Neurons dissociated from rat myenteric plexus retain differentiated properties when grown in cell culture

Neuroscience Vol. 16, No. I, pp. 201-21 I, 1985 Printed in Great Britain

0306-4522/85 $3.00 + 0.00 Pergamon Press Ltd 0 1985 IBRO

NEURONS DISSOCIATED FROM RAT MYENTERIC PLEXUS RETAIN DIFFERENTIATED PROPERTIES WHEN GROWN IN CELL CULTURE II. ELECTROPHYSIOLOGICAL PROPERTIES AND RESPONSES TO NEUROTRANSMITTER CANDIDATES A. L. WILLARD* and R. Nrs~lt Department of Neurobiology, Harvard Medical School, 25 Shattuck Street, Boston, MA 02115, U.S.A. Abstract-We have used intracellular recordings to study the electrophysiological and pharmacological properties of neurons that have been grown in cell cultures after having been dissociated from the myenteric plexus of the small intestine of newborn rats. Studies of action potential mechanisms revealed that all of the neurons could generate Na+-dependent action potentials in the presence of Ca2+-channel blockers and that about 70% could generate Ca *+-dependent action potentials when Na + channels were blocked with tetrodotoxin. No neurons generated long afterhyperpolarizations after single action potentials but about 50% of neurons did so following trains of action potentials. Over 95% of the neurons tested accommodated rapidly to sustained depolarization. The effects of several enteric neurotransmitter candidates were studied by superfusing or pressureejecting test solutions while recording neuronal responses. All of the cultured neurons tested had nicotinic responses to acetylcholine. Subsets of neurons responded to muscarinic cholinergic agonists (slow depolarization and increased excitability), serotonin (fast depolarization or slow depolarization and increased excitability), y-aminobutyrate (fast depolarization), substance P (slow depolarization, biphasic fast and slow depolarization or increased excitability without a change in membrane potential), vasoactive intestinal peptide (slow depolarization and increased excitability), or [Metlenkephalin (slow hyperpolarization and/or decreased action potential duration). We conclude that myenteric neurons grown in cell culture retain many of the physiological and pharmacological properties that they have in situ. Such cultures will permit detailed biophysical and pharmacological studies of the mechanisms of action of enteric neurotransmitter candidates.

nists, y-aminobutyrate (GABA), neurotensin, opiates and opioid peptides, serotonin, somatostatin, substance P and vasoactive intestinal peptide (VIP) (reviewed in Refs 15 and 37). In the preceding paper35 we described conditions for the long-term growth of myenteric neurons that have been dissociated from the small intestine of newborn rats. We presented morphological and immunohistochemical evidence that neurons with properties similar to those of their in vivo counterparts survived and were maintained in our cultures.35 In this paper we describe some of the electrophysiological properties and responses to neurotransmitter candidates of myenteric neurons grown in dissociated cell cultures. Some of the results in this paper have been reported in abstract.47.4*

Neurons in the myenteric plexus of the small intestine of the adult guinea-pig,***36in explant cultures of the myenteric plexus of the cecum of the juvenile guineapig’824 and in cultures of proliferating neurons derived from the myenteric plexus of fetal human intestines29 can be classified according to whether they have one or more of the following properties: (1) the ability to fire action potentials in solutions containing tetrodotoxin (TTX), (2) a long (greater than 500 ms) afterhyperpolarization (LAH) following a single action potential or (3) the ability to fire repetitively during sustained depolarization. Myenteric neurons of adult guinea-pig are also heterogeneous with respect to their responses to neurotransmitter candidates. One or more types of response to each of the following have been observed in subsets of myenteric neurons: acetylcholine (ACh), adrenergic ago-

EXPERIMENTAL

*Present address: Department of Physiology, Medical Research Building 206H, University of North Carolina, Chapel Hill, NC 27514, U.S.A. tTo whom all correspondence should be addressed. Abbreviations : ACh, acetylcholine; AH, afterhyperpolarization; GABA, y-aminobutyrate; 5-HT. 5-hydroxytryptamine (serotonin); LAH, long afterhyperpolarization; SP, substance P; TEA, tetraethylammonium; TTX, tetrodotoxin; VIP, vasoactive intestinal peptide.

PROCEDURES

Culture preparation is described in the preceding paper.)’ All experiments described in this paper were performed on cultures 2-13 weeks old. Electrophysiological

procedures

Cultures were placed on the heated stage of an inverted microscope and were continuously superperfused by a peristaltic pump that maintained constant bath volume (0.1-0.2 ml) and flow rate (1 ml/min). A detailed description of the heating apparatus and of the arrangement of the perfusion inlet and outlet ports is given by O‘Lague et aL4’ 201

