Intracortical transmission of specific evoked activity across the cerebral cortex

281

Electroencephalography and Clinical Neurophysiology, 1975, 38:281-293 "{3 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

INTRACORTICAL

TRANSMISSION

ACROSS THE CEREBRAL

OF SPECIFIC EVOKED

ACTIVITY

CORTEX 1

J. A. ENNEVER 2 Department of Physiology, University College, London W.C.1 (Great Britain} (Accepted for publication: October 14, 1974)

There appear to be three main types of purely intracortical tangential propagation of activity described in the literature on the cortex. One type is characterized by a very slow speed of conduction, 0.001 m/sec-0.001 m/min, and includes the Jacksonian march (Jackson 1931), possibly the movement of ophthalmic migraine scotomas (Milner 1958), and spreading depression. Le~o first described the latter: a wave of intense activity spreads with a velocity of about 0.005 m/rain, in a gradually expandin~ circle across the cortex, leaving the tissue electrically inactive for about 1 min afterwards (Le~o 1944). Burns (1958) described a similar slow-spreading phenomenon in slabs of isolated cortex, which supports the idea that this type of spread is a purely intracortical phenomenon. The transmission is thought to involve synaptic excitation, facilitated by the accumulation of extracellular potassium ions (Grafstein 1956; Vysko~il et al. 1972), and is triggered by the simultaneous activation of an abnormally large number of cells. Thus, this first type of spreading activity is highly abnormal. The second, and faster, type of transcortical transmission travels at a velocity of 0.1-4).2 m/sec in a network of cells 1 mm or more beneath the surface. Burns (1951, 1958) evoked this type of spread in chronic isolated slabs of unanaesthetized cortex by a weak direct stimulus, which was not strong enough to produce spreading depression. This "burst response" spread unattenuated throughout the 20 mm long slab, and Burns This research was supported by a Medical Research Council Training Grat~t. 2 Present address: Brain Research Institute, University of Zurich, August-Forel-Strasse 1, CH-8029 Zurich.

concluded that this demonstrated a potential for unlimited-synaptic-tangential transmission inside the normal cortex. Another example of this type of spreading activity was described by Cobb et al. (1955) in isolated cortex, but in the presence of anaesthetic. In this case the spreading "activity" was a strychnine spike, which does not usually "spread" intracortically more than about 2 mm (Dusser de Barenne and McCulloch 1938, and personal observation), but Cobb et al. induced the intracortical transmission by strychninization of the cortex at 5-7 mm intervals. The same experiments in intact cortex were complicated by a much faster spread, via corticocortical fibres. However, only intracortical, not intercortical, transmission is being considered here. Also included in the second group, should be the apparent spread of activity evoked by weak repetitive direct stimulation of the cortex (Adrian (1936) in anaesthetized brain; Penfield and Boldrey (1937) in unanaesthetized patients; Burns and Smith (1962) in unanaesthetized cat cortex). The exact pathway for these spreading influences is not known, but it seems likely that the repetitive stimulation increases the excitability of a deep cortical network in a paroxysmal fashion, and gradually induces the second type of transmission. The third, and fastest, type of intracortical transmission is that evoked by a single weak direct stimulus to the cortex, again a highly abnormal stimulus. This evokes a surface-negative wave which travels up to 4-5 mm from the stimulus site in layer I (Chang 1950; Ochs and Suzuki 1965). Burns (1958) detected the response up to 10 mm away. The velocity was estimated as 1 m/sec in cat and 0.6--0.7 m/sec in monkey by

282

C h a n g (1950) a n d 2 m/sec in cat by Burns (1958). T h e p r o p e r t i e s o f these three types of s p r e a d i m p l y t h a t there a r e p o t e n t i a l t r a n s c o r t i c a l p a t h ways from a n y one cortical p o i n t to a n y other. H o w e v e r , it is i m p o r t a n t to notice that all these types of s p r e a d have only been seen as the result of highly a b n o r m a l cortical stimulation. H o w these p r o p e r t i e s of cortical circuitry are used in the t r a n s m i s s i o n of n o r m a l cortical activity is not known. T h e e x p e r i m e n t s d e s c r i b e d here were designed to investigate w h e t h e r the short latency r e s p o n s e e v o k e d by w e a k s o m a t i c stimulation, which was c o n s i d e r e d to be a relatively n o r m a l response, i n v o l v e d a n y o f the three types of i n t r a c o r t i c a l s p r e a d d e s c r i b e d above.

J.A. ENNEVER

METHODS B l a c k h o o d e d rats of the Lister strain a n d albin o W i s t a r rats were used. O n l y the y o u n g e r rats, 170-200 g, were used when a n extensive craniot o m y was p e r f o r m e d , since the fibrous connections between t h e d u r a a n d c r a n i u m were so t o u g h on o l d e r rats that the r e m o v a l o f a large a r e a o f b o n e caused s u b d u r a l h a e m o r r h a g e s . T h e rats were a n a e s t h e t i z e d with 1.6-1.9 g / k g b o d y weight of urethane, a d m i n i s t e r e d i.p. as 3 6 ~ solution. U r e t h a n e ( e t h y l - c a r b a m a t e ) was used since it p r o d u c e s a s t e a d y d e p t h o f a n a e s thesia for 6-10 h ; a n d secondly, B i n d m a n et al. (1964a) have s h o w n that u n d e r u r e t h a n e the

4 J

2

...

eOgOOQOQ

Fig. 1. Topographical maps of the cerebral hemisphere of the rat. A : Right side: localization of sensory and motor function in the rat (Woolsey 1958). Left side: cytoarchitectonic map of the cortex (Krieg 1963). Dotted lines represent the sutures of the skull which are used for landmarks in the rat, in the absence of sulci. B: Drawn to same scale (each division = ! mm). Dotted lifie: average boundary of the cortical somatosensory receiving area of the contralateral forepaw (CRA), in urethane anaesthetized rats. Shaded region: central area. Central dot: position at which the largest responses were recorded. Boundary responses were usually recorded from the posterior boundary of the CRA, just inside or on the dotted line.

