Electrophysiology of adult rat facial motoneurones: the effects of serotonin (5-HT) in a novel in vitro brainstem slice

Journal of Neurascience Methods, 28 (1989) 133-146 Elsevier

133

NSM 00930

Electrophysiology of adult rat facial motoneurones: the effects of serotonin (5-HT) in a novel in vitro brainstem slice P.M. L a r k m a n , N . J . P e n i n g t o n a n d J.S. K e l l y Department of Pharmacology, Universityof Edinburgh, Edinburgh (U.K.) (Received 7 November 1988) (Accepted 9 November 1988)

Key words: Intracellular recording; Facial motoneuron; Brainstem slice; Serotonin Studies of adult rat motoneurones using in vitro slice preparations are rare. We here describe a novel brainstem slice of the adult rat containing the facial motor nucleus (FMN). Data obtained for facial motoneurones (FM) by intracellular recording indicate that they display several passive and active properties seen in other rat cranial and spinal motoneurones. Bath appfication of serotonin (5-HT) evokes a reversible depolarization of FMs which is associated with an increase in input resistance due to a reduction in potassium permeability. This effect is unaffected by tetrodotoxin indicating a postsynaptic site of action.

Introduction The in vitro brain slice technique has facilitated ever more detailed studies of the electrophysiology of a wide range of neuronal types in the adult mammalian central nervous system (CNS). However, the study of adult rat motoneurones using this technique has lagged behind, studies being largely restricted to immature neurones of neonatal (rat) hemisected or sliced spinal cord (Fulton and Walton, 1986; Walton and Fulton, 1986; Takahashi, 1978). Fears that intracellular recording in the brain slice m a y be restricted to cell types with compact dendritic fields have been allayed by recent studies of adult m a m m a l i a n cranial motoneurones. Rat ocular motoneurones (Gueritaud, 1988) and both guinea-pig hypoglossal (MosfeldtLaursen and Rekling, 1987) and vagal (Yarom et al., 1985) motoneurones have been successfully investigated using coronal brainstem slices despite

Correspondence: P.M. Larkman, Department of Pharmacology, University of Edinburgh, 1 George Square, Edinburgh EH8 9JZ, U.K.

their extensive dendritic morphologies. Similarly adult cat spinal cord slices have enabled intracellular studies of sympathetic preganglionic neurones (SPN) whose dendritic fields extend at least 2 m m in the longitudinal direction (Yoshimura et al., 1986). Motor nerves supplying the superficial and some deeper muscles of the head and neck form the major branches of the seventh cranial (facial) nerve, and their cell bodies constitute the facial m o t o r nucleus ( F M N ) in the medulla oblongata of the brainstem in the rat. Chromatolytic and retrograde tracing, as well as electrophysiological techniques have shown a topographic representation of the facial muscles and nerve branches on 4 or 5 subdivisions of the F M N (Courville, 1966; Martin and Lodge, 1977; Watson et al., 1982; Friauf and Herbert, 1985; Semba and Egger, 1986). Despite detailed morphological studies in a variety of species, the electrophysiology of the F M N has been confined to in vivo studies in the rat and cat which have principally been concerned with the excitability and synaptic activation of facial motoneurones (FM) (Iwata et al., 1972; McCall and Aghajanian, 1979; 1980; Fanardjian et al.,

0165-0270/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

134

1983a, b; Fanardjian and Manvelyan, 1987a, b). Virtually no data have been published on the specific membrane properties of FMs. Only recently has a preliminary report on in vitro studies on guinea-pig FMs been published (Muhiethaler et al., 1987). Pharmacological studies have shown FMs to be potently excited by the amino acids, glutamate, aspartate and N-methyl-D-aspartate (NMDA), while T-aminobutyric acid (GABA) and glycine act to depress excitatory amino acid (eaa) evoked firing (Martin et al., 1977). Additionally, the rat FMN has been shown to be innervated by 5-hydroxytryptamine (5-HT) and catecholamine containing fibres (McCall and Aghajanian, 1979). Indeed the density of 5-HT innervation is greater here than in any other cranial motor nuclei (Takeuchi et al., 1983). In vivo, iontophoresis of 5-HT facilitated eaa or afferent nerve stimulation evoked excitation of FMs by inducing a slow subthreshold depolarization associated with an increase in cell input resistance (McCall and Aghajanian, 1979; VanderMaelen and Aghajanian, 1980, 1982). The present study describes intracellular recordings from FMs in a slice through the adult rat brainstem. Passive and active membrane properties of the FM are characterised and the action of 5-HT evaluated with the aim of using this preparation for in vitro pharmacological studies. Preliminary accounts of these results have been reported (Larkman et al., 1988; Larkman and Kelly, 1988).

Methods

Adult Cob Wistar rats (160-250 g) were decapitated using a small animal guillotine and the dorsal part of the cranium removed. The brain was sectioned at the level of the inferior colliculus, the cranial nerves cut, and the brainstem with attached cerebellum quickly removed and placed in preoxygenated artificial cerebrospinal fluid (aCSF) at 4°C (Llinas and Yarom, 1981). The tissue was placed on aCSF moistened filter paper, the cerebellum removed and the rostral end of the brainstem glued to a plastic block using cyano-

acrylate glue, with an agar block supporting the dorsal side. After attaching the plastic block to an aCSF (4 ° C) filled plexiglass chamber the brainstem was sliced coronally in the ventral to dorsal direction using a Vibroslice (Cambden Inst.). Since the facial nuclei of a 200 g rat measures approximately 2.3 mm mediolaterally, 1.7 mm dorsoventrally, and 1.5 mm rostrocaudally, 3-4 slices ( - 350 gM) can be obtained. The nuclei are located near the ventral surface of the brainstem on either side of the midline (Fig. 1). Their rostral boundaries are marked by the descending fibres of the facial nerve and the superior olivary complex, while caudally they are associated with the rostral boundaries of the nucleus ambiguus and the inferior ohvary complex. The slices were quickly transferred to a Haastype interface chamber (Haas et al., 1979) and perfused at a rate of 0.5-1.0 ml/min with aCSF at 37°C under a humidified atmosphere of 95% 02/5% CO 2. The whole procedure lasts about 10 rain and slices were incubated for at least 1 h before attempting intracellular recording. The composition of aCSF in mM was: NaC1 124, KC1 5, MgSO4 2, CaCI 2 2, NaHCO 3 26, HEPES 1.25, D-glucose 10, and pH was 7.4 after equilibration with a 95% 02/5% CO 2 gas mixture. In early experiments, to provide protection against anoxia, hydrogen peroxide (0.003% w/v) was added to the aCSF (Llinas et al., 1981; Walton and Fulton, 1983) however this was later abandoned without prejudicing FM viability. The slices were illuminated with fibre optics and the FMN identified as a darker grey area in the ventral brainstem. Identification was often aided by observation of ascending nerve fibres climbing towards the genu of the facial nerve in the dorsal aspect of the slice (Fig. 1). Intracellular recordings were made using microelectrodes pulled on a Flaming-Brown horizontal puller (Sutter Inst.) using 1.2 mm outside diameter, thin walled fibre, glass capillaries (Clark Electromedical). They were filled with 3 M KCI and had DC resistances of 15-50 MI2. The microelectrode was moved through the slice in 2 gm steps and FMs were impaled by briefly increasing the capacity compensation. 'Sealing' was often aided by passage of constant hyperpolarising current (about 1 nA) through the