202

A. L. WILLARDand R. NISHI

Neurons were viewed with phase-contrast optics at 250- or 400-fold magnifications and were impaled with intracellular microelectrodes (IO&200 Ma when filled with 4 M potassium acetate at pH 7.0 or 70-150 MR when filled with 3 M KCI). The electrodes were connected to an amplifier that had an active lo-kHz filter and an active bridge circuit to allow simultaneous current injection and voltage recording (Getting Microelectronics, Iowa City, IA; Model 5). Signals were displayed on an oscilloscope and a chart recorder. Drug application Responses to transmitter candidates were tested by recording from neurons while applying test compounds either by adding them to the perfusion medium or by using a solenoid controlled pressure source to “puff’ them from a blunt micropipette (5-20pm OD) as described by Choi and Fischbach.’ Fast green (20pgg/ml) was frequently included in the test solutions to permit visualization of the flow of solutions out of the micropipette. When testing different concentrations of a particular compound, the same micropipette was used for each of the test solutions; a grid in an eyepiece of the microscope was used to insure that the tip was placed in the same position relative to the neuron being tested after the solution in the micropipette had been changed. During attempts to block responses, antagonists were included both in the puffer pipette solution and in the perfusion fluid. Perfusion medium consisted of Hanks Balanced Salt Solution (79% v/v), Eagles Minimal Essential Medium (20% v/v), rat serum (1% v/v), 3.6mM CaCI,, I .8 mM choline chloride, 2 mM glutamine, 20 U/ml penicillin and 2 mg/ml streptomycin. In some experiments the rat serum was omitted from the perfusion medium. When testing peptide-transmitter candidates, 2.5 mg/ml of bovine serum albumin (Sigma; Fraction V, essentially fatty acidfree) was included to minimize adsorption of the peptide to the glass micropipette. In addition the inside of the pipette was preincubated with the bovine serum albumincontaining solution for I-IOmin before filling it with the peptide-containing solution. Materials

All drugs were from Sigma except VIP (Boehringer), substance P, [Metlenkephalin, somatostatin (Peninsula), methysergide maleate, lysergic acid diethylamide, mianserin hydrochloride (gifts from Dr E. Kraviti), bufotenine (gift from Dr R. Goyal) and tetraethylammonium chloride

(Eastman). RESULTS Resting membrane properties

The results in this paper were obtained almost exclusively from the larger (diameters greater than lO-12pm) cells in the cultures. Impalements were considered acceptable if the neurons had stable resting potentials larger than -45 mV (range -45 to -75) and could fire action potentials in response to injection of depolarizing current pulses. The frequency of obtaining acceptable impalements depended on the solution used to fill the microelectrodes; successful impalements were obtained about 3 times more frequently with electrodes filled with 4 M potassium acetate than with those filled with 3 M KCl. Although most neurons had stable membrane properties when impaled with potassium acetate electrodes, about 20% of neurons impaled with potassium acetate electrodes spontaneously hyperpolarized 10-20 mV during the first l-4 min after having been impaled; this hyperpolarization was ac-

companied by a substantial drop in input resistance. possibly due to activation of a K+ conductance by Ca2+ leaking into the ce1l.‘4 This hyperpolarization decreased over the next 5-15 minutes (presumably because of the membrane sealing around the microelectrode and intracellular buffering of the Ca2+). after which resting potentials were stable (+ 3 mV) for up to 8 h. Input resistances ranged from about 7&500MR, apparently depending on the quality of impalement. When the same cell was impaled with 3 different microelectrodes, the input resistance and resting potential were stable and different: presumably each reflected the summed result of the relative leaks and damage due to that particular microelectrode. The largest values of the input resistance were assumed to be the ones that most closely approximated those of the unimpaled cells. Recent reports of intracellular recordings obtained with patch electrodes have shown that the largest values obtained with conventional intracellular microelectrodes may underestimate the true input resistance by as much as 5-10 fold.12 Active membrane properties Action potential mechanisms. The relative contributions of Na+and Ca2+ currents to the action potentials of 53 cultured enteric neurons were examined by testing their ability to fire action potentials in solutions that block Na+ and/or Ca*+ channels. Most cells (37/53) fired action potentials in solutions containing either Na + or Ca2+ channel blockers (Fig. 1). The other 16 neurons did not fire action potentials when Na+ channels were blocked with tetrodotoxin (TTX) (Fig. 2). Thus all of the cultured enteric neurons had enough Na + channels to generate action potentials in solutions of 500pM CdCl, or 10mM CoCl, and most also had enough Ca2+ channels to allow action potentials in 1 FM TTX. Seven of the 16 neurons with TTX-sensitive action potentials were tested for their ability to fire action potentials in solutions including both TTX and 5 mM tetraethylammonium (TEA), a blocker of K+ channels that repolarize most neurons. Five of the 7 neurons were able to fire action potentials in such solutions (Fig. 3). No neurons could fire action potentials in solutions containing both TTX and Cd2+ or Co*+, regardless of whether TEA was added. The potassium conductances that cause repolarization of the action potential were not examined in detail, but both TEA (20/20) and 4-aminopyridine (15/l 5) prolonged action potentials. In several cells, Ca*+-channel blockers (Co*+ or Cd*+) caused a slight prolongation of the action potentials (Fig. lC), suggesting that a Ca*+-activated potassium current may participate in action potential repolarization as has been reported for bullfrog sympathetic neurons.2*22 Long afterhyperpolarizations. No neurons of over 700 tested were observed to have a long (greater than

203

Membrane properties of cultured myenteric plexus neurons

A- /r-

RTTX I

A. cant rol

_/ C. m

n !J\

2’ 40mV

L

20 msec

Fig. 1. Mixed Na+/Ca2+dependent action potential in a cultured myenteric neuron. (A) An intracellular microelectrode was used to record the response of the neuron to injection of depolarizing current. The addition of the Na+thannel blocker ‘ITX (1 PM; trace B) or the Ca2+-channel blocker CdCl, (SOOpM; trace. C) to the perfusion fluid did not completely block the action potential of this neuron. When both blockers were added, the action potential was prevented (trace D).

-\,

\--.

E.WShl 40mV L 10msfx Fig. 3. Ca’+dependent action potential unmasked by K+ channel blockade. The action potential of a cultured myenteric neuron was blocked by 1 PM TTX (trace B). When 5 mM TEA was added to the perfusion fluid, the neuron could fire action potentials in TTX (trace C). This ‘RXresistant action potential was prevented by 5OOpM CdCl, (trace D).

