Brain Research, 521 (1990) 161-166 Elsevier
161
BRES 15657
Spinal dorsal horn neurons in elevated extracellular calcium: cell properties and spontaneous discharges E. Bernard, L. Urb~in* and G.G. Somjen Divisions of Neurosurgery and Physiology, Duke University Medical Center, Durham, NC 27710 (U. S.A.)
(Accepted 9 January 1990) Key words: Seizure; Interictal discharge; Spinal dorsal horn; Spinal root potential; Membrane potential; Neuron excitability
Recordings were made from neurons in the dorsal horn (DH), and from dorsal and ventral roots (DRs and VRs) of isolated spinal cords of infant mice. Raising calcium concentration ([Ca2+]) in the organ bath from 1.2 to 2.4 mmol/l resulted in a slight hyperpolarization, elevation of threshold current (rheobase), and augmentation of excitatory postsynaptic potentials (EPSPs). In many cells EPSPs acquired a much prolonged late phase. Orthodromic stimulation evoked in some DH neurons an action potential that had the same threshold as, and coincided in time with, the 'dorsal horn response' (DHR) recorded from DR. In spinal cords bathed in elevated [Ca2+], DR recordings showed irregularly recurring spontaneous waves, and DH neurons generated spontaneous EPSPs, often with spikes. Some neurons fired irregularly timed spontaneous action potentials that did not appear triggered by EPSPs. In less than 50% of the neurons the spontaneous EPSPs coincided in time with the spontaneous DR waves. The action potentials that appeared without EPSP were fired independently from DR activity. These observations confirm that elevation of interstitial free calcium concentration results in strong enhancement of excitatory transmission, especially of an EPSP of much extended duration. Virtually all neurons showed increased spontaneous activity in high [Ca2+], but only a minority appeared recruited into the synchronized discharges that are detectable as spontaneous waves in DR and VR recordings.
INTRODUCTION In two recent studies of the effects of changing interstitial divalent cation concentrations on isolated mouse spinal cords w e 6 found that moderate elevation of interstitial calcium, [Ca2÷], results in augmentation of synaptically transmitted responses. Orthodromic monosynaptic population discharges of both dorsal horn (DH) and ventral horn (VH) neuron populations were only slightly increased, but the G A B A - d e p e n d e n t dorsal root reflex ( D R R ) and dorsal root potentials (DRP) were strikingly enhanced. In addition, irregularly recurring synchronized spontaneous discharges, resembling 'interictal' bursts, appeared in D R and V R recordings 7. Raising the Ca 2+ concentration in the bath from the control level of 1.2 mmol/l to 1.8 mmol/l induced such spontaneous activity in some spinal cords and raising it to 2.4 mmol/1 had this effect in all the others. The spontaneous activity was completely suppressed by low concentrations of the G A B A A antagonist drugs bicuculline and picrotoxin 7. These findings were unexpected in two ways. First, because one is used to thinking of low, not high, [Ca 2÷] as a possible cause of tetany and of abnormal excitability
of the C N S 1'11'18. Second, even though bicuculline and picrotoxin are, in higher dose, convulsant drugs, they depressed the spontaneous activity. This unusual combination of facts may be explained if, in high [Ca2+], G A B A had an excitatory instead of an inhibitory effect. Normally, G A B A is believed to inhibit transmission from primary afferent fibers by causing depolarization of presynaptic terminals 8,14. U n d e r the influence of excess [Ca2+], G A B A - m e d i a t e d presynaptic depolarization may become intense enough to cause the release of excitatory transmitter. The normally inhibitory feedback loops would in this way be transformed into reverberating excitatory connections, causing the intermittent recruitment of neuron populations, resulting in spontaneous discharges. All recordings in our previous study were made with extracellular electrodes. Moreover, the spinal cords were kept at a subnormal temperature. Here we report intracellular recordings made in D H of spinal cords maintained near physiological temperature. We concentrated our attention to the D H , because our earlier results suggested that the high [Ca2+]-induced spontaneous activity is paced from this region 7. Some of our findings have already been reported in an abstract 4.