135

A

f.--; "

" ......

"

,i .,'

..-RMg-..,,

,

.- ......

Fig. 1. A: a line drawing of a coronal section through the rat brainstem adapted from Paxinos and Watson (1982) showing the location of the facial motor nucleus (FMN (7)) in relation to other brainstem structures. Abbreviations: asc 7, ascending fibres facial nerve; bas, basilar artery; FMN (7), facial motor nucleus; gT, genu facial nerve; icp, inferior cerebellar peduncle; LVe, lateral vestibular nucleus; MVe, medial vestibular nucleus; PGi, nucleus paragigantocellularis; PrH, prepositus hypoglossai nucleus; py, pyramidal tract; RMg, raphe magnus; RPa, raphe pallidus; Sp50, nu. spinal tr. trigeminal nerve, oral; sp5, spinal tr. trigeminal nerve; SuVe, supravestibular nucleus; VCo, ventral cochlear nucleus; 4V, fourth ventricle. B: morphology of facial motoneurones in the rat brainstem. A montage of a coronal section through the FMN. Neurones were labelled by retrograde transport of HRP applied to the ipsilaterai buccolabial branch of the facial nerve. Note labelling of dorsally directed axon (arrowed). Calibration is 50 ~m.

136

electrode. A high input resistance bridge amplifier (Axoclamp IIA, Axon Inst.), in current clamp mode, enabled simultaneous measurement of membrane potential and intracellular current injection through the recording electrode. Electrode resistance was monitored continuously and balanced when necessary. Output was observed on a digital storage oscilloscope (Gould 1425) and stored on video-tape, using an analog interfaced digital audio signal processor (Sony PCM 701ES)-video cassette recorder (Sony SLF 30) system (Lamb, 1985) for future computer analysis using a CED 502 interfaced to a PDP 11 (Crunelli et al., 1983). A DC pen recorder (Linear Corder MkV Watanabe) was used for a continuous record of resting membrane potential and responses to intracellular current pulse injection. Membrane potentials were calculated by withdrawal of the microelectrode from the cell after recording. 5-HT creatinine sulphate complex (Sigma) 200 /~M in aCSF was bath-applied after being diluted from a 1 mM stock solution in aCSF stored at -20°C.

Identification of motoneurones in the F i N

In a preliminary study motoneurones in the FMN were identified by retrograde transport of horse radish peroxidase (HRP) (Type VI, Sigma), from peripheral branches of the ipsilateral facial nerve. Rats were anaesthetised with a fentanylfluanisone (Hypnorm), midazolam (Hypnoval), water mixture (1:1:1) (3.3 ml/kg i.p.) and the peripheral branches of the facial nerve exposed. Branches were dissected away from underlying and overlying tissue and transected. The proximal stump was washed with sterile water, dried thoroughly, placed on a piece of plastic sheet and an HRP-agar pellet applied to the cut end (Enevoldson et al., 1984; Semba et al., 1984). This was sealed in place using cyanoacrylate glue and the wound sutured. The animals were left for 36 h before being deeply anaesthetised with methohexitone sodium (Brietal Sodium) and transcardially perfused with the following solutions (Rosene and Mesulam, 1978):

(1) Buffered saline solution (g/l): NaH2PO 4 • 2H20 2.03, Na2HPO 4 12.32, NaCI 5. Heparin 1 ml/1 (heparin injection 5000 U/ml, Weddel Pharm Ltd.) 100 ml. (2) Fixative solution: 1.25% glutaraldehyde, 0.4% paraformaldehyde, 5% sucrose in buffered saline solution. 500 ml rate adjusted to last 30 rnin.

The brain was then removed, placed in a fixative solution with 10% sucrose at 4°C for 2-4 h and then soaked overnight in buffered saline solution with 10% sucrose at 4°C. Coronal sections (50 /~m) through the brainstem were cut using a freezing microtome and then processed for HRP using the 3,3', 5,5'-tetramethylbenzidine method of Mesulam (1978). Sections were mounted on chrome-alum subbed slides and coverslipped with DPX. Fig. 1B shows labelled FM's in the lateral, intermediate and dorsal subdivisions of the FMN after HRP application to the buccolabial branch of the facial nerve. This confirms the work of Semba and Egger (1986) and also Watson et al. (1982) who injected HRP into the nasolabialis muscle bed and showed a similar localisation of labelled motoneurones. Resdts

Intracellular recordings from 43 facial motoneurones with stable membrane potentials, in all subdivisions of the FM'N, lasting 0.3-3 h, were obtained. Records from motoneurones with resting potentials exceeding - 6 0 mV and with short duration overshooting action potentials were selected for further analysis. The mean resting potential was - 74.9 + 6.8 mV (+ S.D.). The passive membrane properties of the motoneurones were determined by examining voltage responses to intracellular injection of current pulses 100-120 ms in duration (Fig. 2A). Small hyperpolarising current pulses (0.1-0.5 nA) produced linear ohmic current voltage responses. As hyperpolarising current amplitude was increased, voltage responses peaked after a delay of 10-20 ms and then displayed a time-dependent sag; the magnitude of the sag increased as the magnitude of the test pulse increased. This type of

137

'-57-

"67

-77---

,

t

•"87-

'

