Electrical properties of embryonic rat brainstem motoneurones in organotypic slice culture

ELSFMER

Developmental Brain Research 86 (1995) 187-202

Research report

Electrical properties

of embryonic rat brainstem organotypic slice culture I. Eustache, J.P. Gueritcud

motoneurones

in

*

Accepted 7 February 1995

Abstract Organotypic

co-cultures

of embtywic

E18-19

rat brainstem

slices and tongue

muscle maintained

in vitro for more than 16

properties of developing embryomc brainstem motcmeurones derived from oculomotor, used to study the electrogenic facial or hypoglossal nuclei. This preparation offers a unique opportunity to study the development of prenatal mammalian embryonic motoneurones. Our results show that embryonic rat brainstem motoneurones grown in organotypicculture with tongue muscle develop electrogenic membrane properties that can be compared with those described in newborn and dult animals in slice preparation. These motoneurones displayed a variety of sodrum and calcium mnductauces. Two types of sodh d mnductances were present in all the recorded motoneurones. An Hodgkin-Huxley TlXsensitive sodium conductance was involved in the spike potential, whereas a TTX-insensitive high threshold sodium conductance was uncovered when potassium and calcium currents were suppressed. Prominent calcium potentials contributed to the large delayed dep&rizations foliing the cpik~ potentials. High threshold calcium potentials were triggered with a slightly lower threshold than the sodium spike and contributed to the control of the pattern of discharge of the cell in response to slight shii of membrane potentials. Low threshold calcium potentials were seen in 16% of the recorded motoneurones in hyperpolarized conditions. These calcium currents underlie the triggering of doublet spikes and rebound responses. Brainstem motoneurones in culture did not display differences that could be correlated with the origin of the motoneurone pool explanted and retained undiierentiated features suggesting that development of electrophysiological properties specific of each motoneurone pool are determined by presynaptic networks and target properties.

days were

KeyworcLr: Organotypic culture; Development; Embryonic motoneuron: motoneuron; Ionic conductance; Calcium potential

1. Introduction

Development of motoneurones offers a good opportunity to study the relationship between the differentiation of a complex specific morphology and the onset of well defined functional properties. The relationship between morphology and functional properties is of particular importance for the understanding of signal processing by central neurones of vertebrates. Survival and maturation of motoneurones is closely dependent on the development of functional relatiozships between the neurone and its target. A growing

-. Corresponding author. Fax: (33) 91 26 u) 38 OM-3l?06/95/SO9.50 8 1995 Elxvier SSDI G165.3806(95)00031-3

Science B.V. All rights reserved

Hypoglossal motoneuron:

Facial motoneuron;

Ocular

body of evidence demonstrates the existence of prominent trophic interactions between motoneurones and muscle fibers. Survival and differentiation of ,notoneurones are dependent on muscle derived neurotrophic factors [5,16,34,54]. The proper diierentiation of muscle fibers also depends on the establishment of neuromuscular connections and on the electrophysiological properties of the connected motoneurones [15]. Only innervated myotubes develop further and become diiferentiated muscle fibers, slow and fast muscle fibez becoming specified during development by their innervation 132.37.661. The proper differentiation of motoneuronal properties depends also on the presynaptic drive impinging on the motoneurone [Sli. Numerous studies have described the electrical

I. Eurtoche.J.P Guerimd/Deuclqmmtacrl Bmin Bemwrh 86 (1995) 187-202

prowrties in the

of postnatal or late immature m6toneuronca

acute slice preparation, both at spinal [11,17,33,70]and cranial levels [12,26,31,~,63-65,6, and show the existence of excitability changes which may be of importance in shaping the organization of motor patterns. However, early embryonic chicken and rat motoueurone properties have only been studied in short term dissociated

cultures [18,45,46,52]. Although

these studies pointed out the probable importance of synaptic inputs and func:ional connections with the target for the iong term survival and proper development of motoneurones, dissociated cultures may hnpose substantial changes to the normal differentiation of neurones and lack most of the tissular interactions and trophic effects necessary for the proper diierentiation of motoneurones [38,39,62]. Organotypic co-cultnres of nerve and muscle tissues have been used as a valuable tool to study long term neuromnsctdar development [9,55,56].In this type of preparation, the diierent cellular types constitutive of the nerve and muscle tissues are present and maintain their specific morphology and funaional relationships. In long term spinal cord/skeletal muscle organotypic co-cultures, motoneurones survive, differentiate, and receive synaptic inputs. Furthermore, functional motor units, incorporating reflex arcs with dorsal root ganglion cells, are established [7,10,!%,593. Since the introduction of the slice culture technique by GPhwiler [19],it is possible to obtain a flat, optically accessible nervous tissue well suited for the study of the intricate relationship between the morphology and the function of developing central neurones. We have recently described an organotypic co-culture of embryonic rat brainstem and tongue muscle. Zn this preparation, motoneurones from the cranial motor nuclei differ-.ntiate and become functional, controlling the contraction of striated muscle fibers [14,28]. Tlte aim of the present study was to describe the electrical properties of cranial motoneurones grown in organotypic cultures to study the early phase of their diierentiation process. Part of this work has been published in abstract form 1291. 2. Materials and methuds 2.1. cultures Cultureswere done as previouslydescribed. Briefly, brainstemexplantswere obtained from rat embryos or! embryonic day 18 to 20. At this stage, motoneurones have migrated to their fmd locations and motor nuclei can be clearly recognized on brainstem sections, but the neurones are not yet fully differentiated. Time-mated Wistar female rats were anaesthetized

with ether. After thoroughly cleaning the abdomen with 7(P a!cohol, the animal was decapitated. The two horns of the uterus, each containing4 to 8 embryos, were removed to a sterile Petri-dish. All further preparations were carried out under a laminar flow hood with sterile proce-dures. The wall of the uterus was opened and the intact amniotic sacs were freed and transferred to a new Petri-dish containing Gey’s Balanced Salt Solution (GBSS) (Gibco 24260-010)cooled to WC. The embryowas exposed by tearingopen the sac, immediatelydecapitated and the head was removed to fresh cooled GBSS. Under a stereomicroscope, the cranium was careMy opened along the midline and the brain gently lifted out with a small

smooth spatula and transferred to another Petri-dish containing fresh cooled GBSS. Using a razor blade, a block containingthe brainstemwas isolated with two transverse.sectionsmade at the level of the caudalpart of the cerebral hemispheres and the caudal medulla. With fine forceps, the meninges and blood vessels were peeled away under the stereomicroscope. Great care was exercised not to damage the brainstem. The brainstem was removed to fresh GBSS. Brainstems were pooled and arranged on the base plate of a McIllwain tissue chopper. Transverse sections, 275 pm thick, were obtained. Sections were collected wnh a moist spatula and dispersed in cool GBSS. Sections contaituag oculomotor nuclei, facial and abducens nuclei or hypoglossal nuclei together with dorsal nuclei of the vagus nerve and ambiguus nuclei were identified under the stereomicroscope and explants obtained by cutting away unnecessary tissue. Sections were made in order to prc%notecellular and neuritic outgrowth [l] and to re&.~ccAe size of the explant, The final explants were removed to Patri-dishes identified accordiig to the motor nucleus explanted. Muscle explants were obtained from the tongues of the same embryos. Explants were obtained by chopping the tongues twice on a tissue chopper. After the first passage, the base plate was rotated 90” and a second passage was made to obtain small muscular fragments. Chopping was performed with 408 pm steps. Explants were collected and dispersed in fresh GBSS. A brainstem explant and a small piece of muscle selected under the stereomicroscope. were embedded together in a plasma clot on a perfectly cleaned glass coverslip according to the procedure described by ahwiler 1191.Explants were placed in 40 ~1 of reconstituted chicken plasma (Sigma P3266). The plasma was then coagulated by adding 20 ~1 of thrombine (Sigma T4648,O.S me/ml) and carefully mixing. Before coagulation started, the muscle explant was positioned 0.7 to 1.0 mm from the ventral part of the brainstem explant.

