Bicuculline epileptogenesis in the rat

Bruin Research, 323 ( I t,~84) 23t}- 246

239

Elsevier BRE 1041f,

Bicuculline Epileptogenesis in the Rat ANGELA M. CAMPBELL and O. HOI.MES h~stitute o/'Phys'ioh)gv. University qf Glasgo,'. Glad'co,' (;12 8QQ ( U. K. )

(Accepted March 27th. 19841 Key it'ords. focal intcrictal epileptiform discharges - - cerebral cortex - - penicillin - - bicucullinc --

GABA - - electrophorcsis ~ e p t h profile

diffusion - - rat

Bicuculline has been applied clectrophoretically from a fluid filled microelectrode at different depths within thc primary somatoscnsorv area of the cerebral cortex of rats anaesthetizcd with urethanc. The delay between onset of drug application at a constant rate and onset of spontaneous focal intcrictal epileptiform discharges (FIEDs), detected by a nearby recording microelectrodc, was least whcn bicuculline was applied at a depth of 0.65 mm below the pial surface. The subsequent frequency of FIEDs and their voltagc excursion were also greatest at this depth. The relationship between the delay of onset of epileptiform spiking and the depth of drug application was very similar to that previously determined for penicillin. This similarity of the sensitivity profiles suggests that the epileptogenic actions ol the two agents may be attributable to a common mechanism. At low concentrations, both agents specifically block GABAergic inhibitory synaptic transmission in brain tissue. This is likely to be the mechanism of their cpileptogenic effects. Other synaptic and non-synaptic mechanisms cannot, however, be ruled out because of the high concentrations which arc achieved locally whcn a chemical is applied from a point source,

I N T R O I ) UCTION

perficial or deep to the sensitive layer can be quantitatively a c c o u n t e d for by the time t a k e n for penicillin

The topical application of penicillin is widely used to induce paroxysmal h y p e r s y n c h r o n o u s electrical activity in the cerebral cortex of e x p e r i m e n t a l animals. In the a n a e s t h e t i z e d p r e p a r a t i o n , the h a l l m a r k of this activity is the focal interictal epileptiform discharge or FIED~L This a b n o r m a l i t y occurs in a limit-

to diffuse to this level a n d reach a threshold c o n c e n tration t h r o u g h o u t a critical mass of tissue-'5: F I E D s are then g e n e r a t e d there. Chart and E b e r s o l e 7 and Ebersole and Chart 12,13 have shown similar htminar sensitivity gradients to the e p i l e p t o g e n i c effects of penicillin and have identified layer 4 as the site of-

ed region of cortex a n d consists of a simple potential transient lasting 5 0 - 1 0 0 ms. It resembles interictal spiking recorded in the E E G of certain h u m a n epileptic patients.

maximal sensitivity in the striate cortex of cats. With larger e l e c t r o p h o r e t i c currents ( - 1 5 0 to - 3 0 0 n A ) , latent periods were shorter at all levels. T h e same layer r e m a i n e d the most sensitive but o t h e r lay-

In the p r i m a r y s o m a t o s e n s o r y cortical area of the anaesthetized rat, there are wide differences in the sensitivity of different cortical l a m i n a e to the epileptogenic effects of penicillin .s". W h e n penicillin is applied electrophoretically from a fluid-filled microelectrode, the latent period for p r o d u c t i o n of F I E D s is least when the electrode tip lies at a depth of 0.7 ram. With low penicillin fluxes (electrophoretic currents o f - 5 0 t o - 1 0 0 n A ) , the increase in latent period as the tip of the electrode is placed at a distance su-

ers (from layer 2 superficially to layer 5 at depth) g e n e r a t e d their own interictal spikes .~. This d e m o n strates that all these layers possess the m e c h a n i s m for spike g e n e r a t i o n , although the threshold for locallyg e n e r a t e d spikes is least at a depth of ().7 ram. A l t h o u g h penicillin at low c o n c e n t r a t i o n p r o b a b l y exerts its e p i l e p t o g e n i c effects by specific blockade of inhibitory G A B A e r g i c synaptic transmission jl,3e,~, at higher c o n c e n t r a t i o n s other effects of the agent may c o n t r i b u t e to epileptogenesis. For instance, the

Corre.~pondence: O. Hohnes, Institute of Physiology, University of Glasgow, Glasgow G 12 8QQ. U.K.