I

I

•

-g7 mV -107 -1.5

.O.9

-0~

oa

~

1~

~k

Fig. 2. A: superimposed records of membrane potential (upper traces) during intracellular injection of hyperpolarizing and depolarizing current pulses (100 ms) (lower traces). Resting potential is -82 mV. Note the prominent inward 'anomalous' rectification with the larger hyperpolarizing current pulses and the overshoots at the end of the current pulse. Note also the small depolarizing prepotential on the maximal depolarizing voltage trace (arrowed) which preceeAs the action potential evoked by a larger depolarizing pulse of current. The time constant for this cell was 3.3 ms. B: current-voltage relation for the cell shown in A. Input resistance measured at the peak of the voltage deflection (open squares) is 15.7 Mf~ and is linear throughout the range + 20 mV around resting potential. Steady-state input resistance (asterisks) in the hyperpolarizing direction is 10.5 Ml2 while in the depolarizing direction rectification is less prominent.

response is characteristic of inward ' a n o m a l o u s ' rectification of the kind seen in several other cell types (see Discussion). Current-voltage relations measured at the peak of the voltage response were linear over a wide voltage range ( + 20 mV in Fig.

2B, open squares) around resting potential, and gave cell input resistances of 3.1-17.6 M~2 with an average of 8.9 + 3.4 MfJ (n = 43). The linearity of steady-state current voltage plots was variable from cell to cell and in general restricted to a narrower voltage range near the resting potential. Input resistances measured over the linear region ranged from 1.8 to 15.8 M~2 and the mean was 7.1 ___3.4 M~2 (n = 27) (Fig. 2B, asterisks). The prominence of the inward rectification also varied from cell to cell but was greater in cells with more negative resting potentials. In general, rectification in response to depolarising current pulses was of smaller amplitude and the steady-state slope input resistances were different in the hyperpolarising and depolarising directions (Fig. 2A). Hyperpolarising and depolarising current test pulses, were followed by depolarising and hyperpolarising potentials, respectively (Fig. 2A). The measurement of the m e m b r a n e time constants from hyperpolarising voltage deflections was complicated in m a n y cells by the presence of inward rectification. In cells where small amplitude voltage deflections did not evoke an inward rectifying response, time constants ranged from 2.3 to 4.5 ms with a mean of 2 . 9 7 + 0 . 6 9 ms (n = 16). Action potentials (AP) could be evoked by increasing the amplitude of rectangular injected depolarising current pulses (Fig. 3A, B). In some cells a slowly depolarising prepotential was observed at sub-threshold levels (Fig. 2A, arrow) upon which an action potential was evoked when the current amplitude was increased (Fig. 3A). Mean AP amplitude measured from the resting potential was 75 + 6 m V (n = 24), the mean overshoot 5 mV and the mean threshold - 5 5 + 8 mV. Duration of APs was 1 ms or less and their N a + dependence inferred b y their sensitivity to tetrodotoxin (TTX). When evoked by a relatively brief current pulse the amplitude, overshoot and duration of the AP were similar to those evoked by longer current pulses (Fig. 3B). The directly evoked AP was followed by a fast and slow afterhyperpolarisation (AHP) and a delayed depolarisation (DD) (Fig. 3A). The fast A H P appeared to be a direct continuation of the repolarising phase of the AP and was brief in

138 A.

Effects of 5-hydroxytryptamine

20m~

B.

Fig. 3. A: action potential evoked by a depolarizing current pulse of 1.4 nA, 100 ms. It is followed by fast and slow AHPs and a delayed depolarization (DD). Action potential amplitude is 86 mV from resting potential and overshoots by 4 mV. Slow AHP amplitude measured from threshold potential level is 13 mV and has a duration of 60 ms. B: action potential evoked by a brief depolarizing current pulse of + 2.1 hA, 2 ms in a different cell. Amplitude is 98 mV with an overshoot of 21 mV. The delayed depolarization is more prominent here but may be obscured by passive electrotonic potential decay, while the slow AHP has a smaller amplitude of 2 mV and a duration of 30 ms.

Superfusion with 200 # M 5-HT evoked a monophasic slow depolarisation in FMs (24/24) (Fig. 4). The amplitude of the depolarising response ranged between 3 and 16 mV at resting potentials between - 6 5 and - 8 5 mV and was fully reversed by washing with 5-HT free aCSF. The mean response was 7.8 + 3.2 mV. Effects of superfusion with 100 # M 5-HT were either very small or not present at all (not shown). The depolarisation was associated with an increase in input resistance (Figs. 4, 5) which could be as great as 144% or as little as 9%. The 5-HT evoked increase in peak input resistance for each of 24 responses is plotted against the control values in Fig. 6. A linear relationship was established using a least squares regression analysis. Further analysis is required to evaluate the significance of this relationship. At the plateau of the depolarisation in response to 200/~M 5-HT the peak input resistance represented an increased cell input resistance from an average resting level of 7.8 + 3.5 MI2 to 12.2 + 5.9 MI2 (n = 24). If the membrane potential was clamped manually to the resting level with constant hyperpolarising current at the plateau of the 5-HT response the change in cell input resistance was maintained indicating the increase in cell input resistance was not due to membrane rectification. However, in some cases where this manoeuvre was performed the magni-

1lOmV duration. The slow AHP peaked later and had an average amplitude of 8.8 + 2.5 mV from threshold voltage level and a duration of 57 _+ 14 ms measured from where the repolarising phase of the AP crossed the threshold voltage. APs evoked by brief current pulses were followed by slow AHPs at resting potentials which were of smaller amplitude but equivalent duration to those measured at threshold (Fig. 3B). The delayed depolarisation was identified as a small hump of short duration and small amplitude between the fast and slow AHP's (Fig. 3A).

it,- ~",L ]lnA

IIIF '

roll '

]'~U L

1ram

Fig. 4. Slow depolarization evoked by 5-HT (200 ~M). A continuous record of membrane potential and responses to hyperpolarizing current pulses at a frequency of 1 Hz, showing the development of a slow depolarization associated with increasing input resistance during bath application of 200 /~M 5-HT. This cell reversibly depolarized 8.5 mV from a resting potential of - 8 3 mV. Voltage, upper trace; current, lower trace.

139

A.

ll~mo~s

~

(2~

B. "63-

-71 -

/ z/~

-87-

-95mV -103

,

-3.5

-2.5

,

- 1.5

. -0.5

.