I. Ewoche,

J P. Gummud

/Lkelopmenml

2.2. Maintenance of cultures The cultures were maintained according to the procedure described previously l7,2S]. After coagulation of the plasma clot, the coverslips were introduced into of conical vented plastic tubes containing medium. The tubes were placed on the drum of a tube roller, tilted Y from the horizontal plane and set to 120 revolutions per hour, inside an incubator. Incubation took place at 36°C in a dry atmosphere with an initial concentration of 5.5% CO,. Carbon dioxide concentration changes were made according to Braschler et al. [7]. Medium changes were performed on day 5, 9, 12, 14 and 16. Old medium was replaced by 1.5 ml of fresh medium. After the third week, changes were made every second day and medium replaced by 2.0 ml of fresh medium. Culture medium consisted of 48.6% of Dulbecco’s modified Eagle’s medium with 0.03% glutamine (Gibco 21885). 24.4% of Hank’s Balanced Sa!t So!ution (Gibco 2.4020).8.1% ot

11 VI

heat inactivated fetal calf serum, 2.3% of 20% glucose solution and 16.6% of ultra-purified water. Osmolarity of the medium was measured and kept in the range 290-300 mGsm/kg H,O. To promote the growth and differentiation of cholinergic cells [20,21], nerve growth factor was added to a Gnal concentration of 26 ng/nri during the first week and 5 ng/ml thereafter. Antibiotics were added to the first medium and suppressed m further medium changes. Antimitotics were added for 24 hours at 5 days in vitro @IV) to control glial outgrowth. Great care was exercised to avoid pH changes (alkaline shift) due to CO, uptake by plastic tubes. New empty tubes were saturated with CO, in the incubator for 2 weeks prior to use.

2.3. Intracellular

recording technique

Intracellular recordings were perfomied on cultures between 16 and 48 DIV. Cultures selected for recording displayed a muscle region with strong irregular contractions, a brainstem region with a clear organotypic structure and a flat monolayer between the

brainstem motor region and the newly formed muscle fibers. The coverslip was taken out of the culture tube, fixed to the bottom of a 35 mm plastic Petri-dish (Falcon) 1 ml of solution ml/min.

with a drop of paraffin wax and covered with GBSS. Drugs were added and the bathing renewed by perfusion of the dish at a rate of 1 To minimize tissue movements due to muscle

fibers contractions and osmotic changes due to evaporation, recordings were made at room temperature (25 f 2°C). Standard GBSS contains fmM): NaCI 137, KCl 5, CaCl, 1.5, MgCl, I, MgSO, 0.3, KHsPO, 0.2, NaH,PO., 0.8, NaHCO, 2.7, Glucose 5.5. The dish was placed on the stage of an inverted microscope (Nikon)

Emrn Research 86 (1995) 187-202

189

and the culture was observed with phase contrast or Hoffman Modulation contrast optics. Recordings were performed using fine tip microelectrodes. Pipettes were pulled from standard 1.2 mm o.d. filamented glass capillaries (Clark, Reading, Eng-

ind) on a programmable micropipette puller (model F 87, Sutter Instr. Co., USA) and filled with 1 M rL-Acetate. These electrodes had a tip resistance of 50-90 MI1. Membrane potentials were recorded using a conventional bridge amplifier alknving for the simultdneous injection of constant currents through the electrode. Recordings were digitized at 44 kHz and stored with PCM coding on video tapes (VRllX-A, Instrutech Corp.) for later analysis or digitized at lo-20 kHz and displayed on-line and stored on an IBM 386 compatible micro computer using a CED-:401 digital interface and dedicated software. Extracellular microst++“-_I... llli”,, of presumed motoneurones was performed using glass pipettes with

20-50 wrn ups diameter filled with GRSS. Trains of short CO.!-03 ms) voltage pulses of varying amplitude (0.1-10 V) were delivered by a conventional stimulatcr. These large pipettes activated only one motoneurone from a single stimulus location.

3.

Results

3.!. Ident~fican~n of motoneurones In a previous paper [2S], we showed that brainstem motoneurones II? organotypic culture tend to migrate out of ttc explant, into the monolayer region where they can be recognized as large ce’ls with a clear, excentrared nucleus and one dark nuc!eole, located between thi mmcle and the brainstem in a position

directly dertved from the original motor nucleus. Using these criteria, we further de.nonstrased with a microstimulation procedure that these neurones establish functionnal connections with muscle fibers. A stimulation micropipette was positioned directly above the soma of a presumed motoneurone at a distance of 10 to 50 pm and the muscle fibers were observed \vlitl a x40 phase-contrast objective and recorded on video tape. Upon stimulation, contractions of one or a few muscle tibers were often observed. We consider that this stimulation was a direct stimulation of the motoneurone and not a p&synaptic activation since muscle contractions were observed to follow high fre-

quency stimulation. These contractions ceased if the e!ectrode was moved less rhan one cell diameter away. This demonstrated the specificitv of the stimulation and allowed for clear functional identification of motoneurones. Muscle contractions were observed with all the tested neurones. * The ctitltures used in this study were grown for 16 to

I. Etutache,J.P. Gucnkuui/Dewkwnenta! Brain Rcsmnh 86 (I!.%) 187-202

,

Mmuc

,

F@. 1. A: wotttanwus synapticpotentials with werridmp spikes. 8: membranercspomcs to depolarizing current pulsesof straddling thrcsbotd intensity indicated by bar at&r tbe traces. C: wpcrbnpcecd rcbottttd v to hypctpolarizing current pulses indicated by bar under the traces. D: action pntmtial triggered by a 150 ms intmccllular current i+ctiott pulse of threshold intettsity indicated by bar under the trace. E: sttWimp@ed action potcntittls triggered front diercttt resting potentials. Amw points to Delayed Dcpolariz.ation. Dashed line shows mctttbratit potential of dqmlarkd trace. F: spike potcnthd followed by a short early AHP and a DD &ow) superimposedon the late AHP. Dasbed line indicates mcmbmne potczttial. Blrsir bar indicates current stimulation. Membrane potentials arc indited on the right of each reoxding.