Present address: Department of Anatomy, Medical University of Debrecen, Debrecen, Hungary. Correspondence: G. Somjen, Division of Physiology, Box 3709, Duke University Medical Center, Durham, NC 27710, U.S.A. *
0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
162 MATERIALS AND METHODS
A Our adaptation of the technique of isolating the spinal cords of infant mice2J6 has been reported earlierSJ'L Briefly, 9- to 16day-old mice were anesthetized with diethyl ether, and their spinal cord was removed under rapidly flowing chilled artificial cerebrospinal fluid (ACSF) and split lengthwise. Departing from the previous practice of submerging the preparation in ACSF, one hemicord was placed in an "Oslo' type interface tissue slice chamber where it could be maintained at 34.5 °C. Control ACSF contained (in mmol/l) NaCI 130, KCI 3.5, CaCI2 1.2, MgSO4 1.2, NaH2PO4 1.25, NaHCO 3 24, glucose 10, saturated with 95% O2 and 5% CO2, pH 7.4, flowing at 2-3 ml/min. The same gas mixture, warmed and moistened, perfused the air space of the chamber. Tightly fitted suction electrodes were used to stimulate L5 dorsal root (DR) with constant current pulses, and to record from L 4 DR and L5 VR. Glass micropipettes filled with 3 M potassium acetate, of 60-100 M.Q tip impedance, were used to record from neurons in DH. The spinal cord was laid either with the cut medial surface, or with the lateral surface facing up. In the latter position the pia was opened with sharpened watchmaker forceps. The electrodes were aimed at the dorsal region of the spinal cord. The electrophysiological properties of these cells resembled those identified in an earlier study as located in Rexed's laminae III-VI of the DH ~3, All recordings were from segments L 4 o r L 5. Electrical activity was recorded with a high input impedance bridge amplifier (Neurodata IR-283), which allowed stimulation of cells by current injected through the microelectrode. IntraceUular potentials were referred to an indifferent electrode in the (grounded) bath. DR and VR potentials were recorded from the suction electrodes, also referred to 'ground', by AC coupled amplifiers (A-M Systems 1700). Recordings were stored on videotape (Neurocorder DR-484) and displayed on a digital oscilloscope (Nicolet 3091). Continuous recordings were made by a pen recorder (Grass RPS7C 8A) and shorter segments of recordings were copied by digital plotter (Hewlen-Packard 7470A). [Ca2÷] was raised by perfusing the chamber with ACSF containing 2.4, rarely 3.6 mmol/l CaCI2. RESULTS
Changes in membrane properties and evoked responses Acceptable intracellular recordings were obtained from 34 n e u r o n s in 18 spinal cords. The criteria for acceptance were: at least -55 mV resting potential (except one cell with a durably stable resting potential of - 5 2 mV) and overshooting action potentials for at least 10 min. In 9 cases recordings were obtained in both control and elevated bath calcium ([Ca2+]) from the same n e u r o n for a long enough period of time ( > 30 min) to allow tissue [Ca2+]o level to approach the bath concentration after changing the solution. In 3 of these cells the initial recording was in control A C S F followed by elevated [Ca 2÷] whereas in the others the sequence was reversed. From 15 other cells recordings were made only in high [Ca 2÷] (either 2.4 or 3.6 mmol/1), and in 10 cells only in control ACSF. The m e a n resting potential measured in 2.4 mmol/l [Ca 2÷] was slightly more negative than that measured in control ACSF (mean = -68.4 mV, S.E.M. = + 1.4 mV, n = 13; compared to 61.7 + 1.4 mV, n = 18; P < 0.003). Three cells studied in 3.6 mmol/1 [Ca 2+] had a mean
¸¸¸50
,,U
1.2 mM [Co 2+] 10 msec
62
mV
B 50
2.4 mM ~Co 2+]
[/
.i
J,
c, 1.2 mM ECo2+ ,,',v~,~'.,~
mV
Fig. 1. Responses of a DH neuron to DR stimulation in normal and in elevated [Ca2+]. A: the average of 5 responses in each condition. Note that only single spikes were fired in 1.2 mM [Ca2+], and the two spikes appearing on the trace were combined in the course of averaging, as the latency varied. B: single sample sweeps from the same experiment, shown at a faster timebase to emphasize detail. DR stimulus pulse: 500 #A, 0.02 ms.
resting potential of 61.3 mV (n too small for statistical evaluation). These statistics refer to m e a s u r e m e n t s made shortly after cell penetration when the m e m b r a n e potential settled to a steady level, and do not include measurements made from the same cell after changing bathing solution. In the 9 cases where recordings were made from the same cell at two levels of [Ca 2+] the resting potential was more negative when [Ca 2+] was elevated in 6 cells, and unchanged or very slightly less negative in 3 cases (mean of 67.0 + 1.7 mV at 2.4 mM [Ca2+]0 compared to 63.3 + 2.1 m V at 1.2 m M [Ca2+]o). More important from the point of view of cellular function, the threshold current for direct stimulation (rheobase) was 0.78 n A (S.E.M. + 0.15 n A ) in high [Ca2+], compared to 0.5 n A (S.E.M. _ 0.03 n A ) in control ACSFI measured in the 5 cells exposed to two different levels of [Ca2+]. Correspondingly, the threshold voltage at which an action potential could be triggered by orthodromic (DR) stimulation was also slightly higher in elevated [Ca 2÷] in those cells where recordings were made at two different concentrations of the ion (Fig. 1).