0.5) afterhyperpolarization (AH) following single action potentials. The longest AH we observed following a single action potential lasted 120 ms; it was rare to observe an AH longer than 80 ms. The average AH following a single action potential lasted 48 ms (n = 200). About half of neurons tested (187/394) had an LAH following a train of action potentials (Fig. 4). Such LAHs ranged in size from 2 to 17 mV and lasted from 1 to 55 s. The length of the LAH depended on the length and frequency of the train and reached a maximal length after about 64 action potentials at 10 Hz. There appeared to be two phases to the LAHs that followed trains. The initial phase was accompanied by a decrease in membrane resistance (Fig. 4) while the later (and often longerlasting) phase was not accompanied by a detectable

Fig. 2. Largely Na + dependent action potential. The intracellularly evoked and recorded action potential persisted in 500 p M CdCl, (trace B) but was blocked by 1 p M TTX (trace C).

Fig. 4. Chart recordings of long afterhyperpolarizations (LAH) following trains of action potentials. A train of 60 action potentials was elicited by intracellular injection of depolarizing current pulses at 20 Hz for 3 s. At the end of the tram, the neuron remained hyperpolarized for more than 30 s (top trace). Injection of hyperpolarizing current pulses revealed that the initial phase of the LAH was accompanied by a decrease in membrane resistance (middle trace). The decreased resistance persisted when the membrane potential was manually clamped at its resting level (bottom trace). The action potentials in this and other chart recordings have been attenuated by the chart recorder.

204

A. L. WILLARD and R. NEHI

5sec

Fig. 6. Reversal of an LAH. Separate recording and current-injecting electrodes were inserted. Trains of action potentials were evoked before and after the membrane potential had been changed by steady injection of hyperpolarizing current. Top trace: an LAH elicited at resting potential (- 50 mV). Middle trace: both the LAH and the AHs following single action potentials are null when the membrane potential is at -90 mV. Bottom trace: at - 115 mV, the LAH has been reversed to an after depolarization. Fig. 5. Prevention of an LAH by Caz+-channel inhjbition.

An LAH was elicited by firing action potentials at 10Hz for 5-7 s. The LAH was not prevented by I PM TTX (second trace) but was prevented by 500PM CdCl, (third trace). The top and bottom traces show the LAH before and after treatment with the test solutions.

neurons that fire repetitively are more commonly found {Ref. 34 and unpublished data). Responses to neurotrartsmitter candidates Acetylcholine. Brief (50-300 ms) “puffs” of ACh

change in resistance. This is similar to the observations made by Morita er ~1.~’on guinea-pig my enteric neurons. The LAI-I was sensitive to removal of extracellular Ca*+ and to Ca2+-channel blockers (Fig. 5), suggesting that Ca*+ entry was required for its activation. It was not blocked by TTX (Fig. 5). Four neurons that had an LAH were impaled successfully with separate recording and current passing electrodes. In each of these neurons, it was possible to reverse the LAH and in each case, the estimated reversal potential was around - 90 mV (Fig. 6), the same potential at which the undershoot following the action potential reversed. Repetitivefiring. Most of the cultured neurons that we tested (378/394) accommodated rapidly to long depolarizing pulses, firing only one or a few action potentials at the beginning of the pulse. The number of action potentials they fired at the beginning of a depola~~tion seemed to be unrelated to whether the neuron had an LAH. Under other culture conditions,

(lO-100pM)

caused all neurons tested (236/236) to

5,uM Hex.

50pM Hex.

4

Fig. 7. Nicotinic responses to ACh. Brief (200 ms) pulses of pressure were used to “puff” ACh from a pipette containing 100 pM ACh at the times indicated by the arrows. When the pipette was positioned near the neuron, rapid depolarizations were elicited. Large depolarizations were followed by h~~oia~~tions. The depolarizations were diminish~ by 5 FM and were blocked by 50 p M hex~ethonium. They were not si~i~cantly changed by atropine (1 pM).

205

Membrane properties of cultured myenteric plexus neurons depolarize rapidly by 5-50 mV (Fig. 7). The responses were larger when the cells were hyperpolarized and were smaller when the cells were depolarized. They were accompanied by a large increase in conductance (Fig. 8C). When relatively large depolarizations were elicited by application of AC4 they were often followed by a hyperpolarization (Fig. 7). The response to ACh desensitized when ACh was allowed to leak out of the pipette or when pulses were given in rapid succession. These responses were blocked by 1~lOO~M hexame~onium or curare and were unaffected by 1 FM a&opine or scopolamine (Fig. 7). When longer (OS-2 s) puffs were applied, a small number of cells (22/159) had biphasic responses to ACh (Fig. 8). The later, longer-lasting component was blocked by the muscarinic antagonists atropine or scopolamine (Fig. 8C) and could be elicited independently of the nicotinic response by muscarine or oxotremorine. The muscarinic reponse was accompanied by a conductance decrease (Fig. 8B). During muscarinic responses, the neurons often discharged many action potentials and were capable of repetitive firing. &Y&L&Z. Two effects of puffed serotonin (5-HT) (100 ~1M) were observed. One was a fast depolarization accompanied by a large conductance increase (Fig. 9). Such responses were observed in 21/93 neurons tested. These responses became larger when the neuron was hy~~larized and became smaller

