The electrical excitability of chronically isolated cortex studied by means of permanently implanted electrodes

244

ELECTROENCEPHALOGRAPHY AND CLINICAL NELIROPHYSIOLOGY

THE

ELECTRICAL ISOLATED

EXCITABILITY CORTEX

PERMANENTLY

STUDIED

IMPLANTED

OF

CHRONICALLY

BY MEANS ELECTRODES

OF 1

SETH K . SHARPLESS 2 a n d LAWRENCE M . HALPERN 3

Deparmlent of Pharmacology, Albert Einstein College of Medicine, New York, N.Y. (U.S.A.) (Received for publication: September 29, 1961)

A slowly developing increase in excitability seems to be an almost universal phenomenon in excitable tissue deprived of nervous influences (Cannon and Rosenblueth 1949). Neurally isolated slabs of cerebral cortex appear to conform to this principle by gradually developing an increased susceptibility to experimentally induced epileptiform activity (Echlin and McDonald 1954; Grafstein and Sastry 1957). This increase in the excitability of chronically isolated cortex has been attributed either to denervation sensitization similar to that which occurs in effector organs (Echlin 1959), or to the proliferation of axon-collaterals (Purpura and Housepian 1961). In the present study, the change in electrophysiological properties of chronically isolated cortex in cats has been studied by means of permanently implanted stimulating and recording electrodes. The use of indwelling electrodes has the advantage that the same area of cortex may be studied over long periods of time without the variability introduced by accidents of electrode placement and surgical manipulation. This technique is especially valuable when the effects of long-lasting drugs on neurally isolated cortex are being studied. The present communication is intended to describe some data ob1 Supported by U. S. Public Health Service Grants B-2583 and 2M 6418. " U.S. Public Health Service Senior Fellow, National Institute of Neurological Diseases and Blindness. a Present address: Lederle Laboratories, Pearl River, N.Y. 4 This study is based in part on an analysis of control data obtained during a study of the effects of anticonvulsant drugs on chronically isolated cortex. The latter will be published separately. See L. HALPERN, Ph.D. dissertation, Albert Einstein College of Medicine, 1961.

tained this way, which show the time course of the emerging increased susceptibility to epileptiform discharge in isolated cortex during the first few weeks following the isolation procedure. 4 METHODS

Using aseptic technique, a wide craniectomy was performed on cats anesthetized with pentobarbital, and the dura was reflected to expose the dorsolateral aspects of both cerebral hemispheres. A narrow, blunt spatula was passed through a stab wound in the posterior portion of the suprasylvian gyms and manipulated until the gyrus was undercut and circumsected. The edge of the spatula was brought up along the margin of the isolated area until it was clearly visible against the pia mater, care being taken to avoid damaging the pial circulation. The cortical slab thus isolated comprised most of the suprasylvian gyms to a depth of about 5 mm, and occasionally the adjacent part of the marginal and ectosylvian gyri. An effort was made to avoid excessive manipulation of the cortex during the operation, even at the risk of incomplete isolation in some instances. To allow for the swelling of brain tissue due to edema without occluding the superficial vessels under the electrode plate, the dura was not replaced. In the same operation in which the isolation procedure was carried out, plate electrodes were implanted over the isolated suprasylvian gyms and over the homologous intact area on the opposite hemisphere. The plates were made of a polyester resin in which were imbedded four platinum-platinum chloride discs, approximately 1 mm in diameter, arranged as indicated in Fig. 1. Two discs, 3--4 mm apart functioned as stimulating electrodes; the other two discs, Electroenceph. clin. Neurophysiol., 1962, 14:244-255

245

CHRONICALLY ISOLATED CORTEX

4 - 5 m m anterior and posterior to the stimulating electrodes, were used for recording. A wire fixed in the skull over the frontal sinus or in the occipital protuberance was used as a ground lead, or as an indifferent electrode during monopolar recording. The electrodes

lating electrodes, and was varied systematically during the course of an experiment. A stepwise attenuator which provided approximately equilog increments of current was incorporated in the stimulating circuit. In what follows, stimulus strength is represented in two

REFERENCE LEAD

RECORDING ELECTRODES IS OL ATED CORTICAL SLAB

•

STIMULAT~ ELECTRODES

•

_~mm

P8 mr.