283

INTRACORTICAL TRANSMISSION OF EVOKED POTENTIALS

potential level of the cortex reflected the state of excitation of the subjacent cells. Thus it was convenient to use the potential level of the cortex as an estimate of cortical activity. A trephine hole, 4 mm in diameter, was made over the forepaw receiving area (Fig. 1), the dura was removed and a small ring cemented with dental wax round the hole. This enabled the cortex to be covered with liquid paraffin. The rectal temperature was maintained at 36-37 ° C through a transistor controlled heating device. The sensory stimulus used was a weak electrical pulse, 0.05 msec duration, delivered to the palmar skin of the contralateral forepaw, through a pair of fine needles. The stimulus strength was always subthreshold for movement. The surface records were taken from the cortex with 0.5 mm silver-ball electrodes, coated with AgC1, or glass-pipettes, 4-8 #m tip diameter, filled with 10 % NaC1. The latter were also used for monitoring multi-unit activity during polarisation. The indifferent electrode was placed on the eye. The recording electrodes were connected to Tektronix 3 A 3 amplifiers in the direct-coupled mode, or with a 300 msec time constant, via cathode followers. Measurements were taken from a continuous film of the oscilloscope display. To change the excitatory state of various elements in the cortex a piece of filter paper 2 mm square soaked in 2 % strychnine was applied directly to the cortex for 1 min ; or 2.5 % solutions of GABA were placed on the cortex and washed away after 10 min. Intravenous injection of 0.3 ml of 2 % strychnine solution, which usually produced convulsions, was also used. A fourth method of artificially altering the excitation state was by polarization. Voltages, connected in parallel with a 100 pF capacitor, could be applied between a saline soaked cotton wick on the cortical surface and a second electrode in the mouth, through a 2 Mf~ potentiometer. Currents of 50-300 pA/mm 2 were used. Assuming a linear functional relationship between the electrocorticograms (ECoGs) at two different recording sites, the cross-correlation coefficient, rxy (T) (where the mutual displacement in time T was always zero, i.e., zero timelag) was estimated in order to measure the degree of synchrony of the two signals x (t) (electrode 1) and Yi (t) (successive electrodes 2, 3, 4, 5

or 6, (i), more posterior). The correlation coefficient was calculated with a Packard Bell 440 from the cross-variance (lagged product-moment correlation), with zero mean ; Cxy(T) = E xt,)" yr, + T)

-

-

X~n)" ~ Ytn + T) N

where n is the sample index at time t; 0 < t < 30 secs; 1 < n < N, the sample rate was 2500/sec, thus N = 2500 x 30; and normalized by the autocovariance functions : Cxy rx,= x / C x x . Cyy giving:

=(

r~y

n=l

x ~o,.y,o, - [ ~n 1 x ,°,. n=Es yi,n,] fi1

N q2~l~( N I- N 72~_~ / I1=1 Z x:,-[l- n=lEx,.,J _J()Y~ ~n=l y,~.,-L._E Yi ,o,J

The six silver ball electrodes were spaced at 1 mm intervals in a line across the cortex, in an anteriorposterior orientation. The simultaneous ECoGs, recorded between each electrode and ground were fed into a digital data acquisition system, which converted the signal amplitude to 6 bit sample values (bandwidth 10,000-12 c/sec). Since the apparatus could only read 4 E C o G channels simultaneously, the data for Fig. 6 were recorded in two sets of four (i.e., electrodes 1, 2, 3, 4; then electrodes 1, 4, 5, 6). Each correlation coefficient was calculated from a 30 sec E C o G sample and averaged with the values obtained in 3 immediately subsequent trials. RESULTS

Under most anaesthetics, short latency responses evoked by a shock to the contralateral forepaw can be recorded over about half of the hemisphere in the rat. This is called here the cortical forepaw area " C R A " (see area enclosed by the dotted line in Fig. 1, B). The hatched area, in Fig. 1, B encloses the central region of the CRA where the responses, plotted under urethane anesthesia, are greater than half maximum amplitude; this is the region usually referred to as the primary sensory receiving area of the

284 forepaw in the rat (Libouban-Letouz6 1963: Ennever 1971). However, outside this central region smaller short latency responses can still be recorded• The smallest responses, on the boundary of the CRA (marked by the dotted line in Fig. 1, B), will be referred to as boundary responses. Boundary responses are not the result of electrotonic spread from the central area (Ennever 1971). They are accompanied by spike activity in the subjacent cortex, and are not abolished by cuts entering layer VI of the cortex, which would block intracortical transmission from the centre of the CRA. Theretore boundary responses are not themselves simply a result of intracortical spread; they are independent short latency responses, and are used here to study intracortical transmission. Evidence for the presence of intracortical transmission of evoked potentials would be demonstrated if purely cortical changes could influence the position of the CRA boundary, i.e., induce responses in a region previously outside the CRA or reduce the size of the CRA. No such evidence, however, could be found in the attempts to do this described below• The estimation of spread in terms of mass responses in the centre of the CRA is considered in the second half of the Results section. 1. Boundary responses In an attempt to induce or enhance intracortical spread of the evoked responses, or some component of them, the posterior boundary regions of the CRA (those regions of the cortex on the posterior portion of the dotted line in Fig. 1, B) were treated with strychnine. This appears to increase the excitability of some cortical elements that are activated by the afferent volley. Topical application to about 2 mm 2 produced spontaneous strychnine spikes in an area of 3-4 mm 2 local to the application; but the shape of the boundary evoked responses was unchanged (Fig. 2) and no change in the location of the CRA border was ever found. A second method of locally increasing the activity of some cortical elements in order to induce or just alter the intracortical transmission of primary evoked activity, was to pass a small, steady current (50-500/zA/mm 2) across the cortex (polarization). Purpura and McMurtry (1965)