A. Control

15mV

n

control

\

lllJ-

4

5-HT

-JL

curare

Hex

4

20 mV

L 10 set

Fig. 9. Rapid depolarizations of a cultured myenteric neuron caused by serotonin. Top: brief puffs of 1OOpM 5-HT (arrows) caused a rapid depolarization and a decrease in resistance. Middle: the response was not noticeabIy changed by hexamethonjum (1 mM) but was substantially diminished by &tubocurare (50 pM). Bottom: the response recovered after the curare was washed out.

when the neuron was depolarized. These fast response to 5-HT seemed similar in duration, conductance change and apparent equilibrium potential to the fast nicotinic responses. However, they could be distinguished from nicotinic responses by their pharmacological properties: they were not blocked by the nicotinic antagonist hexamethonium, even at high concentrations but could be blocked by d-tubocurare (Fig. 9). other agents that did not block the response LSD and included methysergide, mianserin, bufotenine. A different type of response to 5-HT was observed in 26f93 neurons. This was a slow depolarization accompanied by a small conductance decrease (Fig. 10). This effect could be partially blocked by meth-

‘I 10 SBC

Fig. 8. Mixed nicotinic and muscarinic cholinergic responses. (A) In control perfusion fluid, a l-s puff of 100 p M ACh (arrow] elicited a rapidly rising, long-lasting depolarization, accompanied by spiking of the neuron. Changes in the size of the voltage responses to current pulses of constant amplitude revealed a resistance decrease during the initial part of the response while the later part of the response was accompanied by a resistance increase. (B) In 50 ~1M hexamethonium (a nicotinic antagonist), ACh elieited a slow depolarization, a resistance increase and intense spiking. (C) In 500 nM atropine (a muscarinic antagonist), ACh elicited a rapid depolarization and a resistance decrease. (D) No response was observed in the presence of a combination of 50pM hexamethonium and 5OOnM atropine.

N.S.C. 1611-J

20 mV I

15 set

Fig. 10, Slow depolarization caused by serotonin. Brief puffs of 100 PM S-HT (arrows) elicited a small, long-lasting depolarization, accompanied by a resistance increase and spiking activity. The response was diminished reversibly by 10 p M methysergide maleate.

206

A. L. WILLARDand R. NISHI

C Fig. 11. Responses to GABA. (A) Puffs of 100 PM GABA (arrows) elicited a rapid depolarization and a resistance decrease at resting potential (-63 mV, bottom trace). The depolarization was annulled when the membrane potential had been moved to -41 mV (middle trace) and was reversed at - 28 mV (top trace). (B) The response of a different neuron was reversibly diminished by 1OOpM picrotoxin @tx).

ysergide (Fig. 10B) but was not blocked by hexamethonium, curare, bufotenine, lysergic acid diethylamide or mianserin. It could be desensitized by 5-HT. Forty-six of the 93 neurons tested did not respond to S-HT. y-Aminobutyrate. A third transmitter candidate that caused large conductance changes in a significant proportion of neurons (22/26) was GABA. At resting potential GABA caused a rapid depolarization that was easily reversed by injection of depolarizing current (Fig. 11A). This was accompanied by a large conductance increase, apparently to Cl- ions; injection of Cl- from KC1 electrodes caused the reversal potential to move to more positive potentials. This response appeared to be of the GABAA type since it could be blocked by bicuculline or picrotoxin (Fig. 11B) and mimicked by muscimol or l-amino-propane sulfonic acid. Peptides (i) Substance P

Brief puffs of 1 PM substance P (SP) caused several different responses (Fig. 12): a slow depolarization accompanied by a small conductance decrease (18/42), a slow depolarization accompanied by a small conductance increase (4/42) or a biphasic depolarization that consisted of an initial fast response and conductance increase followed by a long depolarization and small conductance decrease (6/42). In 8/32 neurons, a different effect of SP was observed: it caused no change in the resting membrane potential or in the voltage response to small hyperpolarizing current pulses (Fig. 13) but when depolarizing pulses that were just subthreshold were given, substance P caused these pulses to become suprathreshold (Fig.

2 ..-.A _

c

4 set A2, C2 2Osec rest

Fig. 12. Different responses of cultured myenteric neurons to substance P. During the times indicated by the bars, SP (I PM) was puffed onto 3 different neurons. (A) SP elicited a slow depolarization and a resistance increase. (B) In a second neuron, SP elicited a slow depolarization and resistance decrease. (C) In a third neuron, SP elicited a biphasic depolarization, consisting of an initial rapid phase and resistance decrease and a slower, longer lasting phase and resistance increase. Traces (A2) and (C2) are faster recordings of responses from the same neurons as in (Al) and (Cl), respectively.

l-7--7-l l&L-/ 1-u

2@L

5 sec,l5 msac Fig. 13. Altered excitability caused by substance P. Top: a chart-recording of the voltage responses of a neuron to depolarizing and hyperpolarizing current pulses before, during and after a puff of 100 nM SP (indicated by the bar). Middle and bottom traces: oscilloscope records of the current monitor and of the voltage responses indicated by the arrows pointing to the top trace. Although the puff of SP did not cause a change in the membrane potential or in the voltage response to the hyperpolarizing pulses, it did cause a change in the responses to the depolarizing pulses.

Membrane properties of cultured myenteric plexus neurons

iii-P

‘.