191

48 DIV (mean = 26 f 1 DIV). Seventy four motoneuand the data base rones were recorded intracellularly for the present study was constructed from a sample of 54 motoneurones in which stable intracellular recordings were obtained for up to 90 minutes with membrane potentials below -50 mV and overshooting spikes. This population subdivided in 38 cells from the caudal brainstem region containing hypoglossal, ambiguus and dorsal vagal nuclei, 14 ce!ls from the midbrainstem region containing facial and abducens nuclei and 2 cells from the oculomotor nucleus region. me were unable to identify significant morphotogical or electrophysiological differences between these motor populations in organotypic culture. Although the 20 neurones excluded from the data base were presumed to be damaged by impalment, we could not exclude that they were undifferentiated cells. 3.2. Passive membrane properties 3.2.1. Membrane potential Upon penetration of the microelectrode, a voltage drop of -50 to -70 mV was observed. Tnis often tended ro decrease uniess a ilyperpoiarizing current (- 0.2 to - 1 uA) was applied for a few minu:es. This resulted in an improvement of recordings and a stabilization of the membrane potential. Injection of current was then discontinued. Values of membrane potential ranged from -51 to -85 mV with a mean of -65.5 f 1.4 mV (n = 47). No correlation was found betwtzn m;mbranc potential values and age of the cultures.

Input

3.2.2. resistance The whole neurone

input

Y esistance

CR,)

was meato underthreshold depolarizing and to hypcrpolarizing current pulses of low intensity (0 to - 1 nA) applied through the recording microelectrode. Measurements were made 25 msec after the onset of the pulse and plotted. R, was taken as the slope of the computed regression iine through data points. ivicali “nr,, v&e for the complete data set was 53.9 f 5.5 MR (n = 47). In fact, most (n = 35) of the values are distributed between 10 and 60 MLJ with a mean R, of 34.6 f 2.5 MR, but 12 neurones had R, ranging from 62 to 200 MR. Correlation of R, with age in vitro showed that R, tended to be smaller in older cultures but this may not be

sured using the transmembrane responses

significant due to the small number of cultures maintained for more than 40 days. 3.2.3. Rectijications Prolonged hyperpolarizmg curr,nt pulses were applied to the motoneurones and transmembrane voltage measurements were made at the onset and near the end of the pulse. No anomalous or time dependent inward rectifications were observed in the range of

currents used. Plots of the two measured values gave superimposed linear graphs.

stimuiating

Tire

time constant of the neuronal membrane was measured using the passive charging of the neuronal membrane in response to hyperpolarizing current steps. The voltage curve could be closely fit by a single time constant did expwentia~ and the corresponding nor differ sigmticantly from the values measured using the simplified procedure previously described [XI which was thus retsined as standard procedure. Values ranged from 3 to 47 msec with a mean of 13.2 f 1.6 msc (n = 45!. 3.3. Action potential Spontaneous action potentials were often observed in brainstem motoneurones in culture. They were triggered on large spontaneous depolarizing potentials of large amplitude and irregular shape (Fig. lA), pre-

sumed to be synaptic potentials since they were sup pressed by bath application of tetrodotoxin CITXI. Action potentials were also triggered by depolarising current pulses (Fig. 1B) or as rebound responses at the break of an hyperpolarising current pulse (Fig. 10. With depolarizing currents of increasing intensity, a prepotential, 0.5 to 5 mV in amplitude and lasting up ms was first seen on the eleetrotomc depolarization before the spike potential was triggered (Fig. 1B). The depolarizing nature of this potential is demonstrated by the lack of rectification of the I/V curve. According to cells, it displayed different shapes, ranging from an early rounded hump to a slow, long lasting, depolarization. Threshold for spike generation was -49 f 1 mV (n = 48). Spike potentials had an amplitude of 58 f 1 mV (n = 48) and were fast rising without inflexion points on the rising phase, except in two cases where a clear IS /SD dissociation was observed (not illustrated). Mean overshoot value was 8.4 t 0.8 mV tn = 48). Spike duration measured at 50% of spike amplitude was 3.7 f 0.3 ms (n = 48). The spike potential was followed by after potentials of different shapes. In 25 cases, the spike was followed by an afterhyperpolarization @HP) without any Delayed-Depolarization (DD) superimposed on the AHP or the repolarizing phase of the spike (Fig. 1D). At resting potential, the AHP had a mean amplitude of 4 f 1 mV tn = 25) and a duration of 51.3 f 8.8 m.s (n = 24) but could be as long as 290 ma. The time course of the repolarizing phase of this AHP was linear or gently convex. The AHP amplitude was voltage dependent, being larger when the cell was depolarized and suppressed by hyperpolarization.

to IM

I. Eustache, 3.P. Gwimd/Decelqmetuol

Brain Research 86 (1995) 187402

+

43.8 nA

1.1 aA

-.-

Fu. 2 Low voltage activatedpotentiials;upper traces,voltage rawdings. Membrane resting potential is indicated on the beghmi~gof each trace. Lover trace: current stimulatioo. DC value is inducted oo the beginnieg of the trace. Black bar indicates current iqiection pulse with oonesDoadingintcnsily shown above.A: after depolamtion following UKspike potential. B:when triggeredfrom a $liitly more polarized level, this after depolarizationtriggersa doobkt spike. C when tbe all is @rized to - 106 mV a depolarization Carrow)is seen ttoder the spike. D: in soother motooeumnc, the LVA potential is sea in isolation. When the stibmdusis increased,the LVA potential (first arrow) triggers a full spike fotkwed by P slow rchrm to baselime(sezond anow). Voltaee cab%ntions:SOmV: time calibration: SOms.

In some c&k, the base of the spike was enlarged and the AHP had a very low amplitude that could only be worked out from the baseline by depolarisation (Fig. 1E). In these cases, hyperpolarizing the membrane uncovered a depolarization overimposcd on the base of the spike (Fig. lE, arrow). In 29 cases, the spike potential was followed by a distinguishable DD. In 18 cases, this RD couid be seen as a small depolarization overimposcd on the AHP @lg. lF, arrow). This sizJak1 wan similar to that

described on adult oculomotor neurones in vitro [26] where spikes are followed by a fast AHP, a delayed depolarization and a slow, long lasting AHP. In embryonic motoneurones, this DD had a variable latency, a duration of 30 to 60 ms and wss separated from the spike by a short AHP of 5.6 f 0.9 mV (n = 18) in amplitude. The long duration of this potential often offset the late AHP. In 11 casts, the DD was more prominent, and no AHP cuuld be Seen (Fig. 2A). These cells always dis-

9 2.

1

-_

f’tt ni-A3: t~arr~mbrros uar+x~~~(u-r tru;i w UI_Q currsnr pukes iir L.JXZ: _f Li?~a;lin~ intznS?,*.Notice post bunt AHP. A,: iostanuu~ws ioterspike f&~.~nciea for the successive interspike iotetvals as P fttoctiott of the injected correot itttmsity. llte rank of the iotet~al is iodiited by ttttmbcm.%ttw ncutunc as &A,. Et-B,: Modiitkxt of the responseof another motoncuronc (upper trace) to the same 1 nA dclaluizintt curcot pulse &wer trace) applied at three different membrane potentials indiuted on the left. B,: supccrimpwed rewonw of the samemohmeurone as B,-ESto a just undertbresboldcttrrettt pttlse at two diierettt membrane potentials.