163 The membrane resistance was not consistently different in high compared to normal [Ca 2÷] (mean of 34 + 3.8 MI2, n = 12, in high; 33 + 4.1 MI2, n = 10, in control
[Ca2+]0). The most striking change seen was, however, in the configuration of excitatory postsynaptic potentials (EPSPs). Especially conspicuous was a very prolonged late phase of the EPSP evoked by D R pulses of 0.02 or 0.04 ms duration seen in most cells exposed to elevated [Ca2+]. These EPSPs usually exceded 100 ms in duration and sometimes lasted for more than i s. In cells examined in both normal and high [Ca 2+] it was evident that EPSPs also rose faster and to a higher amplitude in high [Ca2+]. In the example of Fig. 1 the elevated threshold in high [Ca2+]0 prevented the cell from firing, but usually the augmentation of the EPSP more than offset the threshold change. In some cases the long EPSPs triggered trains of up to 8 spikes, where the same volley triggered but one action potential in control ACSE Stimulus pulses of 0.2 ms duration did, in some instances, evoke long EPSPs with spike trains even in control ACSE Even these already long responses were augmented in amplitude and duration in high [Ca2+]. In our previous study 6, we identified the potential waves recorded from a D R in response to stimulation of a neighboring D R as follows. In control ACSF, the responses consist of, in succession: a directly evoked fiber volley; a postsynaptic response interpreted as having been led from D H by volume conduction, termed the 'dorsal horn response' (DHR); and a late 'slow' wave, the dorsal root potential (DRP). In elevated [Ca 2÷] an additional discharge was grafted on the rising phase of the DRP, which we identified as the 'classical' dorsal root reflex (DRR). As the D H R has not previously been described by other authors, it was important to verify that our interpretation of its generation was correct. Fig. 2 shows the D R potential together with the simultaneous intracellular response of a neuron in 2.4 [Ca2+]. In this cell a weaker pulse evoked a DRP in the D R and an EPSP with a delayed action potential in the neuron (not illustrated). With stronger pulses the D H R and D R R appeared in the D R recording, and the cell now fired two spikes of which the earlier one coincided in time with the D H R (Fig. 2). Many but not all neurons responded to D R stimuli with a spike of latency similar to that of the D H R (less than 2.5 ms from stimulus artefact to D H R onset).
Spontaneous activity Initially, while the preparation remained in control ACSF, the D R and VR recordings showed no spontaneous activity. When [Ca 2+] was raised to 2.4 mM, spontaneous discharges erupted after a variable delay, as
described earlier 7. After resuming perfusion with control ACSF following exposure to high [Ca2+], spontaneous activity subsided very slowly so that it usually took more than 30 min, sometimes more than 1 h, for all spontaneous activity to disappear. Of 17 neurons observed in control ACSF, the resting potential was quite constant without any spontaneous activity in 10 (these numbers include 4 cells not included in membrane potential statistics because Vm < 55 mV). The other 7 showed spontaneous depolarizations resembling EPSPs which sometimes triggered action potentials. Of the 24 cells observed in elevated [Ca2+] only two were completely quiescent, the others showed spontaneous EPSPs with variable frequency with or without spikes or spike bursts (Figs. 3, 4 and 6). Of the 9 cells seen in both normal and elevated [Ca2+]o, 6 were silent in the control state, but they were spontaneously active in high [Ca2+]o. In all 3 cases when a normally spontaneously active cell was observed in both low and high [Ca2+], both the frequency and mean amplitude of the spontaneous EPSPs were greater when [Ca 2÷] was high (mean frequency in normal [Ca2+]0: 10 events/min, range 2-18 / min; in elevated [Ca2+]o: mean = 27/min, range 6-60 / min). The high [Ca2+]-induced cellular spontaneous EPSPs were, to varying degrees, partially synchronized with D R discharges. In six neurons every spontaneous EPSP
DR
IC
10
msec
Fig. 2. Responses recorded simultaneously from the DR and from a DH neuron (tracing marked 'IC') in 2.4 mM [CaZ+]. FV, directly evoked fiber volley; DHR, dorsal horn response, recorded by volume conduction from gray matter into dorsal root; DRR, the GABA-dependent dorsal root reflex; the unlabelled slow wave is the 'classical' dorsal root potential (DRP) (for further description of DR recordings see Ref. 6). At this stimulus intensity (600/~A, 0.02 ms) the cell consistently fired two spikes, the first of which coincided with the DHR.