TTX

207

neurons also retain a range of electrophysiological properties and of responsiveness to neurotransmitter candidates. To date we have tested only the larger cells (those with diameters greater than 10-12 pm) in the cultures and thus it is likely that there are classes of neurons that we have not yet studied. The properties of the smaller cells in the cultures will be examined by using the whole-cell patch-clamp technique.” In the remainder of this discussion, we will compare the electrical excitability and responses to neurotransmitter candidates of myenteric neurons in our cultures to those of myenteric and other autonomic and central neurons in vivo and in culture. Electrical excitability of myenteric neurons

20 mV

L 1oz

Fig. 14. Excitation by VIP. At the time indicated by the bars, VIP (10 nM) was puffed onto a neuron. Top: the VIP caused a small, long-lasting depolarization and a discharge of action potentials. Bottom: in the presence of TTX (1 PM), which blocked the spiking, a small depolarization and resistance increase were observed.

13). This effect will be explored more fully with whole-cell, patch-clamp techniques. (ii) Vasoactive intestinal peptide Puffed VIP (10 nM) caused a slow depolarization of 2-30mV and a small conductance decrease (Fig. 14) in 23/37 neurons tested. This effect lasted 1S-300 s and aften caused the neuron to discharge a train of action potentials. During a depolarization caused by VIP, neurons could often fire repetitively. (iii) [Metlenkephalin Two inhibitory effects were observed when 10 PM [Metlenkephalin was puffed onto neurons. The first was a 3-14 mV hyperpolarization (Fig. 15A) that was accompanied by a small increase in conductance. This effect was observed in 4/9 neurons tested and could be reversibly diminished by 1 PM naloxone (Fig. 15A). In a different experiment it caused a decrease in the duration of action potentials that had been prolonged by TEA or 4-aminopyridine in 616 neurons tested (Fig. 15B). The pharmacological properties of the receptors causing this effect were not investigated.

Previous studies of neurons from the myenteric plexus of guinea-pigs and humans have found that myenteric neurons can be categorized according to one or more of the following criteria: (1) ability to generate action potentials in TTX, (2) ability to fire repetitively during sustained depolarization and (3) whether or not their action potentials are followed by a long (greater than 500 ms) afterhyperpolarization (LAH). The first and most thoroughly studied enteric neurons have been those of the myenteric plexus of the small intestine of the adult guinea-pig. In that system most neurons can be placed into either of two categories, type 1 (S) or type 2 (AH).22”6J7 Type 1 cells (also called S cells because they receive conspicuous nicotinic synaptic potentials) fire repetitively, do not fire action potentials in TTX and do not have an LAH following single action potentials. Type 2 cells accommodate rapidly to sustained depolarization, fire action potentials in TTX and generate an LAH after firing even a single action potential. This LAH, which is due to the activation of a Ca*+-dependent K*+ current, can list l-30 s30 When myenteric neurons in explant cultures from juvenile guinea-pigs were categorized according to their response to prolonged depolarizations, approximately two-thirds accommodated rapidly while the remainder fired repetitively.15 Very few (2/117) neu-

A.

control

naloxons

wash

DISCUSSION

In the first paper in this series,35 we showed that myenteric neurons retain many of their morphological and immunohistochemical properties when they are grown in dissociated cell cultures. Here we have presented evidence that cultured myenteric

Fig. 15. Responses to [Metlenkephahn. (A) At the times indicated by the bars, 10 PM enkephalin (ENK) was puffed into a neuron. Naloxone (1 PM) reversibly diminished the response. (B) In a different neuron, whose action potential had been prolonged by the addition of TEA (5 mM) to the perfusion fluid, 10 p M enkephalin caused a reversible shortening of the action potential.

208

A. L. WILLARD and

rons in the explants had an LAH following single action potentials; their TTX sensitivity was not reported. In cultures of neurons derived from the myenteric plexus of fetal human intestines, Maruyama29 recorded both TTX-sensitive and TTXresistant action potentials. The relative frequencies of the two categories and whether or not any of them had an LAH was not reported. Excitability of cultured rat myenteric neurons Tetrodotoxin-sensitivity of action potentials. About 70% of the neurons that we tested had TTX-resistant action potentials that could be blocked by solutions containing Ca 2+-channel blockers or lowered extracellular Ca2+. Since some of the neurons with TTXsensitive action potentials could fire action potentials in TTX when Kc-channel blockers (TEA or 4-aminopyridine) were added to the perfusion fluid, the differences in the Ca2+ conductances of the cultured neurons appear to be graded rather than all-or-none. Long afterhyperpolarizations. The longest AH we observed following a single action potential in a cultured neuron lasted about 120 ms and AHs lasting more than 80 m were rare. About half of the cultured neurons generated an LAH following a train of action potentials. Such LAHs were similar to those recorded from myenteric neurons in situ: they were blocked by Ca 2+-channel blockers or by lowering extracellular Caz+, they were accompanied by an increase in conductance, their reversal potential was close to the equilibrium potential for K+ and their duration increased when longer trains of action potentials were used to evoke them. Similar observations of myenteric neurons which lacked an LAH following a single action potential, but which could generate one after having fired a train of action potentials, have been made both in situ and in explant culture.‘8.22*36 Repetitive firing. Fewer than 5% of cells tested could fire long trains of action potentials during sustained injection of depolarizing current; most accommodated after firing fewer than 10 action potentials. In contrast, about 30-60x of guinea-pig myenteric neurons, either in situ or in explant culture, fire repetitively.‘s.22*36,52 Possible explanations for dtjierences. There are two quantitative differences between the electrophysiological properties of myenteric neurons in our cultures and of those in situ. (1) A lower proportion of the cultured neurons continue to fire action potentials during sustained depolarization. (2) More than a single action potential is required to produce an LAH in the cultured neurons. Because successful intracellular recordings are more frequently obtained from larger cells, a relatively trivial (but perhaps the most likely) explanation for the low proportion of cultured neurons that fire repetitively may be that, under these culture conditions, cells that fire repetitively tend to be small.