0006-8993/84/$03.00 .@ 1984 Elsevier Science Publishers B.V.

240 properties of the neuronal membrane are altered in this circumstance3.22. Bicuculline produces convulsions when administered systemically or when applied topically to the cerebral cortex 27,>. The changes in cerebral blood flow, arterial blood gas concentrations and extracellular ionic concentrations have been extensively studied 5,20.>. This agent is widely accepted as a more specific blocker than penicillin of GABAergic transmission 9.3~.42, although in cortical neurones, some authors have found bicuculline to be of doubtful value as a specific antagonist of GABA1S.23. If bicuculline application to neural tissue produces changes in activity, this is at least suggestive evidence that GABAergic transmission may be involved. Until an agent is discovered which specifically antagonizes the synaptic effects of G A B A , bicuculline remains the best chemical available for this purpose L,4~. In the present work, the cortical depth sensitivity profile for bicuculline epileptogenesis is investigated and a comparison with the penicillin profile is made. Rats were used because they are lissencephalic; this simplifies interpretation of plots of depth sensitivity 25. MATERIALS AND METHODS Sprague-Dawley albino rats weighing 180-220 g were anaesthetized with 25% urethane soluton administered intraperitoneally. The depth of anaesthesia was adjusted to be such that the animal exhibited a mild withdrawal reflex when a hind limb was pinched (urethane dosage around 1.9 g/kg). The trachea was cannulated, a craniotomy was performed, the dura reflected and the cortex covered with liquid paraffin at 37 °C. The head was held at 3 points: two metal rods with conical ends were pushed one into each external auditory meatus and the snout was gripped with a screw clamp. For recording, two electrodes were used. A fluidfilled glass micropipette (outside tip diameter 0.5-1 ~m, electrical resistance 0.5-2 Mffa) sampled activity at depth and a chlorided silver ball electrode, diameter 0.25 mm, sampled the surface electrocorticogram. Bicuculline was applied from a second fluid-filled glass microelectrode. Bicuculline solution was freshly prepared for each experiment. Immediately be-

fore the first drug application, / + ) bicuculline (Sigma) was weighed and added to a volume of distilled water to achieve a concentration of 5 raM. A few drops of glacial acetic acid were added to the water to promote the solution of the bicuculline; this brought the pH to about 4. The bicucuUine did not dissolve completely; 5 mM is an upper limit to the concentration of the bicuculline solution, The sensitivity profile of the cerebral cortex was independent of the order of challenging the different cortical layers and this demonstrates that bicuculline in the electrode remained active for the duration of the experiment. The bicuculline solution was introduced into a glass micropipette, outside tip diameter 3-6 .urn. The resistance, when an electrophoretic current of +350 nA was first applied, was in tile range of 5-10 M~2. This resistance often drifted either up towards 20 M ~ or down towards 2 Mff~ during the first minute or two of electrophoresis and thereafter remained fairly constant. At most cortical sites, bicucu[[ine electrophoresis rapidly induced epileptiform activity. Electophoretic application of acetic acid solution at the same concentration as that used to dissolve the bicuculline reduced no interictal epileptic spikes nor other detectable alteration of the on-going electrical activity of the cerebral cortex. Before it was lowered into the cortex, the bicuculline electrode was liberally washed in water to remove any' adhering bicuculline. A backing voltage was applied to the electrode before it made contact with the cortical surface, in ordeT ro minimize passive effiux of bicuculline. As soon as the electrode touched the cortical surface, the 'constant current' circuit which was attached to the bicuculline electrode applied a backing current o f - 100 nA. The electrode was advanced to lie closely adjacent to the recording electrode. The two electrode tips lay at the same depth, as indicated by the electrode drive vernier. The electrophoretic current of +350 nA was passed through the bicucultine electrode until interictal epileptic discharges appeared at the recording microelectrode. Bicuculline flux was terminated by reapplying the backing current and the bicuculline electrode was withdrawn from the cortex. The electrode was subsequently reinserted along the same track either to the same or to a different depth; in the latter case the depth of the recording microelectrode was

241

also adjusted. This procedure minimized unwanted passive efflux of bicuculline. AI least 1 h was allowed to elapse between applications of bicuculline. In different experiments the order of challenging different cortical depths was randomized in order to avoid any systematic effects attributable to bicuculline accumulation and possible cortical damage. All experiments were p e r f o r m e d on the primary somatosensory area of the cortex.

