishing the whole-cell configuration. This early increase may be the result (Jones, S., personal communication) of the recovery of I, from the inhibition by ATP (Akasu et al., 1983) leaking from the patch pipette. This might be the case since this early increase is virtually absent when the seals are formed under continuous perfusion, thus diluting any leakage from the pipette. In contrast to the muscarinic and peptidergic responses, I, inhibition by 2 mM Ba was unaffected by GDPPS (not shown). Finally, if GDPPS had any effect at all on the additional inward current response to muscarine and D-Ala6-LHRH, it was in the direction of a less frequent and less intense response of this type.

UTP and substance P. The recovery of I, following inhibition with UTP (50 pM, n=4, or 100 pM, n=3) and SP (50 pM, n=9) was greatly reduced or completely abolished by dialysing the cells with 50 pM GTPyS (not tested at other GTPyS concentrations), while in control cells I, always recovered at least to 80% of control from both UTP (50 pM, n=2, or 100 pM, n=4) and SP (n=4). Examples are shown in the records of Figure 9. The results were very consistent from cell to cell, and indistinguishable from those obtained for muscarine and D-Ala6-LHRH at equally effective concentrations of GTPyS.

A control D-AIa'-LHRH 5 pM

mus 1 pM

35 mln

--L,zj=- FIG. 8. (A) The inclusion of 500 pM GDPDS in the patch pipette significantly reduced the extent of I, inhibition by D-Ala6-LHRH (bottom records), as indicated by the smaller inward current at the holding potential, and larger slow inward and outward relaxations as well as larger instantaneous step at the onset of the command pulse. Top records show responses of control cells for comparison. Similar results were obtained for muscarine.

(B) Attenuation of the inhibitory effect of muscarine over time in a single neuron dialysed with 100 pM GDPPS. In this neuron, the application of 1 pM muscarine at 35 min of whole-cell recording produced much less I, inhibition than the same concentration of agonist at 15 min of recording. In addition, the amplitude of I,,, gradually increased over time as judged from the larger current slow relaxations, and larger holding current at the holding potential (the horizontal mark at the beginning of the records in the second row indicate previous level of holding current). Similar results were observed for D-Ala6-LHRH. In contrast (Fig. 3 ) . the extent of I, inhibition by muscarine or D-Ala6-LHRH over time remained constant in control cells.

recovery was much slower than in control cells. The slow recovery of the agonist-induced leakage increase in GTPyS-dialysed cells meant that if the recording was not sufficiently prolonged, the effect appeared to be irreversible (see statement in Lopez et al., 1987). If the recovery period was sufficiently long (up to 40 min on a few occasions) the leakage effect completely reversed whereas the inhibition of I, showed an irreversible component, whose size depended on the GTPyS concentration.

GDPPS. A number of experiments tested the effect of dialysing the neurons with GDPPS on the muscarinic and peptidergic inhibition of I,. At high concentrations (100 or 500 pM) GDPPS caused a consistent decrease in the sensitivity of I, to both muscarine and D-Ala6-LHRH that manifested itself in two ways. Firstly, after a dialysis time that varied from cell to cell over a wide range (5 -60 min, mostly (8/12) within 20 min) the amount of I, inhibition by 1 p M muscarine (7/10 neurons) and 5 pM D-Ala6-LHRH (5/5 neurons) was smaller in GDPPS-dialysed neurons than in control neurons. The records in Figure 8A illustrate this point. In that example, the inhibition

Intact ganglion The data presented are from 22 cells from the intact bullfrog ganglion voltage-clamped with a single microelectrode using conventional intracellular recordings. These cells tolerated well the double


A UTP 100 pM -

* * *

UTP 100 pM - - I

0.5 sec f

B

i l *

s I /

I _

I

I r i /

.-- I P-

I 0.5 sec

FIG. 9. Effect of GTPyS on the I,, response to UTP (A) and SP (B). In control cells, I, was reversibly inhibited by 100 pM (or SO pM) UTP (A. top) and 50 gM SP (B, top). In neurons dialysed with S O pM GTPyS, UTP (A. bottom) and SP (B. bottom) responses did not reverse. or did so very little, even with prolonged washout (up to 40 min)

impalement and iontophoretic current during GTPyS injection, and looked healthy, as judged by their resting membrane potential (negative to -30 mV after the second impalement) and ability to fire spikes at the end of the experiment. The effect of muscarine on 1, was tested in the same neuron before and after intracellular injection of GTPyS, and thus each neuron served as its own control. Not included in the results are neurons in which, during the control stage, 1, was either unresponsive to muscarine, or did not recover from muscarine to more than 75 % of control.

The upper panel of Figure 10 shows a control muscarinic response in a B cell impaled with a single KCI-filled microelectrode. As described in the previous section, I, was recorded as slow current relaxations elicited by a 1 s command voltage step to -50 mV from a holding potential of -30 or -40 mV. A brief (20-40 s) bath application of 1 pM muscarine partially inhibited I, causing the usual reduction in amplitude of the slow current relaxation, and of the instantaneous current step at the onset of the command voltage pulse, as well as an inward shift of the holding current. On average. I, was reduced 53% ( + 19 SD, n=22), although the extent of inhibition in individual cells was quite variable (24-79%). All these effects readily reversed following a 30-120 s washing with frog Ringer (Fig. 10). The amplitude of the inward relaxations recovered on average to 94% of control ( & 13 SD, n=22).