I

--I

0.4nA

A2 11 U[ ~~~ ”

___..A___-______________~

I

0.7nA

\

!_-----

_1

1nA

I

1. Euttache,J.P. Gutrftaud/L&+mental

played anodal break rebound depolarizattons which eventually triggered full spikes (Fig. 10 and a typical feature was the frequent firing of doublet spikes on top of the DD (Fig. 2B). The second spike had a higher threshold, lower amplitude and slower time course than the first one. When the membrane was hyperpolarized to levels below -90 mV, low voltage activated (LVA) potettt%?!rrtiggerea by depolarizing current pulses (Fig. 2C,D) were shown to underly the spike. Fig. 2D shows two superimposed traoss recorded in a motoneurone with a membrane potential of -75 mV hyperpolarized to -93 mV by a constant current of -0.4 nA. An intracellular stimulation of 0.1 nA induced a depolarization, 23 mV in amplitude. overlasting the current pulse (lower trace). A 0.2 nA cur-

Brain Rexarch 66 (1995) 187-202

rent stimulation triggered a fast spike. This sp;ke can be seen to start on an underlying prepotential (left arrow) and to repolarize without an AHP. At the break of the current pulse, the potential returned slowly to base line (right arrow). In these cases, the .QD foltowing the spike could thus be identified as a low voltage activated (LVA) potential (Fig. 2A-0. 24. Rqetitiue firing Increasing stimulating current intensity increased the amplitude and duration of the DD until a second action potential was triggered on top of this depolarization and repetitive firing of the motoneurone occured, A spike train was then triggered for the whole

A

I= TTX

Jl

--

F~4.A:rm,superimpasedtracesshowing responw to a 0.2 nA current pulsebefore and after (arrow) TIX superfitsion(10-6M). Dashed line ittdiit~s ekctmtonic potential. 8: two superimposedtraces (lo-*M) shwhtg rcspw~s to a 0.4 nA current putsebefore and after (arrow) ‘bariumsuperfwion). Membrane potenttat of controt responseis indicated by dashed line. Barium trace is 5 mV hyperpolarized compared to control. Spikes triggered during Ba*+ superfusionare indkated by dots. Note shorter AHP and increased frcqttettcy.Post bunt AHP (lower arrow) is suppreued by Ba’+ superfusion.C: phttcru wpoase obtained atIer barium bra acted for 3 minutes.D: superimposedresponses(upper tracej to an h&ted &pokizing current pAx of constant intensity applied at different membrane potentials with TTX (10~shlul)and Ba” (10~‘MI superfusii to show vo!tye sensitivecalcium potentiat ~~~akd by barium appliition and enhanced by depolarization. Duration of current iqjectii is indicated by bar under the correspondingrcmrdh~gs.Current intensityas indicated. Calibrations for all tracts: 25 mV, 25 ms.

I. Eastache, J.P. Gwntaud

/ Lkelopmcnrol

duration of the current pulse (Fig. 3 Al-AZ+). Progressive adaptation of the discharge rate was ah++ pitsent. The interspike intervals during the spike train changed progressively from a steep ramp to a slower s&+::j r~‘Tidz3 t’i.T”8” course. Increasing the currer,: intensity increased the frrquency of firing (Fig. 3 Al -

Brarn Research R6 (1995) 187-202

195

A3). Instantaneous frequency for the first interspike interval could reach values above 100 Hz, whereas the maximum steady state discharge uxld reach SO Hz. Fig. 3 A4 shows the frequency/current curves where the instantaneous frequency for the displayed intervals is plotted against the injected current. The cmves are

TEA+lTX

B

F

C TEAflTX

WASH +

Fig. 5. Upper Itaces: voltagerecordings.Lower traces:curteat stimulatuxts. A: control responseto an intracellular current pulse of 0.15 aA. 8: response13 the sartte stimulus after TEA (lo-‘M) superfusion. Notic. widenmgof the acoonpotential and diippc~nnce of AHP. C: resqonse to the samestimulus after addition of ITX f10-6M) to the superfusion medium. D: responsesto current stimulations of imxcasedintensity as indicatedon the right of eachrecord.E: samesituation as D, after addition of cob.& LOthe (IO-%) petfksiin medium.Notice bloc& of slow potential and appearance of a spikelet response.F: ~ecovetyafter return to control medntm.

I. Eu.mche, J.P. Gwitoud/Dc~~ntal

1%

Brain Baeanh 86 (1995) LV.202

parallel with an average slope of SOsp/nA. Although these curves are slightty convex and the frequency of the Srst spikes tended to increase faster with low intensities, the curves do not show more than one range. The firing pattern of motoneurones in response to an intracellular stimulation of constant intensity was

modified by slight changes in resting potential. Fig. 4 Bl-B4 illustrates a neurone in which a continuous discharge was obtained with a 1 nA current pulse when the membrane potential was held at -58 mV (Fig. 3 Bl). Slight polarization of the membrane to -60 mV changed the firing pattern to the same stimulation. The cell displayed an early high frequency burst of action

D

A

1 Xhnscc

1nA

-l

1.2nA

E

B

-I -58mV .-_I

1OOmsec

Trx

TEA

L 1IlA s0nw-c -

C +

TTX TEA

A

25mV

P--7-__

I

f-----T

4SmV

-I-

-6OmV

I .2nA

I

laomsec

-

ZltA

1 5Omscc -

-

Fig. 6. A! control response (upper trace) to a depolarizing current pulse &wcr trace). Same cell as Fig. 4 B,-B,. B: Blockade of spikes by TtX (10-6M) SUPC?~UGOO.C: ~nki~m ptenthts tri#ped by an inmead currat pulseafter TIX (10v6M) and TEA (lo-‘M) superfusion. D: bkckadc of calcium ~~4eatials atIer Cobalt (10-3M) was ad&d to the supetfusii me&urn. Notice rctngining spihkt. E: same situation in another motoncuronc. Supcrintpmed tnnsmcmbnw responsea (upper tracts) with increasing current stimulations &wcr trace) after ITX (lo-‘M), TEA (10-6W and Co*+ (10-3M) supettttaii. F: suppression of the spikekts when sodium is replaced by choline. Voltage catibratlon in F applies to all traces. Time calibration as indited.

potentials followed by a silent period after which firing resumed at slower frequency (Fig. 3B2). Further polarizing the neurone to -65 mV resulted in the same stimulation inducing only a 200 ms burst of action potentials (Fig. 3B3). Reducing stimulation current to 0.6 ILL shows that this burst is triggered over an underlying prepotential, 5 mV in amplitude of same duration as the burst
Upon cessation of current injection, firing stopped, membrane potential returned to base line and a postburst AHP was seen (Fig. 3 Al-A9 The amplitude of this potential increased with the number of spikes trlggered on the pulse until a maximum was reached after which it remained constant with further increases of stimulation.