~j,,~/,,

O. 55ram 40

45

5(1 soc.

145 15(1~ec. Duration of bicuculline flu×

RESULTS

E~}'~,cts of bicuculline In rats anaesthetized with urethane, bicuculline was ejected electrophoretically from a fluid-filled micro-electrode the tip of which lay in the primary somatosensory area of the cerebral cortex. This resulted in the onset of focal interictal epileptiform discharges ( F I E D s ) detected by a nearby recording microelectrode. Fig. 1 shows a typical record of the first F I E D s produced by bicuculline electrophoresis; the F I E D s are prominent in the trace from the depth electrode (trace m a r k e d 0.7 mm) and are just detectable in the trace from the surface electrode. In both traces the F I E D s are p r e d o m i n a n t l y negative going. The latent period of the first F I E D was 24 s in this trial. A f t e r the end of a bicuculline trial, both the frequency and the amplitude of the interictal spiking declined but spiking seldom stopped completely. Subse-

Surface 0.7 nlm

Surface gs,, ~ ~ ~ , / ~ ¢ ~ f f ~ ,

~

lmV F

if5 J sec. 30 Duration of bicueuUine flux

Fig. I. Cortical field potentials measured simultaneously from a chlorided silver ball electrode lying on the cortical surface and a fluid filled glass microelectrode at a depth of 0.7 mm below the pial surface. Bicucu/line was injected from a separate fluid filled glass microelectrodc at a depth of 0.7 mm close to the recording microelectrode. This was the first trial in this rat. The lapse of time from onset of bicuculline electrophoresis is indicated. This acts as a time base. In this and the next figure, an upward deflection indicates that the active recording electrode has become positive to an indifferent electrode touching the skull far from the epileptic focus.

Fig. 2. A later trial in the same rat as Fig. 1, The challenged depth was 0.55 ram. The upper panel shows the first spike to stand out from background activity. Fhc lower panel shows a length of record after a gap of 85 s to illustrate the ew)lution of the waveform of the FIEDS.

quent applications of bicuculline had to be made despite the persistence of ' b a c k g r o u n d spiking. This complicated the interpretation of the results as exemplified by Fig. 2. Fig. 2 shows the results of challenging a superficial cortical layer later in the same experiment from which Fig. 1 was derived. The bicuculline flux was the same. ' B a c k g r o u n d ' spiking is present. The first spike to stand out from the background noise and spiking is indicated by the arrow; it is much less prominent than the first spike of Fig. 1. The reasons for accepting it as the first spike of the trial in progress rather than as due to residual bicuculline spiking are described in the next section. The lower panel in Fig. 2 shows that, although the voltage of the F I E D s increased as bicuculline electrophoresis continued, their amplitude at this cortical depth never reached the magnitude of those al the (l.7 mm depth of Fig. 1. "The ong-going electrical activity of the cortex is different in Figs. 1 and 2. This is known to influence the probability of occurrence of FIEDs2. These and other differences related to the non-stationarity of the animal p r e p a r a t i o n were c o m p e n s a t e d by randomizing the o r d e r in which different depths of the cortex were challenged in different experiments. It turned out that the spatial sensitivity gradients to the epileptogenic effects of bicuculline were d o m i n a n t : other variables such as the n u m b e r of previous trials and nature of on-going cortical activity and possible

242 Depth

cortical depths in o n e rat are plotted. Each circle rep-

0.55 mm

, •

,b!,+%~o**+,,~ ** *, +,,,+,

5 0.7 mm

+

%

°

?

p e r i m e n t (0.8 m m d e p t h ) , there is no b a c k g r o u n d

i ° +~

i.

i?

spiking so the identification of the first spike ( s h o w n by the arrow) is u n e q u i v o c a l . At the o t h e r e x t r e m e ,

0.8 IBm

-20

-10

resents a spike; its vertical height above the time axis is the spike's a b s o l u t e voltage. Z e r o time indicates the onset of bicuculline flux. In the first trial of the ex-

/

1-

'op ~ 10

,1

~0

50

DUFation of

a@

50

at the (t.55 m m d e p t h there is o n g o i n g spiking due to

• fi@

bicucull|ne

78

8(}

qo

!00 10~,

f l u x in sec.

