Significance of slow synaptic potentials for transmission of excitation in guinea-pig myenteric plexus

Neuroscience Vol. 14, No. 2, pp. 661-672, 1985 Printed in Great Britain

0306-4522/85 $3.00 + 0.00 Pergamon Press Ltd c, 1985 IBRO

SIGNIFICANCE OF SLOW SYNAPTIC POTENTIALS FOR TRANSMISSION OF EXCITATION IN GUINEA-PIG MYENTERIC PLEXUS K. MORITA* and R. A. NORTH? Neuropharmacology

Laboratory,

Massachusetts

Institute U.S.A.

of Technology,

56245,

Cambridge,

MA 02139,

Abstract-Intracellular recordings were made from neurones in myenteric ganglia of the guinea-pig ileum in vitro. Synaptic potentials were evoked by electrically stimulating presynaptic fibres as they entered the ganglion, using a small focal electrode. Slow synaptic depolarizations (excitatory postsynaptic potentials) were evoked in most myenteric neurones of both types. A single stimulus was more likely to evoke a slow excitatory postsynaptic potential in cells with nicotinic synaptic input (S cells; 50%) than in cells with long-lasting after-hyperpolarizations following the soma action potential (AH cells; 20%). Two pulses often evoked a slow excitatory postsynaptic potential in AH cells when one pulse was ineffective. The optimally effective time between the pulses was about looms. Ten pulses resulted in slow excitatory

postsynaptic potentials even when delivered at frequencies as low as 0.5 Hz. For the same frequency of presynaptic stimulation, the duration of the slow excitatory postsynaptic potential was greater in AH cells than in S cells and the amplitude of the slow excitatory postsynaptic potential was slightly greater in S than AH cells. Spontaneous depolarizations were observed which had time-courses and amplitudes similar to the evoked slow excitatory postsynaptic potential. They were not blocked by tetrodotoxin or atropine. The calcium-dependent after-hyperpolarization which follows one or more action potentials in AH cells was reduced or even abolished during the slow excitatory postsynaptic potential. Presynaptic nerve stimulation at intensities lower than those required to cause a slow excitatory postsynaptic potential caused a reduction in the calcium dependent after-hyperpolarization. It is concluded that the slow excitatory postsynaptic potential is generated by an intracellular intermediate process which is sensitive to the intracellular calcium concentration. The results suggest that the postsynaptic action of the synaptic transmitter is to interfere with the intracellular process which couples the entry of calcium to the increase in potassium conductance.

Transmission of information in autonomic ganglia has been considered typically to involve the release of acetylcholine (ACh) which acts for a few milliseconds on nicotinic cholinoceptors on the postsynaptic ce11.25 It is known that, additionally, a variety of slow synaptic potentials occur in autonomic ganglia which have durations in the order of several seconds to a few minutes. The slow excitatory postsynaptic potential (EPSP) and the late slow EPSP, first described in bullfrog sympathetic ganglia,‘8,26, have subsequently been described in several mammalian ganglia.28 However, the physiological role of the slow EPSP has been doubted because the potentials were not observed during in vivo recording (ref. 33 and V.I. Skok, personal communication); furthermore, repetitive synchronous activation of presynaptic fibres has generally been used to evoke the slow EPSP and this may not occur in the whole animal. *Present address: Department of Autonomic Phvsioloav. Medical Research institute, Tokyo Medical anh De&l University, 2-3-10, Kanda-Surugadai, Chiyoda-ku, Tokyo, 101 Japan. tTo whom all correspondence should be addressed. Abbreviations: ACh, acetvlcholine; AH cell. cell with a long-lasting after-hypebolarization following the soma action potential; EPSP, excitatory postsynaptic potential; S cell, cell with nicotinic synaptic input. 661

In earlier papers, we described a slow EPSP in neurones of the guinea-pig myenteric plexus.‘3~‘4~30 This slow EPSP appears to result from the action of at least two transmitters. The cholinergic slow EPSP results from activation of muscarinic cholinoceptors and typically lasts for 5-15 S.30 The non-cholinergic slow EPSP appears to result from the action of substance P4,13 and has a longer duration; this is presumably analogous to the peptide-mediated late slow EPSP observed in bullfrog ganglia.‘2,‘5,‘7,26 One objective of the present experiments was to investigate the dependence of the slow EPSP on the pattern and intensity of presynaptic nerve stimulation, in order to assess the likelihood that the slow EPSP could transmit significant excitation in the intact organism. A second purpose of the present study related to the postsynaptic action of transmitters mediating the slow EPSP. Two types of neurones can be distinguished electrophysiologically in the myenteric plexus (Type 1 S cells and Type 2 or AH cells).“,27 They differ in one important respect: AH cells have a long-lasting after-hyperpolarization following a single action potential, “,*’ whereas S cells have nicotinic cholinergic fast EPSPs. The after-hyperpolarization in AH cells results from calcium entering the neurone during the action potential leading to an increase in

K. Morita and R. A. North

662

membrane potassium conductance.“,22.27 Because the slow EPSP is caused by a reduction in membrane potassium conductance,” it seemed important to determine whether the transmitter released by presynaptic nerve stimulation reduced the same potassium conductance which was increased by calcium ions entering the AH neurone. This was approached by comparing the slow EPSP of S and AH cells and by examining the interaction between the slow EPSP and the calcium-dependent after-hyperpolarization in AH cells.