Following the iontophoretic intracellular injection of GTPyS (-2 to -4 nA, 2 min), a brief bath application of I pM muscarine inhibited I, as before. However, in most cells I, did not recover, or did so very poorly, even after prolonged washout (up to 20 min) (Table 1). The bottom panel of Figure 10 illustrates this effect of GTPyS in the same cell which responded to muscarine in a reversible way (upper panel) before the GTPyS injection. In most cells (15/22) the percentage of I, recovery upon washout was 20% or less after the injection of GTPyS, in contrast to a percentage of recovery above 60% in most cells (19/22) before the injection of GTPyS (Table I ) . With a few exceptions, in each of the 22 cells tested the muscarinic response became irreversible after the injection of the nucleotide. In addition, the muscarinic inhibition of I, was more variable (13- 114% of control) after the GTPyS injection, and several cells showed no response at all, although they did before the injection. Nevertheless, there was no recovery of I, in virtually all cells showing any significant response to muscarine.

GTPyS appeared to reduce I, by itself, independently of muscarinic agonist, as suggested by the smaller slow current relaxations and smaller holding current observed after the injection. However, a possible contribution to I, reduction by non-specific cell damage caused by the double impalement and iontophoretic injection precluded any quantification of this effect.


Control

mV 1 "A]

1 sec

After ionophoresis of GTP-'Y-S

1 rnin

FIG. 10. Effect of 1 pM muscarine on I, before and after the iontophoresis of GTPyS in a single B cell of the intact sympathetic bullfrog ganglion. The voltage clamp protocol was similar to that in previous figures. In the control stage (top panel), I, fully recovered from muscarine inhibition. GTPyS was then iontophoretically injected (-2 nA, 2 min) in the same cell. After the injection (bottom panel) muscarine irreversibly inhibited I,. An additional leakage response to muscarine lagged behind I, inhibition.

TABLE 1. Effect of GTPyS on the recovery of I, inhibition by rnuscarine in neurons from the intact sympathetic ganglion

Number of cells Percentage of recovery Before GTPyS After GTPyS ~ ~~

0-20 21 -40 41 -60 61 -80 81-100

1 0 2 8

11

15 1 2 0 4

In addition to interfering with the recovery of I, from the inhibition by muscarine, GTPyS injections enhanced the voltage-independent leakage current response to muscarine in an analogous way to that described for isolated cells. Thus, after GTPyS iontophoresis, some neurons responded to muscarine with a larger and faster increase in the leakage inward current instantaneously appearing at -50 or -60 mV. The cell in Figure 10, for example, showed a pure M current response before the injection of GTPyS, but after the injection the response to muscarine included an additional inward current component, although moderate. This additional inward current still developed with a slower time course relative to I, inhibition, and eventually reversed with washout, although more slowly than in control cells.

The effect of double impalement and iontophoretic current themselves on the reversibility of muscarinic responses was tested following a

similar recording protocol, but using a 3 M KC1-filled electrode for the second impalement and for injecting -4 nA for 2 min. No effect on either the sensitivity of I, to muscarine or the recovery of I, from muscarinic inhibition was detected in three such control experiments (not shown).

Discussion

We have investigated the transduction mechanism of the inhibition of the voltage-dependent potassium M current in bullfrog ganglion cells by muscarine and an LHRH analogue, two agonists that emulate slow synaptic responses, as well as UTP and SP. We conclude here that a G protein is a component of the transduction mechanism coupling all these agonists to the M conductance. The involvement of a G protein is inferred from the way the presence of intracellular hydrolysis-resistant analogues of GTP and GDP affect the inhibition and recovery of I,,, following bath application of agonists.

Intracellular GTPyS reduces or completely abolishes I, recovery following inhibition by muscarine, D-Ala6-LHRH, SP and UTP. Normally, I, fully recovers from inhibition when the agonists are washed out from the perfusate, and hence receptor stimulation ceases. Intracellular GTPyS reduces, in a dosedependent manner, the recovery of I, after the agonists have been completely washed out and consequently are no longer acting on their receptors. Thus, the effector pathway leading to M channels closure becomes operationally


uncoupled from muscarinic, peptide and nucleotide receptors. This persistent, and now agonist-independent, inhibition of 1, is the result expected if the effector pathway is regulated by a G protein (Breitwieser and Szabo, 1985; Gilman, 1987). In basal conditions, mostly GDP is bound to the alpha subunit in the heterotrimeric G protein, while regulation of the activity of the effector pathway requires binding of GTP to the same site upon release of GDP. G protein-effector interaction ceases upon hydrolysis of GTP to GDP and P by the G protein itself, restoring the basal state with bound GDP. Activation of G protein-coupled receptors greatly accelerates this cycle of nucleotide interchange (Gilman, 1987). Because GTPyS cannot be hydrolysed, G proteins remain in a state of permanent activation once GTPyS, competing with native GTP, occupies the guanine nucleotide binding site. A mechanism of this sort is the simplest explanation of the irreversible inhibition of I, by receptor-coupled agonists in the presence of intracellular GTPyS.