Fig. 3. Plot of the absolute voltage of FIEDs (ordinate) as a function of time (abscissa). Zero on the time scale indicates the onset of the constant bicuculline flux. The results from 3 trials in one rat are shown; the depths of bicuculline and the recording microelectrode are indicated. At the 0.7 mm depth, the bicuculline flux was discontinued after 58 s and hence the truncation of this graph at this time. The arrows show the FIEDs judged to be the first due to the trial in progress; see text for details.

changes in the biological activity of the bicuculline solution e x e r t e d n o d e m o n s t r a b l e effects in these exp e r i m e n t a l circumstances.

previous bicuculline trials. T h e first spike to stand out from b a c k g r o u n d was selected as the first of the trial in progress: it is again s h o w n by the arrow. The spikes following it were larger t h a n those p r e c e d i n g it. Also, the f r e q u e n c y of spiking s h o w e d an a b r u p t increase at (or, in the case illustrated, just before) the arrow. T h e s e changes are m o r e o b v i o u s in Fig, 4 in which the c u m u l a t i v e spike voltage is plotted. At the 0.55 m m d e p t h , the b a c k g r o u n d spiking results in a sloping baseline. T h e r e is an a b r u p t increase in this slope at 52 s which clearly r e p r e s e n t s an increase in the a m o u n t of spiking. This increase in slope is very

240 -

The identification of the initial FlED O u r selection of the first spike to stand out from b a c k g r o u n d activity was subjective a n d the decision was s o m e t i m e s difficult. T h e plots s h o w n in Figs. 3 and 4 were m a d e to c o n f i r m the validity of spike selection. T h e results are typical of the series of e x p e r i m e n t s . In Fig. 3 the spikes o c c u r r i n g in 3 trials at different

o.,o

1+1 T6,

o. yj

200

160

+

¢

12o

8O

%"./

40 1!

0

i

-20 -10

I i l l l l ' ~ l l

0

I

0

°'~ I

10

20

30

40

~f)

60

70

8C)

9C 100 110

Duration of blcucull|ne flux In sec.

Fig. 4. Cumulative spike voltage plot as a function of time after the onset of bicuculline flux. The same 3 trials are shown as in Fig. 3. a indicates the delayed increase in slope at 0.8 mm referred to in the text.

0.2

0.4

0.6 Depth (mm)

0.8

1.0

Fig. 5. The symbols are pooled from 4 rats. Abscissa. perpendicular depth beneath pial surface Ordinate. delay of onset ot interictal spiking elicited by bicucuUine. Separate trials from a single experiment are plotted with one symbol. The calculated curve is the solution to the diffusion equation for a unit point source lying a, he depth indicated by the abscissal axis. The ordinate, shows the time taken for the concentration of bicucultine to reach a threshold at the circumference of a disc. radius 0.16 mm, parallel with the cortical surface and centred on the line through the source and perpendicular to the cortical surface, Depth of sensitive stratum = 0.65 mm.

243 close to the spike selected subjectively as being the first induced by the trial in progress.