J.A. ENNEVER

TOPICAL APPLICATION OF STRYCHNINE TO POSITION 4. before

after

"?00LLV '30 msec

1234 Fig. 2. Responses evoked by stimulation of the contralateral forepaw, recorded at 1 mm intervals simultaneously from the surface of the cerebral cortex [diagrammatically illustrated below). Electrodes I and 2 are in the central CRA. electrodes 3 and 4 in the more peripheral region. Point 4 is on the CRA boundary. A 2 mm 2 piece of filter paper soaked in 2 o~, strychnine was placed for 1 min on the CRA boundary; spontaneous strychnine spikes occurred at point 4. and could also be detected at 3. No detectable difference could be found in the boundary evoked responses at this point: the relative pattern of responses was unchanged. On this and all other figures• upward deflections signal positivity of the focal electrode.

have shown that surface-positive current (SP) tends to increase the spontaneous firing of some neurones deeper than 0.7 mm4).9 mm in the cortex, whereas surface-negative current (SN) inhibits it• The current has the opposite effect on more superficial cells• The standard changes in the shape of the mass response were seen: SP increased negative components, SN increased positive components (Bindman et al. 1964b). (However, it was noticed that these changes were reversed in the mass responses recorded 3 mm away from the polarizing wick,

INTRACORTICALTRANSMISSIONOF EVOKED POTENTIALS

285

EFFECTS OF CORTICAL POLARIZATION

fire spontaneously; these are interspersed with periods when m a n y cells can be found to fire ON EVOKED RESPONSES. spontaneously. The active bursts are accompanied by a positive D C shift of the surface pobefore during tential (see Fig. 7). These periods of activity are 500lay also related to an increased state of arousal, since any alerting stimulus will evoke such a burst, SN and they occur more frequently and eventually 10reset 5oopv become continuous as the anaesthetic depth is reduced. However, it was found that these bursts SP of generalized cortical activity do not affect the 10msec 5oopv position of the boundary of the CRA. Neither did sub-convulsant and supra-convulsant doses SP of strychnine which also affect the central excita10msec tory state in some way (Fig. 5) (Smolin and Samko Fig. 3. 1: Effect of surface-negative (SN) polarization of 50 1968). The convulsant did increase the amplitude #A/mm2 on cortical responses evoked by contralateral fore- of the primary potentials, but as with polarization paw stimulation. Potentials are recorded from the centre of the change in amplitude was proportional to the CRA. Notice typicalincrease in surface-positivecomponents. original size of the response, and no change in 2: Effect of surface-positive(SP) current of 200/~A/mm2 on a response recorded more peripheral in the CRA than 1. the position of the CRA boundary was found in Note typical increase in surface-negative components. any experiment. 3: Boundary response: 200/zA/mm2 SP increases a negative Thus none of the direct or induced manipulacomponent but the change is very small, proportional to the tions of cortical activity had any effect on the original amplitude.

I

__l

_A

presumably because the direction of the effective current is reversed in that part of the field.) A microelectrode monitoring the activity of deep neurones showed that even when deeper structures were facilitated or inhibited by the current (which was far greater than the current intensities that maximally affect the surface mass response), no change in the boundary of the CRA could be detected, and no component of the evoked responses could be induced beyond the original boundary. The current only changed the shape of the existing evoked potentials in the usual manner (Fig. 3). The amplitude of the change produced by a supramaximal current was found to be directly proportional to the original peak-to-peak amplitude (see Fig. 4) and extrapolation of the graph to zero confirms the observation that polarization could not induce a response where previously none existed (i.e., peak-to-peak a m p l i t u d e = 0 ) . This is not the effect that would be expected if the afferent volley evoked activity in an unlimited network, whose excitability was increased by the polarizing current. Under urethane anaesthesia, there are quiescent periods in the cortex during which n o cells

-~ IO

:.

~ I.O

,

o • lirao

o.~

:...-

,+

•

o•

.; o.o~

~

I

Ob

• I

I

I

O.Ol 0.1 .o IO Peak-to-peak amplitudeof evokedpotentials (my) Fig. 4. Notice direct proportionality ofchange in peak-to-peak amplitude of the cortical evoked responses, plotted on a log-log scale, with the amplitude of the response before passing the polarizingcurrent. The potentials were evoked by electrical stimulation of the contralateral forepaw.The results shown on the graph include changes induced by surfacepositive (SP) and surface-negative (SN) current (current densities of 50-300 gA/mm2, and durations of 2-60 min). The direct relationship revealed by the graph is therefore independent of the dimensions of these large currents.