The specificity of the GTPyS effect on a G protein component in the transduction mechanism is given further credence by the dependency of the extent of reversibility of I,,, inhibition on the intracellular concentration of the GTP analogue. It is reasonable to argue that full or partial recovery of I, in cells dialysed with lower concentrations of GTPyS results from successful competition of native GTP for the G protein binding site so that none or only a fraction of M conductance- coupled G protein is locked in the form G protein-GTPyS. However, as the form G protein-GTPyS builds up with subsequent agonist applications (Fig. 4, middle trace) the recoverable I, is progressively smaller. It is worth pointing out that this recoverable I, is coupled to ligand receptors in a completely normal way as it is normally inhibited by subsequent agonist application, regardless of how small a current it is. Thus, uncoupling of surface ligand receptors to the M conductance develops in parallel with agonist application as the fraction of G proteins in the form G-GTPyS presumably builds up with each agonist exposure. These are the results expected for a competitive irreversible activation of G proteins by GTPyS, and rule out the possibility that partial recovery of I, is related to an additional non-specific blocking action of the nucleotide on the channels or any other site in the transduction pathway. It is interesting to note that GTPyS has no effect whatsoever on the recovery of I,,, if it is inhibited by a means which makes no use of the transduction pathway normally coupled to muscarinic or peptide receptor activation. Thus, I, fully and rapidly recovers following its inhibition by Ba, which is thought to directly block the M channels from the outside (Constanti et al., 1981). Ba, in contrast to other metal ions (Gilman, 1987), has no known effects on G proteins themselves. In these same neurons, however, I, recovers poorly or not at all following its inhibition by muscarinic or D-Ala6-LHRH. This contrast between receptor-coupled inhibition and direct blockade of M channels in GTPyS-dialysed neurons is in accord with a selective action of GTPyS on a component of the transduction mechanism physiologically activated by agonists.

Although nucleotide interchange at the binding site of G proteins is speeded up by the interaction of agonists with G protein-linked receptors, a steady-state agonist-independent basal interchange of GDP and GTP still occurs (Northup, 1985; Gilman, 1987). Because GTPyS cannot be hydrolysed, an agonist-independent build up of the form G protein-GTPyS would thus be expected to occur spontaneously in GTPyS-dialysed neurons, provided G proteins are present in the membranes of bullfrog sympathetic neurons. In turn, the existence of G protein-M conductance coupling would reveal itself as a corresponding spontaneous decrease of I, over time. This is precisely what we observe when the intracellular solution contains GTPyS (only

50 pM tested). I, slowly decreases over time, and very little is left after 20-40 min of whole cell recording. Thus GTPyS mimics the effect of inhibitory agonists on I,, albeit with a slower time course, suggesting that the nucleotide is activating the effector pathway for I, inhibition. The slow time course of this effect most probably reflects the slow nucleotide interchange in the absence of agonist. where the dissociation of GDP is the postulated rate limiting step (Ferguson et al., 1986; Cassel and Selinger, 1977). Such an effect of GTPyS by itself occurs equally slowly even when diffusional barriers are essentially absent such as in the activation of K channels from excised membrane patches with high concentrations of GTPyS (100-200 pM) (Kurachi et al., 1986; Yatani et al., 1987). The agonist-independent. irreversible activation of an effector pathway by non-hydrolysable GTP analogues is characteristic of G protein involvement (Schramm and Rodbell, 1975; Cuatrecasas et al., 1975; Northup, 1985; Gilman, 1987) and has recently been used to demonstrate the coupling of hormone and neurotransmitter receptors to K currents via G proteins in a number of cell types (Andrade et al., 1986; Kurachi et al., 1986; Yatani et al., 1987; Breitwieser and Szabo. 1988).

The dependency of G protein-linked ionic channel regulation on intracellular GTP has proved elusive in intact cells as compared to isolated membrane preparations (Northup, 1985; Gilman. 1987). In our work, as well as in a number of other reports using whole-cell recordings (Breitwieser and Szabo, 1985; Holz et al., 1986), membrane current responses shown, by other means, to be coupled to extracellular ligands by G proteins could consistently be recorded with pipettes containing GTP-free solutions. Demonstration of GTP dependency has required the use of isolated membrane patches (Kurachi et al., 1986), very small cells patched with low resistance, nucleotide-free pipettes (Pfaffinger et al., 1985), or harsh metabolic inhibition, in addition to omitting GTP in the patch pipette (Breitwieser and Szabo, 1988). Our use of the larger cells in the dish (usually 4 0 x 6 0 pm, and sometimes larger) obscured the dependency on GTP, which could presumably be effectively replenished despite diffusional loss to the pipette. Our dose - response data indicates an intracellular GTP concentration greater than 10 pM (see Figs 4 and 5), assuming equal affinities of GTP and the non-hydrolysable analogue for the G protein binding site. The 10 pM lower limit seems reasonable since in neurons dialysed with 10 pM GTPyS I,,, fully recovers even after repeated agonist applications, which favour the build-up of the form G protein-GTPyS. In addition, 10 pM compares reasonably well to the estimated 25 -50 pM steady-state cellular concentration of GTP in bullfrog cardiac myocytes (Breitwieser and Szabo, 1988). Given an estimated affinity for nucleotide in the 50-500 nM range (see refs in Breitwieser and Szabo. 1988), a GTP intracellular concentration of at least one order of magnitude higher would presumably be able to support GTP-dependent G protein activation despite diffusional loss to the pipette, and this would especially be so if lost GTP can be effectively replenished.