Parameters of presynaptic stimulation

The slow EPSPs observed in the present series of experiments were generally similar in their properties to ihose reported previously.‘4,37 The amplitude of the slow EPSP was increased in a graded manner when the stimulus voltage was increased. Slow EPSPs evoked by a single pulse were constant in amplitude provided that the pulse was repeated at intervals of at least 30s. Most neurones (80% AH cells and 50% S cells) showed no response to a single pulse stimulus, even though they showed large siow EPSPs in response to a brief tram of pulses. Most strikingly, EXPERIMENTAL PROCEDURES many cells showed large slow EPSPs in response to Intracellular recordings were made from neurones in two or three pulses when one pulse was without isolated myenteric ganglia from the ileum of adult guineaoies. The techniaues have been described in detail.‘4.27 effect. Therefore we made a more systematic study of the influence of changing the number and frequency k&ording electrodes contained potassium chloride (3 M) and had resistances of SO-100 Ma. Potentials were recorded of the pulses applied to the presynaptic nerves. The and current injected with a WP Instruments (M701 or largest stimulus which was used was 90 pulses M707) preamplifier, and displayed on an oscilloscope and a (30 Hz,‘3 s).

pen recorder [Gould, pen response 50 mm (usually 50 mV) in less than 5 ms]. Presynaptic nerves were stimulated by applying a glass micropipette (tip dia IO-20pm) to the surface of the nerve fibres which entered one or other pole of the ganglion. The micropipette contained the same solution as that which superfused the ganglion. This comprised (mM): NaCl 11’7,KC1 4.7, NaH,PO, 1.2, CaCf, 2.5, MgCI, 1.2, NaHCO, 25, glucose 11, gassed with 95% 02 and 5% CO,. The sunerfusing solution was heated so that its temperature as it-passed over the ganglion was maintained at a value between 35 and 37°C throughout a given experiment. In many experiments, hy~rpola~zing electrotonic potentials were evoked by passing current pulses of fixed amptitude across the cell membrane. The amplitude of these potentials was used as a measure of input resistance (R). The fractional conductance change during the slow EPSP was calculated from [(R/R’) - 11, where R is the input resistance prior to nerve stimulation and R’ is the input resistance during the slow EPSP.‘~22R’ was measured after restoring the membrane potential to its control level by passing sufficient hy~~ola~zing current. In some experiments, atropine sulphate (Sigma) and tetrodotoxin (Sankyo) were added to the superfusing solution. RESULTS

These results are based on intracellular recordings from more than 160 neurones in myenteric ganglia from 57 guinea-pigs.

EfSect of increasing both the number andjiiequency ofpulses. The slow EPSP became progressively larger,

but particularly longer, as the number of pulses to the presynaptic nerve was increased. The typical recordings from an AH cell are shown in Fig. I; the pooled data from several neurones are shown below in Fig. 4. EfSect of varying the number of pulses at a fixed frequency. When the train length was increased from

1 to 60 pulses, the slow EPSP continued to grow larger and longer. These expe~ments were performed with frequencies of stimulation between 1 and 10 Hz. The total number of pulses delivered to the presynaptic nerves thus appears to be an important determinant of the postsynaptic response. l@ect of varying frequency of a fixed number of pulses. The slow EPSP evoked by a fixed number of

presynaptic stimuli was very much dependent on the frequency. Stimuli comprising 10 pulses were almost always effective in evoking slow EPSPs, even when delivered at frequencies as low as 0.5 Hz (as illustrated in Fig. 2). At 0.5 Hz a substantial membrane depoIarization (about 13 mV) occurred by the time of the 10th stimulus, suggesting that this low frequency of presynaptic action potentials could be significantly

Fig. 1. Effect of increasing the number and frequency of pulses delivered to presynaptic nerves. Train length was 3 s; pulse frequency increased from 0 to 30 Hz. Total number of pulses is given beneath each trace. Larger slow excitatory postsynaptic potentials gave rise to action potentials; in this and other figures the full height of the action potential is not shown. AH neurone. Resting potential, -58mV.