164
:~"
DR
IC
t
I:
~
mV
B
5 sec
C
+
20 mV J
k
IC~
200 msec
coincided with a spontaneous discharge of the D R (Figs. 3, 4 and 6A), but spontaneous D R waves did occur, at times, without intracellular EPSP, indicating that the cell under observation did not participate in all the population discharges. In other cells the spontaneous membrane potential activity was temporally more or less independent from D R or VR events. In a sample of 18 cells (those in which spontaneous waves were discrete enough to be counted) the mean frequency of spontaneous EPSP-Iike waves was 10.5/min (range 0-30; counts during either 5 or 10 min epochs, depending on the cell's stability). The mean frequency of spontaneous D R discharges counted during the same time intervals was 23.3/min (range 18-39). The EPSPs that were synchronized with D R discharges sometimes did and sometimes did not trigger action potentials (Figs. 3 and 4). The onset of an intracellular EPSP sometimes slightly preceded and sometimes succeeded the onset of the D R wave; the precedence could reverse even as successive events were recorded from a single cell. Remarkably, 4 cells fired action potentials that did not appear to be triggered by an EPSP (Figs. 3 and 5). In some cases these spikes arose from a 'flat' membrane
t..k_k
Fig. 3. Spontaneous activity recorded simultaneously in DR and in a DH neuron (tracing marked 'IC'), in 2.4 mM [Ca2+]. All tracings from the same cell. A: pen recordings on chart paper; note that EPSP-based bursts coincide with DR wave complexes, while single spikes (single upstrokes of pen) do not. B: example of EPSP-based burst with DR discharge recorded by digital oscilloscope. C: examples of single spikes not triggered by EPSPs and 'miniature spikelets', without corresponding DR activity.
A
B +
DR
A
÷
"
'
2° "v I
t
2o ,,v I
lO0 msec
L
'[
,
IC
!
icl -60 -70[
B •
2 msec
i '
~
C
' *
t
t
0
5 s~c C
20 mV I
I
Fig. 4. Partially synchronized spontaneous activity in DR and a DH neuron. A: recorded by pen-writing instrument; 3 arrows mark responses evoked by DR stimulation, all other activity is spontaneous. B and C show examples of individual discharges recorded by digital oscilloscope; note EPSP-like waves on intracellular trace without action potentials. [Ca2.1 was 2.4 mM.
C
2 ,sQc
Fig. 5. Examples of spontaneously fired action potentials, not triggered by EPSPs. A and B from one cell, C and D from another. In A the DR trace shows spontaneous D R wave in time independent from action potentials in IC trace. C shows brief, miniature 'spikelet', D shows an action potential with prepotential that is not an EPSP. [Ca 2+] was 2.4 mM in all cases.
165
+ 0R
-60.
IC
,
,
,
mY
30 sec ~[+ DR >
I
IC ' .
//
/
.
~//
/
.. . . . . .