286

J.A. ENNEVER

l o c a t i o n o f the e v o k e d potential. It s h o u l d be e m p h a s i z e d t h a t a decrease in C R A w o u l d also have p r o v i d e d positive evidence for the existence of lateral t r a n s m i s s i o n of the e v o k e d response. H o w e v e r , the results all i m p l i e d t h a t the b o u n daries o f the f o r e p a w a r e a at a n y one time a r e rigid. It is p o s s i b l e t h a t the r i g i d i t y is the result of a n a t o m i c a l limitations, since a t t e m p t s a t altering the p h y s i o l o g i c a l state o f the cortex d i d n o t induce a n y lateral t r a n s m i s s i o n o f the response. Very o c c a s i o n a l l y a n e v o k e d negative wave w o u l d a p p e a r successively a l o n g a r o w o f electrodes, even a l o n g t h o s e o u t s i d e the C R A b o u n dary. H o w e v e r , n e i t h e r their a m p l i t u d e o r freq u e n c y was c h a n g e d b y the a t t e m p t s to alter cortical excitability.

EEG Electrodes lmm

apart

2 mm apart

f f

Z LU

lmvLf 1sec

ii ii U.I 0 C.) Z

g _J

INTRAVENOUS

W

INJECTION

OF STRYCHNINE

0 0

before

after

I

I

I

I

2

3

/.

5 mm

Fig. 6. The ECoG recorded from points l mm apart yield very similar traces lupper set of records, retouched). If the electrodes are moved 2 mm apart differences marked by arrows become obvious. Graph: data from experiments in which 6 recording electrodes were used. Each point is a quantitative estimation of the variation between two simultaneous ECoG records. The correlation coefficient is a measure of the synchrony between activity from electrodes placed 1-5 mm apart. The graph shows the relative changes over 5 mm.

|

i•

i

1

]5oot.tv i

30 msec

1234

Fig. 5. Responses evoked by stimulation of the contralateral forepaw recorded at 1 mm intervals, simultaneously from the surface of the cerebral cortex. After a convulsant dose of strychnine (30. mg/kgl, the amplitude of the central responses, 1 and 2, increased due to a change in the excitation state of the brain: however, the peripheral responses. 3 and 4. were not significantly enhanced.

2. Central responses E v o k e d responses r e c o r d e d at two p o i n t s near the centre o f the C R A . a b o u t 2 m m a p a r t , are usually similar in c o n f i g u r a t i o n , a n d their s h a p e varies in a n identical m a n n e r . T h e variations in s h a p e have been s t u d i e d by B i n d m a n et al. (1964a), w h o have shown t h a t the shape of a r e s p o n s e varies with the s p o n t a n e o u s activity in the c o r t e x o n the a r r i v a l o f the afferent volley. T h e p e a k - t o - p e a k a m p l i t u d e o f the responses a r e r o u g h l y p r o p o r t i o n a l to the a m p l i t u d e o f the surface-positive D C shift which a c c o m p a n i e s the s p o n t a n e o u s activity (i.e.. E C o G a m p l i t u d e ) . E x p e r i m e n t s were designed to d i s c o v e r w h e t h e r the changes in s h a p e of the t w o s i m u l t a n e o u s l y r e c o r d e d responses were closely c o u p l e d because of i n t r a c o r t i c a l spread, o r the result o f identical fluctuations in the cortical activity at each point.

INTRACORTICALTRANSMISSIONOF EVOKED POTENTIALS

a. The instantaneous synchrony of the cortical activity at two points E E G activity was recorded simultaneously from six surface electrodes, placed in a line at 1 m m intervals across one hemisphere, and with n o specific relationship to the CRA. The sync h r o n y between the E C o G at the first electrode and each of the other five was measured a n d expressed in terms of a correlation coefficient. The results (Fig. 6, graph) show that there is a high correlation between the instantaneous activity at the two points, which decreases with increasing distance over a range of a b o u t 2 m m before reaching a constant level. Therefore, it is possible that the closely coupled variations in the shape o f the evoked responses were due only to the high correlation of the s p o n t a n e o u s activity at each point.

b. Shape of two simultaneously recorded responses when cortical excitation states at the two recording sites differ W h e n a burst of s p o n t a n e o u s activity occurred

287

at one point a n d n o t the other, and it was arranged for a stimulus to be delivered at such an instant, the shape of the responses was completely different. E a c h evoked response always assumed the appropriate s h a p e for the level o f s p o n t a n e o u s activity at its recording site (Fig. 7). T h u s the Shape of the p r i m a r y evoked response is mainly determined by the state of a local area o f cortex less than a b o u t 2 m m in radius, a n d it is not strongly affected by activity 2 m m or m o r e away.

c. Measurement of the dependence of evoked response on remote cortical excitation A l t h o u g h the responses 2.5 m m apart (Fig. 7) appeared to behave independently, it was possible that the burst o f s p o n t a n e o u s activity at one point still had a small influence on the evoked activity in a neighbouring area. Therefore, the peak-topeak amplitude (a) of between 30-90 responses, at one point, was correlated with the coincident levels of s p o n t a n e o u s cortical activity at points 1.5-2:5 m m away. In fact, since the level of s p o n t a n e o u s activity is directly p r o p o r t i o n a l to

I0 m s e c Fig. 7. Two top traces: ECoG recorded at two points 2.5 mm apart in the. central CRA. I, II and III are taken from the same records at times indicated by arrow, but on an expanded time scale. They show the responses evoked by weak electrical stimulation of the contralateral forepaw, more clearly. The responses in III are typical of evoked potentials taken when the cortex is "quiescent" (DC level=0). This stimulus also generates a burst of activity after a cortical response (see top 2 traces). Responses in II are entirely negative, both were evoked during a burst of activity, when the potential level of the cortex was positive with respect to the baseline. I shows the positive and negative forms occurring together. It can he seen from I that there is a burst of activity (DC level positive) at one point and not at the other, and the evoked potentials at this instant vary completely independently of each other. Voltage calibration = 1 mV.