663

Slow synaptic potentials

1 Hz

Fig. 2. Effect of increasing the frequency of stimulation with a fixed number (10) of presynaptic pulses. Even stimulation at 1 Hz for 10 s produced a slow excitatory postsynaptic potential which reached threshold for action potential generation. Downward deflections are hyperpolarizing electrotonic potentials evoked by repeated passage of a current pulse (about 80ms duration); the pulse amplitude was constant throughout each trace but was adjusted between traces. S neurone, in atropine (1 PM). Resting

potential, - 56 mV.

excitatory. The main effect of increasing the frequency of presynaptic pulses above 1 Hz was to increase the rate of postsynaptic depolarization, and to increase the number of action potentials generated (Fig. 2). Experiments were also performed with two pulses. As the interval between the pulses was reduced from 1 to 200 ms the resulting slow EPSP became larger. The maximum amplitude was achieved for a pulse separation of SO-100 ms, corresponding to 10-20 Hz. A further reduction in the interval between the pulses progressively reduced the slow EPSP amplitude. Two pulses separated by 5 ms gave the same response as a single pulse. In brief, two pulses were often effective when delivered at lO-20Hz even though when

shorter or longer intervals separated the pulses no postsynaptic response was evoked. Comparison between S and AH neurones It was previously reported that a single pulse stimulus would elicit a slow EPSP in some S neurones but not in AH neurones.14 The response to a single pulse stimulus is often abolished by atropine, suggesting that the synaptic transmitter is ACh.30 In the present series of experiments, these observations on the differences between S and AH cells were extended in two ways. First, a single pulse stimulus was found to elicit a slow EPSP not only in S cells but also in a small proportion of AH cells (Table 1). The absolute proportion of neurones affected is rather approxi-

Table 1. The amplitude and time course of the slow excitatory postsynaptic potential recorded in myenteric neurones following one or 30 shocks to the presynaptic nerves Peak amplitude

(mV)

Time to half decav (s)

Single pulse stimulus 4.3 + 3.3 5.4 + 1.0 3.8 k 1.3 2.2 f 0.4 Repeated pulse stimulus (10 Hz/3 s) 9.4 + 1.3 s cell 19.9 * 2.1 25.0 + 9.1 12.6 + 1.5 AH cell

s cell AH cell

Peak fractional conductance change (AK)

n

9.3 k 3.8 11.8 k 3.4

0.27 k 0.06 0.1 * 0.04

17 6

37.3k4.1 82.5 + 13.7

0.46 k 0.08 0.38 + 0.07

15 13

Total duration (9

664

K. Morita and R. A. North

A

c

I -40 potentfal

-10

(mV)

potential

fmV)

L

Fig. 3. Reversal of slow excitatory postsynaptic potential. (A) Slow excitatory postsynaptic potential evoked by a single pulse at various membrane potentials (indicated in mV beside each trace). (B) Slow excitatory postsynaptic potential evoked by 30 pulses (10 Hz) in a different neurone. Both (A) and (B) are recordings from S neurones; in some parts of (A) the fast excitatory postsynaptic potential is truncated. In (B) the fast excitatory postsynaptic potential sometimes gave rise to an action potential, the full height of which was not reproduced by the pen recorder. (C) Amplitude of slow excitatory postsynaptic potential evoked by single pulse as function of membrane potential in 3-7 S neurones. (D) Amplitude of slow excitatory postsynaptic potential evoked by 30 pulses (10 Hz) as function of membrane potential in 3-7 neurones (both Sand AH cells). In (C) and (D) the horizontal and vertical bars through the points indicate the S.E. of the mean values. Lines filled by eye; reversal in each case occurred at -95mV.

mate, because it was our impression that this proportion was greater when the tip diameter of the stimulating electrode was larger, which may have increased the likelihood of exciting presynaptic fibres. Second, even after the addition of atropine (1 PM) slow EPSPs could be observed in some neurones of both types in response to a single pulse applied to the presynaptic nerves. In many of the present experiments, atropine was not added to the solution (be-

cause of the possibility of blockade of presynaptic inhibition)23 and it therefore could not be determined whether the synaptic potential was mediated by ACh, substance P,4,‘3 or a mixture of these and other ~ansmit~rs. It was therefore necessary to determine first whether the ionic mechanism of the slow EPSP was the same when it was evoked by one or many pulses, and whether it was evoked in S or AH cells. In both S and AH neurones, slow EPSPs were

665

Slow synaptic potentials

associated with a conductance decrease, and the slow EPSPs reversed at the same membrane potential whether evoked by one or many pulses (Fig. 3). Furthermore, no difference in reversal potential was found for the slow EPSPs in S and AH cells. Two differences were observed in the slow EPSP between S and AH neurones in the present study. The first was that the synaptic potential evoked by a given number of pulses was, on average, slightly larger in S cells than in AH cells (Figs 4 and 5). A second difference which was more striking was that the duration of the slow EPSP in AH cells was three to four times longer than that in S cells (Figs 4 and 5). The slow EPSPs with the very longest durations (2-10min) were always observed in AH cells.