The effect of dialysing the isolated neurons with GDPOS, a metabolically stable analogue of GDP, also accords with the participation of a G protein in the inhibition of I,. In GDPPS-dialysed neurons muscarine and D-Ala6-LHRH were much less effective in inhibiting I, as compared to control cells. In contrast, GDPpS did not appear to modify the response to Ba. These results are fully compatible with the ability of GDPpS to competitively inhibit the binding of GTP to the guanine nucleotide binding site of G proteins (Eckstein et al.. 1979) and thus block the GTP-dependent interaction of G proteins with the effector pathway (Cassel et al., 1979; Jakobs, 1983; Lemos and Levitan, 1984; Holz et al., 1986; Andrade et a]., 1986; Breitwieser


Wang, 1987). In addition, neither we (unpublished data) nor others (Pfaffinger, 1988) found substrate for ADP-ribosylation in isolated membranes from bullfrog ganglion. The absence of a PTX effect could also be related to lack of surface receptors for the toxin in intact cells (Ui, 1984) or simply poor catalytic activity for ADP-ribosylation. On a speculative basis, however, the fact that PTX-sensitive G proteins are involved in a number of somatostatin-induced processes such as stimulation of K channels (Yatani et al., 1987) and reduction of calcium influx (Reisine et al., 1988) in pituitary cells, as well as inhibition of adenylate cyclase (Katada et al., 1984), lessens the appeal of this subclass of G proteins as candidates for mediating the inhibition of I, by muscarine and other agonists, as somatostatin is itself an enhancer of I, in two neuronal types of the rat brain (Moore et al., 1988; Jacquin et al., 1988).

The data in this study establish further distinctions between the M current component and the additional inward current component of the responses to all tested agonists. The latter, non-M current response, involves an increase in membrane conductance and has been recognized in a number of studies (Kuba and Koketsu, 1976; A d a m et al., 1982b; Katayama and Nishi, 1982; Kuffler and Sejnowski, 1983; Akasu et al., 1984: Jones et al., 1984; Jones, 1985). Pharmacological receptor blockade, when possible, showed that both I, inhibition and the increase in conductance are produced by the activation of the same receptors (Jones et al., 1984; Jones and Adams, 1985). However, the coupling mechanisms distal to the receptor stage seem different since, in contrast to I, inhibition, the increase in conductance always reversed in GTPyS-dialysed cells or in cells of the intact ganglion when the recordings were sufficiently long. The cause of the observed increase in the frequency of occurrence and in the magnitude of the additional inward current response in neurons dialysed with the highest concentrations of GTPyS (50- 100 pM) is unclear, although this effect appears to be indirectly related to the activation of G proteins.

Currently available data allow us to envisage a likely, although possibly fragmentary, coupling mechanism for the inhibition of I, by neurotransmitters and other inhibitory agonists. Here we show that activation of muscarinic, LHRH, UTP and SP receptors is coupled to M channels through a common step involving a G protein. The definitive identity of such G protein(s), however, has yet to be resolved, although it is likely to be of the type insensitive to PTX. It seems safe to rule out a mechanism whereby each ligand receptor physically interacts with M channels either by permanent or transient association causing them to close. G proteins provide a transduction link common to all receptors. This pattern of transduction, whereby several neurotransmitter receptors regulate a single ionic conductance via a G protein, has been reported in a variety of neuron types (Kurachi et al., 1986; Holz et al., 1986; Andrade et al., 1986; Sasaki and Sato, 1987). Distally to G protein activation, the sequence of steps leading to M channels closure is poorly understood. Protein kinase C (PKC) may play a role, as its activation by phorbol esters (Nishizuka, 1984) partially mimics the action of muscarine and LHRH in bullfrog sympathetic neurons (Adams and Brown, 1986; Brown and Adams, 1987; Pfaffinger et al., 1988), and inhibits I, in NG10815 neuroblastoma-glioma hybrid cells (Higashida and Brown, 1986). In this regard, G proteins can regulate PI turnover with subsequent activation of PKC (Gilman, 1987; Fain et al., 1988), and muscarinic receptor activation is known to increase PI turnover in neural tissue (Fisher et al., 1983; Horwitz et al., 1984; Fisher, 1986; Irvine and Berridge, 1986; Worley et al., 1987). However, activation of PKC alone fails to inhibit I, in full and so, if physiologically relevant, it may require the conjoint action of other factor(s). Although obvious

and Szabo, 1988). In addition to an initially poor response to agonist, further insensitivity develops over time in GDPPS-dialysed neurons. The latter may reflect either the time required for achieving an increasingly effective GDPPS intracellular concentration through dialysis or a decrease in intracellular GTP if replenishment of GTP depends on GDP, as GDPPS is a much poorer substrate than GDP for nucleoside triphosphate regeneration systems (Ekstein et al., 1979). Either is consistent with a G protein involvement. An estimation of the time constant (tau) for the diffusion of the nucleotides into the cells was obtained from a simplified expression that closely agrees with numerical calculations (Oliva et al., 1988). Assumin a spherical cell and a diffusion coefficient for nucleotides of cm' sec-', extreme tau values would be 6 and 12 min for a cell of 60 pm in diameter with access resistance of 3 MR, and a cell of 65 p n diameter with access resistance of 5 MR, respectively.

The very low increase over time of I, in several GDPPS-dialysed neurons was an unexpected finding. A tentative explanation is that nucleotide competition results in increasingly less G proteins in the minority form G protein-GTP (the active form), thus unmasking a small amount of steady-state I, inhibition in the basal state. Indeed, an extra, 'hidden', available I, shows up in the early stage of recovery from inhibition by agonist, in the form of a transiently increased I,. This phenomenon is more frequent and much more prominent in isolated neurons under whole-cell recording than it is in neurons of the intact ganglion recorded with microelectrodes. Interestingly, the M conductance in isolated neurons, which is about half of that measured in the intact ganglion (Adams et al., 1982a), grows considerably at the peak of the transient overshoot during washout. In this regard, experiments in progress show that I, in isolated neurons can grow to values close to those in intact ganglion cells following dialysis with 1 mM GDPPS (twice the highest concentration used in this study). This suggests that in neurons under whole-cell recording the amount of available I, in steady-state may be controlled, in addition to membrane potential, by a separate modulatory mechanism involving G proteins. Whether this has any physiological significance remains to be explored.