288

J.A. ENNEVER

the DC level of the cortex (Bindman et al. 1964a), in this preparation the DC level (B) was used in the correlations as the measure of the level of spontaneous activity. It was shown above that there is a high degree of correlation between the ongoing spontaneous activity at two adjacent points: therefore since (a) is very highly correlated with the DC level (A) at that point, (a) will also be highly correlated with (B), the DC level 1 or 2 mm away.However, the aim of the present calculation was to try to uncover some relationship between (B) and [a) independent of this correlation. To do this the following analysis was performed. The correlation coefficient ra.A for (a) on (A) was calculated, and similarly the coefficients ra.B and rA. B" Using these three correlation coefficients, the partial correlation coefficients were calculated for the dependence of (a) on (A) when the variations due to (B) are excluded, i.e., when (B) is effectively kept constant (r~A.B), and (a) on (B) when (A) is effectively kept constant (r~B.A). The significance of these coefficients was found from standard statistics tables. The results of eight analyses are shown in Table I. In none of the cases where (A) and (B) vary significantly independently (i.e., raA,B has P<0.001) was there a significant correlation

between (a) and (B), independent of (A) (i.e.. raB.A= not significant). Thus no evidence could be found for any lateral influence of spontaneous activity in one area of cortex on evoked activity in another area about 1.6 mm or more away, The high degree of correlation between the spontaneous activity at points less than about 2.0 mm apart (see Fig. 6 and Table I) prevented a more accurate estimation of the extent of lateral transmission being made. DISCUSSION

These results have shown that the manoeuvres. such as polarization and topical or intravenous application of strychnine, altered the physiological state of some cells in the cerebral cortex. and yet neither they. nor the spontaneous changes in cortical excitation, could induce any detectable tangential spread of the specific evoked responses beyond the boundaries of the cortical forepaw receiving area (CRA). Any activity that had spread from the CRA faster than 0.03 m sec and with an amplitude greater than 100 pV, recorded from the surface, would have been detected in these experiments. Therefore no evidence was found for the propagation of evoked activity by the second and third types of trans-

TABLE 1 Results of a statistical analysis to show the effect of excitation gradients on evoked potentials. Distance apart of A and B (mm) 1.5 1.5 1.6 1.6 2.0 2.0 2 ", - _5

No. of observations

33 33 34 34 89 89 49 49

Correlation coefficients and probabilities ..................... ra. A P ra. R

P

r~,.,

P

Partial correlation coefficients and probabilities r,,A., P r~n,A P

-0.53 -0.53 -0.54 -0.69 -0.59 -0.50 -0.67 -0.71

0.001 0.01 0.001 0.01 0.001 0.001 0.001 0.001

0.68 0.68 0.60 0.60 0.57 0.57 0.41 0.41

0.001 0.001 0.001 0:001 0.001 0.001 0.0l 0.01

-0.12 - 0.31 -0.22 -0.58 -0.46 0.33 - 0.58 -0.65

0.01 0.01 0.001 0.001 0.001 0.001 0.001 0:001

-0,64 --0A9 -0,69 -0,45 -0,44 --0,43 -0,51 -0,45

not sig. not sig. not sig. 0.001 0.001 0.001 0.001 0,001

-0.44 - 0.21 -0.55 -0.07 -0.15 --0.02 --0.35 -0.24

"

0.01 not sig. 0.001 not sig. not sig. not sig. not sig. not sig.

The E C o G was recorded at two sites, A and B, in the CRA. The DC level of the E C o G immediately proceeding an evoked response at A was correlated with the amplitude of the response itself (a), giving the correlation coefficient (co) ra. A. (a) was "similarly correlated with the excitation at B, c c = r , l w Using ra.A, ra. n and rA,n the partial cc raA.B and raa,A were calculated. It can be seen that as the distance between A and B increases, rA,s decreases; i.e., B varies more independently of A; however, no significant correlation could be found between the excitation a t B and (a), when B varies independently of A: i.e., r,B,A is not sig. for larger A - B separations.

INTRACORTICAL TRANSMISSION OF EVOKED POTENTIALS

cortical spread, 0.1 and 1.0 m/sec respectively, which were described in the introduction. The first type of spread takes 12 sec to travel 1 mm, which excludes it from participation in all the fast information processing of evoked responses, and since it also is followed by a period of electrical inactivity never seen in these experiments, it almost certainly cannot participate in a natural propagation of weak somatosensory evoked responses. The three possibilities to account for these negative results that will be considered in turn are : (1) The cortical elements involved in the evoked response can transmit activity tangentially, but the techniques used to induce the change were not effective. (2) The techniques were successful in inducing the spread of evoked activity, but changes in the spread were not detected. (3) The neural elements involved in the specific evoked response cannot transmit activity laterally in the cortex.