R.

NISHI

This possibility will be investigated by using whole cell tight seal patch clamp electrodes28 to study the properties of the smaller cells in the cultures. Under different culture conditions, which cause several immunohistochemically distinguishable classes of neurons to become larger, neurons that fire repetitively are found much more frequently (Nishi and Willard. manuscript in preparation; also see Ref. 34). A second possibility arises from the fact that the ability of a myenteric neuron to fire repetitively can be controlled by the synaptic inputs it receives.53 If the cultured neurons do not receive the same mixture of synaptic inputs that they would have received in situ, then they may not be able to fire repetitively. Our observations that several neurotransmitter candidates (muscarinic agonists, SP, 5-HT and VIP) could cause the cultured neurons to become capable of repetitive firing also suggests that a lack of ability to fire repetitively may reflect a lack of necessary synaptic inputs. It is also possible that the ability of the cultured neurons to fire repetitively is being inhibited by synaptic inputs or by components of the culture medium. The difference in the number of action potentials required to evoke an LAH in cultured myenteric neurons capable of producing LAHs could be due to less Ca2+ entry during single action potentials, to different intracellular Ca2+ buffering characteristics or to a difference in the activation kinetics of the Ca2+-dependent K+ channels in these cells. Any of these differences could be due to a species difference (rat vs guinea-pig), to the relative immaturity of the neurons at the time the cultures are made or to an effect of being removed from the intestine and grown in culture. With respect to the latter two possibilities, it is noteworthy that a train of action potentials was usually required in order to evoke an LAH in neurons in explant-cultured myenteric plexus from juvenile guinea-pigs.” Responses to neurotransmitter candidates

We have found that a variety of neurotransmitter candidates have effects on the cultured rat myenteric neurons that are qualitatively similar to their effects on guinea-pig myenteric neurons in situ. Acetyicholine. Ail of the cultured myenteric neurons had nicotinic responses to ACh. These responses are similar to those of myenteric36.43and other autonomic [email protected] The sensitivity of individual neurons varied, but was not obviously correlated with whether the neuron could produce an LAH. Type 2 (AH) neurons in situ are either insensitive37.43or much less sensitiveI than are Type 1 (S) neurons to nicotinic agonists. About 15% of the cultured neurons also had muscarinic responses to ACh. This percentage is similar to the percentage of enteric nerve cell bodies that are labelled in explant culture by tritiated propylbenzilylcholine mustard (an irreversible ligand of muscarinic receptors).5 The muscarinic responses,

Membrane properties of cultured myenteric plexus neurons which are excitatory and accompanied by a conductance decrease, are similar to those of guinea-pig myenteric neurons3’ as well as of other peripheral” and central’9.39 neurons. We observed, as did Morita et uZ.,~’that muscarinic responses could be obtained even at very hyperpolarized membrane potentials. Thus it is likely that at least part of the response is due to a mechanism other than suppression of an M current.’ We did not determine the properties of the action potentials of the neurons with muscarinic responses. In situ, both types of myenteric neurons have mus~nic ACh responses?’ Serotonin. The fast, curare-sensitive excitatory effects of 5-HT are similar to those reported in guinea-pig enteric neurons2’*25and also in sensory’9*20 and sympathetic”,45 neurons. The slow excitatory responses to 5-HT are similar to those seen in myenteric neurons by Johnson ei uf.2s and by Wood and Mayer.5’,54 We did not observe any hyperpolarizing effects of S-HT; in contrast Johnson et af.25 observed hyperpolarizing responses to application of 5-HT in about 30% of myenteric neurons, most of which were Type 2 (AH). ~-~~i~ob~t~rate. The fast increase in Cl- conductance in response to GABA is similar to that reported by Cherubini and North6 and is similar to the GABA+, responses that have been widely found elsewhere (reviewed in Ref. 11). However, we observed such responses in neurons with and without the ability to generate an LAH, whereas Cherubini and North observed GABA responses only on type 2 (AH) neurons. Substance P. The responses to SP included slow depolarizations accompanied by either increased or decreased conductance. Katayama et af.26 and Hanani and Burnstock” also observed slow depolarizations of myenteric neurons in response to SP. However, the slow depolarizations observed by Katayama et al. were always accompanied by a conductance decrease. Our finding that the depolarizations can be accompanied by either an increase or a decrease in conductance is similar to the observations of Dun and Minota” on the effects of SP on guineapig sympathetic neurons. We did not observe isolated fast responses to SP: all fast responses occurred as the initial phase of a biphasic depolarization (Fig. 12, traces 4 and 5). In contrast, Hanani and Burnstock” occasionally observed fast responses in the absence of slow ones. We have not seen reports of SP increasing the excitability of enteric neurons without changing

209

membrane potential, but such effects have been reported in other neurons (e.g. Ref. 33). The action ol SP, which may be due to suppression of a voltage dependent potassium conductance, will be in vestigated more fully with whole-cell tight-seal. recording techniques. Vusoactiue intestinal peptide. There have not been previous reports of intracellularly recorded response: of enteric neurons to VIP, but Williams and North,s using extracellular recordings, observed that it strongly excited myenteric neurons. This finding is consistent with our observation of the excitatory effects of puffed VIP. Jeftinija et ~1.‘~ observed that VIP caused long-lasting depolarizations of dorsal horn neurons of rats. [Metlenkephafin. The inhibitory effects of enkephalin were similar to those that have been reported for myenteric and other neurons. North et uL3* reported that enkephalin caused a naloxone-sensitive hyperpolarization of myenteric neurons. Several groups4*32,36 have observed that enkephalin causes a shortening of Ca’+-dependent action potentials in sensory neurons.