Fig. 1 Locus of isolated cortical area and position of implanted electrodes. Electrodes on right suprasylvian gyrus were used as controls. This area also showed a slight increase in susceptibility to epileptiform activity several weeks after isolation of opposite side. were brought out through a stab wound in the skin and terminated in a small plastic socket screwed or bolted to the skull. Since it was planned to stimulate the same site many hundreds of times during the course of an experiment, it was considered desirable to use paired, opposite-going, approximately square pulses. It has been reported that this type of stimulation is less likely to produce tissue damage than monophasic pulses (Lilly 1960; Rowland et al. 1960). Pulse duration was 0.5 msec, and opposite-going pulses were separated by intervals of 0.5 msec. During stimulation, trains of pulse pairs were presented at the frequency of 30/sec for periods of 2.5 sec. All parameters of stimulation were held constant except peak current per pulse. The latter was determined by measuring the voltage drop across a known resistor in series with the stimu-

ways, in terms of peak current per pulse (for threshold measurements), and in terms of the number of equilog increments of current over threshold, each increment being approximately 1.6 db. Comparison of stimulus strength in different animals in terms of peak current per pulse would not be justified in every case, since there was evidence that different proportions of the total current flowed through the cortical tissue in different experiments. Expressed as the logarithm of the proportion of threshold current, stimulus strength is independent of variations in shunt resistance among different cats and different experimental sessions. Testing was begun in some cases within 24 h after the operation. At the beginning of each test session, thresholds for epileptiform afterdischarge in isolated and intact cortex were

Electroenceph.clin.Neurophysiol.,1962, 14:244-255

246

S. K. S H A R P L E S S AND L. M. HALPERN

determined. Isolated and intact cortex were stimulated alternately in such a way' that a rest period of 10 rain elapsed between successive stimuli to the same site. Stimuli were presented at successively higher values, until an after-discharge was obtained. The strength (in peak current per pulse) of the first stimulus to produce an after-discharge lasting more than 1 sec was taken as the threshold value. Stimulation was then continued, with successive increments in current, as before, until the stimulus exceeded a value 2-2.5 times threshold (a value usually insufficient to produce generalized seizures on the intact side). This procedure was repeated until each cortical area had been stimulated 5-10 times at each of three or four stimulus intensities. The present report is based on 31 test sessions with twelve cats. With five of these animals, the change in susceptibility to after-discharge following repetitive electrical stimulation of the isolated cortical area was studied systematically at weekly intervals for the first postoperative month with adequate controls to permit statistical evaluation of the results. The remaining cats were studied at various times up to 11 months after the operation. RESULTS

Histological examination at the end of the experiment revealed that the isolation was often incomplete, a bridge of tissue remaining intact in the anterior corner of the isolated gyrus. Even in those cases where isolation appeared to be complete, it was possible that some fibers in the molecular layer escaped destruction. The histological changes resembled those described by Echlin (1959). The isolated area appeared somewhat pale in Nissl preparations, in contrast to adjacent cortex. Cell counts were not carried out, but apart from the largest pyramidal cells, the cellular layers were densely populated with neurons, and there was evidence of laminar organization. In the weeks following the operation, a thick fibrous membrane formed between the exposed cortical surface and the overlying muscle, filling in the skull defect, and eventually encapsulating the polyester electrode plate. The occurrence of this fibrous reaction appeared to

depend primarily on the absence of dura, although there may have been some specific reaction to the electrode plate as well. In most cases, the fibrous membrane replacing the dura could easily be dissected away from the surface of the cortex at autopsy. However, there were occasional adhesions and invasions of the molecular layer. The presence of the electrode assembly did not appear to affect the neuronal layers of the underlying cortex. The development of fibrous tissue under the plate resulted in the elevation of the electrodes above the cortical surface, however, and in the formation of a pathway through which some of the current from the stimulating electrodes could bypass the cortical tissue. This was reflected in a gradual decrease in the effective strength of the stimulating current, and in a decrease in the amplitude of the recorded spontaneous activity both in the isolated cortical slab and in intact cortex. These changes were relatively small compared with the marked changes in excitability produced by the isolation procedure, as will be described below. During the first few days following the operation, the isolated cortex usually exhibited little spontaneous activity, although in some animals, periodic bursts of activity were observed similar to the "after-bursts" described by Burns (1954). By the 2rid week, a distinctly abnormal form of spontaneous activity had developed in the isolated slab, which differed considerably from one cat to another, but was usually characterized by occasional spikes, slow waves, and short bursts of high frequency activity (cf. Henry and Scoville 1952). In subsequent weeks, the frequency of the spontaneous activity tended to increase, and the spikes diminished in frequency, until, in some cases, the activity resembled that recorded from normal cortex in the waking animal.