The laminar sensitivity prq~'le The relationship between the latent period for the production of FIEDs and the depth of application of bicueulline is shown in Fig. 5. The results from 4 rats are superimposed, one symbol being used for the trials on one rat. In all these experiments the electrophoretic current was +35(I hA. There are steep gradients in the laminar sensitivity of the cerebral cortex to the epileptogenic effects of bicueulline. The most sensitive depth is 0.7 mm. Since Fig. 5 is so similar to the plot obtained for penicillin (ref. 25, Fig. 1), the same curve-fitting procedure was adopted for the smooth curve in Fig. 5 as in the previous work. The curve is derived from the diffusion equation with the assumptions that, in order to induce FIEDs, bicuculline must diffuse to a sensitive lamina in the cortex and reach a threshold concentration throughout a critical mass of cortical tissue. Evidence for this latter assumption comes from the work of Reichenthal and Hocherman 40 and Gabor and Scobey 1~. The smooth curve in fig. 5 is the solution to the diffusion equation in a form appropriate to the exposed cerebral cortex 2s. The point estimators and 95(>f confidence intervals for the parameters in the diffusion equation were calculated by a least squares best fit procedure described by Holmes and Lockton zs. They were respectively: sensitive depth 0.65 mm (0.63, 0.67): radius of critical mass 0.16 mm (0.15, 0.17): and diffusion coefficient 3 x 1 0 -6 cm-2/s (1×1(I -~, 15x 11)-6). A 'goodness of fit' test indicates that there are no consistent deviations between the experimental points and the calculated curve. DISCUSSION

GA BA ergic blockade There is evidence that, at low concentrations, penicillin specifically blocks synaptic transmission by the inhibitory neurotransmitter GABA-~2,3-L At higher concentrations, penicillin exerts other actions, such as blocking inhibitory synapses where neurotransmitters other than G A B A act and exerting non-synaptic actions3.2~..~3•34,~:. An accumulating body of evidence suggests that G A B A e r g i c blockade is the basis of the

epileptogenic action of penicillin ¢',u~.a-L When a chemical is ejected from a point source, regions of brain at different distances from the source are subjected to widely differing concentrations of the chemical. So electrophoretic application of penicillin 25,s¢~,sl may produce mixed effects. This seemed unlikely, since Macon and King ~4,3s have shown that, with electrophoretic currents below 71t0 nA, penicillin seems to act specifically by G A B A - r e c e p t o r antagonism. Only with application rates in excess of this did the specificity of G A B A e r g i c blockade decline, with penicillin also antagonizing the inhibitory action of the neurotransmitters glycine and alanine, probably by blocking the chloride ionophore and thus interfering with a common step in the inhibitory mechanisms of the different inhibitory synapsesS,~. Nevertheless, it seemed worth checking whether the sensitivity profile for penicillin epileptogenesis would be matched by the profile for another G A B A blocker. The present investigation consisted in repeating the measurement of laminar sensitivitv with bicuculline, which is widely accepted as a more specific blocker than penicillin of G A B A e r g i c transmission 9,3s,42. The most striking result is that, with bicuculline, the sensitivity profile is the same as with penicillin. Even bicuculline has non-specific effects at high concentration I ~,z-~-ssso this evidence is not proof that GABAergic blockade is thc basis of the epileptogenesis. Nevertheless, the close similari'tv between the laminar sensitivity profiles suggests that the two agents produce their effects through a common mechanism and the most likely candidate for this mechanism is the blockade of G A B A e r g i c synaptic transmission.

Dijfusion in the cerebral cortex Diffusion of drugs in the cortex has long been known to play an important role in determining which layers of cortex are influenced. Bindman et al. 4 reported that G A B A applied to the surface of the cortex inhibited electrical activity and that deeper laminae were successively inlfuenced as the drug diffused and reached them• Noebels and Pedley 3¢, applied penicillin to the cortical surface and measured the penetration of the agent at a time when its epileptogenic effects appeared• Gabor and Scobeyl~, and Ebersole and Chatt 12 applied penicillin locally into cat's striate cortex to localize the effects of the them-

244 ical. Holmes and Lockton 25 estimated the concentration field produced by such electrophoretic application in their study of the laminar sensitivity profile in the somatosensory cortex of the rat. This estimation of penicillin concentration was made possible by the expedient of dissolving Sodium Penicillin G in distilled water so that the carriage of the electrophoretic current by ions other than sodium and penicillin was minimized; the penicillin concentration could then be estimated from the magnitude of the electrophoretic current. With bicuculline it was necessary to add acetic acid to promote its solution. Since we could not estimate the proportion of electrophoretic current carried by the various ions present in the bicuculline electrode, we could make no estimate of bicuculline concentrations in the brain tissues. Diffusion places a limit on the precision with which the site of action of an epileptogenic agent can be ascertained with the techniques of the present work. Nevertheless there is a depth which shows maximal sensitivity to the epileptogenic effects of bicuculline and this is the same (within experimental error) as the sensitive layer in penicillin epileptogenesis. A similar determination has been made in vitro in visual cortex 17.