Myenteric neurons retain many of their electrophysiological and pharmacological properties when grown in dissociated cell culture. Because cell culture permits improved experimental access to neurons, detailed biophysi~l and pha~acolo~cal studies of the receptors responsible for the different responses of myenteric neurons to enteric neurotransmitter candidates should now be relatively straightforward. Furthermore, because the cultured myenteric neurons form synaptic connections that are similar in several ways to those in situ,49it is now possible to record sim~taneously from pre- and postsynaptic neurons while testing the effects of the various enteric neurotransmitter candidates on synaptic transmission. Acknowledgements-These experiments were performed in the laboratories of Drs D. D. Potter and E. J. Furshpan, to whom we are very grateful for their advice, support and encouragement. We thank D. Sah and C. Jahr for their comments of this manuscript. This work was supported by postdoctoral fellowships from the NIH (R. Nishi), the Muscular Dystrophy Association (R. Nishi and A. L. Willard) and the Massachusetts Afftliate of the American Heart Association (A. L. Willard) by NIH Training Grant NS07112 (R. Nishi), by NIH grants NSI 1576 and NS18316 to D. Potter and by NIH grant NSZOO74to A. L. Willard, who is an Aifred P. Sloan Research Fellow.

REFERENCES 1. Adams P. R., Brown D. A. and Constanti A. (1982) Pharmacological inhibition of the M current. J. Pkpiol., Land. 332, 223-262. 2. Adams P. R., Constanti A., Brown D. A. and Clark R. B. (1982) Intracellular Ca2+ activates a fast voltage-sensitive K+ current in vertebrate sympathetic neurones. Nature 296, 746-749. 3. Benardo L. S. and Prince D. A. (1982) Ionic mechanisms of cholinergic excitation in mammalian hippicampal pyramidal cells. Brain Res. 249, 333-344. 4. Bixby J. L. and Spitzer N. C. (1983) Enkephalin decreases calcium action potentials in Rohon-Beard neurons in uivo. J. Neurosci. 3, lOlClOl8.

A. L. WILLARDand R. NISHI

210

5. Buckley N. J. and Burnstock G. (1984) Distribution of muscarinic receptors on cultured myenteric neurons. Brain Res.

310, 133-137. 6. Cherubini E. and North R. A. (1984) Actions of GABA on neurones of guinea-pig myenteric plexus. Br. J. Pharmac. 82, 93-100.

I. Choi D. W. and Fischbach G. D. (1981) GABA conductance of chick spinal cord and dorsal root ganglion neurons in cell culture. J. Neurophysiol. 45, 605-620. 8. Dennis M. J., Harris A. J. and Kuffler S. W. (1971) Synaptic transmission and its duplication by focally applied acetylcholine in parasympathetic neurons in the heart of the frog. Proc. R. Sot. Ser. B 177, 509-539. 9. Dodd J., Dingledine R. and Kelly J. S. (1981) The excitatory actions of acetylcholine on hippocampal neurons of the guinea-pig and rat maintained in uifro. Brain Res. 207, 109-127. 10. Dun N. J. and Minota S. (1981) Effects of substance P on neurones of the inferior mesenteric ganglia of the guinea-pig, J. Physiol., Lond. 321, 259-271.

11. Enna S. J. (1983) Biochemical and electrophysiological

characteristics

of mammalian GABA receptors. Inr. Rev.

Neurobiol. 24, 18 1-2 12.

12. Fenwick E., Marty A. and Neher E. (1982) A patch-clamp study of bovine chromaffin cells and their sensitivity to acetylcholine. J. Physiol., Lond. 331, 571-597. 13. Freschi J. and Shain W. G. (1980) Slow muscarinic depolarization in neurons of dissociated rat superior cervical ganglia can be evoked by iontophoresis of acetylcholine. Brain Res. 185, 429-434. 14. Galvan M. and Adams P. R. (1982) Control of calcium current in rat sympathetic neurons by norepinephrine. Brain Res. 244, 135-144. 15. Gershon M. D. (1981) The enteric nervous system. A. Rev. Neurosci. 4, 227-272. 16. Grafe P., Wood J. D. and Mayer C. J. (1979) Fast epsps in AH (type 2) neurons of guinea-pig myenteric plexus. Bruin Res. 103, 349-352. 17. Hanani M. and Burnstock G. (1984) Substance P evokes slow and fast responses in cultured myenteric neurons of the guinea-pig. Neurosci. Left. 48, 19-23. 18. Hanani M., Baluk P. and Burnstock G. (1982) Myenteric neurons express electrophysiological and morphological diversity in tissue culture. J. Aufon. New. Sysr. 5, 155-164. 19. Higashi H. (1977) 5-Hydroxytryptamine receptors on visceral primary afferent neurones in the nodose ganglion of the rabbit. Nature 267, 448450. 20. Higashi H. and Nishi S. (1982) 5-Hydroxytryptamine receptors of visceral primary afferent neurones on rabbit nodose ganglion. J. Physiol., Land. 323, 543-567. 21. Hirst G. D. S. and Silinsky E. M. (1975) Some effects of 5-hydroxytryptamine, dopamines and noradrenaline on neurones in the submucous plexus of guinea-pig small intestine. J. Physiol., Lond. 251, 817-832. 22. 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. 23. Jeftinija S., Murase K., Nedeljkov V. and Randic M. (1982) Vasoactive intestinal peptide excites mammalian dorsal horn neurons both in tivo and in vitro. Brain Res. 243, 158-164. 24. Jessen K. R., McConnell J. D., Purves R. D., Burnstock G. and Charnley-Campbell J. (1978) Tissue culture of mammalian enteric neurons. Bruin Res. 152, 573-579. 25. Johnson S. M., Karayama Y. and North R. A. (1980) Multiple actions of 5-hydroxytryptamine on myenteric neurones of the guinea-pig ileum. J. Physiol., Land. 304, 459-470. 26. Katayama Y., North R. A. and Williams J. T. (1979) The action of substance P on neurons of the guinea-pig small intestine. Proc. R. Sot. Ser. E 206, 191-208. 27. MacDermott A. B. and Weight F. F. (1982) Action potential repolarization may involve a transient, Ca*+-sensitive outward current in a vertebrate neuron. Nature SOB, 185-188. 28. Marty A. and Neher E. (1983) Tight-seal whole-cell recording. In Single Channel Recording (ed. Sakmann B. and Neher E.), pp. 107-122. Plenum Press, New York. 29. Maruyama T. (1981) Two types of spike generation of human Auerbach’s plexus cells in culture. Neurosci. Lett. 25, 143-148.