Threshold for epileptiform after-discharge With the stimulus parameters used (paired, opposite-going pulses, 0.5 msec pulse duration, 30 pulse-pairs/sec for 2.5 sec), the mean threshold for epileptiform after-discharge was 4.5 mA with a standard deviation of 2.5 mA. These values were calculated from all determinaElectroenceph. clin. Neurophysiol., 1962, 14:244-255

CHRONICALLY

ISOLATED

tions, including those made on isolated cortex. The average quantity of electricity per pulse at threshold was therefore more than 2 I~C, which is much greater than that reported by Lilly (1960) to be a typical threshold value for motor effects. On the other hand, it was much less than that required to produce histologically apparent tissue damage - - , 20 , C / p u l s e according to Rowland et al. (1960). The relatively high threshold values observed in these experiments were attributed to the fibrous membrane under the electrodes which provided a pathway for current return bypassing the excitable elements of the cortex. The variation in threshold from one experiment to another may be attributed in some cases to unavoidable changes in the stimulating circuit. Measurements of electrode resistance made during each recording session by monitoring voltage during the passage of pulses of known current value indicated that changes in the resistance of the electrode system were associated with threshold changes. High resistance was generally associated with low threshold values (Spearman rank-order coefficient, ,, z - 0.58; t ' < 0.01). Since constant (peak)

INTACT

CORTEX •

247

CORTEX

current stimulation was used, the observed relation between resistance and threshold may be accounted for by the development of low resistance shunts in the stimulating circuit (other than the fibrous membrane under the electrodes, since the resistivity of the latter would probably not be lower than that of cortex), e.g., moisture in the electrode socket, or a layer of fluid interposed between the electrode plate and cortical surface. When only those cases in which the resistance was in the same range on the two sides were considered, the threshold of the isolated slab tended to be lower than that of the intact cortex after the 1st week. In Fig. 2, the mean thresholds are plotted as a function of time after operation, after excluding threshold determinations made when the resistance of either side exceeded that of the other by more than 50 per cent. The figure shows a fall in threshold from the 1st to the 2nd week:, and then a gradual rise in the threshold of both sides as the electrodes became encapsulated by connective tissue. Unfortunately, the data do not permit a statistical evalution of the changes in threshold from one week to the next, but if threshold determinations made after the 1st week are averaged (again excluding cases where the resistance of the two sides differed by more than 50 per cent), in nine out of ten cats, the average threshold on the isolated side was lower than that of the intact side. It may be concluded that the threshold for epileptiform after-discharge falls in isolated cortex, and that the change is already evident in the second postoperative week.

Duration of after-discharges I

I

I

I

z

3

F 4

I 5

WEEKS POST-OPERATiVE

Fig. 2 Mean threshold for epileptiform after-discharge during successive postoperative weeks. Week 1 includes tests carried out on the first to the sixth postoperative days. Cases in which resistance of stimulating electrodes on the two sides differed from one another by more than 50 per cent were excluded. The rise in threshold in both the isolated and intact cortex after the 3rd week may be attributed to the growth of fibrous tissue around the electrodes. Different animals are represented in averages for different weeks (see text).

The duration of the after-discharges recorded from either isolated or intact cortex increased with increases in the strength of the stimulating current. In freshly isolated cortex, the maximum duration of after-discharge rarely exceeded 15 or 30 sec. In intact cortex, if the stimulus were increased beyond a certain value (usually more than two or three times the threshold value), the paroxysmal activity spread to involve the ipsilateral motor cortex and usually became a generalized seizure, which might last for several minutes. If stimulus intenElectroenceph. clin. Neurophysiol.,

1962,

14:244-255

248

S. K. SHARPLESS AND L. M. HALPERN

sities were limited to values that would not produce generalized seizures, the duration of after-discharge was approximately the same in freshly isolated and intact cortex. An attempt was made to limit stimulus intensities to moderate values, but occasionally, an after-discharge elicited in intact cortex spread to become a generalized seizure. On such occasions, the experiment was terminated or interrupted until the animal had recovered from its post-ictal depression. In Fig. 3, the average duration of afterdischarge is plotted as a function of time after operation at different intensities of stimulation. Each point represents 25 after-discharges, five

(x: 80 < "i" _uv)7o

S" K log "~T

-

,

i't

C~

fl:

60 '

<

50

ISOLATED

4() z

o ~ 30 -

g 2o z

I

2

3

POSTOPERATIVE WEEK

Fig. 3 Mean duration of after-discharge (in seconds) as a function of time after operation. S is stimulus strength in terms of number of standard equilog increments of current over threshold. The duration of afterdischarges in isolated cortex increases five-fold or more during the first 3 weeks. A slight increase is also seen in intact cortex, which has been deprived of callosal projections.