Our initial report of a G protein involvement in the transduction mechanism for I,,, inhibition (Lopez et al., 1987) is confirmed and extended in this study. While for the most part we used isolated cells under whole-cell recording so as to allow for better control of the intracellular milieu, the fact that we obtained essentially the same results in the more physiological condition of conventional microelectrode recording in the intact ganglion neurons gives us further confidence in our conclusions. Simultaneous studies by other laboratories found comparable effects of nucleotide analogues on I, and its inhibition by agonists in isolated sympathetic neurons from the bullfrog (Jones, 1987), grass frog (Pfaffinger, 1988) and rat (Marrion, 1987). Neither we nor others have so far been able to identify which subtype of G protein (Northup, 1985; Graziano and Gilman, 1987; Gilman, 1987; Lochrie and Simon, 1988) is associated with the M conductance. However, in contrast to cardiac K channels activated by muscarine (Breitwieser and Szabo, 1985; Pfaffinger et al., 1985; Kurachi et al., 1986), it does not seem to be a pertussis toxin-sensitive type (Katada and Ui, 1981, 1982; Ui, 1984) such as Go (Sternweis and Robinshaw, 1984), several Gi-like subtypes (Lochrie and Simon, 1988) and other newly reported sensitive forms (see Milligan, 1988). We have seen no effect on I, itself or attentuation of the inhibition of I, by muscarine after treating either dissociated cells or the intact ganglion with activated pertussis toxin. Equally ineffective was injection of the toxin into experimental animals (Andrade et al., 1986; Aghajanian and


candidates, it appears that neither C a nor IP3 are involved (Adams and Brown, 1982b; Brown and Adams, 1987; Pfaffinger et al., 1988). In contrast, PKC activation does not affect I, in hippocampal pyramidal cells, where IP3, independently of Ca , may be the second messenger (Dutar and Nicoll, 1988). These differences may reflect different mechanisms, or a still poor understanding of the complexities involved, such as the individual or conjoint participation of less familiar metabolites of inositol like IP4, IPS and IP6 (Vallejo e t al., 1987; Morris et al., 1987; Downes, 1988). Other cellular signalling systems in the control of I,, such a s cyclic A M P and arachidonic acid metabolism, have been investigated with negative results (Adams and Brown, 1982b; Brown and Adams, 1987; Villarroel and Adams, unpublished). It may well be that, like heart K channels (Kurachi et al., 1986; Condina et al., 1987), the M conductance is directly regulated by interaction with G proteins, with no need for soluble messengers. In this regard, it is interesting that the latency of the slow epsp in bullfrog cells (200-300 ms; Adams et al., 1987) is similar to that of the muscarine-induced, K current-mediated hyperpolarization in heart (Purves, 1976; Hartzell, 1981; Momose et a l . , 1984).

Acknowledgements

This work was supported by the Howard Hughes Medical Institute and the N.I.H. (N.S. 19846). H. S. Lopez is a research fellow from CONICET (Consejo Nacional de Investigaciones Cientificas y Tecnicas), Argentina. We thank Harriet Sheridan for technical assistance, and David Brown for his participation in preliminary intracellular experiments.

Abbreviations ADP ATP dSEVC EGTA

ePsP GDPpS GTPyS GTP G,

G”

IP3 IP4 IP5

I”,

Hepes

IP6

LHRH PKC

PI PTX SP UTP D-Ala6-LHRH G protein

adenosine diphosphate adenosine triphosphate discontinuous single electrode voltage-clamp ethylene glycol bis(b-aminoethyl ether)-N,N’-tetraacetic acid excitatory postsynaptic potential guanosine 5’-0-(2-thidiphosphate) guanosine 5’-0-3-thiotriphosphate guanosine triphosphate PTX-sensitive G protein that couples inhibitory receptors to adenylate cyclase PTX-sensitive G protein abundant in brain membranes 4-(2-hydroxyethyl)- 1 -piperazineethanesulfonic acid inositol ( I ,4,5)-trisphosphate (1,3,4,5)-tetrakisphosphate (1,3,4,5,6)-pentakisphosphate inositol hexakisphosphate potassium M current luteinizing hormone-releasing hormone protein kinase C (Ca*+-phospholipid dependent protein kinase) phosphatidylinositol pertussis toxin substance P uridine triphosphate peptide analogue of mammalian LHRH guanine nucleotide binding protein

References

Adams. P. R. and Brown, D. A. (1980) Luteinizing hormone-releasing factor and muscarinic agonists act on the same voltage-sensitive K+-current in Bullfrog sympathetic neurons. Br. I . Pharmac. 68: 353-355.

Adams, P. R., Brown. D. A. and Constanti, A. (1982a) M-currents and other

potassium currents in Bullfrog sympathetic neurons. J . Physiol. 330: 537 -572.

Adams, P. R., Brown. D. A. and Constanti. A. (1982b) Pharmacological inhibition of the M-current. J . Physiol. 332: 223-262.

Adams, P. R. and Brown, D. A. (1982) Synaptic inhibition of the M-current: slow excitatory post-synaptic potential mechanism in Bullfrog sympathetic neurones. J . Physiol. 332: 263-272.

Adams, P. R., Brown, D. A. and Jones, S. W. (1983) Substance P inhibits the M-current in Bullfrog sympathetic neurons. Br. J. Pharmac. 79: 330-333.

Adams, P. R. and Brown, D. A . (1986) Effects of phorbol ester on slow potassium currents of Bullfrog ganglion cells. Biophys. J . 49: 215A.