1. Effectiveness of techniques used to change cortical excitation The origin of the small mass evoked responses near the boundary of the CRA is still unclear. However, it has been shown that these responses were not dependent on an intact intracortical pathway to the centre of the CRA, since their shape was independent of both manipulations of the central CRA responses alone (topical application of GABA, Ennever 1971), and spreading depression induced in the contralateral hemisphere. Also, evoked spikes could be recorded from the subjacent cortex. The boundary responses were therefore not the result of electrotonic spread from the central area--a conclusion supported by the demonstration of the independence of both evoked and spontaneous cortical activity just 2 and 2.5 mm apart respectively (Fig. 6 and 7). These figures also emphasize the low space-constant of mass recordings at the surface of the cortex. Therefore, the CRA responses involve elements in the directly subjacent cortex and should be affected by manipulations of these cortical areas. However, since it is unclear exactly which are the elements involved in the response, it is impossible to claim that the manoeuvres used here, in an attempt to induce cortical excitation changes, affected them. To overcome this problem several methods were

289

used to produce the excitation changes. They included the systemic injections of strychnine, local application of current and strychnine, also observations during spontaneous excitation cycles, which occur under urethane anaesthesia. It is considered unlikely that from such a wide range of methods, each known to alter excitation of neural elements (Li 1959; Purpura and McMurtry 1965; Ennever 1971; Towe and Mann 1973) none affected the elements involved in the response either in an excitatory or inhibitory manner. After topical application of strychnine to the cortex, the presence of spontaneous strychnine spikes indicated that the excitation of some cortical elements had been affected (Towe and Mann 1973), but the boundary responses were unchanged. The intravenous strychnine clearly attected the central nervous system, since it produced an increase in the size of the large central evoked potentials; however, again the boundary of the CRA did not change. The spontaneous changes in cortical activity have been shown in 21 of 47 cells to be accompanied by changes in response probability (Ennever 1971), but these bursts of activity did not produce a detectable change in the CRA boundaries ; they had no effect on lateral transmission. Towe (1966) showed that the major part of the somatosensory evoked response was the result of potential changes in "s" elements that are concentrated in the superficial regions of the cortex, and that the cells of layers V and VI contribute very little to the response. If this hypothesis also applies to the small boundary responses of the CRA, then any attempt to induce the spread of the evoked activity should affect these superficial cells. The currents used were shown from simultaneous microelectrode recordings to be strong enough to change the firing rate of deep cells and would also, therefore, affect the superficial ones. However, the effect of the current on a cell is hard to estimate because it depends critically on its morphology--that is, whether a large enough dipole is created for the resultant current flow to affect the trigger-zone. It appears unlikely then, that none of the manoeuvres, either inhibited or excited the neural elements involved in the response, yet none revealed a change in the lateral transmission of the activity.

290

2. Sensitivity of recordings The current modified the shape of the surface mass responses in a predictable manner. It was found that the changes were often independent of current densities above 50 p A / m m 2 (subthreshold for affecting most cells), but were directly proportional to the original size of the response. Since the size of a response is a function of the number of elements involved, this result implies that the size of the induced change depends on the original number of elements excited by the afferent volley, a n d thus the current is recruiting no new elements into the recorded response. One explanation of this negative result is that the evoked activity is not labile in the superficial network and therefore excitation changes do not increase the number of active elements. A second is that the surface records are not sensitive enough to detect the changes in excitation although they detect other accompanying changes like those in the evoked response. In support of the latter possibility, Burns (1951. 1958) showed that a network of elements 1 mm beneath the surface had to be activated before spreading activity could be initiated, while Towe (1966) showed the activity 1 m m deep in the cortex contributed very little to surface evoked potentials. In addition, the changes in surface potentials with polarization are certainly different from those of the deep cortical cells (Creutzfeldt et al. 1962) where the change in discharge rate was directly proportional to the amplitude of the polarizing currents. It is therefore considered possible that changes in excitation, and thus spread, in deeper layers of the cortex could occur, but would not be detected by surface recordings. A study using more sensitive recording techniques such as single cell or intracetlular analyses coula well reveal more subtle changes. The technique of averaged evoked responses was used in the experiments investigating CRA during spontaneous changes in cortical activity (Ennever 1971) in an attempt to increase sensitivity of the surface recordings. But even with these methods no change in the CRA boundary could be found. 3. Absence of tangential transrmssion of the evoked response Szent~igothai (1964) commented on the striking lack of horizontal elements in the cerebral

J.A. ENNEVER cortex : in layers I I - V I in the cat they spread only 100 ~m. Jacobson (1965) found evidence for the generalization that in the rat parietal cortex the neural elements in layers I IV tended to have axons that run vertically to deep cells. He found horizontal processes only in layer V. spreading about 2 m m Boundary changes of at least 0.5 m m would have been detected with the electrodes used here. but were never seen. It is concluded therefore that the neural elements that support the primary evoked responses do not transmit positive or negative components of activity synaptically, in a lateral direction, in the superficial layers of the cortex. Although most of the evidence for this comes from studying the small boundary evoked potentials, it appears to be true for the generators of the large potentials in the centre of the primary receiving area too, which were found to be almost completely independent of any activity greater than 2 mm away. This conclusion does not exclude the possibility of transmission either in the deeper cortical layers, the transmission of activity other than the primary evoked activity in the upper cortical layers, or even transmission of a component of primary evoked potentials which is eliminated by the urethane anaesthesia. However, Andersen et al. (1967b) showed barbiturate spindles recorded 2 m m or more apart in cat cortex to be clearly different. and their activity was not tangentially propagated. The synchrony of the spindles over wide areas of cortex was the result of synchronous activity, projected from the lateral thalamic nuclei (Andersen et al. 1967al. Waves were occasionally noticed to appear at consecutive electrodes, giving the impression of a 2 m/sec travelling wave, but such an event could not reliably be related to the peripheral stimulus. Another observation was that the absolute size of the CRA would gradually increase throughout the course of the experiment. Since the effect could never be produced by controlled manipulations of the cortex, the change is unlikely to be a cortical phenomenon The reason for the change is unknown: however, it might be a reflection of the strong thalamic control of cortical activity, suggested by Andersen et al. (1967a). The experimental conditions here are identical