Action ootentials of brainstem motoneurones of the rat in organotypic cultures displayed the classical sodium and potassium dependence. Spontaneous action potentials and spikes evoked by depolarizing current pulses were reversibly suppressed by lTX (10W6M) into the bath (Fig. 4A). Full spikes were then absent regardless of the strength of the stimulation. Fu!: spike potentials were also absent when sodium was replaced

A I-TX

C

--

D

0 61tA

-

I-TX

O.&IA

5amsec

Fig. 7. Low-threshold calcium potentials. A: two supenmpcwd responses(upper traces) to a depdting current ptk of 0.6 nA (kwer trace) applied in an hyperpolarizcdnwtoneurone before and after TTX (lo-%I) superfus~on.-\rrow points to TTX insenkive depolarization. 8: thrcz superimposedresponses(upper traces) to the sane depolatizing current pulse applied at three different membrattepotentials to showMilppe depettdsnceof the TIX-insensitive depolarization. Same cell as A. C: two superimposedresponsw(upper aceS) to a dcpdaririw cttmOt puk (lower trace) before and after (atrow) addition of TEA (iO-‘M) to the superfusronmedium. D: black& of the potential when Co*+ (lo-“M) is added.

hy choline chloride in the bathing solution. These two procedures also resulted in the suppression of all synaptic noise. All the tested neurones showed evidence for the existence of TIX-resistent sodium potentials. These potentials appeared as spikelets (Fig. 5E, Fig. 6D,E) when potassium and calcium conductances were blocked by tetraethyiamonium @EAT and cobalt respectively. These spikelets were suppressed when

sodium was replaced by choline chloride (Fig. 6F). Blockade of potassium currents by partial substitution of Ca2+ by Ba*+ (10-4M) in the bathing medium @ii. 4C) or addition of TEA (10-3M) into the bath (Fig. 5B) resulted in the widening of the action potentials and suppression of the early AHP. The presence of calcium dependent potassium conductances can be demonstrated since these are blocked hy Ba*+, resuhing in a shortening of the AHP, an increase in firing frequency (Fig. 4B) and the suppression of the pmburst AHP (Fig. 4B, arrow). Effect of Ba*+ on late AHPs was difficult to observe due to the low amplitude of these potentiais. 3.51. Colcim potentials Calcium dependent potentials were observed in 28 motoneurones. These potentials belonged to two di’. ferent families: high voltage activated (EVA) and low voltage activated (LVA) potentials. 3.5.1.1. High thre&told ptentiols. When spikes were suppressed by ‘ITX, a slow depolarization remained in response to depolarizing current pulses (Fig. 4A, arrow, Fig. 6B). When potassium conductances were blocked by TBA (10-3M), z vobagc dependent, ‘ITXresistent potential was uncovered by increased intracellular stimulation (Fig. SC,D). At depolarized levels, a succession of progressively inactivating spikes was obtained (Fig. SC). Blockade of calcium currents hy Co*+ (Fig. SE and 6D) shows that these potentials were mainly due to calcium currents. The contribution of calcium currents could be further demonstrated when Ba*+ partially replaced Ca*+ in the perfusing medium (Fig. 4D) since Ba2+ was shown to enhance. current through Ca *+ channelc. In these conditions, a broad, voltage dependent, low amplitude spike was obtained. It was triggered at a membrane potential closeto -40 mV. Further membrane depolarization :riggered the spike earlier and increased its amplitude to a maximum of 18 mV. For potentials above -25 mV, it started to inactivate. 3.5.1.2. Low thresbld potentials. The calcium dependence of LVA potentials described above is demonstrated on Fig. 7. In this neurone, a doublet of spikes is triggered upon depolarization of the oz.11from -100 mV. The first spike starts on a prepotential which can

hc Seen in isolation when spikes are suppressed by TTX (Fig. 7A). This depolarization is voltage dependent, since.it is only triggered at levels below - 72 mV. It is maximal when triggered from -90 mV (Fig. 7B). Addition of TEA to the bath increases the amplitude of this depolarization (Fig. 7C) which can be cotnpletely suppressed when Co*” replaces Car+ in the medium (Fig. 7D). This demonstrates that this depolarization corresponds to the activation of LVA calcium conductances. In 4 motoneurones, EVA and LVA caicium potentials were demonstrated in the same neurone, controlmg different discharge properties according to membrane polarization. 4. Discussion The main goal of our work was to describe electrical properties of differentiating brainstem motoneurones grown in organotypic culture with skeletal muscle fibers. In a previous pager [281,we showed that the presence of muscle tissue is necessary for the survival and differentiation of neurones derived from the explanted motor population. Motoneurones were identiBed on the basis of morphological and topographical criteria. Although these cells were derived from the explanted motor population and newly differentiated motor end plates where observed, it is necessary to demonstrate that these neurones establish functiomml commctions with muscle fibers. In the present study, we demonstrated the motor nature of these neurones using microstimulation techniques. Although this procedure could only be applied to migrated and well accesstble motoneurones and thus does not demonstrate that all the neurones are actually connected, it shows that the majority of motoneurones accessible for intraceUular recording are ftmctionrd. The results reported in this work demonstrate that brainstem motoneurones of rat embryos maintained in organotypic culture in vitro for 2 to 6 weeks display a complex electrical activity, especially with regard to sodium and calcium potentials. 4.1. Passive properties We excluded from our data base neurones with membrane potentials above -50 mV, since they may he considered as damaged neurones. We may thus have excluded from this study undifferentiated motoneurones with small membrane potentials, but this was justified since they could not he distinguished from damaged cells. Measured membrane potentials were in good agreement with values given for cranial motoneurones both in vivo [2,13,221and in vitro [12,26,41,48,50]and for embryonic motoneurones [3,70].

199

Whole neurone input resistance are in close agreement with those described for adult ocular motoneurones in vitro [26] or neonatal hypoglossal motoneurones [63], but larger than that given for facial [SO]or hypoglossal motoneurones 1481.Whole neurone input resistance measured in the soma is an index related to neuronal size and geometry of t:re dendritic tree [3,24,25,44,57]. It may be expected that brainstem motoneurones developing a flat morpnotogy m culture wouiJ have larger input resistances. A decrease of R,

occurs with the differentiation of the neurone [17,65,70]. We noted lower R, values for cultures older than 40 days, but most R, values were usually scattered without correlation with age of the cultures. This suggests that indtvidual neurones may ryach different maturation states in a given culture. The monoexponential time course of the membrane voltage transient to hypetpolariiing current pulses sug-

gests a compact electrotonic structure with little redistribution of charges in distal dendrites. These results are in good agreement with Viana et al. [6S] who also described a compact electrotonic structure and noted that input resistance decline was weak during the first in two postnatal weeks for hypoglossal motoneurones

slices. The membrane of the motoneurones had an ohmic behaviour in the range of hyperpolarizing or depolarizing currents examined and did not display either anomalous rectification or time dependent inward rectification. Although this is in agreement with earlier studies on adult cats in vivo 181, lack of rectification was often described in embryonic or newborn animals in

vitro [12,17,31,65]whereas non linear changes of motoneurone membrane resistances with polarizing currents [49] are often described in vitro in adult animals [26,41,48,50,6S]. Rectifications are attributed to slow ionic currents [4] and their role in cell behavior is not clear. Since they seem to develop only very late, they may contribute to some elaborate aspects of signal processing. 4.2. Acthe properties Brainstem motoneurones in culture display an intense synaptic noise and spontaneous action potentials. Although a detailed study of synaptic input to these cells was beyond the scope of this paper, our results show that these motoneurones receive a powerful excitatory drive from the surrounding structures. Although no hypetpolarixing potentials were seen, the presence of inverted IPSPs cannot be ruled out [59,60]. 4.3. Action potentials Spike potentials had thresholds and amplitudes in good agreement with other studies. In our cultures,