each from five cats. The mean duration of after-discharge elicited in isolated cortex by stimuli of moderate intensity increased markedly during the first 3 weeks. The increase was statistically significant in the case of each animal (I' < 0.01 for four cats, and < 0.05 for the fifth; nonparametric tests with stimuli matched for strength and order of occurrence; Wilcoxon 1949). The data are not sufficient to determine whether an asymptote is attained. The greatest increase occurs from the first to the second postoperative week, but there is evidence that the duration of after-discharge in isolated cortex continues to increase after the 3rd week. In several preparations tested from 3-11 months after the isolation procedure, a single stimulus to the isolated slab initiated an "after-discharge" lasting several hours. In such cases, this afterdischarge often degenerated into a series of epileptiform spikes which continued to occur irregularly at intervals of a second or so. When the spiking finally disappeared, a second stimulus was capable of initiating the same sequence of events. As time passed after the operation, there was a change in the character of the after-discharge recorded from the isolated cortical slab. Fig. 4, 5 and 6 show representative after-discharges obtained from a single animal with identical stimulating and recording parameters at various times during the first three postoperative weeks. There is a striking increase in the "tonic" phase of the after-discharge. In normal or freshly isolated cortex, this phase, characterized by synchronous rhythmic discharges at more than 20 c/sec, occupies a small part of the afterdischarge. In the prolonged after-discharges

ItNl?ml ISOLATED

SUPRASY/VlAN G.

~)1~

C' I,-, rr'tl,iIIiI CAT # 88 STIMULUS= 2mA 1 DAY POST -OR

.... )

]looHv

sec

Fig. 4 See legend for Fig. 6.

Electroenceph. clin. Neurophysiol., 1962. 14:244 255

CHRONICALLY

ISOLATED

observed in chronically isolated cortex, on the other hand, the "tonic", high frequency activity tends to dominate the response. It may be interrupted by spike-slow wave activity but often recurs in bursts persisting for many seconds. The duration of after-discharges recorded

, bQ)LAIP_ID .

.

.

.

.

.

.

.

.

~UI-'NAbYL .

.

.

.

.

.

IN

.

//!lf//tt!I/lt!i'/

~4~,~ ~1'!,,",:i~1[~

249

CORTEX

The increase in the susceptibility of the intact gyms to epileptiform after-discharge is especially interesting. Since the homologous area on the opposite hemisphere had been undercut, callosal fibers would have degenerated, depriving even the intact suprasylvian gyms of ,, ,-,, ....... ~ . . , . ~ . . , , m ~

. . . . . . . . . . . . .

CAT ~ 88 STIMULUS= ~) DAYS

2mA

POST - OR

,

]~oo~v

Isec

Fig. 5 See legend for Fig.

from the intact gyrus also increased slightly as a function of time following operation (Fig. 3). The increase in the duration of after-discharges recorded from the intact side during the first 3 postoperative weeks was small compared w;th the increase observed in the isolated gyms. Nevertheless, it was statistically significant for three of the cats ( P < 0.05) and just short of significance for a fourth (Wilcoxon 1949). There was also evidence that as time passed after the operation, after-discharges induced in the intact suprasylvian gyms spread more readily to become generalized seizures. Thus, in preparations tested from 3-11 months after the operaration, it was difficult to obtain a localized after-discharge in the intact gyms. An increase of stimulus intensity to slightly above threshold was sufficient to initiate a generalized seizure with motor concomitants.

6.

part of its innervation. Thus, some of the cells in the intact gyms were partially denervated and may have undergone denervation sensitization. If so, the degree of sensitization was much less than that which appeared on the side which had been completely undercut, although it was present and measurable as early as the second postoperative week. Since the intact area was subjected to repeated stimulation during the experimental procedure, the possibility must also be considered that its functional properties were altered as a result of frequent exposure to paroxysmal activity. Morrell (1960) has described an increase in susceptibility to epileptiform activity in a cortical area receiving callosal projections from an epileptogenic focus induced by the topical application of ethyl chloride to the opposite hemisphere. Electroenceph. clin. Neurophysiol.,

1962,

14:244-255

250

S . K. S H A R P L E S S

AND L. M. H A L P E R N

The e~/ects o~ repeated stimulation According to Rosenblueth and C a n n o n (1942) and other investigators, an electrical after-discharge produced by repetitive stimulation can be reproduced after a rest pause of 15 sec - 5 min. This was so in pilot experiments

ges could be repeatedly elicited in both intact and isolated cortex for periods of 4 5 h without any sign of depression. However, most preparations exhibited at least some depression during the course of an experimental session. In eighteen out of the 25 test sessions

IS©I_ATFD SUPRASYLVIAN

'

tlttl'

I

v'~)l

, I

CAT "~ 8 8 STIMULUS = 2 m A 17 DAYS

POST-OR

.... -.......

I Sec

Fig.

-.