mY 0.2 nA
Fig. 6. Simultaneous recordings from DR and a DH neuron in 2.4 mM [Ca2+]. A: without stimulation. B: during injection of 0.2 nA DC depolarizing current into the neuron through the intracellular electrode; note increased activity in neuron, and increased amplitude and frequency of partially synchronized discharges in the unstimulated DR. potential (Fig. 5A,B), in others there was a brief prepotential (Fig. 5D). In the latter type there also occurred small 'spikelets' that were too brief to be EPSPs (Figs. 3C and 5C). Such 'anomalous' non-EPSP-triggered spikes always appeared to be fired at random, independently from DR waves (Figs. 3, 5A). Injection of current by way of the intracellular electrode into DH neurons generally left the rhythm and amplitude of the spontaneous discharges of the DR unaffected. In one case, however, prolonged depolarizing current changed markedly the pattern of the discharge not only of the cell, but also of the DR (Fig. 6). DISCUSSION It is well known that raising interstitial calcium concentration, [Ca2+]o, augments synaptic potentials while also raising the threshold of postsynaptic neurons. Previously, using extracellular recordings, we found that synaptically transmitted responses are enhanced when [Ca2+]o is increased, indicating that the augmented EPSPs more than offset the effect of decreased postsynaptic excitability3,6. In the present experiments, when the input volley was weak, the firing probability of some cells became smaller in high [Ca 2÷] than it has been in control ACSF (Fig. 1), demonstrating the elevation of threshold. At stronger inputs, however, the EPSP increased to the point where it overcame the raised threshold, resulting in a larger population output 6. In the intracellular recordings of this study the enhancement of the late phase of EPSPs was especially striking, particularly those evoked
by short (0.02-0.04 ms) volleys which activate only A-fibers. Such prolonged EPSP components have been attributed by other authors to activation of NMDA receptors 1°'12. In our previous study we found that the high [Ca2+]-induced spontaneous activity was depressed, but not abolished, by NMDA antagonist drugs, while it was reliably suppressed by G A B A A blocking agents 7, which also block dorsal root reflexes (DRR), 'slow' DRPs and the delayed ventral root reflex (VRR2) 6. In most respects the DR of the spinal cords maintained in this series of experiments at 34.5 °C in an interface chamber behaved similarly to those described previously, when spinal cords were submerged in cool (21-24 °C) ACSF 6'7. One difference, however, between these and our previous experiments, was the relatively small amplitude of orthodromic and spontaneous VR responses. This could be due to a depression of VH neurons at the higher temperature, or to less favorable arrangement of electrodes. Since DR responses recorded now were comparable to those seen in the submerged, cool, spinal cords previously, and since our attention was focussed on cells in the DH, we largely ignored VR behavior in this study. Another minor difference was the very slow subsidence of spontaneous activity when ACSF with elevated [Ca 2+] was replaced by control ACSF. This may be the result of the slow removal of the excess Ca 2+ since, unlike submerged preparations, the spinal cord in the interface chamber is washed by flowing ACSF at only one of its surfaces. Neurons observed in elevated [Ca 2+] generally displayed a higher level of spontaneous activity than those in control ACSF, but only a minority of these were discharging in synchrony with the DR. We have concluded earlier 7 that, in elevated [Ca2÷], definite assemblies of neurons linked by a network of connections preferentially discharge together. Results now reported confirm limited recruitment of neuronal elements into synchronized discharges, but they also indicate that high [Ca 2÷] induces spontaneous, if not always synchronous, activity in almost all DH neurons. Some of the 'synchronized' cells are likely to supply axo-axonic junctions capable of releasing G A B A on primary afferent terminals, which then is detected by electrodes on the DR as spontanoeus DR potential waves. Other neurons may receive excitatory input directly or indirectly from GABA-depolarized axon terminals. In one cell the injection of depolarizing current caused the firing of the DR to change markedly, and in this case we may have chanced upon a 'driver cell' of the spontaneous DR discharges. If so, then the profusely branching axon of this unit must have supplied G A B A to numerous DR fiber endings. The other 'synchronized' cells seem to be recipients rather than suppliers of excitatory drive.