spike potentials had a larger width than scribed. Large width of spike potential

generally

de-

is an index of immaturity ]47], and is known to decrease during the first postnatal weeks [40,65,7OJ,consistent with an increase in the membrane density of sodium channels [36]. Low temperature of the recording medium also results in the widening of spike potentials (611. Further studies, where other means of functional identification will allow for the blockade of muscle contractions by curarization, will have to determine the relative contribution of these two factors. In accordance with previous studies on embryonic mo:oneurones [33,63], our results show the important contribution of calcium events to the triggering and shape of the action potential. A typical feature of many brainstem motoneurones in culture was the presence of underthreshold potentials and delayed-depolatizaticms of large amplitude related to the presence of voltage dependent calcium potentials. Afterdepolarizations or delayed-depolarizations have been descriid in all studies on spinal or brainstem motoneurones and are attributed to calcium events in the dendritic tree L23.30.331. 4.4. Repetitwe finng Brainstem motoneurones in culture display repetitive firing properties similar to those already descriid for motoneurones in vivo and in vitro. This includes con:inuous firing, frequency adaptation and post-burst hyperpolariz;tion. The main differences are due to a slower adaptation and a lower firing range. Of potential functional significance is the moduiation of the firing pattern with slight shifts of membrane potential. This effect is due to an underlying voltage dependent calcium conductance activated at depolarized potentials. Our results show that this conductance may play an important role in shaping the firing pattern of the neurone since it produces a burst of action potentials and this shii of pattern from tonic to phasic discharge may result in important changes in the speed and intensity of muscle contractions. Similar results were described in postnatal hypogfossal motoneurones in the slice preparation [63]. 4.5. loruc conductances The membrane of brainstem motoneurones in culture contains different types of ionic conductances. Our results demonstrate the existence of two types of sodium conductances, one being the classical Hodgkin-Huxley Tl’X-sensitive conductance which is present in all neurones whereas the other one is an inactivating high threshald ‘fTX-insensitive sodium

conductance. This type of sodium conductance has

200

been described in dorsal root ganglion cells at all stages from embryos to adulthood [53,69] and in striatal and hippocampal neurones [35] where it may conmbute to modulation of neuronal firing. Our results also show the existence of a fast inactivating _potassiumconductance wkich ls blocked by TEA application and of a calcium dependent potassium conductance which underlies the late AHP and the postburst hyperpolarixation and is suppressed by barium application. Barium was shown to be a better charge carrier through calchun channels and thus suppresses the activation of calchnn dependent potassium conductances which were shown to underly the slow long-lasting AHP and the post-burst hyperpolarixation [3,11]. The small amplitude of these potentials in our preparation might be due to the large amplitude of simultaneous calcium potentials, but it could also tend to demonstrate that this type of potassium conductance is not yet completely developed. When sodium entry is blocked and the balance between potassium and calcium currents mod&d by partial block of potassium currents by TEA or increase.

of charge transfer through calcium channels by barium, large amplitude calcium potentials where observed. We were able to demonstrate the presence of at least two types of calcium potentials. HVA calcium potentials were descriid either as cobalt sensitive prepotentials with a slightly lower threshold than the spike contributing in a significant way to the pattern of discharge of the motoneurone or as slow calcium spikes uncovered by ‘FIX blocade of sodium potentials with increased intracellular stimulation. Although precise characterisationof these ca!cim channels requires further studies with voltageclamp techniques, it is possible that these potentials cormspond to dllerent maturation states. LVA calci~ potentials were observed in 16% of the recorded motoneurones. T&se potentials were first descriid in the olivary complex [42,43].LVA calcium currents enhanced by serotonin were recently descriid in neonatal rat spinal motoneuroncs ,[6] and low threshold transient responses activated from holding potential of about -90 mV were seen in adult rat oculomotor neurones in vitro [271.Altho~gh these potentials are demonstrated at hyperpolarixed levels, they may contriiute significantly to the control of firing of differentiated motoneurones in response to specific ncurotransmitter activation. It may be concluded from this study that organotypic cultures of embryonic rat brainstem with muscle Ebers offers the possrbility to study the early phase of the differentiation process of motoneurones. Recordings obtained from cultures 3 weeks in vitro display more signs of immaturity ihan those obtained in vitro on slices from animals of equivalent age [26]. This shows that diierentiation proceeds more slowly than in

vivo, yet motoneurones wcome functional and display a large set of membrane conooctances that can be compared to those already described in adult or late immature motoneurones recorded in vivo or in slice preparation [11,12,26,31,33,4&63,64,67,70]. Our results show that motoneurones in the same culture may display different membrane properties and specific calcium currents. This may correspond to the fact that, within the same culture, individual motoneurones reach different maturation states. But it may be significant to notice that we did not observe differences in membrane properties that could be correlated with the anatomical origin of the motoneurone pools present in

our cultures. Although the motoneurones grown in our cultures were derived from different cranial motor nuclei, the different populations displayed strikingly similar electrical properties. This may mean that adult motoneurones in vivo display the same set of complex membrane properties, but that these are differently modulated by the synaptic inputs from the surrounding neural network, or that synaptic activity induces changes in the distribution and properties of the ionic conductances present in the membrane. McCobb et al. [45,46] have shown that large changes in density of several voltage dependent ionic currents take place durhrg maturation of chick motoneurones. It may also result from neuromuscular interactions with a specific muscle target. Porter and Hauser [S6] have recently shown the importance of appropriate motoneurone/ muscle fiber coupling for the differentiation of neuromuscular function. Further studies will have to address the role of presynaptic networks and specific muscle fiber types in the indiidualiition of each motonearone pool and determine the specific distribution of membrane currents.

This work was done with the technical collaboration of N. !Seyfritx.We want to thank Prof. H-R Llischer, Dr. J. Streit and Dr. S. I)+Dumont for reviewing the manuscript and for many fruitful discussions. Thomas Launey helped with illustrations. We thank Michele El&m for her help wth the manuscript. The work was supported by a grant from Fondation pour la Recherche Mbdicale. Isabelle Eustache was supported by a grant from Ministbre de la Recherche et de la Technologie.

RefeI¶?Alces [I1 Alkrmtd, CD., Patteros of neuronal differentiation in develop ing adttues of tteonatnl matsc cercbcUum: a liip and sihrer imprecation study, I. Camp. Newof., 142 (1971) 167-204.