Adams, P. R., Jones, S. W., Pennefather, Brown, D. A,, Koch. C. and Lancaster, B. (1986) Slow synaptic transmission in frog sympathetic ganglion. J . Exp. Biol. 124: 259-285.

Aghajanian, G. K. and Wang, Y.-Y. (1987) Common alpha2 and optiate effector mechanisms in the locus coeruleus: intracellular studies in brain slices. Neuropharmacology 26: 793 -799.

Akasu, T., Hirai K. and Koketsu, K. (1983) Modulatory actions of ATP on membrane potentials of Bullfrog sympathetic ganglion cells. Brain Res. 258:

Akasu, T.. Gallagher, J. P., Koketsu, K. and Shinnick-Gallagher, P. (1984) slow excitatory post-synaptic currents in Bullfrog sympathetic neurones. J . Physiol. 351: 583-593.

Andrade, R., Malenka, R. C. and Nicoll, R. A. (1986) A G protein couples serotonin and GABA, receptors to the same channels in hippocampus. Science 234: 1261 -1265.

Blackman. J . G., Ginsborg, B. L. and Ray, C. (1963) Synaptic transmission in the sympathetic ganglion of the frog. J. Physiol. 167: 355-373.

Breitwieser, G. E. and Szabo, G. (1985) Uncoupling of cardiac muscarinic and beta-adrenergic receptors from ion channels by a guanine nucleotide analogue. Nature 317: 538-540.

Breitwieser, G. E. and Szabo, G . (1988) Mechanism of muscarinic receptor- induced K+ channel activation as revealed by hydrolysis-resistant GTP analogs. J . Gen. Physiol. 91: 469-493.

Brown, A. M. and Birnbaumer, L. (1988) Direct G protein gating of ion channels. Am. J. Physiol. 254: H401-H410.

Brown, D. A. and Adams, P. R. (1987) Effects of phorbol dibutyrate on M currents and M current inhibition in Bullfrog sympathetic neurons. Cell. Mol. Neurobiol. 7: 255-269.

Cassel, D., Eckstein, F., Lowe, M. and Selinger. Z. (1979) Determination of the turn-off reaction for the hormone-activated adenylate cyclase. J. Biol. Chem. 254: 9835-9838.

Cassel, D. and Selinger, Z. (1977) Catecholamine-induced release of [3H]-Gpp(NH)p from turkey erythrocyte adenylate cyclase. 1. Cyclic Nucleotide Res. 3: I 1 -22.

Codina. J . . Yatani. A, , Grenet, D., Brown, A. M. and Birnbaumer, L. (1987) The alpha subunit of the GTP binding protein G, opens atrial potassium channels. Science 236: 442-445.

Constanti, A,. Adams, P. R. and Brown, D. A. (1981) Why do barium ions imitate acetylcholine? Brain Res. 206: 244 -250.

Cuatrecasas, P. , Jacobs, S. and Bennett, V. (1975) Activation of adenylate cyclase by phosphoramidate and phosphonate analogs of GTP: possible role of covalent enzyme-substrate intermediates in the mechanism of hormonal activation. Proc. Natl. Acad. Sci. USA 72: 1739- 1743.

Dolphin, A. C. and Scott, H. (1987) Calcium channel currents and their inhibition by (-)-badofen in rat sensory neurones: modulation by guanine nucleotides. J . Physiol. 386: 1 - 17.

Downes, C. P. (1988) Inositol phosphates: a family of signal molecules‘? TINS

Dunlap K.. Holz, G. G. and Rane, S. G. (1987) G proteins as regulators of ion channel function. TINS 10: 241 -244.

Dutar, P. and Nicoll, R. A. (1988) Stimulation of phosphatidylinositol (PI) turnover may mediate the muscarinic suppression of the M-current in hippocampal pyramidal cells. Neurosci. Lett. 85: 89-94.

Eckstein, F., Cassel, D., Levkovitz, H. , Lowe, M. and Selinger, Z. (1979) Guanosine 5’-0-(2-Thiodiphosphate). An inhibitor of adenylate cyclase stimulation by guanine nucleotides and fluoride ions. J . Biol. Chem. 254: 9829-9834.

Eiden, L. E. and Eskay, R. L. (1980) Characterization of LRF-like immuno- reactivity in the frog sympathetic ganglia: non-identity with LRF decapeptide. Neuropeptides 1: 29-37.

Eiden, L. E., Loumaye, E. , Sherwood, N . and Eskay. R. L. (1982) Two chemically and immunologically distinct forms of luteinizing hormone-

313-317.

1 1 : 336-338.


potential in Bullfrog sympathetic ganglion cells. Jap. J. Physiol. 26: 651 -669. Kuba, K. and Koketsu, K. (1978) Synaptic events in sympathetic ganglia.

Progress Neurobiol. 11: 77- 169. Kuffler, S.W. (1980) Slow synaptic response in autonomic ganglia and the pursuit

of peptidergic transmitter. J. Exp. Biol. 89: 257-286. Kuffler, S. W. and Sejnowski, T. J. (1983) Peptidergic and muscarine excitation

at amphibian sympathetic synapses. J. Physiol. 341: 257-278. Kurachi, Y. , Nakajima, T. and Sugimoto, T. (1986) On the mechanism of

activation of muscarinic K + channels by adenosine in isolated atrial cells: involvement of GTP-binding protein. Pflugers Arch. 407: 264 -274.

Lemos, J. R. and Levitan, I. B. (1984) Intracellular injection of guanyl nucleotides alters the serotonin-induced increase in potassium conductance in Aplysia neuron R15. J. Gen. Physiol. 83: 269-285.