Voltage of F1EDs Our records were from a surface electrode and a single depth microelectrode and so we cannot localize the generators of the FIEDs, as would be possible with the techniques of Gumnit ~9, Chatt and Ebersole 7.s and Ebersole and Chatt/3. We cannot, for instance, discriminate amongst different possible explanations for the observation that the FIEDs were of greatest amplitude when recorded from the layer in which they appeared with shortest delay. This may be due to the extracellular current field being greatest when epilepsy is induced by bicuculline applied to that level. It could equally be due to the fact that when bicuculline is applied to an insensitive layer, the agent initiates spiking in a distant sensitive layer. If this is the case the amplitude of recorded spikes is merely a reflection of the different relative position of the recording electrode and the current generator field. We propose to discriminate between these alternatives by current density analysis.

The initial increase in slope o f the accumulated voltage plot (Fig. 4) is unambiguously interpreted as the onset of epileptic spiking attributable to the trial in progress. In many trials this initial increase in slope was followed by later increase (see for instance point st, Fig. 4). This was a reflection of an increase in the amplitude of individual FIEDs rather than an increase in frequency. This may be due to recruitment of cortical columns adjacent to the column initially affected25. Alternatively, there may be recruitment of neuronal elements in layers of the cortex neighbouring the most sensitive layer as the focus evolves. Such a recruitment has been observed in penicillin epileptogenesis by Holmes and Lockton 24 and Lockton and Holmes 3j.

Substrate for epileptogenesis In striate cortex, Chatt and Ebersole 7 and George and Connors 17 have identified layer 4 as the site of maximal sensitivity in penicillin epileptogenesis. In primary sensory cortex, this layer is prominent and is the site of termination of the majority of the specific thalamic input. The laminar density distribution of specific thalamocorticat terminals in the sensory cortex of rats 26 is closely matched by the sensitivity profile of the cortex for penicillin epileptogenesis 31 and bicucuiline epileptogenesis, as shown in the present work. This mainly excitatory input must presumably be controlled by inhibitory mechanisms, probably GABAergic. Endogenous G A B A concentrations are high in this layer of cortex ~4. The conclusion of the present work is that the primary somatosensory cortex exhibits essentially the same depth sensitivity profiles to penicillin and bicuculline. These two agents probably generate epileptiform discharges by a common mechanism, it is likely that blockade of GABAergic transmission is involved since both agents are specific GABA blockers at low concentration. ACKNOWLEDGEMENTS We acknowledge generous support from the Scottish Epeilepsy Association, the We!lcome Trust and the Physiological Society's Dale Fund. We thank Dr. M. N. Wallace for helpful assistance and discussion.