291

INTRACORTICAL TRANSMISSION OF EVOKED POTENTIALS

to those of Holmes and Houchin (1966) (i.e., rat cerebral cortex, at the same depth of anaesthesia ---deeper level of Holmes and Houchin). They found that "synchronous with the periods of frequent (neural) discharge the ECoG exhibits rapid oscillations". These are the oscillations correlated in the tests for synchrony in this report, in order to discover if the ECoG synchrony at two points could account for their high correlations in evoked potential changes. Holmes and Houchin, however, described some of the factors which could affect the absolute values of these correlation coefficients, for example anaesthetic depth, anaesthetic type, electrode placement, and how these vary as a function of temporal displacement. It was found that peaks in the temporal displacement (or cross intensity) function could be accounted for by postulating a spreading influence of 0.1 m/sec. However, these peaks were only visible in the deeply anaesthetized preparation. There is, therefore, a lot of evidence for the spread of either spontaneous, non-specific, and abnormal activity in the cortex; however, the results here show that it is unlikely that the short latency evoked response spreads in an analogous manner, possibly because it represents activity in a different set of neural elements. It is not possible from the results obtained here to comment on the mechanisms giving rise to the shape of graph in Fig. 6. However, it is interesting that as a result of a study on spindle activity in cats Andersen and Andersson (1968) suggested that synchrony between areas more than about 0.8 mm apart (the diameter of a cortical column, Mountcastle 1957) is impressed on the cortical circuitry through synchronous activity in thalamic projections. In conclusion, it is possible to say that lateral trans-synaptic spread of activity may play a role in the normal functioning of the cerebral cortex ; but it is likely that any such transmission, if it does occur, is confined to the deeper layers of the cortex. The primary evoked potential, which predominantly reflects the activity of the superficial cortical elements, appears to have no component in this anaesthetized preparation, that could be transmitted laterally when the neighbouring network is facilitated by several manoeuvres that are known to alter excitability of

some neural elements. It is therefore concluded that this rigidity is the result of a limitation of the cortical network involved in this primary evoked potential. SUMMARY

The intracortical transmission of abnormally generated activity, like spreading depression and the direct cortical response, across the cortex is well known. In the experiments described here, the tangential transmission of a more normal type of cortical activity, the short latency response evoked by contralateral forepaw stimulation, was investigated. No evidence for tangential transmission was found from surface recordings. Various methods, such as passing polarizing currents, intravenous injection and topical application of strychnine, were used to alter the state of cortical excitation in order to induce a change in tangential transmission; none was found. The boundaries of the contralateral forepaw receiving area (CRA) appeared to be rigid, since the responses could not be induced beyond the perimeter by these manipulations of the cortex. Evoked responses in the centre of the CRA were shown to be unaffected by the excitation state of the cortex 2 mm or more away. The data support the hypothesis that the neural activity represented by the short latency surface evoked response is not transmitted laterally across the cortex and does not appear to be influenced by activity, beyond a range of about 2 mm or possibly less. The effectiveness of the methods used to alter cortical excitation, and to detect a spread of the evoked activity are discussed. RESUME TRANSMISSION INTRACORTICALE

DE

EVOQUI~ SPECIFIQUE AU TRAVERS

L'ACTIVITE DU

CORTEX

CEREBRAL

La transmission longitudinale au travers du cortex d'activit~s 6voqu6es ~ l'aide de moyens artificiels, telles la d6pression propag6e et la r6ponse corticale directe, est bien connue. On envisage ici la transmission tangentielle d'une activit6 corticale d'un type plus ordinaire, c'est/~-dire de la r6ponse/~ courte latence 6voqu6e par la stimulation du membre ant6rieur contralat6ral.

292

Aucune transmission tangentielle de cette activite n'a pu 6tre decel6e dans des enregistrements de surface. Diverses mhthodes (courants de polarisation, injection intraveineuse et application topique de strychnine) ont 6t6 utilishes en vue de changer l'excitabilit6 corticale et de modifier la transmission intracorticale mais sans succ+s. Les limites de l'aire de r6ception du membre ant6rieur (CRA) ont donc paru tr6s stables, puisque malgr6 ces manipulations du cortex il n'a pas 6t6 possible d'induire de r6ponses en dehors de cette aire. Les rhponses 6voquhes en son centre n'ont pas 6t6 affect~es par l'6tat &excitation du cortex situ6 ~ 2 mm plus loin. Ces r6sultats sugg6rent que l'activit6 nerveuse que repr6sente la r6ponse 6voqu6e ",icourte latence sur la surface, n'est ni transmise lat6ralement au travers du cortex, ni affect6e par l'excitabilit6 du cortex situ6 it une distance de 2 mm au moins. La discussion est centrde sur la sensibilit6 des methodes utilisbes pour modifier l'excitabilit6 corticale et pour d6tecter une transmission intracorticale de l'activit6 6voqu6e. The author wishes to thank Mr. Gary Harding of University of Washington, Seattle. U.S.A.. for the computer analysis, and Dr, O. C. J. Lippold for his help and advice. The majority of the work was carried out under a Medical Research Council (England) Scholarship and was presented in part for a Ph. D. Thesis.