Spontaneous

depolarizations

Spontaneous depolarizations occurred in 10-20x of myenteric neurones of both types. These spontaneous depolarizations had rise times, amplitudes and decay times strikingly similar to the evoked slow EPSPs in the same neurone, although the actual amplitudes varied widely from cell to cell and at times within the same cell (Fig. 6). The mean amplitude of the spontaneous depolarization in 10 cells was 11.8 + 3.2 mV (mean f SE). The spontaneous depolarizations were associated with conductance decreases. It is possible that these depolarizations are not due to spontaneous release of transmitter from pre-

A

synaptic terminals, because synchronous activation of several presynaptic fibres is presumably necessary to evoke a slow EPSP of comparable amplitude (Fig. 6). Tetrodotoxin (1 p M) and atropine (1 PM) did not change the amplitude of spontaneous depolarizations. There is another reason for considering that these depolarizations occur independently of a synaptically released transmitter, and that is the observation that the likelihood of their occurrence could be influenced by postsynaptic factors. A prolonged period of antidromic stimulation of AH neurones occasionally triggered the appearance of spontaneous depolarizations. It was also observed that spontaneous depolarizations were more likely to occur in neurones which had previously been stimulated to produce large and long-lasting slow EPSPs. The similarities in time course between the spontaneous depolarizations and the evoked slow EPSPs suggest that they may result from the same intracellular process which is normally triggered or initiated by synaptically released transmitter. Spontaneous depolarizations were also more likely to be observed in the period following washout of a substance which caused a hyperpolarization of the cell membrane, either clonidine2’ or normorphine. The large after-hyperpolarization which follows a burst of action potentials in AH neurones was also sometimes followed by a membrane depolarization with many of the characteristics of the slow EPSP. The input resistance was increased and the calcium-

s cell

_1

20 mV

10

8

Fig. 4. Differences between slow excitatory postsynaptic potentials in S and AH cells. Typical slow excitatory postsynaptic potentials are shown from an S cell (A) and an AH cell (B) in response to various frequencies of presynaptic nerve stimulation. Trace (B) also shows to spontaneous depolarization occurring after the two-pulse stimuli. Atropine (500 nM) present in (B).

666

K. Morita and R. A. North

0 Ol

I

I

2

5

-

Frequency

5

I I J 10 2030

10

2030

(Hz)

Fig. 5. Parameters of slow excitatory postsynaptic potentials in S and AH neurones. Each point represents the mean value from 9-13 neurones; l S cells, 0 AH cells. Bars are SE. of means, they are omitted on (B) and (D) for clarity. Points to the left of the ordinate indicate the response to a single stimulus. Abscissae is frequency of pulses in a stimulus train lasting for 3 s (note logarithmic scale). (4) Mean amplitudes. Slow excitatory postsynaptic potentials in S cells were slightly larger than in AH cells. (B) Peak conductance change (See Experimental Procedures) reflected the amplitude difference. (C) Total duration of slow excitatory postsynaptic potential and (D) time to half-decay, were much longer in AH cells than in S cells.

dependent after-hyperpolarization of a single spike was reduced (see below). When calcium ions were from removed the medium, the afterhyperpolarization and the slow depolarization were depressed. This suggests that the slow depolarization may be a “rebound” response to the large amount of calcium entering the cell during the action potentials or that it is due to recurrent release of the synaptic transmitter which mediates the slow EPSP. A similar slow depolarization following a tetanus was described

by Dun and Minota6 in cells of the inferior mesenteric ganglion

and the superior

cervical

ganglion.

Effect of presynaptic nerve stimulation dependent after -hyperpolarization

on calcium-

The after-hyperpolarization which follows the action potential may reduce the frequency of action potential discharge. During the slow EPSP the rate of action potential discharge is often very high (up to 100 Hz) indicating that the action potential after-

667

Slow synaptic potentials

k+J+-J _+LL&JL

4

nerve

4

4

4

JlOmV

4

Fig. 6. Spontaneous depolarizations of similar time course to the slow excitatory postsynaptic potential. Arrows indicate times of nerve stimulation (10 Hz/3 s). S cell, in atropine (1 PM). Before and during the

second evoked slow excitatory postsynaptic potential and the subsequent spontaneous depolarization, constant current pulses were passed across the membrane. The downward deflections are the resulting electrotonic potentials. Resting potential, - 54 mV.

hyperpolarization was depressed. We therefore examined more closely the effects of presynaptic stimulation on the after-hyperpolarization. Strong stimuli. A typical effect of the slow EPSP on the calcium-dependent after-hyperpolarization following a single action potential is shown in Fig. 7. The time course of the after-hyperpolarization was considerably shortened during the slow EPSP. There was a small reduction in the amplitude of the afterhyperpolarization, despite the fact that membrane depolarization of this degree would be expected to increase slightly the after-hyperpolarization amplitude.22 The after-hyperpolarization slowly returned

to its normal duration as the slow EPSP passed off (Fig. 7). Similar results were obtained when the afterhyperpolarization was induced by several action potentials (Fig. 8). Presynaptic nerve stimulation caused a shortening of the after-hyperpolarization with only a small postsynaptic potential change (Fig. 8A). When a large slow EPSP was evoked by strong presynaptic stimulation, the after-hyperpolarization was completely eliminated (Fig. 8B). Under these circumstances, an increase in the number of action potentials could overcome the effect of the slow EPSP and the after-hyperpolarization reappeared.

A

.

mV

Fig. 7. The after-hypeqolarization is shortened during the slow excitatory postsynaptic potential. Arrows indicate time of presynaptic nerve stimulation (single pulse stimulus). Action potentials were evoked by passing brief depolarizing currents. (A) The after-hyperpolarization duration is much reduced. Peak amplitude is also depressed. (B) Action potentials were evoked at different times during the slow excitatory postsynaptic potential. (C) Repeated action potentials during the slow excitatory postsynaptic potential have much reduced after-hyperpolarizations. AH neurone: Resting potential, - 64 mV.