30. Morita K., North R. A. and Tokimasa T. (1982) The calcium-activated potassium conductance in guinea-pig myenteric neurones. J. Physiol., Land. 329, 341-354. 31. 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. 32. Mudge A. W., Leeman S. E. and Fischbach G. D. (1979) Enkephalin inhibits release of substance P from sensory neurons in culture and decreases action potential duration. Proc. mtn. Acad. Sci., U.S.A. 76, 526530. 33. Nicoll R. A. (1978) The actions of TRH, substance P and related peptides on frog spinal motor neurons. J. Pharmac. exp. Ther. 207, 817-824.

34. Nishi R. and Willard A. L. (1983) Conditioned medium alters distribution of peptides in enteric neurons in cell culture. Sot. Neurosci. Abstr. 9, 304. 35. Nishi R. and Willard A. L. (1985) Neurons dissociated from rat myenteric plexus retain differentiated properties when grown in cell cultures. I. Morphological properties and immunocytochemical localization of transmitter candidates. Neuroscience 16, 187-199.

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

37. North R. A. (1982) Electrophysiology of the enteric nervous system. Neuroscience 7, 315-325. 38. North R. A., Katayama Y. and Williams J. T. (1979) On the mechanism and site of action of enkephalin on single myenteric neurons. Bruin Res. 165, 67-77. 39. Nowak L. M. and MacDonald R. L. (1983) Ionic mechanism of muscarnic cholingeric depolarization of mouse spinal cord neurons in cell culture. J. Neurophysiol. 49, 792-803. 40. Obata K. (1974) Transmitter sensitivities of some nerve and muscle cells in culture. Bruin Res. 73, 71-88. 41. G’Lague P., Potter D. D. and Furshpan E. J. (1978) Studies on rat sympathetic neurons developing in cell culture. I. Growth characteristics and electrophysiological properties. Devl Biol. 67, 384-403.

Membrane properties of cultured myenteric plexus neurons

211

42. Rang H. P. (1981) The characteristics of synaptic currents and responses of acetylcholine of rat submandibular ganglior cells. J. Physiol., Lond. 311, 23-55. 43. Tokimasa T., Cherubini E. and North R. A. (1983) Nicotinic depolarization activates calcium dependent gK ir myenteric neurons. Brain Res. 263, 57-62. 44. Wallis D. I. (1981) Neuronal 5-HT receptors outside the central nervous system. Life Sci. 29, 2345-2355. 45. Wallis D. I. and North R. A. (1978) The action of 5-hydroxytryptamine on single neurons of the rabbit superior cervica ganglion. Neuropharmacology 17, 1023-1028. 46. We& M. A. and MacDonald R. L. (1982) Opioid peptides decrease calcium-dependent action potential duration 01 mouse dorsal root ganglion neurons in cell culture. Brain Res. 239, 315-321. 47. Willard A. L. and Nishi R. (1982) Fast cholinergic and slow non-cholinergic synaptic interactions between enteric neurons grown in dissociated cell culture. Sot. Neurosci. Abstr. 8, 779. 48. Willard A. L. and Nishi R. (1983) Physiological and pharmacological properties of cholinergic enteric neurons in cell culture. Sot. Neurosci. Absrr. 9, 504. 49. Willard A. L. and Nishi R. (1985) Neurons dissociated from rat myenteric plexus retain differentiated properties wher grown in cell culture. III. Synaptic interactions and modulatory effects of neurotransmitter candidates. Neuroscience 16, 213-221. 50. Williams J. T. and North R. A. (1979) Vasoactive intestinal polypeptide excites neurons of the myenteric plexus. Brain Res. 175, 174-177. 51. Wood J. D. and Mayer C. J. (1978) Slow synaptic excitation mediated by serotonin in Auerbach’s plexus. Nature 276.

836837. 52. Wood J. D. and Mayer C. J. (1978) Intracellular study of electrical activity of Auerbach’s plexus in guinea-pig small inestine. Pfltigers Arch. 374, 265-275. 53. Wood J. D. and Mayer C. J. (1979) Intracellular study of tonic-type enteric neurons in guinea-pig small intestine. J. Neurophysiol. 42, 569-58 1. 54. Wood J. D. and C. J. Mayer (1979) Serotonergic activation of tonic-type enteric neurons in guinea-pig small bowel. J. Neurophysiol. 42, 582-593. (Accepted 29 March 1985)