. 0.5

. 1.5

2.5

3.5

nA

Fig. 5. A: membrane potential records (upper traces) during intracellular current injection (lower traces) illustrating the actions of 5-HT (200 #M). The cell depolarized 9 mV from a resting potential of - 83 mV and peak input resistance reversibly increased from 3.1 to 6.0 M0 (94%). Note the increase in firing in response to previously subthreshold current injection, in the 5-HT condition. B: current-voltage relations from peak voltage deflections for the cell shown in A. The reversal potential shown by the point of intercept of the extrapolated lines of regression is -92.5 inV. Control, asterisks; 5-HT, open squares.

tude of the increase in input resistance was reduced indicating that rectification may contribute to the increase in resistance in some cells. As shown in Fig. 5A the depolarisation and increase in cell input resistance evoked by 200 # M 5-HT was accompanied by an increase in excitability.

Extrapolated linear, peak, current voltage plots from the control and 5-HT superfused conditions intersected at - 92.7 + 7.5 mV (n = 24) (Fig. 5B). Although we do not know the intracellular ion concentrations, under these experimental conditions the reversal potential estimated from the

140

5-HT EVOKED RESISTANCE INCREASE (n=24) 40-

o (~j 30J J [k Z

p 3

[]

oo o

W 13_

~o

5

lO

15

CONTROL PEAK INPUT RESISTANCE (MQ)

Fig. 6. X - Y plot showing the relationship between the peak input resistance before (abscissa) and after (ordinate) 5-HT application. The line was drawn using a least squares regression analysis and fits the equation y = M x + C where M = 1.48+0.16 and C = 0.57_+1.4 ( + S.D.).

Nernst equation suggests that the event is mediated by K + ions, in this case a 5-HT induced reduction in K + conductance. Application of T T X (1 /~M) in the superfusing aCSF abolished Na+-dependent action potentials, did not alter the 5-HT evoked depolarisation (9 + 2.8 mV), increase in peak input resistance (9.2 + 3.6 to 11.2 + 4 M~2), or the extrapolated reversal potential ( - 9 4 . 3 + 0.7 mV) (n = 3) (not shown). This TFX insensitivity indicates that 5-HT acts directly on FMs and not via a presynaptic mechanism.

Discussion

Characterisation of the facial motoneurone (FM) The results show intracellular records from rat FMs in vitro to come from a relatively homogeneous population Of motoneurones-when characterised in terms of their membrane properties and responses to bath-applied 5-HT. Morphological studies in the rat and cat have suggested that the F M N contains few if any interneurones (Courville, 1966; McCall and Aghajanian, 1979). In the cat recordings from neurones in which an antidromic spike could not be elicited by stimulation of the facial nerve have been made (Fanardjian et al., 1983b) but the frequency of these observations

was not stated and thus difficult to compare with previous morphological data. Values of peak input resistance for FMs in vitro are more than two-fold greater than those measured in vivo in the rat and cat (VanderMaelen and Aghajanian, 1982; Fanardjian et al., 1983b). Higher input resistances appear to be a characteristic of in vitro slice preparations and have been attributed to a number of different factors. Reduced cell membrane damage due to finer electrode tips and subsequent improved 'sealing' of m e m b r a n e around the electrode have both been suggested. Additionally input resistance is related to soma size and the extent of dendritic branching. Facial motoneurones in the rat have been shown to have soma dimensions of on average 20 × 36 /~M in their shortest and longest diameters, with 5-11 primary dendrites (Friauf, 1986). Similar values were seen in our labelling studies. Moreover, dendritic branching is extensive and can continue for up to 600 gM. These values suggest that significant dendritic damage must take place in a coronal slice 350 g m thick. This amputation of dendrites may also contribute to an increase in input resistance. Finally, replacement of the normal extracellular environment with an artificial medium and the severing of synaptic connections will also play a part in altering cell input resistance. The resistance values we have obtained compare well with the values obtained in rat ocular motoneurones by intrasomatic recordings (Gueritaud, 1988) in vitro, but are much lower than those of rat and guinea-pig vagal motoneurones which have considerably less dendritic branching and smaller cell bodies (Yarom et al., 1985; Fukuda et al., 1987). By days 9-11 in the neonatal rat spinal cord in vitro l u m b a r motoneurones had lower input resistances than those of adult FMs even when measured at 2 0 - 2 2 ° C (Fulton and Walton, 1986). Presumably soma growth and dendritic expansion will lead to lower input resistances of rat spinal motoneurones in the adult. Consistent with differences in cell size and factors mentioned previously, values of input resistance for adult cat spinal motoneurones in vivo are much lower than those measured here (Coombs et al., 1959; Ito and Oshima, 1965; Burke

141 and Ten Bruggencate, 1971; Barrett and Crill, 1974). Similarly frog motoneurones at room temperature and below, both in situ (Magherini et al., 1976) and in vitro (Kubota and Brookhart, 1963) have lower input resistances than FMs. Interestingly, despite being on average larger than FMs, motoneurones in the turtle spinal cord in vitro have greater input resistances (Hounsgaard et al., 1988). The linearity of current voltage relations in FM~ over a wide voltage range above and below resting potential contrasts with that seen in cat motoneurones by Ito and Oshima (1965) although Barrett and Crill (1976) did show linearity between resting potential and threshold. Similar observations have been made in frog spinal motoneurones (Magherini et al., 1976), whereas recent observations from rat ocular motoneurones in vitro have shown linearity over a narrower range (Gueritaud, 1988). The time constants for membrane charging, despite higher cell input resistances, are comparable with those obtained from neonatal rat spinal motoneurones, adult rat ocular motoneurones and adult cat spinal motoneurones (Coombs et al., 1959; Barrett and Crill, 1974; Fulton and Walton, 1986; Gueritaud, 1988) but are considerably less than those observed in frog and turtle spinal motoneurones (Magherini et al., 1976; Hounsgaard, 1988). The presence of time-dependent inward rectification (IR) was described in cat motoneurones by Ito and Oshima (1965) and subsequently studied by Nelson and Frank (1967). In fact prior to this Coombs et al. (1955, 1959) had observed that values of cell input resistance were lower if measured after several seconds of current injection than when using short (15-20 ms) current pulses. Linearity of steady-state current voltage plots varied considerably between cells and often calculated slope resistances were different in the hyperpolarising and depolarising directions, similar to observations made in FMs. Studies of IR in FMs in vivo have not been made so there are no comparative data. Barrett et al. (1980) showed inward rectification in voltage clamped cat motoneurones in response to hyperpolarising steps, to be due to a slowly developing inward current reversing near the resting potential. Depolarisa-