Lochrie, M. A. and Simon, M. I . (1988) G protein multiplicity in eukariotic signal transduction systems. Biochemistry 27: 4957 -4965.

Lopez, H. S . , Brown, D. A. and A d a m P. R. (1987) Possible involvement of GTP-binding proteins in coupling of muscarinic receptors to M-current in Bullfrog ganglion cells. Neurosci. Abstr. 13: 532.

Marrion, N. V. (1987) Probable role of a GTP-binding protein in mediating M-current inhibition by muscarine in rat sympathetic neurones. J . Physiol. 396: 87P.

Milligan, G. (1988) Techniques used in the identification and analysis of function of pertussis toxin-sensitive guanine nucleotide binding proteins. Biochem. J. 255: 1-13.

Momose, Y. , Giles, W. and Szabo, G. (1984) Acetylcholine-induced K + current in amphibian atrial cells. Biophys. 1. 45: 20-22.

Moore, S. D. , Madamba, S. G . , Joels, M. and Siggins, G. R. (1988) Somatostatin augments the M-current in hippocampal neurons. Science 239: 278-280.

Morris, A. P., Gallacher, D. V., Irvine, R. F. and Petersen, 0. H. (1987) Synergism of inositol trisphosphate and tetrakisphosphate in activating Ca*+-dependent K + channels. Nature 330: 653 -655.

Nicoll, R. A. (1988) The coupling of neurotransmitter receptors to ion channels in the brain. Science 241: 545-551.

Nishizuka, Y. (1984) The role of protein kinase C in cell surface signal transduction and turnover promotion. Nature 308: 693 -698.

Northup, J . K. (1985) Overview of the guanine nucleotide regulatory protein systems, N, and N,, which regulate adenylate cyclase activity in plasma membranes. In: Cohen, P. and Houslay, M. D. Molecular Aspects of Cellular Regulation, vol. 4, Molecular Mechanisms of Transmembrane Signalling pp 91 - 116. Elsevier Science Publishers, B.V. (Biomedical Div.), Amsterdam.

Oliva, C. , Cohen, I. S . and Mathias, R. T. (1988) Calculation of time constants for intracellular diffusion in whole-cell patch-clamp configurations. Biophys.

Pfafinger, P. J., Martin, J. M., Hunter, D. D., Nathanson, N. M. and Hille, B. (1985) GTP-binding proteins couple cardiac muscarinic receptors to a K channel. Nature 317: 536-538.

Pfaffinger, P. J., Leibowitz, M. D., Subers, E. M., Nathanson, W. A., Almers, W. and Hille, B. (1988) Agonists that suppress M-current elicit phosphoinositide turnover and Ca2+ transients, but these events do not explain M-current suppression. Neuron 1: 477 -484.

Pfaffinger, P. J. (1988) Muscarine and t-LHRH suppress M-current by activating an 1AP-insensitive G protein. J. Neurosci 8(9): 3343 -3353.

Purves, R. D. (1976) Function of muscarinic and nicotinic acetylcholine receptors. Nature 261: 149- 151.

Reisine, T., Wang, H.-L. and Guild, S . (1988) Somatostatin inhibits CAMP- dependent and CAMP-independent calcium influx in the clonal pituitary tumor cell line AtT-20 through the same receptor population. J . Pharmac. Exp. Ther. 245: 225-231.

Sasaki, K. and Sato, M. (1987) A single GTP-binding protein regulates K+-channels coupled with dopamine, histamine and acetylcholine receptors. Nature 325: 259-262.

Schramm, M. and Rodbell, M. (1975) A persistent active state of the adenylate cyclase system produced by the combined actions of isoproterenol and guanylyl imidodiphosphate in frog erythrocyte membranes. J. Biol. Chem. 250:

S h e n v d , N., Eiden, L., Brownstein, M., Spiess, J., Rivier, J . and Vale, W . (1983) Characterization of a teleost gonadotropin-releasing hormone. Proc. Natl. Acad. Sci.'USA 80: 2794-2798.

Sternweis, P. C. and Robinshaw, J. D. (1987) Isolation of two proteins with high affinity for guanine nucleotides from membranes of bovine brain. J . Biol. Chem. 259: 13806-13813.

J. 54: 791 -799.

2232-2237.

releasing hormone are differentially expressed in frog neural tissues. Peptides 3: 323-327.

Fain, J. N., Wallace, M. A. and Wojcikiewicz, R. J . (1988) Evidence for involvement of guanine nucleotide-binding regulatory proteins in the activation of phospholipases by hormones. FASEB J. 2: 2569-2574.

Ferguson, K. M., Higashijima, T., Smigel, M. D. and Gilman, A. G. (1986) The influence of bound GDP on the kinetics of guanine nucleotide binding to G proteins. J. Biol. Chem. 261: 7393-7399.

Finkel, A. S. and Redman, S. (1984) Theory and operation of a single microelectrode voltage clamp. J . Neurosci. Methods 11: 101 - 127.

Fischer, S. K . , Klinger, P. D. and Agranoff, B. W. (1983) Muscarinic agonist binding and phospholipid turnover in brain. J. Biol. Chem. 258: 7358-7363.

Fisher, S. K. (1986) Inositol lipids and signal transduction at CNS muscarinic receptors. TIPS, Feb. Supplemt: 83 - 87.

Gilman, A. G. (1987) G proteins: transducers of receptor-generated signals. Annu. Rev. Biochem. 56: 615-649.