6

Fig. 4, 5 and 6 show representative after-discharges obtained from isolated cortex by means of implanted electrodes on 1st, 9th and 17th day, respectively, after the operation. Identical stimuli (2 mA peak current) and amplifier gains. Top trace: bipolar. Bottom trace: monopolar, to bone overlying frontal sinus. Notice appearance of spontaneous activity (preceding stimulus artifact) on successive occasions, and the relative increase of the "tonic" phase of the afterdischarge. The decrease in the amplitude of the recorded signal from the 9th to the 17th day may be attributed to the growth of fibrous tissue around the electrode. on acute preparations. But in quantitative studies on chronic preparations, both the isolated and intact gyri were repeatedly stimulated at 10 min intervals for a period of several hours. U n d e r these conditions, the cortex often became somewhat depressed as indicated by a rise in threshold and a shortening of the average duration of after-discharge following successive stimuli. There were marked individual differences in the degree to which this depression developed. In some preparations, after-dischar-

suitable for analysis of threshold changes, the threshold of one of the two cortical areas was observed to rise by at least one standard increment (current up 1.6 db) before more than fifteen successive after-discharges had been produced. The isolated slab was more susceptible to depression than the intact cortex. Fig. 7 shows the average duration of afterdischarge initiated by a test stimulus of moderate intensity (one standard increment over threshold) on successive presentations of this Electroenceph. olin. Nettrophysiol.,

1962,

14:

244-255.

251

CHRONICALLY ISOLATED CORTEX

stimulus during a single experimental session. Since stimuli of different intensities were presented in rotation, 40 rain or more intervened between successive presentations of the same stimulus. Thus, in Fig. 7, there was a period of 40 min or so between successive test stimuli, INTACT CORTEX (10 CATS)

z

o |I z'~

15~

(28 EXPERIMENTS)

o

I

t05

~

2

,.-oWEEK3

3

4

5

TEST STIMULUS

~75 \ z 60 o

cortex at the 1.2 per cent, 0.2 per cent and 0.2 per cent levels of confidence respectively for weeks 1, 2 and 3. Many hours were often required for isolated cortex to recover completely to its previous state of excitability after an experimental session during which it had been subjected to repeated after-discharges. In Fig. 8, the duration of after-discharge in isolated cortex is plotted as a function of successive trials on the 21st and 23rd days following the operation. In this case, the chronically isolated cortex still showed evidence of partial depression 36 h after an experiment during which it had been subjected repeatedly to paroxysmal discharge.

ISOLATED CORTEX

PERSISTENCE OF DEPRESSION

{Io CATS) (28 EXPERIMENTS)

CAT S - 4 6 (21 DAYS) ISOLATED CORTEX 5O

2

30 DECEMBER

U

<=30

40-

WEEK 3 ~5

-

WEEK 2 c.2

3

--o 4

oWEEK I 5

TEST STIMULUS

Fig. 7 Decrease in duration of after-discharge with successive test stimuli. During the test session, the animal was stimulated at 10 rain intervals, every fourth or fifth stimulus being a "test" stimulus (one standard increment, approx. 1.6 db, over threshold). The mean duration of the after-discharges elicited by successive test stimuli decreases significantly in isolated but not in intact cortex.

and each area was stimulated several times during this period. In plotting these curves. cases in which the threshold rose by two or more increments were excluded - - i.e., only the responses to suprathreshold stimuli were averaged. It is clear that even after eliminating cases in which there were large changes in excitability, the duration of after-discharges elicited by stimuli of constant intensity declined exponentially with successive trials. This effect occurred chiefly in isolated cortex, and was most pronounced 3 weeks after isolation. A statistical comparison of the first and fifth responses to stimuli of each intensity revealed a significant depression of responsiveness only in isolated

Z

~_3o <

c3 2 0 Z < ~ I0-

! I I

TRIALS-(BLOCKS OF 4)

Fig, 8 The depression of responsiveness produced in isolated cortex by repeated stimulation during a single experimental session persists for at least 36 h, as indicated by a shorter duration of after-discharges elicited after a day's rest. Each point represents the average of four successive after-discharges elicited by stimuli ranging in strength from 0 4.8 db over threshold. DISCUSSION

The prolongation of paroxysmal after-discharge in chronically isolated cortex has been attributed to denervation supersensitivity by Echlin (1959). The destruction of afferent projections to the cortex results in the partial denervation of many neurons and in a general decrease in the activity of the isolated area. In similar circumstances, excitable peripheral structures (effector organs, ganglion cells, etc.) which have been partly or wholly deprived of Electroenceph. clin. Neurophysiol., 1962, 14:244-255