tion-activated inward current probably distinct from the former was important for IR with depolarising steps. Prominent IR has been recorded in rat ocular motoneurones (Gueritaud, 1988) and turtle spinal motoneurones (Hounsgaard, 1988) in vitro where the steady-state current voltage relation linearity is often affected. IR was only rarely seen in frog motoneurones between rest and threshold. The presence of IR was not observed in neonatal rat spinal motoneurones or in cultured embryonic rat and chick motoneurones (Fulton and Walton, 1986; O'Brien and Fischbach, 1986; Fruns et al., 1987). Thus there appears to be a developmental difference between species, nuclei and studies. For instance, cells in rat dorsal root ganglia develop IR concurrent with axonal myelination during early neonatal development (Fulton, 1987). The end of current pulse resting potential overshoots observed in FMs were also observed in cat motoneurones (Ito and Oshima, 1965). Action potentials evoked by pulses of constant depolarising current into the FM soma show a sequence of afterpotentials similar to those observed in adult rat, cat and frog motoneurones (Coombs et al., 1955; Granit et al., 1963; Barrett and Barrett, 1976). The action potential is of short duration and overshoots zero potential. The amplitude of this overshoot varies and while it may be overestimated by the continued presence of depolarising current, action potentials evoked by very brief current pulses, which end before the action potential reaches a peak, have also been shown to overshoot. The action potential is blocked by TTX inferring similarity to the Na + dependent inward current described in cat motoneurones (Barrett and Crill, 1980) and so far seen in all mammalian central neurones (Crill and Schwindt, 1983). The action potential is followed by a fast and a slow AHP and a delayed depolarisation (DD). In cat, frog and neonatal rat motoneurones the repolarising phase of the action potential is contributed to by an increase in a voltage dependent K + conductance which is said to be responsible for the fast AHP observed in these cells (Barrett and Barrett, 1976; Magherini, 1976; Barrett et al., 1980; Walton and Fulton, 1986). The slow AHP peaks

142 later and is of longer duration. The duration of the slow AHP is comparable to those of adult rat spinal (Bradley and Somjen, 1 9 6 1 ) ocular (Gueritaud, 1988) motoneurones in vitro and facial motoneurones, in vivo (Fanardjian et al., 1983a) but shorter than those of neonatal rat spinal motoneurones in vitro (Fulton and Walton, 1986). Interestingly, the slow A H P observed in cultured embryonic rat motoneurones is also of equivalent duration though they were not always observed and the temperature at which recordings were made was not given (Fruns et al., 1987), while the slow AHP in cultured embryonic chick motoneurones at 2 4 ° C was longer (O'Brien and Fischbach, 1986). Frog spinal motoneurones in situ have similar duration AHP's (Magherini et al., 1976) while in vitro they appear to be longer (Barrett and Barrett, 1976). Cat spinal motoneurones have longer slow AHP's than detailed here (Bradley and Somjen, 1961) as do turtle spinal motoneurones in vitro (Hounsgaard, 1988). Despite this variation the underlying current appears to be outward and calcium dependent to some extent (Barrett and Barrett, 1976; Barrett et al., 1980; Walton and Fulton, 1986; O'Brien and Fischbach, 1986; Hounsgaard, 1988). Burke and Rudomin (1977) suggested a correlation between short duration AHP's and innervation of fasttwitch type muscle. This has support here as rat facial muscles are predominantly of the fast-twitch type (Lindquist, 1973). Slow AHP amplitude has been demonstrated to be clearly voltage dependent as is the case here comparing amplitudes at spike threshold and resting potential. AHP amplitude in rat spinal motoneurones has been shown to be smaller than in cat motoneurones (Bradley and Somjen, 1961) and said to be related to the faster repetitive firing properties of rat motoneurones. The DD observed as a hump separating the fast and slow AHP's has been noted in about 50% of cat spinal motoneurones and more frequently in rat, frog and turtle spinal motoneurones (Granit et al., 1963; Nelson and Burke, 1967; Barrett and Barrett, 1976; Magnerini et al., 1976; Hounsgaard et al., 1988). It was also prominent in neonatal rat spinal motoneurones, adult rat ocular motoneurones, and rat and cat facial motoneurones in vivo

(Fanardjian et al., 1983b; Walton and Fulton, 1986; Gueritaud, 1988; Iwata et al., 1972). Its presence has been attributed to passive or active passage of the soma action potential into the dendrites (Granit et al., 1963; Nelson and Burke, 1967) while Traub and Llinas (1977) in their model of neuronal conductivity suggested its root to be in a localised density of Na conductance in the dendrites. However, variation may occur as in neonatal rat motoneurones the D D is Ca 2 ÷ dependent (Walton and Fulton, 1986) while in turtle spinal motoneurones it is insensitive to Ca 2+ antagonists (Hounsgaard et al., 1988).

5-HT evoked depolarisation Depolarisation of FMs in vitro in response to bath applied 5-HT is similar to that seen with iontophoretic application in vivo (VanderMaelen and Aghajanian, 1980, 1982). The associated increase in input resistance is in all probability due to a decreased conductance of potassium ions which is 5-HT activated and not secondary to membrane rectification. The type of potassium channel mediating this effect has not been identified but it is apparent that channels mediating IR are not closed by 5-HT (Fig. 5). The actions of 5-HT are direct on the motoneuronal membrane and appear to lack a presynaptic component, persisting in the presence of TTX. Indeed the presence of 5-HT containing axosomatic and axo-dendritic synaptic boutons on FMs has been shown using autoradiographic electron microscopy, supporting a proposed role of 5-HT in modulating FM excitability (McCall and Aghajanian, 1980). The source of these serotonergic connections in the rat is uncertain. In the cat efferent projections to the F M N from both the medullary and dorsal raphe nuclei have been identified suggesting that serotonergic neurones of these nuclei may be responsible for 5-HT containing synapses of the F M N (Holstege et al., 1984; Takada et al., 1984; Pogosyan and Fanardzyhan, 1986; Isokawa-Akesson and Komisaruk, 1987). A 5-HT evoked excitation has been observed on rat and cat spinal motoneurones in a variety of preparations. Intravenous administration of the 5-HT precursors 5-hydroxytryptophan and tryptophan increased the discharge of a- and 7-