Graziano, M. P. and Gilman, A. G . (1987) Guanine nucleotide-binding regulatory proteins: mediators of transmembrane signaling. TIPS 8: 478-481.

Hamill, 0. P., Marty, A, , Neher, E., Sakmann, B. andsigworth, F. J. (1981) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391: 85- 100.

Hartzell, C. H. (1981) Mechanisms of slow postsynaptic potentials. Nature 291: 539-544.

Higashida, H. and Brown, D. A. (1986) Two polyphosphatidylinositide metabolites control two K+ currents in a neuronal cell. Nature 323: 333 - 335.

Holz, G . G., Rane, S . G. and Dunlap, K. (1986) GTP-binding proteins mediate transmitter inhibition of voltage-dependent calcium channels. Nature 3 19: 670 -673.

Horwitz, J., Tsymbalov, S. and Perlman, R. L. (1984) Muscarine increases tyrosine 3-monooxygenase activity and phospholipid metabolism in the superior cervical ganglion of the rat. J. Pharmacol. Exp. Ther. 229: 577-582.

Irvine, R. F. and Berridge, M. J. (1986) Inositide metabolism in the brain: its potential role in complex neuronal pathways. In: Iversen, L. L. and Goodman, E. (eds) Fast and Slow Chemical Signalling in the Nervous System pp 185-204. Oxford University Press, Oxford.

Jacquin, T. , Champagnat, J., Madamba, S . , Denavit-Saubie, M. and Siggins, G . R. (1988) Somatostatin depresses excitability in neurons of the solitary tract complex through hyperpolarization and augmentation of Im, a non-inactivating voltage-dependent outward current blocked by muscarinic agonists. Proc. Natl. Acad. Sci. USA 85: 948-952.

Jakobs, K. J. (1983) Determination of the turn-off reaction for the epinephrine- inhibited human platelet adenylate cyclase. Eur. 1. Biochem. 132: 125- 130.

Jan, L. Y. and Jan, Y. N. (1982) Peptidergic transmission in sympathetic ganglia of the frog. J. Physiol. 327: 219-246.

Jan, Y. N., Jan, L. Y. and Kuffler, S. W. (1979) A peptide as a possible transmitter in sympathetic ganglia of the frog. Proc. Natl. Acad. Sci. USA 76: 1501 - 1505.

Jan, Y. N., Jan, L. Y. and Kuffler, S . W. (1980) Further evidence for peptidergic transmission in sympathetic ganglia. Proc. Natl. Acad. Sci. USA 77:

Jones, S . W., Adams, P. R., Brownstein, M. J. and Rivier, J. E. (1984) Teleost lutenizing hormone-releasing hormone: action on Bullfrog sympathetic ganglia is consistent with role as neurotransmitter. J. Neurosci. 4: 420-429.

Jones, S . W. and Adams, P. R. (1985) Modulation of potassium currents of vertebrate neurons. In: Levitan, I. B. and Kaczmarek, L. K. Neuromodulation pp. 1-38, Oxford University Press, Oxford.

Jones, S. W. (1985) Muscarinic and peptidergic excitation of Bullfrog sympathetic neurones. J. Physiol. 366: 63 -87.

Jones, S . W. (1987) GTP-g-S inhibits the M-current of dissociated Bullfrog sympathetic neurons. Neurosci. Abstr. 13: 533.

Katada, T., Bokoch, G . M., Smigel, M. D., Ui, M. andGilman, A. G . (1984) The inhibitory guanine nucleotide-binding regulatory component of adenylate cyclase. J. Biol. Chem. 259: 3586-3595.

Katada, T. and Ui, M. (1981) Islet-activating protein: a modifier of response- mediated regulation of rat islet adenylate cyclase. J. Biol. Chem. 256: 8310-8317.

Katada, T. and Ui, M. (1982) Direct modification of the membrane adenylate cyclase system by islet-activating protein due to ADP-ribosylation. Proc. Natl. Acad. Sci. USA 79: 3129-3133.

Katayama, Y. and Nishi, S. (1982) Voltage-clamp analysis of peptidergic slow depolarizations in Bullfrog sympathetic ganglion cells. J . Physiol. 333: 305 -313.

Kuba, K. and Koketsu, K . (1976) Analysis of the slow excitatory postsynaptic

5oO8 -501 2.


Ui, M. (1984) Islet-activating protein, pertussis toxin: a probe for functions (1987) Cholinergic phosphatidylinositol modulation of inhibitory. G protein-linked, neurotransmitter actions: electrophysiological studies in rat hippocampus. Proc. Natl. Acad. Sci. USA 84: 3467-3471.

Yatani. A., Codina, I., Sekura, R. D., Birnbaumer, L. and Brown, A. M . (1987) Reconstitution of somatostatin and muscarinic receptor mediated stimulation of K + channels by isolated G , protein in clonal rat anterior pituitary cell membranes. Mol. Endocrinol. 1 : 283-289.

of the inhibitory guanine nucleotide regulatory component of adenylate cyclase. TIPS July: 277-279.

Vallejo, M., Jackson, T., Lightman, S. and Hanley. M . R. (1987) Occurrence and extracellular actions of inositol pentakis- and hexakisphosphate in mammalian brain. Nature 330: 656-658.

Worley, P. F., Baraban, J. M., McCarren, M., Snyder, S. H. and Alger. B. E.

A G Protein Mediates the Inhibition of the Voltage-Dependent Potassium M Current by Muscarine, LHRH,...

Documents

Transcript of A G Protein Mediates the Inhibition of the Voltage-Dependent Potassium M Current by Muscarine, LHRH,...