668

K. Morita

and R. A. North

A DC

I

10

mV

Fig. 8. Presynaptic nerve stimulation shortens or abolishes the after-hyperpolarization. (A) Fifteen action potentials were evoked by passing depolarizing current pulses (5 Hz/3 s). Stimulation of presynaptic nerves (30 Hz/3 s, arrow) evoked a small (3 mV) slow excitatory postsynaptic potential. [The focal stimulation also excited a cell process. The first two pulses caused action potentials which propagated into the soma (upward deflection at arrow) and these action potentials were followed by a small after-hyperpolarization, which precedes the slow excitatory postsynaptic potential.] The duration of the after-hyperpolarization evoked by 15 action potentials was reduced during the slow excitatory postsynaptic potential. This was still apparent after hyperpolarizing the membrane (DC). Last part of record is control after a break of -53 mV. (B) Another neurone. The presynaptic nerve stimulus was 20 Hz/3 s 40 s. Resting potential, (arrow). This caused a slow excitatory postsynaptic potential which gave rise to a few spontaneous action potentials. The after-hyperpolarization was almost eliminated during the slow excitatory postsynaptic potential and slowly recovered. Note that depolarization of the cell to the same membrane potential as the peak of the slow excitatory postsynaptic potential (DC) did not depress the after-hyperpofarization. Resting potential, -61 mV.

The depression of the calicum-dependent afterhyperpolarization was not dependent on the membrane potential change. The after-hyperpolarization is depressed even when it is evoked at the control resting potential level. (Fig. 8). Also, the slow afterhyperpolarization often remained depressed for 2&60 s after the potential change of the slow EPSP had passed off (e.g. Fig. 9A). The fast after-hyperpolarization following the action potential, which is not eliminated in calcium-free solutiAns,27 was not reduced in amplitude or duration during the slow EPSP. In fact, the amplitude of the fast after-hyperpolarization was usually increased as a result of the increased driving force for its generation.‘4,27 The clear difference between the effects on the fast (calcium independent) and the slow (calcium dependent) after-hyperpolarizations is illustrated in Fig. 9. Weak stimuli. During the course of experiments involving exogenous application of ACh3’ and substance P,29 it was observed that low concentrations of these agonists caused a shortening of the afterhyperpolarization following one or more action potentials. Only when the concentrations were increased did the substances cause a depolarization and conductance decrease. This suggested that it may be possible to detect the release of the transmitter from presynaptic nerves by observing its effects on the

after-hyperpolarization. The result of such an experiment is shown in Fig. 10. Stimulation of presynaptic nerves with a low voltage caused no change in the membrane potential or input resistance of the impaled neurone. But the same stimulus to the nerves reduced the duration of the after-hyperpolarization. The after-hyperpolarization following six or more action potentials typically had two components in its decay.22 Presynaptic nerve stimulation depressed particularly the second later component of the afterhyperpolarization (Fig. 10). Similar results were obtained in four other neurones when the intensity of presynaptic nerve stimulation was reduced to a value which did not cause any change in postsynaptic membrane potential or input resistance. It is not likely that the transmitter was affecting the amount of calcium entering the cell during the action potentials because the nerve stimulation caused no change in the peak amplitude of the afterhyperpolarization. Morita et al.” and North and Tokimasa” found that a reduction in calcium entry with cobalt reduced the peak amplitude rather than the duration of the after-hyperpolarization. More importantly, the effect of presynaptic nerve stimulation on the late part of the after-hyperpolarization was still observed even when it was applied after the action potentials had occurred (Fig. 10). This experiment suggests that the postsynaptic action of the

669

Slow synaptic potentials

transmitter is to interfere with an intracellular process which couples the entry of calcium to the increase in potassium conductance. The shortening of the after-hyperpolarization can be a significantly excitatory form of synaptic transmission, if the after-hyperpolarization itself normally limits the frequency of action potential discharge. Spontaneously firing cells are occasionally observed in the isolated ganglia of the myenteric plexus. An experiment in which presynaptic nerve stimulation Led to a marked increase in the frequency of the spontaneous discharge is illustrated in Fig. 11. When the strength of the presynaptic stimulus was increased, the synaptic excitation also increased and it was then accompanied by a clear membrane depolarization. DISCUSSION