143

motoneurones in cat ventral roots (Myslinski and Anderson, 1978). Iontophoresis of 5-HT onto cat and rat spinal motoneurones enhanced excitatory responses to glutamate (White and Neuman, 1980; White, 1985). Depolarising ventral root potentials in hemisected neonatal rat and frog spinal cord in response to 5-HT were shown to persist in TTX or high Mg 2+ containing perfusing medium (Neuman, 1984; Connell and Wallis, 1987) while in adult rats in vivo stimulation of the raphe obscurus in the brainstem evoked a depolarisation in lumbar motoneurones which was absent in 5,7 dihydroxytryptaminelesioned animals even though in these animals motoneurone responses to direct iontophoretic application of 5-HT were potentiated (Roberts et al., personal communication). Thus a large body of evidence, supporting a role for 5-HT in excitatory synaptic activation of both spinal and cranial motoneurones, exists. Excitation by 5-HT in other mammalian central neurons has been seen in a sub-population of somatosensory cortex pyramidal neurons of the rat (Davies et al., 1987) and also in lateral horn sympathetic preganglionic neurones (SPN) of neonatal rat and adult cat spinal cord (Yoshimura and Nishi, 1982; Ma and Dun, 1986). In the latter case some SPN responses to 5-HT are reduced in TTX or low Ca 2+ high Mg 2÷ containing medium indicating that an additional presynaptic component may be important. The depolarisation by 5-HT reported here contrasts strikingly with the hyperpolarising response observed in dorsal raphe (Rainnie et al., 1987; VanderMaelen and Aghajanian, 1983) and lateral septal neurons (Joels et al., 1987) in vitro. The responses here are mediated by an increase in potassium conductance probably through a 5-HTtA receptor subtype. Different again is the complex triphasic response seen in hippocampal pyramidal cells in vitro. Here an initial hyperpolarisation is associated with an increase in a resting potassium conductance which is 5-HTIA mediated (Segal, 1980; Andrade and Nicoll, 1987; Colino and Halliwell, 1987). A subsequent depolarisation is associated with increased membrane resistance due to a decrease in K + conductance, possibly IM, (Colino and Halliwell, 1987) which is not prevented by the antagonists which block the hyperpolarisation

(Andrade and Nicoll, 1988). The pharmacology of this depolarisation appears to be different to that seen in rat FMs in vitro (Larkman and Kelly, 1988) and in vivo (VanderMaelen and Aghajanian, 1980) as well as in cortical neurones (Davies et al., 1987), SPNs (Ma and Dun, 1986) and spinal motoneurones (Connell and Wallis, 1987) which are all antagonised by methysergide while the hippocampal depolarisation is not (Andrade and NicoU, 1987). Finally, the slow AHP (I~¢~c~2+)) is depressed, an effect which outlasts the changes in membrane potential and results in decreased accommodation. In conclusion, the results presented here show that FMs in vitro have membrane properties which are comparable with those of other vertebrate cranial and spinal motoneurones both in vivo and in vitro. Thus it is hoped that the FMN slice will allow a detailed and relevant study of adult rat motoneurones. The presence of the 5-HT evoked depolarisation in the slice preparation, despite some dendritic damage, is consistent with the observation of axo-somatic as well as axo-dendritic 5-HT containing synaptic boutons, and will allow us to pharmacologically characterise, using intracellular recording methods, the receptor type mediating this response.

Acknowledgements PML is in receipt of a SERC-CASE award in part funded by Glaxo. The work was supported by an award from The Wellcome Trust to J.S.K.

References Aghajanian, G.K. and McCall, R.B. (1980) Serotonergic synaptic input to facial motoneurones: localiTation by electron-microscopic autoradiography, Neuroscience, 5: 2155-2162. Andrade, R. and Nicoll, R.A. (1987) Pharmacologically distinct actions of serotonin on single pyramidal neurones of the rat hippocarnpus recorded, in vitro, J. Physiol. (Lond.), 394: 99-124. Barrett, E.F. and Barrett, J.N. (1976) Separation of two voltage-sensitive potassium currents, and demonstration of a tetrodotoxin-resistant calcium current in frog motoneurones, J. Physiol. (Lond.), 255: 737-744.

144 Barrett, E.F., Barrett, J.N. and Crill, W.E. (1980) Voltage-sensitive outward currents in cat motoneurones, J. Physiol. (Lond.), 304: 251-276. Barrett, J.N. and Crill, W.E. (1974) Specific membrane properties of cat motoneurones, J. Physiol. (Lond.), 239: 301-324. Barrett, J.N. and Crill, W.E. (1980) Voltage-clamp of cat motoneurone somata: properties of the fast inward current, J. Physiol. (Lond.), 304: 231-249. Bradley, K. and Somjen, G.G. (1961) Accommodation in motoneurones of the rat and the cat, J. Physiol. (Lond.), 156: 75-92. Burke, R.E. and Ten Bruggencate, G. (1971) Electronic characteristics of alpha motoneurones of varying size, J. Physiol. (Lond.), 212: 1-20. Burke, R.E. and Rudomin, P. (1977) Spinal neurons and synapses. In J.M. Brookhart and V.B. Mountcastle, (Eds.), Handbook of Physiology, Sect. I, The Nervous System, Vol. I, Cellular Biology of Neurons Part 2, pp. 877-944. American Physiol. Soc., Bethesda. Colino, A. and HaUiwell, J.V. (1987) Differential modulation of 3 separate K-conductances in hippocampal CA1 neurones by serotonin, Nature (Lond.), 328: 73-77. Connell, L.A. and Wallis, D.I. (1987) Effects of 5-hydroxytryptamine (5-HT) in the isolated hemisected spinal cord of the neonate rat, J. Physiol. (Lond.), 390: 44P. Coombs, J.S., Curtis, D.R. and Eccles, J.C. (1957a) The interpretation of spike potentials of motoneurones, J. Physiol. (Lond.), 139: 198-231. Coombs, J.S., Curtis, D.R. and Eccles, J.C. (1957b) The generation of impulses in motoneurones, J. Physiol. (Lond.), 139: 232-249. Coombs, J.S., Curtis, D.R. and Eccles, J.C. (1959) The electrical constants of the motoneurone membrane, J. Physiol. (Lond.), 145: 505-528. Coombs, J.S., Eccles, J.C. and Fatt, P. (1955) The electrical properties of the motoneurone membrane, J. Physiol. (Lond.), 130: 291-325. Courville, J. (1966) The nucleus of the facial nerve; the relation between cellular groups and peripheral branches of the nerve, Brain Research, 1: 338-354. Crill, W.E. and Sehwindt, P.C. (1983) Active currents in mammalian central neurons, T.I.N.S., 6: 236-240. Crunelli, V., Forda, S., Kelly, J.S. and Wise, J.C.M. (1983) A programme for the analysis of intracellular data recorded from in vitro preparations of central neurons, J. Physiol. (Lond.), 340, 13P. Davies, M.F., Deisz, R.A., Prince, D.A. and Peroutka, S.J. (1987) Two distinct effects of 5-hydroxytryptamine on single cortical neurons, Brain Research, 423: 347-352. Enevoldson, T.P., Gordon, G. and Sanders, D.J. (1984) The use of retrograde transport of horseradish peroxidase for studying the dendritic trees and axonal courses of particular groups of tract cells in the spinal cord, Exp. Brain Res., 54, 529-537. Fanardjian, V.V., Manvelyan, L.R. and Kasabyan, S.A. (1983a) Mechanisms regulating the activity of facial nucleus motonenrones-1. Antidromic activation, Neuroscience, 9(4): 815-822.