252

S. K. S H A R P L E S S AND L. M. H A L P E R N

nervous influence show a gradual increase in sensitivity (Cannon and Rosenblueth 1949; Thesleff 1960; Emmelin 1961). This increase in sensitivity following denervation may be disclosed by an exaggerated response to neurohumors, or in the case of partial denervation, by an increase in the response of the partly denervated organ to stimulation of those nerve fibers which remain intact. Within a few weeks after chronic denervation, the smooth muscle investing the nictitating membrane of the cat's eye, for example, undergoes a tenfold increase in its responsiveness to epinephrine, and a comparable increase in responsiveness to other chemical agents, including acetylcholine (Rosenblueth 1932). The same kind of change has been observed in skeletal muscle, cardiac muscle, glands, ganglion cells, and in almost every excitable peripheral structure which has been studied after denervation (Cannon and Rosenblueth 1949). In view of the ubiquity of the phenomenon in peripheral structures, it would be surprising indeed if sensitization did not occur in central nervous elements that had been partly or wholly deprived of nervous input. With regard to time course, the smooth muscle of the nictitating membrane, for example, shows an enhanced sensitivity to epinephrine within 2 days after denervation; this sensitivity increases to a near-maximal level in about 2 weeks (Hampel 1935), and then may continue to increase more gradually for some weeks thereafter (Rothballer and Sharpless 1961). If the muscle is not actually denervated, but rather deprived of input by destruction of preganglionic fibers, sensitization follows about the same time course but it does not attain the same level (Hampel 1935). A similar time course is observed in other autonomic effector organs and skeletal muscle after destruction of motor nerves (Cannon and Rosenblueth 1949). In a recent study by Franken (1960), attention was given to the time course of sensitization in cortical areas which had been partially denervated by brain lesions in various sites, including the homologous cortical area on the opposite hemisphere. It was concluded that a much longer time (2-4 months) was required for the enhanced sensitivity of cortex to become evident than was required for the development

of denervation supersentivity in peripheral structures. In the present experiments, in which the use of permanently implanted electrodes permitted repeated measures of sensitivity at intervals after the operation and the accumulation of sufficient data to attain statistical reliability, marked changes were observed in isolated cortex and significant changes in the homologous area of the opposite hemisphere during the first two postoperative weeks, in conformity with the time course of denervation sensitization in peripheral structures. The accumulated evidence that central nervous structures too undergo denervation sensitization is now certainly very persuasive (see review by Stavraky 1961). Attempts to study the effects of partial denervation of central elements have been complicated, however, by the occurrence of regenerative changes which themselves may affect sensitivity. Thus, some investigators have attributed the increase in sensitivity of central neurons following partial denervation to sprouting of adjacent fibers. It has been suggested that the enhanced sensitivity of spinal motoneurons to supraspinal influences observed by Teasdall and Stravr~tky (1953) following deafferentation of the hind limb may be due to sprouting of corticospinal fibers (Liu and Chambers 1958). Similarly, Purpura and Housepian (1961) have argued that the hyperexcitability of chronically isolated cortex may be due to the proliferation of recurrent axon collaterals, at least in immature animals. Our demonstration that chronically isolated cortex becomes sensitized during the first 2 weeks following isolation does not exclude the possibility that collateral proliferation is involved, since the greatest amount of sprouting and proliferation may also occur in the first few weeks following isolation. Murray and Thompson (1957) have observed collateral sprouting in the partially denervated superior cervical ganglion. Sprouts arise from the remaining intact preganglionic fibers during the 1st week and grow into close apposition with the denervated ganglion cells during the next 6 8 weeks. The ganglion cells exhibit denervation supersensitivity to acetylcholine only during an intermediate stage in this Electroenceph. clin. Neurophysiol., 1962, 14:244-255

CHRONICALLY ISOLATED CORTEX

process (2-4 weeks). As collateral growth continues, and the denervated ganglion cells are reinnervated, there is an actual regression of hypersensitivity to acetylcholine. Echlin (1959) has demonstrated that partially isolated cortex, unlike the partially denervated ganglion, shows a permanent hypersensitivity to acetylcholine. However, this point is also not critical, since it may be that the increased response of chronically isolated cortex to acetylcholine is an indirect effect, dependent on an increased number of connections between adjacent neurons rather than an increased sensitivity to the neurohumor. Thus, the concentration of acetylcholine required to excite any single neuron may not be affected by isolation, but each neuron, once excited, may be more efficient in setting off responses in its neighbors, because of the proliferation of connections between cells. There is another analogy between central and peripheral structures which may throw some light on the mechanisms involved in sensitization. A state indistinguishable from denervation supersensitivity can be produced in peripheral structures by the prolonged use of pharmacologic blocking agents. Sensitization after "pharmacologic denervation" has been demonstrated in peripheral effector organs following treatment with cholinergic and adrenergic blocking agents, ganglionic blocking agents, as well as drugs that deplete the stores of neurohumors in presynaptic endings (e.g., reserpine, botulinum toxin; see review by Emmelin 1961). The prolonged use of central depressants usually leads to an increase in the sensitivity of central nervous tissue, which, in the case of barbiturates, resembles that occurring in chronically isolated cortex - - v i z . , an increased tendency towards paroxysmal activity. This increase in the sensitivity of central neurons is revealed when the drug is rapidly withdrawn ("abstinence syndrome", Wulff 1959). The parallel between this and the increase in responsiveness of effector organs when blocking agents are withdrawn after long use - - e . g . , the nictitating membrane at the end of several weeks of continuous adrenergic blockade (Nickerson and House 1958) - - is striking. Indeed, Emmelin, who is responsible for much of the work on sensitization by phar-