Presynaptic stimuli

Single pulse stimuli are often effective in eliciting slow EPSPs in myenteric neurones, particularly S cells. In many cells where one pulse is without effect, two pulses separated by 50400 ms produced a postsynaptic response. This finding increases the likelihood that the slow EPSP plays a physiological role

in transmission from cell to cell in the myenteric plexus. On the other hand, it must be kept in mind that a single electrical pulse presumably initiates action potentials almost synchronously in several presynaptic fibres. It is desirable to study the effect of stimulating a single presynaptic fibre. Neurones of the guinea-pig submucous plexus also exhibit slow EPSPs in response to a single pulse presynaptic stimulus, 34 but in other parts of the autonomic nervous system repetitive stimulation is required to elicit the slow potentials.‘~9*‘p~z5~26 The dependence of the slow EPSP amplitude on the frequency of presynaptic stimulation, particularly the effectiveness of two pulses as opposed to one pulse, is difficult to interpret. It is not clear whether two presynaptic pusles release more than twice as much transmitter as one pulse, or whether the postsynaptic cell shows a strong non-linearity with respect to its response to given amounts of transmitter which reach it. The observation of spontaneous depolarizations of a very similar time-course to the evoked synaptic potential could be interpreted in terms of spontaneous release of packets of transmitter. This seems unlikely-because in some cells the spontaneous events could be mimicked only by repetitive stimu-

A

-J

10 mV 10 8

Fig. 9. Slow after-hyperpolarization, but not fast after-hyperpolarization, is depressed during slow excitatory postsynaptic potential. (A) A single action potential (evoked by passing a brief depolarizing pulse) was followed by a fast after-hyperpolarixation and a slow after-hyperpolarization. The fast after-hyperpolarization appears as a single downward deflection, but was fully resolved by the pen recordet. The record illustrates the effects of presynaptic nerve stimulation (arrows, 5 Hz/l s) on two occasions. On the second occasion, the resulting stow excitatory postsynaptic potential was eliminated after its onset by passing current through the recording electrode. The broken line indicates the outline of the slow excitatory postsynaptic potential which would have otherwise occurred. (B) A similar experiment in which the after-hyperpolarixation followed 6 or 7 action potentials at 2 Hz. All three traces are continuous. Downward deflections on final part of trace are hyperpolarixations caused by repeatedly passing a brief current across the cell membrane to monitor input resistance. Resting potential, -61 mV. N.S.C. 14124

670

K. Morita and R. A. North

“resting” or “leak” conductance, because, insofar as can be determined by voltage recording, it is essentially independent of membrane potential between -60 and - 110 mV. This conductance has been referred as g,,ca because it requires calcium entry during the action potential to trigger its increase. However, it is not known whether its long duration in AH cells results from a prolonged period during which intracellular calcium is elevated, or from another long lasting intracelllular metabolic process which is simply triggered by the entry of calcium. The difference in the amplitude of the slow EPSP between S and AH cells was largest for the single pulse stimulus but was present at all frequencies.

20 mV

6

VI

Fig. 10. Low intensity stimulation of presynaptic nerves reduces after-hyperpolarization without changing resting membrane potential or resistance. (1) Nerve stimulation (single pulse, arrow) had no effect. (2) Afterhype~ola~zation following 9 action potentials at 10 Hz was clearly biphasic. (3) The slower component of the afterhyperpolarization was reduced by nerve stimulation. (4) Nerve stimulation had a similar effect when applied after the action potentials, implying that it does not result from a change in calcium entry. (5) Control response, repeated. Resting potential, - 59 mV.

lation of several presynaptic fibres. This suggests the view that the spontaneous depolarization occurs postsynaptically, and that their time-course is a “programmed” property of the postsynaptic cell. According to this scheme, the effect of the synaptic transmitter may be to initiate or trigger this programme. Postsynaptic

mechanisms

Several of the present findings implicate the potassium conductance activated by intracellular calcium (gKca) in the generation of the slow EPSP. First, the symmetrical reversal of the slow EPSP (Fig. 3) is matched by the reversal of afterthe hyperpolarization. z2 Second, there are differences in the slow EPSP between S and AH cells, and these cell types differ also in their expression of gK,Ca. Third, presynaptic stimuIation directly reduces the calciumafter-hy~rpo~ari~tion. These three dependent points will be discussed in turn. The conductance which is increased by calcium entering during the action potential and that which is reduced during the slow EPSP is best described as a

Fig. 11. Shortening of after-hyperpolarization induced by presynaptic nerve stimulation increases firing frequency of postsynaptic neurone. This AH neurone was firing spontaneously at an apparent resting potential of - 55 mV. Presynaptic nerves were stimulated at 10 Hz/3 s (arrows). With 6V intensity, an acceleration of firing results from the shortened after-hyperpolarization. When the stimulus intensity was increased to 12 V a membrane depolarization was also apparent. The stimulation at 10 and 12 V also excited a cell process; the action potential propagated into

the soma and caused the initial after-hyperpolarization.