245 REFERENCES I Andrews, D. R. and Johnston, O. A, R.. G A B A agonists and antagonists, Biochem. Pharmacol,, 28 (1979) 2697-27O2. 2 Atsev, E. M., Dynamics of electrical manifestations of brain in cats and rabbits with an experimental epileptogenic focus. In Z. Servit l e d . ) ('ornparative and Cellular Pathophwiologv of Epilepsy, Proceedings of Symposium in Lib/ice, Excerpta Medica, Amsterdam, 1966, pp, 221-234. 3 Ayala. G. F., Lin, S. and Vasconetto, C., Penicillin as an epileptogenic agent: its effect on an isolated neuron, Scien(e, 167 11970) 1257-1260. 4 Bindman, Lynn, J.. Lippold, O. C. J, and Redfearn. J, W. T., ]'he non-selective blocking action of gamma-amino butyric acid on the sensory cerebral cortex of the rat, J. Phv,~iol. (Lond.), 162 (1%2) 105-12/). 5 Chapman, A. G.. Meldrum, B. S. and Siesjo, B. K,, Cerebral metabolic changes during prolonged epileptic seizures in rats. J. Neurochem., 28 (1977) 1025-1036. 6 Clarke, G. and ltill, R. ( . , Effccts of focal penicillin lesion on response ol rabbit cortical neurones to putative transmirrors, Brit. J. Pharmacol., 44 (19721 435-441. 7 ('hart, A. B. and Ebersole, J. S., The laminar sensitivity oi cat slriatc cortex to penicillin induced epileptogencsis, Brain Research, 241 11982) 2~2-287. 8 Chart, A. B. and Ebersole. J. S., The differential laminar effects of phenytoin during penicillin-induced epileptogenesis. Epilepsia. 24 11982) 25~. 9 Curtis. D. R., Duggan, A. W., Felix, D. and Johnston, G. A. R., (}ABA. bicuculline and central inhibition, Nature (l, ond. ), 226 (1970) 1222-1224. 10 Curtis, D. R., Game, C. J. A., Johnston, G. A. R., McCulloch, R. M, and Mcl,achlan, R. M., Convulsive action of penicillin. Brain Re.search, 43 119721 242-245. 11 Dichter, M. A., Physiological identification of G A B A as an inhibitory transmission for mammalian cortical neurones in cell culturc, Brain Research, 1911(198/)) 111 - 121. 12 Ebersole, J. S. and Chatt, A. B,, The laminar susceptibility o! cat visual cortex to penicillin-induced cpilcptogenesis, Neurology, 30 (1980) 355P. 13 Ebcrsolc, J. S. and Chatt, A. B., Cortical laminar epileptic foci from penicillin microinjection: differential sensitivity, interdepcndencc and polarity changes with development. Neurology, 31 (1981) 52. I4 Emson, P. C. and Lindwall, O., Distribution of putative neuroiransmittors in the neocortex, Neuroscienee, 4 11979) 1-3O.

15 Elger, (7. E. and Speekmann, E. J., Focal interictal epileptiform discharges (FIED) in the epicortical E E G and their relation to spinal field potentials in the rat, Eleetroenceph. olin. Yeurophysiol,, 37 ( 1974 ) 1310-1327. 16 Gabor, A. J. and Scobey, R. P,, Epileptogenicity and the cortical column, Electroenceph. clin. Neurophysiol.. 43 ( I C177) 461 - 462. 17 George, (7. P. and Connors. B. W,, Initiation of synchronized bursting in neocortex, Soc. Neurosci. Abstr., 9 ( 1983 ) 396. 18 Godfrained, .l.M., Krnjevic, K. and Pumain, R., Doubtful value of bicucullinc as a specific antagonist of G A B A , Nature (Lond.), 298 ( 19701 675-676. 19 Gumnit. R. J., Field potentials in partial s e i z u r e s - - a reappraisal. In E. J. Speckmann and H. Caspers (Eds.), Origin ~[ ('erebral lqeld Potential, s, Thieme, MA, 1979, pp.