253

macologic denervation, has suggested that one may be a model of the other - - that the increase in sensitivity of central neurons subjected to prolonged depression by drugs may be a manifestation of sensitization by pharmacologic denervation (1961, p. 32). There is as yet no reason to suppose that the chronic use of depressant drugs causes collateral sprouting in the central nervous system. In the present study, chronically isolated cortex showed not only a tendency towards paroxysmal activity but also a tendency to be selectively depressed following an experimental session in which it had been repeatedly stimulated. In interpreting this observation, it must be recognized that longer after-discharges were produced in isolated cortex than in the intact cortex, so that the greater tendency towards depression in the slab might be attributed to its greater total activity. It is also possible that partially denervated and chromatolyzed cortical neurons have less metabolic reserve, or that access to essential metabolites is somehow impaired in chronically isolated cortex. There is, however, an attractive alternative hypothesis which deserves consideration: it is possible that the process of sensitization produced as a result of inactivity in the isolated slab was reversed by the intense activity occurring during an experimental session; that the partially denervatcd neurons had become desensitized following an excess of stimulation just as they had previously become sensitized following privation of stimulation. It has been reported that the sensitized, denervated smooth muscle of the nictitating membrane (Simeone 1938) and intestine (Youmans 1938) may be desensitized by chronic treatment with epinephrine. The local application of acetylcholine and the repetitive stimulation of motor nerve are also said to desensitize the chemical receptors of skeletal muscle (Thesleff 1960). We have emphasized the similarity between the changes following denervation and pharmacologic blockade in the brain on the one hand and in peripheral structures on the other - - particularly those autonomic effector organs which have been extensively studied, such as the smooth muscle of the nictitating membrane. Electroenceph. clin. Neurophysiol., 1962, 14:244-255

254

S. K. SHARPLESS AND L. M.I-IALPERN

Much is known about the enzymatic and other changes accompanying pharmacologic or structural denervation in effector organs, but it cannot be said that the mechanism of denervation supersensitivity is understood. Different mechanisms may be involved in the central and peripheral nervous systems, and in different effector organs. However, the impressive number of parallels should not be ignored. Certainly, the hypothesis of a common mechanism ought to be entertained on the grounds of parsimony, until the proof is in. S UMMARY

1. A method is described for the study of chronically isolated areas of cortex by means of permanently implanted electrodes. This method has been used to study the emergence of an increased susceptibility to paroxysmal after-discharge in the suprasylvian gyrus of the cat after neural isolation. 2. At first, isolated cortex tended to show little spontaneous activity (except occasional bursts of abnormal activity in some instances), but continuous activity of an abnormal character commonly developed before the end of the second postoperative week. Histological examination revealed that the isolation was often incomplete; a few strands of connecting fibers in the anterior corner of the gyrus sometimes remained intact. 3. The threshold for epileptiform afterdischarge fell significantly below that of the homologous opposite cortex during the second postoperative week. 4. The average duration of after-discharges elicited by stimuli of moderate intensity increased, often more than fivefold, during the first three postoperative weeks. 5. During the same time period, there was a slight but significant increase in the duration of after-discharges induced in the opposite hemisphere (which had been partially denervated by destruction of callosal fibers during the isolation procedure). 6. The excitability of chronically isolated cortex may be depressed towards normal for a period of several days following repeated afterdischarges. 7. The time required for sensitization of