Slow synaptic potentials

There was a more striking difference in the duration of the synaptic potential between the two cell types (Fig. 5). This finding suggests that the property of the AH cell which results in the long-lasting afterhyperpolarization might also be involved in the generation and decline of the synaptic depolarization. Grafe et al.’ first pointed out that the calciumdependent after-hyperpolarization was reduced during the slow EPSP. They concluded that this resulted from a reduction in calcium entry during the action potential. Our experiments suggest otherwise. Nerve stimulation was effective in reducing the afterhyperpolarization even when applied after the calcium entry has occurred, implying that the synaptic transmitter modulates the process whereby a rise in intracellular calcium leads to an increase in potassium conductance. This interpretation is satisfactory in the sense that it agrees with effects of ACh. Ionophoretic application of ACh to a myenteric neurone immediately after a burst of action potentials effective in afteris still reducing the hyperpolarization.3’ The transmitter mediating the slow EPSP in each of the present experiments is not known with certainty. Focal stimulation might be expected to excite a wide variety of neuronal elements ranging in the myenteric plexus. In S cells, the slow EPSP evoked by a single pulse stimulus is abolished by atropine.30 In AH cells, single pulse stimuli evoked slow EPSPs even in the presence of atropine.4 Repetitive pulse stimulation evokes atropine-insensitive slow EPSPs in most S and AH cells’4 and there is considerable evidence that substance P is the mediator of these potentials.4J3 It is therefore reasonable to assume that ACh and substance P are the transmitters released by the presynaptic fibres in the present experiments. Acetylcholine and substance P have very similar actions on myenteric neurones. At low concentrations, their only effect is to shorten the afterhyperpolarization which results from calcium entering the neurone during the action potential.‘6x29,3’At higher concentrations, they reduce membrane potassium conductance and depolarize the neurone.‘6,24 It was previously noted that the slow EPSP mimicked in time-course and ionic mechanism these actions of ACh and substance P.14 The present studies confirm and extend these findings. They also draw attention to a second similarity between the synaptically released transmitter and exogenously applied ACh or

671

substance P, this is the reduction in the afterhyperpolarization. There have been several recent reports that neurotransmitters can alter calcium-activated potassium Acetylconductances in mammalian neurones. choline,24s3’substance PI6 and 5-hydroxytryptamine3’ all decrease the duration of the afterhyperpolarization which follows the action potential in myenteric neurones. Muscarinic agonists have a similar action in bullfrog sympathetic ganglion cells3’ Clonidine (acting on a,-adrenoceptors) and opioids prolong the after-hyperpolarization in myenteric neurones.36 Similar effects have been described in hippocampal pyramidal neurones; dopamine2’0 increases calcium-dependent the after-hyperpolarization, whereas histamine,” noradrenaline (acting on /I-adrenoceptors)l”*” and ACh (acting on muscarinic receptors)5 all reduce the after-hyperpolarization duration. In a few of these studies, this action of exogenous substances was mimicked by synaptically released transmitters.5~8~35Application of the exogenous transmitter, or stimulation of the presynaptic nerves, resulted in a shortening of the calcium-dependent after-hyperpolarization such as we observed. Nerve stimulation also caused a membrane depolarization which was apparently due to a reduction in the resting potassium conductance of the membrane 3.5,‘4.29.30 These results might be taken to suggest that the most sensitive measure of postsynaptic action is the shortened after-hyperpolarization, and only at higher levels of presynaptic nerve activity do resting potassium channels close in the postsynaptic cell. Whether or not these two actions utilize the same postsynaptic intracellular metabolic paths remains to be shown directly. However, the modulation of afterhyperpolarizations by synaptic transmitters does intoduce a further aspect into neurone-neurone communication because the effect of the presynaptic nerves will become strongly dependent on the level of postsynaptic cell firing. It also implies that the failure to detect slow postsynaptic potential does not necessarily mean that slow muscarinic or noncholinergic transmission is not present. Acknowledgements-This work was supported by U.S. Department of Health and Human Services grants AM32979, DA03 160 and DA03 161. We are most grateful to L. Fillion for preparing the manuscript.

REFERENCES

1. Ashe J. H. and Libet B. (1981) Orthodromic production of non-cholinergic slow depolarizing response in the superior cervical ganglion of the rabbit. J. Physiol., Lond. 320, 333-346. 2. Bemado L. S. and Prince D. A. (1982) Dopamine action on hippocampal pyramidal cells. J. Neurosci. 2, 415-423. 3. Bernard0 L. S. and Prince D. A. (1982) Cholinergic excitation of mammalian hippocampal pyramidal cells. Brain Res. 249, 315-331. 4. Bomstein J. C., North R. A., Costa M. and Furness J. B. (1984) Excitatory synaptic potentials due to activation of neurones with short projections in the myenteric plexus. Neuroscience. 11, 723-732.

5. Cole A. E. and Nicoll R. A. (1983) Acetylcholine mediates a slow synaptic potential in hippocampal pyramidal cells. Science, N.Y. 221, 1299-1300.

612

K. Morita and R. A. North

6. Dun N. I. and Minota S. (1982) Post-tetanic depolarization in sympathetic neurones of the guinea-pig. J. Physiol., Lond. 323, 325-337.