Fanardjian, V.V., Manvelyan, L.R. and Kasabyan, S.A. (1983b) Mechanisms regulating the activity of facial nucleus motoneurones-2. Synaptic activation from the caudal trigeminal nucleus, Neuroscience, 9(4): 823-835. Fanardjian, V.V. and Manvelyan, L.R. (1987a) Mechanisms regulating the activity of facial nucleus motoneurones-3. Synaptic influences from the cerebral cortex and subcortical structures, Neurocience, 20(3): 835-843. Fanardjian, V.V. and Manvelyan, L.R. (1987b) Mechanisms regulating the activity of facial nucleus motoneurones-4. Influences from the brainstem structures, Neuroscience, 20(3), 845-853. Friauf, E. and Herbert, H. (1985) Topographic organization of facial motoneurones to individual pinna muscles in rat ( Rattus rattus ) and bat ( Rousettus aegyptiacus ), J. Comp. Neurol., 240: 161-170. Friauf, E. (1986) Morphology of motoneurones in different subdivisions of the rat facial nucleus stained intracellularly with HRP, J. Comp. Neurol., 253: 231-241. Fruns, M., Krieger, C. and Sears, T.A. (1987) Identification and electrophysiological investigations of embryonic mammalian motoneurones in culture, Neurosci. Lett., 83: 82-88. Fukuda, A., Minami, T., Nabekura, J. and Oomura, Y. (1987) The effects of nor-adrenaline on neurones in the rat dorsal motor nucleus of the vagus, in vitro, J. Physiol. (Lond.), 393: 213-231. Fulton, B.P. (1987) Development of inward rectification in A cells of isolated rat dorsal root ganglia, J. Physiol. (Lond.), 394, 150P. Fulton, B.P. and Walton, K. (1986) Electrophysiological properties of neonatal rat motoneurones studied in vitro, J. Physiol. (Lond.), 370: 651-678. Granit, R., Kernell, D. and Smith, R.S. (1963) Delayed depolarization and the repetitive response to intracellular stimulation of mammalian motoneurones, J. Physiol. (Lond.), 168: 890-910. Gueritaud, J.P. (1988) Electrical activity of rat ocular motoneurons recorded in vitro, Neuroscience, 24: 837-852. Haas, H.L., Sehaerer, B. and Vosmansky, M. (1979) A simple perfusion chamber for the study of nervous tissue slices in vitro, J. Neurosci. Methods, 1: 323-325. Holstege, G.H., Tan, J., Van Ham, J. and Bos, A. (1984) Mesencephalic projections to the facial nucleus in the cat. An autoradiographical tracing study, Brain Research, 311: 7-22. Hounsgaard, J., Kiehn, O. and Mintz, I. (1988) Response properties of motoneurones in a slice preparation of the turtle spinal cord, J. Physiol. (Lond.), 398: 575-589. Isokawa-Akesson, M. and Komisaruk, B.R. (1987) Difference in projections to the lateral and medial facial nucleus: anatomically separate pathways for rhythmical vibrissa movements in rats, Exp. Brain Res., 65: 385-398. Ito, M. and Oshima, T. (1965) Electrical behaviour of the motoneurone membrane during intracellularly applied current steps, J. Physiol. (Lond.), 180: 607-635. Iwata, N., Kitai, S.T. and Olson, S. (1972) Afferent component of the facial nerve: its relation to the spinal trigeminal and facial nucleus, Brain Res., 43: 662-667.

145 Joels, M., Shinnick-Gallagher, P. and Gallacher, J.P. (1987) Effect of serotonin and serotonin analogues on passive membrane properties of lateral septal neurons in vitro, Brain Res., 417: 99-107. Kubota, K. and Brookhart, J.M. (1963) Recurrent facilitation of frog motoneurones, J. Neurophysiol., 26: 877-893. Lamb, T.D. (1985) An inexpensive digital tape recorder suitable for neurophysiological signals. J. Neurosci. Methods, 15: 1-14. Larkman, P.M. and Kelly, J.S. (1988) The effects of serotonin (5-HT) and antagonists on rat facial motoneurones in the in vitro bralnstem slice, J. Neurosci. Methods, 24(2): 199. Larkman, P.M., Penington, N.J. and Kelly, J.S. (1988) The effects of 5-hydroxytryptamine (5-HT) on facial motonenrones in the in vitro brain slice of the rat, Neurosci. Lett. Suppl. 32, $9. Lindquist, C. (1973) Contraction properties of cat facial muscles, Acta. Physiol. Scand., 89: 482-490. Llinas, R. and Yarom, Y. (1981) Electrophysiology of mammalian inferior olivary neurones in vitro: different types of voltage-sensitive ionic conductances, J. Physiol. (Lond.), 315: 549-567. Llinas, R., Yarom, Y. and Sugimori, M. (1981) Isolated mammalian brain in vitro: new technique for analysis of electrical activity of neuronal circuit function, Fed. Proc., 40: 2240-2245. Ma, R.C. and Dun, N.J. (1986) Excitation of lateral horn neurons of the neonatal rat spinal cord by 5-hydroxytryptamine. Dev. Brain Res., 24: 89-98. Magherini, P.C., Precht, W. and Schwindt, P.C. (1976) Electrical properties of frog motoneurones in the in situ spinal cord, J. Neurophysiol., 39: 459-473. Martin, M.R. and Lodge, D. (1977) Morphology of the facial nucleus of the rat, Brain Res., 123: 1-12. Martin, M.R., Lodge, D., Headley, P. Max and Biscoe, T.J. (1977) Pharmacological studies on facial motoneurones in the rat, Eur. J. Pharmacoi., 42: 291-298. McCall, R.B. and Aghajanian, G.K. (1979) Serotonergic facilitation of facial motoneuron excitation, Brain Res., 169: 11-27. McCall, R.B. and Aghajanian, G.K. (1980) Pharmacological characterisation of serotonin receptors in the facial motor nucleus. A microiontophoretic study, Eur. J. Pharmacol., 65: 175-183. Mesulam, M.-M. (1978) Tetramethyl benzidine (TMB) for HRP neurohistochemistry. A non-carcinogenic blue reaction product with superior sensitivity for visualising neural afferents and efferents, J. Histoch. Cytochem., 26: 106-117. Mosfeldt-Laursen, A. and Rekling, J.C. (1987) Hypoglossal motoneurones in slices from guinea-pig have both depolarizing and hyperpolarizing responses to glycine, J. Physiol. (Lond.), 390: 17P. Miihlethaler, M., Serafin, M. and Vidal, P.P. (1987) Electrophysiological properties of guinea-pig facial motoneurons: a combined study in the acute preparation and the in vitro whole brain. Sot:. Neurosci. Abstr., 13; 293.5.