183-194. 20 Gutnick, M. J., Heinemann, U. and Lux. H. D., Stimulus induced and seizure related changes in extr:tcellular potassium concentration in cat thalamus (VP[,), Electroenceph. elin. Neurophysiol., 47 (1979) 329-344. 21 Heyer, E. J.. Nowak, L. M. and Macdonald, R. L., Bicuculline: a convulsant with synaptic and non-synaptic actions, Neuroh,,gy, 31 ( 1981 ) 1381 - 1390. 22 Heycr, E. J.. Nowak, L. M. and Macdonald, R. L., Membrane depolarisation and prohmgation ol calcium-dependent action potentials of mouse neurones m cell culture b,, two convulsants: bicuculline and penicillin. Brain Research, 232 11982) 41-56. 23 Hill, R. G., Simmonds, M. A. and Straughan. P. W,, A comparative study of some convulsant substances as gamma-aminobutyric acid antagonists in the fclinc cerebral cortex, Brit. J. Pharmacol.. 49 ( 1973 ) 37-51. 24 Holmes, O. and Lockton, J. W.. Migration ol thc generators of interictal spikes in penicillin epilepsy, J. Ph.sviol. (Lond.), 319 119811S2P. 25 Holmes, O. and Lockton, J. W., Penicillin epileptogenesis m the rat: diffusion arid the differential laminar sensitivity of the cortex ccrebri, Brain Research. 231 (1982) 131 - 141. 26 Jacobson, S. and Trojanowski, J. Q., Corticothalamic neurons and thalamocortical terminal ficlds: an investigation in rat using horseradish peroxidase and autoradingraphy, Brain Research, 85 ( 1975 ) 3 g5- 4111. 27 Kendall, D. A., Fox. D. A. and Enna. S. J., Effect of gamma-vinyl G A B A on bicuculline induced seizures, Neuropharmacology, 211 ( 19g 1) 351- 355. 28 Kuschinsky, W. and Wahl. M., Local chemical and neurogenic regulation of cerebral vascular resistance. Physiol. Rev., 58 (197g) 656-089. 29 Levin, V. A., Fenstermacher. J. D. and Patlak, (7. S.. Sucrose and insulin space mcasurements of cerebral cortex in 4 mammalian species, Amer. J. l'hyviol., 219 (19701 1528-1533. 3t1 Lockton, J. W. and Holmes, O., Site of penicillin-induced epilepsy in the cortex cerebri of the rat. Brain Researck, 190 (19g0) 301-304. 31 Lockton, J. W. and Holmes, O., Penicillin epilepsy in the rat: the response of different layers of thc cnrtcx ccrebri. Brain Research, 258 (1983) 79-89. 32 MacDonald, R. L. and Barker, J., Pcntvlencterazol and penicillin are selective antagonists of G A B A - m e d i a t e d post-synaptic inhibition in cultured mammalian neurones, Nature (Lond.), 267 ( 19771 72/)-721. 33 Macdonald, R. L. and Barker, J. L., Specific antagonism of GABA-mediatcd post-synaptie inhibition in cultured spinal cord neurons: a common mode of convulsant actJon, Nellrolow, 28 (1978) 325-330. 34 Macon, J. B. and King, D. W., Penicillin iontophorcsis and the response of somatoscnsory cortical neurons to amino acids, Electroenceph. olin. Neurophysiol., 47119761 52-63. 35 Macon, J. B. and King, D. W., Responses of somatosensory cortical ncurones to inhibitory amino acids during topical and iontophoretic application of cpileptogenic agents, Electroenceph, olin. Neurophysiol., 47 ( 1979 ) 41 - 51. 36 Noebels, J. L. and Pedley, T., Anatomical localisation of topically applied (14C) penicillin during experimental focal epilepsy in cat ncocortex, Brain Research, 125 119771 293-303. 37 Noebels. J. L. and Prince, D. A., Presynaptic origin of penicillin afterdischarges at mammilian nerve terminals. Brain

246 Research, 138 (1977) 59-74, 38 Nowak, L. M., Young, A. B. and Macdonald, R. L., GABA and bicucultine actions on mouse spinal cord and cortical neurons in cell culture, Brain Research, 244 (1982) 155-164. 39 Pong, S. F. and Graham, L. T., N-methyl bicuculline, a convulsant more potent than bicuculline, Brain Research, 42 (1972) 486-490. 40 Reichenthal, E. and Hocherman, S., The critical cortical area for development of penicillin-induced epilepsy, Electroenceph, clin. Neurophysiol., 42 (1977) 248-251.

41 Roberts. E. and Hammerschlag, R.. Amino acids transmitters. In G. J. Siegel. R. W Albers. R. Katxman and B. W Agranoff (Eds./, Basic Neurochernistrv, 2nd edn. Little. Brown, Boston. MA. 1976. pp. 218-245. 42 Sillito, A. M.. Inhibitory mechanisms influencing complex cell orientation selectivity and their modification at high resting discharge rates. J. PhvsioL ¢Lond. J. 284 (19791 33-53. 43 Van Duijn, H.. Schwartzkroin. P A. and Prince. D. A.. Action of penicillin on inhibitory processes in the cats cortex, Brain Research. 53 ( 1973/47t)-476.