REFERENCES ADRIAN, E. D. The spread of activity in the cerebral cortex. J. Physiol. (Lond.), 1936, 88:127 161. ANDERSEN, P. and ANDERSSON, S. A. Physioloyical basis O/ 1he alpha rhythm. Appleton C e n t u r y Crofts, New York, 1968. AND~iRSEN, P., ANDERSSDN, S. A. and l~Mo, "F. Some l'actors involved in the thalamic control of spontaneous barbiturate spindles. J. Ph.vsioL (Lond.), 1967a, 192: 257-281. ANDERSEN, P., ANDERSSON, S. A, and L~MO, T. Nature of thalamo-cortical relations during spontaneous barbiturate spindle activity, d. Physiol. fLond, i, 1967b. 192: 283 307. BJNDMAN, L. J.. LWPOLD, O. C. J. and R~DF~;ARN, J. W. "1 The relationship between the size and form of potentials evoked by sensory stimulation and the background electrical activity in the cerebral cortex of the rat. J, Physiol. (Lond.), 1964a, 171:1 25. BINDMAN. L. J., LIPPOLD, O. C. J. and REDFEARN, J. W. T. The action of brief polarizing current on the cerebral cortex of the rat (1) during current flow and (2) in the production of long lasting after-effects. J. Physiol.

J . A . ENNEVER

ILond.L 1964b, 172: 369- 382. BURNS, B. D. Some properties of isolated cerebral cortex in the unanaesthetized cat. J. Physiol. (Lond.). 1951, 112: 156 175. BURNS, B. D. The mammalian cerebral cortex. Edward Arnold Ltd.. London, 1958, BURNS, B. D. and SMITH, G. K. Transmission of information in the unanaesthetized cat's isolated forebrain. J. Physio/ ( Lond.), 19162, 164: 238-251. CHANG, H. T. Dendritic potentials of cortical neurones as produced by direct electrical stimulation of the cerebral cortex. J. NeurophysioL, 1950, 14: 1-21. COBB, W. A., COWAy. W. M., POWELL, T. P. S. and WRIGHT, M. K. Intracortical excitation following strychnine spikes. J. Phl'sioL (Lond.). 1955, 129:316 324. CREUTZFELI)I. 0. D.. FROMM, G. H. and KAPP, H: Influence of transcortical D-C currents on cortical neuronal activity. E.W. NeuroL, 1962, 5: 436~52. DUSSE~R DE BARENNIx J. G. and McCuI,LO('H, W. S. Functional organization of the somatosensory cortex of the monkey (Macaca mulatta). J. Neurophysiol.. 1938, I: 6 9 85. ENNEVER, J. A. l-t:oked eh'clricat aclil:it)' in the SODI¢llO sensory cortev q! the rat. Ph.D. Thesis. University of l_ondon. 1971. GRAI'STEJN. B. Mechanism of spreading cortical depression. J. Neurophysiol., 1956, 19:154 171. HOLMES, O. and Hotx H)N, J: Units in the cerebral cortex of the anaesthetized rat and the correlations between their discharges. J. Ph)'siol. (Loml.), 1966, 187:651 671. JACKSON, J. H. Selected writinqs qf John Huqhlings Jackson. l ol. 1: On ~TihT,sy and epilept(fi~rm convulsions. Basic Books Inc., New York. 1931. JACOBSDN, S. lntrataminar, interlaminar, callosal and thalamo-cortical connections in frontal and parietal areas of the albino rat cerebral cortex. J. comp. Neurol., 1%5, 124:131 146. KRIEG, W. J. S. Connections o! the cerebral corteA. Brain Books, Evanston, IlL 1963. L~/,o, A. A. P. Spreading depression of activity in the cerebral cortex. J. Neurophysiol., t944, 7 : 3 5 9 390. LL C~-L. Cortical intracellular potentials and their responses lo strychnine. J Neurophysiol., 1959, 22: 436-450. LmOUBAN-LI~:rot;ZE, S. Etude ~;lectrophysiologique des structures c&(;brales du rat b&nc. Doctoral Thesis, Paris, 1963. Mn, NI-:R. P. M. Note on a possible correspondence between lhe scotomas of migraine and spreading depression of Lefio. Electroenceph. olin. NeurophysioL, 1958, 10: 705. MOUNTCASTLE, V. B. Modality and topographic properties of single neurones of cat's somatic sensory cortex..1. Neurophysiol., 1957, 20:408-434. ()tits. S. and SI ZUKt, H. Transmission of direct cortical responses. Eh,ctroenceTh. olin. Neurophysiol., 1965, 19: 230 236. PENFIELI), W. a n d Bol DRI Y, [;. Somatic motor and sensory representation m the cerebral cortex of man as studied by electrical stimulation. Brain, 1937, 60: 389-443. PURI'tJRA, D. P. and McMt:R rRV, J. G. Intracellular activities and evoked potential changes during polarization of motor cortex, d Neurophysiol., 1903, 28:166:185, SMOI,IN, L. N. and SAMKO, N. N. Action of strychnine on

INTRACORTICAL TRANSMISSION OF EVOKED POTENTIALS the specific somatosensory projection pathway. Int. J. Neuropharmacol., 1968, 7:155 160. SZENT~GO'rHAI, J. The use of degeneration methods in investigation of short neuronal connections. Progr. Brain Res., 1964, 14:1 30. TowE, A. L. On the nature of the primary evoked response. Exp. Neurol., 1966, 15:113-139. Tow~, A. L. and MANN,~M. D. Effect of strychnine on the primary evoked response and on the corticofugal reflex

293

discharge. Exp. Neurol., 1973, 39: 395-413. VYSKO(~IL, F., K~I~, N. and BUREt, J. Potassium selective microelectrodes used for measuring the extracellular brain potassium during spreading depression and an~xlt depolarization in rats. Brain Res., 1972, 39: 255-259. WOOLSEY,C. N. Organization of somatic sensory and motor areas of the cerebral cortex. In H. F. I"IARLOWand C. N. WOOLSEY (Eds.), Biological and biochemical bases of Behavior. Wisconsin University Press, 1958.