166 Spikes that were not triggered by an EPSP may have originated at axon terminals whence they were conducted antidromically into the cell body. Alternatively, such spikes may have been triggered in a dendrite so far from the recording electrode that the EPSP was not detectable. Subliminal spikelets are compatible with either explanation, as they may represent impulses approaching but not invading the cell soma. Since intracellular recordings were referred to a distant ground and not to the extracellular space near the neurons, the theoretical possibility cannot be excluded that a depolarizing prepotential was concealed by an extracellular potential wave of opposite polarity. This seems unlikely with view of the solitary nature of the prepotential-less spikes. Nonsynchronous action potentials arising straight from a 'flat' resting potential have recently been recorded in hippocampal tissue slices 'kindled' in vitro by Stasheff and Wilson 2°, and antidromic spikes are also generated during penicillin-induced electrographic seizures 9,15. While in these other examples the electrophysiologic details differed and the mechanism of generation may REFERENCES 1 Ajajouanine, "12,Contamin, E and Cathala, H.P., La Syndrome Tetanie, Baill~re, Paris, 1958. 2 Bagust, J. and Kerkut, G.A., An in vitro preparation of the spinal cord of the mouse. In G.A. Kerkut and H.V. Wheal (Eds.), Electrophysiology of Isolated Mammalian CNS Preparations, Academic Press, London 1981, pp. 337-365. 3 Balestrino, M., Aitken, P.G. and Somjen, G.G., The effects of moderate changes of extracellular K+ and Ca2÷ in synaptic and neural function in the CA1 region of the hippocampal slice, Brain Research, 377 (1986) 229-239. 4 Bernard, E., UrNin, L. and Somjen, G.G., Intracellular recording of the spontaneous activity in dorsal horn interneurons induced by elevated Ca2+ in mouse spinal cords in vitro, Soc. Neurosci. Abstr., 15 (1989) 1032. 5 CzEh, G., Obih, J.-A. and Somjen, G.G., The effect of changing extracellular potassium concentration on synaptic transmission in isolated spinal cords, Brain Research, 446 (1988) 50-60. 6 Cz6h, G. and Somjen, G.G., Changes in extracellular calcium and magnesium and synaptic transmission in isolated mouse spinal cord, Brain Research, 486 (1989) 274-285. 7 Cz6h, G. and Somjen, G.G., Spontaneous activity induced in isolated mouse spinal cord by high extracellular calcium and by low extracellular magnesium, Brain Research, 495 (1989) 89-99. 8 Eccles, J.C., The Physiology of Synapses, Springer, New York, 1964. 9 Gutnick M.J. and Prince, D.A., Thalamocortical relay neurons: antidromic invasion of spikes from a cortical epileptogenic focus, Science, 176 (1972) 424-426. 10 Jessel, T.M., Yoshioka, K. and Jahr, C.E., Amino acid mediated
also vary, they have in c o m m o n a hyperactive state of the gray matter. 'Spikes from baseline' have also been recorded by Price et al. ~7, in dorsal horn of normal spinal cords of cats, but their sample differed from ours in that their cells that fired spikes without prepotentials never generated EPSPs or IPSPs. For this reason the authors ~7 considered it likely that their recordings were from axons rather than cell somata, which was not the case with our recordings. Earlier we already discussed the possible significance of spontaneous activity for the normal and pathologic function of the spinal cord 7 and we cannot at this time add new insight to that question. We would, however, repeat one point, that 2.4 m M [Ca 2+] in a bathing solution is too high to be used in experiments designed to investigate the normal functioning of CNS tissue.
Acknowledgements. We thank Mrs. Marjorie Andrews for secretarial help. The work was supported by NIH Grants NS 17771 and NS 06233. transmission at primary afferent synapses in rat spinal cord, J. Exp. Biol., 124 (1986) 239-258. 11 Jesserer, H., Tetanie, Thieme, Stuttgart, 1958. 12 King, A.E., Thompson, S.W.N., Urban, L. and Woolf, C.J., An intracellular analysis of amino acid induced excitations of deep dorsal horn neurones in the rat spinal cord slice, Neurosci. Lett., 89 (1988) 286-292. 13 King, A.E., Thompson, S.W.N., Urban, L. and Woolf, C.J., The responses recorded in vitro of deep dorsal horn neurons to direct and orthodromic stimulation in the young rat spinal cord, Neuroscience, 27 (1988) 231-242. 14 Levy, R.A., The role of GABA in primary afferent depolarization, Progr. Neurobiol., 9 (1977) 211-267. 15 Lothman, E.W. and Somjen, G.G., Functions of primary afferents and responses of extracellular K÷ during spinal epileptiform seizures, Electroencephalogr. Clin. Neurophysiol., 41 (1976) 253-267. 16 Otsuka, M. and Konishi, S., Electrophysiology of mammalian spinal cord in vitro, Nature (Lond.), 252 (1974) 733-734. 17 Price, D.D., Hull, C.D. and Buchwald, N.A., Intracellular responses of dorsal horn cells to cutaneous and sural nerve A and C fiber stimuli, Exp. Neurol., 33 (1971) 291-309. 18 Somjen, G.G., Allen, B.W., Balestrino, M. and Aitken, P.G., Pathophysiology of pH and Ca2+ in bloodstream and brain, Can. J. Physiol. Pharmacol., 65 (1987) 1078-1085. 19 Somjen, G.G. and Cz6h, G., Pathophysiology of the spinal cord studied in vitro, J. Neurosci. Methods, 28 (1989) 35-46. 20 Stasheff, S.F. and Wilson, W.A., Increased ectopic action potential generation accompanies epileptogenesis in vitro, Neurosci. Len., in press.