[Z] Baker, R. and Precht, W.. Electrophysiologicsl properties of trwhlear mntoncurones as revealed by IVth xrve stimulation. .&q~.Bmm Ru.. 14 (1972) 127-157. 131 Barrett. J.N. and Grill. WE., Specific membrane properties of cat moto”eum”es. I. Physio~. 239 (1974) 301-324 141 Barrett, E.F., Barrett. J.N. and Cnll, WE., Voltage-sensitive outward currents in cat motoneuronss. 1. Physrol., 304 (1980) 251-276. I51 Bennett. M.R., Lai, K. and Nurcom’>e, V., Identification of embryonic motoneurons in vitro: their survival is dependent on skelcsl muscle, B&n Res.. 190 (1980) 537-542. 161 Berget. A.J. and Takahashi, T., Serotonin anhances 3 low voltage-activated calcium torrent III rat spinal motoneurons, Eur. 1. Newzxer.. 10 (1990) 1922-1928. [fl Braschlcr, fJ.5, A Fpenger, C, Streit, J and Likcher. H-R., A modified roller tube technique for organotypic urmltures of embryonic rat spinal cord, sensory gangha ano skeletal muscle, J. Nnvosci Merho& 29 (19893 121-129. [S] Coomhs, J.S., Eccles, J.C. and Fatt, P, The electrical pmyrtiec of the m~toneorone membrane, 1. Physrol., I30 (1955) 291-325 191 Grain, SM. and Peterson, E.R., Development of neural mnnections in rultore, Ann. NYAcnd Sci., 228 (1974) 6-34. [lOI Grain, SM., Alfei, L. and Peterson, E.R.. Neuromuscular transmission in cultures of adult human and rodent skeletal muscie after innervation in vitro by tetal rodent spinal cord. 1. Neurobid., l(l97c; 471-489. ill] Grill, W.E. and Schwindt, P.C., Active currents III mammalian central neurons, T.J.N.S., 6 (1983) 236-240. [12] C”rti$ J.C. and Appenteng, K_, The electrical geome!ry. electricat properties and synaptic connections onto rat V motoneurones in vitro, I. Phy&/., 465 (1993) 85- 119. [13] Destombes, J.. Dorand, J., Gozan, P.. Gointaud, J.P.. Herchollc-Bmsavit, G. and 7)wD”mont. S., Ultrastructural .;ld electrophysiological properties of Accesso~ Abducens nucleus motoeeurones: an intracellular honeradish peroxldase study m the cat, Newoscie”ce, 10 (1983) 1317-1332. [14; Eustache, I.. Soyfritz, N. and Gueritaud, J.P.. Effects of insulinlife growth factors on organotypic co-cultures of embryonic rat brainstem slices and rlteletal muscle fibers, Dev. Bram , 81 (1994) 284-292. Da] t+hring, R.C., &pert, G.W. and Munson. J.d.. Motor-unit properties following cross-innelvatlon of cat lateral gastmcnendus and solcus muscles with medial gas:rocnemius nerve. 1. Influence of mOto”euro”s on muscle. J. Neurophysrol., 57 (1987) 1210-1226. I161 Foehring, R.C.. Sypert, G.W. and Munson. J.B., Motor-on” pmperties following cross-innervation of cat lateral gastrocnemius and soleus mrsscles with medial gastrocncmius new. II. influence of muscle on motoneurons. 1. Neurophysid., 57 (1987) 1227-1245. [17] Fulton, B.P. and Walton K., Electrophysiologwal properties of neonatal rat motoneurones studied in vitro, 1. Physd. 370 (19%) 651-678. [18] irons, M., Krieger, C. and Sears, T.A Idmtlftrat+.n md e!cctrophysiological investigations of embryonic mammolia” motoncurorrs in culture, Neumrct. L.&r, 83 (198’7) 82-88. I191 Gihwilcr, B.H., Organotypic monolayer cultures of nervous tissue, 1. Nnrmrci Metho&, 4 (1981) 329-337. I201 GBhwiler. B.H., Enz, A. and Hefti, F., Nerve growth factor promotes development of the rat septo-hippocampal cholinergtc properties in vitro, Newo.x~ Lrrr., 75 (1987) 6-10. [Zl] G&viler. B.H. Rietschin, I.., Kniipfcl. T. and Ena, A., Continuous presence of nerve growth factor is required for maintainance of cholincrgic septal neurons in organotypic siice culture, Neumrciencee. 36 (1990) 27-31. 1221 Gogan. P., Guiritaud, J.P., Horchollc aswit, G. and TycDuntont, S., Electmtonic interaction between motoncurons of

Iannone,

Res

the Abdocens nucletts of the cat. lo G. l_e”“emtrand and P. Bach-Y-Rota (Eds.). Basic Mech&rmr of Cc&r h&n&y and thw Climc4l Im+arfou, Pergamon, oxford, 1975, pp. 461-463. [Ul Gramt. R.. Kern&, D. and Smith, RS, Detayed depolx&fion and repetitive response to intmlhdar stinmlation of mammahan motoneuroas. I Pk@!.. 168 (1%3) 890-910. velocity, I241 Gmntyn. R. Grantyn. A and Schaaf, P., conduaion tnpuf rewtance and size of cat ocular “mtooeoro”cs staiaed wth Proc~on Yellov, Erofn Res, 135 ClQ77) 1677-173. I251 Granty”, R. and Grantyn. A, Morpbobgfal and e lo&r properties of the eat abduceos moloMuTones. Em. Brain Res , 1 I ( 1978) 249 -274. (26) GuOntaud, J F., Electrical activity of tat onrlar moto”eunms recorded I” wtr”. Newoxience, 24 (1988) 837-852. I271 Guentaud. J PI Bariom-induced bistahility in rat ocular motoneurones III v:tro, Newwci. Lmfr.. 170 (1994) 158-162 I281 Guintaud, J P and Seyfria N., A” orgaootypic coxolture of embrynazc rat bra&stem sod tongue or skektat muscle, lxu. I. Neluoscr 4 f I291 Guentaud I P, a”d Seyfri& Y., Electrical activity of cramal motoneonxxs in orynotypie slice co=cultores of embryomc rat brainstem aad striated murk. XVfth Europea” Neuroscience Meeting, Madrid, Eur. I. Navmci., Suppi. 6(1993) 28. [30] Gustafsson. B., Afterpotendats and transdoctii” proper&s i” different types of central newones_ Arrh Ital Bid.. 122 (1983) 17 30 (311 Haddad, G G.. Donnelly, D.F. and Getting, P_A.. Biophyskal p~jpert~es of hrposlorsa “eum”s in vitro: iotracelluLr studies m adult and neonatal rats, J. Appl. physic.. 69 (1990) 3509-1517. [321 Hammarbe.2. C and Kelkrth, 1.0.. The postnatal developmeat of some twtch und fat&e properties of single motor wits i” the ankle mucks of the kitten. Acta Physf& Swnd., 95 (1975) 245-257 (331 Haradn. Y. and Takabashi, T., The calcium ccm?pone”t of the actlo” puteatlal in spinal motoocumnes of the rat, 3. phyad, 335 (1983) 89-100. I341 Hollyday, M. ani Hamburger, V, Reduction of the naturally occunng motor neuron loss by cnlargcmcnt of the periphery, 1. camp Neural , 170(1976) 311-320. 1351 Hoehn, R. Watson. T.W.J. and Ma&ii. B.A., A “owl Tetrodotoua-mxnsitive, slow sodium current in strietal and hippocampal newons, Neuron, 10 (1993) 543-552 !36] Hoguenard. J.R.. Hamil, O.P. and Piinee, DA, Developmental changes m Nat amductanees in rat neomrtical neotws: ap pearance of a slowly inactivating component. 1. Ncwophysiof., 59 (1988) 778-785. I371 Hwar, P., Kuno, M. and M&at.+, Y.. Differentiation of motoneumnes and skeletal muscles in kittens, J. Phyti, 252 (1975) 465-419. (381 Johnson. E M.. Tan&hi, hf. and Di Stefano, P.S., Expression and possible function of nerve growth factor od khan celii. 7’ie”d.r Neumsu., 11 (1988) 299-304. [39] Johnson. MI. Higgins, D a~ A A-i, M.D., Astmcytea induct dendritlc development in cultured sympathetic neurons, Deu. Bram Res.. 47 089~ 289-292. [40] Kellerth. J.O.. Mclktr6m, A and Skoglund, S.. Postnatal escltabdity changes of kitten “~Xoneurones, Acrcr Phy&k! scnnd., 83 (1971) 31-41. [-‘I] Larknun. P.M.. Peningto”. NJ. and Kelly, J.S., Ekctrophystology of adult rat facial motoneorones: the effects of semtonin (5-m) in a novel in vitro brainstem slice. J. NNarci Me!hod& 28 (1989) 133-146. [42] Winas, R. and Yam”& Y.. Elect-ilogy of mammalian inferior olivary mtmxes in vitro. Different types Of vdtase-dependent ionic cwduarnoes, 1. physid, 315 (1931) 549-567. !O] Llinas, R. and Yuom, Y., Properties and distnftttiott cvi ionic

19921616-627 Eu.stack, 1.