Myslinski, N.R. and Anderson, E.G. (1978) The effect of serotonin precursers on a and T-motoneuron activity, J. Pharmacol. Exp. Ther., 204: 19-26. Nelson, P.G. and Burke, R.E. (1967) Delayed depolarization in cat spinal motoneurones, Exp. Neurol. 17: 16-26. Nelson, P.G. and Frank, K. (1967) Anomalous rectification in cat spinal motoneurons and effect of polarizing currents on excitatory postsynaptic potential, J. Neurophysiol., 30: 1097-1113. Neuman, R.S. (1985) Action of serotonin and norepinephrine on spinal motoneurones following blockade of synaptic transmission, Can. J. Physiol., 63: 735-738. O'Brien, R.J. and Fischbach, G.D. (1986) Isolation of embryonic chick motonenrones and their survival in vitro, J. Nenrosci., 6: 3265-3274. Pogosyan, V.I. and Fanardzhyan, V.V. (1986) Afferent connections of cat facial nucleus; a study using retrograde axonal transport of HRP, Nierofiziologiya, 1: 35-45. Rainnie, D.G., Penington, N.J. and Kelly, J.S. (1987) On the hyperpolarizing action of 5-hydroxytryptamine on rat dorsal raphe neurons in vitro, Neurosci. Letts. Supp., 29, $72. Rosene, D.L. and Mesulam, M.-M. (1978) Fixation variables in HRP neurohistochemistry. 1. Effects of fixation time and perfusion procedures upon enzyme activity, J. Histochem. Cytochem., 26: 28-39. Segal, M. (1980) The action of serotonin in the rat hippocampal slice preparation, J. Physiol. (Lond.) 303: 423-439. Semba, K., Sood, V., Shu, N.Y., Nagele, R.G. and Egger, M.D. (1984) Examination of geniculate ganglion cells contributing sensory fibres to the rat facial 'motor' nerve, Brain Res., 308: 354-359. Semba, K. and Egger, M.D. (1986) The facial 'motor' nerve of the rat: control of vibrissal movement and examination of motor and sensory components, J. Comp. Neurol., 247: 144-158. Sprouse, J.S. and Aghajanian, G.K. (1987) Electrophysiological responses of serotoninergic dorsal raphe neurones to 5-HT1A and 5-HT m agonists, Synapse, 1: 3-9. Takada, M., Itoh, K., Yasui, Y., Mitani, A., Somura, S. and Mizuno, N. (1984) Distribution of premotor neurones for orbicularis oculi motoneurones in the cat, with particular reference to possible pathways for blink reflex, Neurosci. Lett., 50: 251-55. Takahashi, T. (1978) Intracellular recording from visually identified motoneurones in rat spinal cord slices, Proc. R. Soc. Lond. Set. B, 202: 417-421. Takeuchi, Y., Kojima, M., Matsuura, T. and Sano, Y. (1983) Serotonergic innervation of the motoneurones in the mammalian brainstem, Anat. Embryol., 167: 321-333. Traub, R.D. and Llinas, R. (1977) The spatial distribution of ionic conductances in normal and axotomized motoneurones, Neuroscience, 2: 829-849. Vandermaelen, C.P. and Aghajanian, G.K. (1980) Intracellular studies showing modulation of facial motoneurone excitability by serotonin, Nature (Lond.), 287: 346-47. Vandermaelen, C.P. and Aghajanian, G.K. (1982) Serotonin-

146 induced depolarization of rat facial motoneurones in vivo: comparison with amino acid transmitters, Brain Res., 239: 139-152. Vandermaelen, C.P. and Aghajanian, G.K. (1983) Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extra-cellularly and intracellularly in rat brain slices, Brain Res., 289: 109-119. Walton, K. and Fulton, B. (1983) Hydrogen peroxide as a source of molecular oxygen for in vitro mammalian CNS preparations, Brain Res., 278: 387-393. Walton, K. and Fulton, B. (1986) Ionic mechanismsunderlying the firing properties of rat neonatal motoneurones studied in vitro, Neuroscience, 19: 669-683. Watson, C.R.R., Sakai, S. and Armstrong, W. (1982) Organization of the facial nucleus in the rat, Brain Behav. Evol., 20: 19-28.

White, S.R. and Neuman, R.S. (1980) Facilitation of spinal motoneuron excitability by 5-hydroxytryptamine and noradrenaline, Brain Res., 188: 119-127. White, S.R. (1985) Serotonin and co-localized peptides: effects on spinal motoneuron excitability, Peptides, 6, Suppl., 2, 123-127. Yarom, Y., Sugimori, M. and Llinas, R. (1985) Ionic currents and firing patterns of mammalian vagal motoneurones in vitro, Neuroscience, 16: 719-737. Yoshimura, M. and Nishi, S. (1982) Intracellular recordings from lateral horn cells of the spinal cord in vitro, J. Auton. Nerv. Syst., 6: 5-11. Yoshimura, M., Polosa, C. and Nishi, S. (1986) Electrophysiological properties of sympathetic preganglionic neurones in the cat spinal cord in vitro, PflUgers Arch., 406: 91-98.