isolated cortex is approximately the same as that required for the development of denervation supersensitivity in peripheral structures. 8. The possible roles of axon-collateral proliferation and denervation supersensitivity are discussed in connection with the increased susceptibility to paroxysmal activity developing in cortex after surgical isolation or after prolonged pharmacologic depression. REFERENCES BURNS, B. D. The production of after-bursts in isolated unanesthetized cerebral cortex. J. Physiol. CLond.), 1954, 125: 427-446. CANNON, W, B. and ROSENBLUETH, A. The supersensitivity o/ denervated structures. Macmillan, New York, 1949, 245 p. ECHLIN, F. A. The supersensitivity of chronically "isolated" cerebral cortex as a mechanism in focal epilepsy. Electroenceph. olin. Neurophysiol., 1959, /1: 697-722. ECHLI~X, F. A. and MCDONALD, J. The supersensitivity of chronically isolated and partially isolated cerebral cortex as a mechanism in focal cortical epilepsy. Trans. Amer. nenrol. A3:~., 1954, 7 9 : 7 5 79. EMMEL~N, N. Supersensitivity following "pharmacological denervation'. Pharmacol. Rev., 1961, 13: 1737. FRANKEN, L. Contribution ?~ I'~;tude expdHrnentale des phdnomOnes de ddnervation air nivea~t de l'&'orce cdrt;brale. Thhse, pr4sent4e 7t la Facult6 de M6decine de I'Universit6 de Bruxelles, 1969. GRaVSTHN, B. and SASTRY, P. B. Some preliminary electrophysiological studies on chronic neuronally isolated cerebral cortex. Electroenceph. clin. Nenrophysiol., 1957, 9: 723-725. HAMPEL, C. W. The effect of denervation on the sensitivity to adrenine of the smooth muscle in the nictitating membrane of the cat. Amer. J. Physiol., 1935, 111:611--621. HENRY, C. E. and SCOWLLE, W. B. Suppression-burst activity from isolated cerebral cortex in man. Electroenceph, clin. Neurophysiol., 1952, 4 : 1 22. IALLY, I. C. Injury and excitation of brain by electrical current. In E. R. RAMI::Y and D. S. O'DOHERTY (Editors), Electrical studies of the unanesthetized brain. Hoeber, New York, 1960, pp. 96-105. Liu, C. N. and CHAMB~:RS,W. W. Intraspinal sprouting of dorsal root axons. A. M. A. Arch. Neurol. Psychiat., 1958, 7 9 : 4 6 61. MORRELL, F. Secondary epileptogenic lesions. Epilepsia (Amst.), 1959/1960, 1:538 560. Mt:RaAY, J. G. and THOMPSON, J. W. The occurrence and function of collateral sprouting in the sympathetic nervous system of the cat. J. Physiol. (Lond.), 1957, 135:133 162. NICKERSON, M. and HoustL H. D. Mechanisms of denervation supersensitivity. Fed. Proc., 1958, 17: 398.

Electroenceph. clin. Neurophysiol., 1962, 1 4 : 2 4 4 - 2 5 5

CHRONICALLY ISOLATED CORTEX PURPURA, D. P. and HOUSEPIAN, E. M. Physiological consequences of axon-collateral proliferation in isolated immature neocortex. Fed. Proc., 1961, 20: 333. ROSENBLUETH, A. The action of certain drugs on the nictitating membrane. Amer. J. Physiol., 1932, 100: 443-446. ROSENBLUETH, A. and CANNON, W. B. Cortical responses to electrical stimulation. Amer. J. Physiol., 1942, 135: 690-741. ROTHBALLER, A. B. and SHARPLESS, S. K. Effects of intracranial stimulation on denervated nictitating membrane of the cat. Amer. J. Physiol., 1961, 200: 901-908. ROWLAND, V., MACINTYRE, W. J. and BIDDER, T. G. The production of brain lesions with electric currents, II. Bidirectional currents. J. Neurosurg., 1960, 17: 55-69. SI~Eoyv:, F. A. The effect of previous stimulation on the responsiveness of the cat's nictitating m e m b r a n e sensitized by denervation. Amer. J. Physiol., 1938, 122: 650-658.

255

STAVRAKY,G. W. Supersensitivity following lesions of the nervous system (an aspect of the relativity of nervous integration). University Toronto Press, Toronto, 1961, 216 p. TEASDALL, R. D. and STAVRAKV, G. W. Response of deafferented spinal neurones to corticospinal impulses. J. Neurophysiol., 1953, 16: 367-375. THESLEFF, S. Effects of motor innervation in the chemical sensitivity of skeletal muscle. Physiol. Rev., 1960, 40: 734-752. WILCOXON, F. Some rapid approximate statistical procedures. American Cyanamid Company, Stanford, Conn., 1949, 17 p. WuLvv, M. H. The barbiturate withdrawal syndrome. A clinical and electroencephalographic study. Electroenceph, clin. Neurophysiol., 1959, Suppl. 14: Munksgaard, Copenhagen, 173 p. YOUMANS, W. B. Similarity of effects of adrenalin and inhibitory sympathin on intestinal motility; sensitization by denervation. Amer. J. Physiol., 1938, 123: 424-431.

Reference : SHARPLESS,S. K. and HALPERN,L. M. The electrical excitability of chronically isolated cortex studied by means of permanently implanted electrodes. Electroenceph. clin. Neurophysiol., 1962, 14: 244-255.