I. Eusbxhe, I.P. Guetimd/Lkwlcpmeml

202

conductawes generating ekctroresponsiveness of mammalian inferior&vary neumnu in vitm, J. P&sfd, 315ft%1)569-584. I441Lux, H.D., Schubert, P. and Kreutzberg,G.W., Direct matching of nmrpbolodal and ekctmphysiological data in cat spiaal motooeumns. In P. Aodersen and JKS. Jansen (Eds.), tiimre+y*n&c Mec/wtitmt, Uniwsitctsforiaget, OsIo, 1970, pp. 189-198. [451 WC&b. D.P., Best, P.M. and Beam, KG., Dcvelom,,ent alters the expression of c&urn currents in chiik lib mototteumns, N-, 2 (1989) 1633-1643. 1461McCobb. D.P., Best, P.M. and Beam, KG., The differentiatiw of excitabilityin embryonicchiik limb motoneorons,J. Ncumvci., 10 wwJ12974-2984. [47l Mecormidr, D.A. and Fliocc, D.A., Post-nataldev:!optttentof ekctmpbysiologicatpropertics of ra: cerebralcortical wantidal oeumnes, J. physid, 363 (1987)143-162. I481 md Reklittg, J.C., Ekarophysiological motoneomns of guinea-pigsstudied in pmpertiesof~ vitro, Ncumwience,30 (1989) 619-637. [491Nelson, P.G. and Fraak, K., Anomalous rectjfkation in cat spinal motooeomocs and effect of polarizbtgcomttts on excitatoty p&synaptic potential, L Neum&iof., 30 Cl%3 10971113. [SO1Niiimttra, Y., Schwiadt, P.C. and Grill, WE., Electrical prop crtks of fid mofoneufons in brainstemstkes from guinea pig, ~min per w2 (1989) 127-142. [SO O’Hat~km,G.M. and Lowtic, MB., Evidence for age dependent cbaoges in motooewone dendritic morphologyfolkwing own*taI mm-crosb in the rat, Neumxiawe, 57 (1993; ‘?95-1ODS. 1521O’Brien, RJ. and Fiscbbrh, G.D., Isolation of embryonicchick tn$meneos and their survival io vitro, 1. N-i., 6 (1986)

Mdddt-Lmrsen, A.

I531Ogata, N. ‘md Tatebwashi, It, Gntogenic davetopmentof the ‘Il’X-sensLii and TI’X-insensitiw Na+ channels in neurons of the rat dorsal mot gaagli& Dcu. Brain Res., 65 w92) 93-100. 1541Oppcnheii R.W., The neumtmpbw:theory and naterally oc12 (1989) 252-255. curing motooeumn death, Tmdc N-L, I551Petersott,E.R. sod Cmbt, SM., Regetteratiooand innervation in arjtum of adult maamuliatt skektal mode coopkd with fetal rodent spinal cord, Erp New& 36 (1972) 136-159. [%I Porter. J.D. and Hauser, KF., SurvivaIof extraoculrr muscle in long-term organotypicculture: diierential influence of appt’opriate and inappropriate motoneumns, Lku. Bid., 160 (19939 39-M.

Brain Rewarch 84 (1995) 187-202

[57l Rail, W., Branchingdendritic tree and motoneuron membrane resistivity, Eqx New&, 10959) 491-527. 1581Spenger, C., Braschler, U.F., Streit. J. and Liischcr, H-R., An organotypic spinal card-dorsal mot ganglion-skeletal muscle mcoltete of embryonic rat. 1. The morphologicalcorrelates of the spinal reflex arc, Eur. J. Neumci., 3 (1991) 1037-1053. I591Streit, I., S:. gcr, C. and Liisch.x, H-R., An organotypicspinal cord-dorsal mot ganglion+skeletalmuscle coculture of embryonic rat. II. Functional evidence for the formation of spinal retlex RIGS in vitro. Eur. L Neumrci., 3 (1991) l&54-1%8 [@j Streit, J. and Liischer. H.-R., Miniature excitatory postsynaptic potentials in embryonic motoneumns grown in slice cultures of spinal cord, dorsal mot ganglia and skeletal muscle, Erp. Bruin Ret., 89 (1992)453-4.58. I611Thompsoo, SM., Masukawa,L.M. and Prince. DA., Temperature dependence of intrinsic membraneproperties and synaptic potentials in hippocampalCA1 neurons in vitro, .I. Neurosc&5 (1985)817-824. I621Tmpaa, M., Johnson, M.I. and Higgins, D., Glial cells promote dendritic developmentin rat sympatheticneurons in vitro, Glir, l(1988) 390-*?2. [631Viana, F., Baylks, DA. and Berger,A.J., Calcium conductances sod their role in the firiog behaviorof neonatal rat hypoglossal motottewons, J. New&.vsio/., 69 (1993)2137-2149. [641Viana, F., Bayliss, D.A. and Berger, AJ., Multiple potassium conductances and their role io act& potential rep&riTztion and repetitive fwing behavior of neonatal rat hypoglossal motoneomtts, I. Neumphjviof.,69 (1993) 2150-2163 1651Vhma, F.. Lyliss, 3.A. and Berger, A.J., Postnatalchanges in rat hypoplossal motceszmn tnetttbrane properties, Newosc&tce, 59 (1994! 131-148. 1661V&ova, G., Gordon, T. and Jones, R., Nenx-mw& InznacChapmanand Hall, London, 1978,226~. l67l Walton, K and Falton, B.P., Ionic mechanismsunderlying the firing pmperties of rat neonatal motoneumns studied in vitro, N-ience, 19 (1986) 669-683. 1681Yamm, Y., Sogimori, M. and Llinas, R., Ionic cwrents and firing patterns of mammalianvagal motoneurons in vitro, Neutusc&mce,16 (1985) 719-737. I691Yoshida, S., Matsuda. Y. and Samejima A., Tetmdotosin-res&ant sodium and calcium potentials in dorsal mot ganglion cells of the adult mouse, X Nnuophysid., 41(1978) 1096-1106. :701Z&kind-Conhaim,L., Ekctrical pmperties of motoneurons in the spinal cord of rat embryos, Deu. Bid., 128(1988)21-29.

km.