7. Ginsborg B. L. (1967) Ion movements in junctional transmission. Pharmac. Rev. 19, 289-316. 8. Grafe P., Mayer C. J. and Wood J. D. (1980) Synaptic modulation of clacium-dependent potassium conductance in myenteric neurones in the guinea-pig. J. Physiol., Lond. 305, 235-248. 9. Griffith W. H., Gallagher J. P. and Shinnick-Gallagher P. (1980) An intracellular investigation of cat vesical pelvic ganglia. J. Neurophysiol. 43, 343-354. 10. Haas H. L. and Konnerth A. (1983) Histamine and noradrenaline decrease calcium-activated potassium conductance in hippocampal pyramidal cells. Nature 302, 432-434. Il. Hirst G. D. S., Holman M. E. and Spence I. (1974) Two types of neurones in the myenteric plexus of duodenum in the guinea pig. J. Physiol. Lond. 236, 303-326. 12. Jan Y. N., Jan L. Y. and Kuffler S. W. (1979) A peptide as a possible transmitter in sympathetic ganglia of the frog. Proc. natn. Acad. Sci. U.S.A. 76, 1501-1505.

13. Johnson S. M., Katayama Y., Morita K. and North R. A. (1981) Mediators of slow synaptic potentials in the myenteric plexus of the guinea-pig ileum. J. Physiol., Lond. 320, 175-186. 14. Johnson S. M., Katayama Y. and North R. A. (1980) Slow synaptic potentials in guinea-pig myenteric neurons. J. Physiol., Lond. 301, 505-576.

15. Katayama Y. and Nishi S. (1982) Voltage-clamp analysis of peptidergic slow depolarizations in bullfrog sympathetic ganglion cells. J. Physiol., Lond. 333, 305-314. 16. Katayama Y., North R. A. and Williams J. T. (1979) The action of substance P on neurones of the myenteric plexus of the guinea-pig small intestine. Proc. R. Sot. E 206, 191-208. 17. Kuffler S. W. and Sejnowski T. J. (1983) Peptidergic and muscarinic excitation at amphibian sympathetic synapses. J. Physiol., Lond. 341, 257-278.

18. Libet B. (1964) Slow synaptic responses and excitatory changes in sympathetic ganglia. J. Physiol., Lond. 174, l-25. 19. Libet B. (1967) Long latent periods and further analysis of slow synaptic responses in sympathetic ganglia. J. Neurophysiol. 30, 494-5 14.

20. Madison D. V. and Nicoll R. A. (1982) Noradrenaline blocks accommodation of pyramidal cell discharge in the hippocampus. Nature 299, 636-638. 21. Morita K. and North R. A. (1981) Clonidine activates potassium conductance in myenteric neurones. Br. J. Pharmac. 74, 419-418.

22. Morita K., North R. A. and Tokimasa T. (1982) The calcium-activated potassium conductance in guinea-pig myenteric neurones. J. Physiol., Lond. 329, 341-354. 23. Morita K., North R. A. and Tokimasa T. (1982) Muscarinic presynaptic inhibition of synaptic transmission in myenteric plexus of guinea-pig ileum. J. Physiol., Lend. 333, 141-149. 24. Morita K., North R. A. and Tokimasa T. (1982) Muscarinic agonists inactivate potassium conductance of guinea-pig myenteric neurones. J. Physiol., Lond. 333, 125-139. 25. Nishi S. (1974) Ganglionic transmission. In The Peripheral Nervous System (ed. Hubbard J. I.). Plenum, New York. 26. Nishi S. and Koketsu K. (1968) Early and late afterdischarges of amphibian sympathetic ganglion cells. J. Neurophysiol. 31, 109-121.

27. Nishi S. and North R. A. (1973) Intracellular recording from the myenteric plexus of the guinea-pig ileum. J. Physiol., Land. 231, 471-491.

28. North R. A. (1985) Mechanisms of autonomic integration. In Handbook OfPhysiology, section I. The Nervous System, vol. 1. (in press). 29. North R. A., Morita K. and Tokimasa T. (1983) Peptide actions on autonomic nerves. In Systemic Role of Regulatory Peptides (ed. Bloom S. R., Polak J.), pp. 77-88. Schattauer, New York. 30. North R. A. and Tokimasa T. (1982) Muscarinic synaptic potentials in guinea-pig myenteric neurones. J. Physiol., Lond. 333, 151-156.

31. North R. A. and Tokimasa T. (1983) Depression of calcium-dependent potassium conductance by muscarinic agonists. J. Physiol., Lond. 342, 253-266.

32. North R. A. and Tonini M. (1977) The mechanism of action of narcotic analgesics in the guinea-pig ileum. Br. J. Pharmac. 61, 54&541.

33. Skok V. I. (1973) Physiology of Autonomic Ganglia. Igakushoin, Tokyo. 34. Surprenant A. (1985) Slow excitatory synaptic potentials recorded from guinea-pig submucosal plexus. J. Physiol., Lond. (in press). 35. Tokimasa T. (1985) Muscarinic agonists depress calcium-dependent gx in bullfrog sympathetic neurones. J. Auton. Nerv. Sys. (m press). 36. Tokimasa T., Morita K. and North R. A. (1981) Opiates and clonidine prolong calcium dependent afterhyperpolarizations. Nature 294, 162-163. 37. Wood J. D. and Mayer C. J. (1979) Serotonergic activation of tonic-type enteric neurones in guinea-pig small bowel. J. Neurophysiol. 42, 582-593. (Accepted 13 September 1984)