- Email: [email protected]
Dynamic interactions of evoked potentials in a polysensory cortex of the cat
0306-4522/8453.00+ 0.00 Pergamon Press Ltd
NeuroscienceVol. 13, No. 3, PP. 645-652, 1984 Printed in Great Britain
IBRO
DYNAMIC INTERACTIONS OF EVOKED POTENTIALS IN A POLYSENSORY CORTEX OF THE CAT J. TOLDI, 0. FE&R and L. FEUER* Department of Comparative Physiology, Jbzsef Attila University, Szeged, Hungary, POB 533, H-6701. *Chinoin Pharmaceutical and Chemical Works Ltd., Budapest, T6 u. l., Hungary, H-1325 Abstract-Interactions of acoustic and somatosensory evoked potentials were studied in the anterior suprasylvian gyrus of the cat. Two kinds of interaction could be observed: occlusion or facilitation. In most cases occlusion was observed. The interactions showed dynamic changes and were susceptible to different kinds of influences. After having determined the control values of interaction over a period of several minutes, reversible enhancement of occlusion was observed after synchronous activation of the acoustic and somatosensory inputs with 2Hz frequency. The same effect could be observed during
stimulation of the mesencephalic reticular formation with 200 Hz frequency. The interactions could also be influenced by amphetamine and y-glutamyltaurine, known as drugs capable of influencing the arousal level of the brain. During treatment with amphetamine the interaction was shifted in the facilitatory direction. The antagonists of amphetamine (haloperidol and reserpine) prevented this effect. The authors suggest that the interactions of acoustic and somatosensory responses are mediated by interneurons (inhibitory and/or excitatory) and particular stimulus situations and drugs are able to modify the equilibrium between excitatory and inhibitory subsystems.
The examination of the interactions of evoked potentials began with the experiments of Bremer,2,3 some 30 years ago. This field of research proved to be very fruitful because it gave insight into problems such as the organization of polysensory structures. In studying the interactions of acoustic and somatosensory evoked potentials Berman has shown that there exists a considerable overlapping of these areas in the anterior ectosylvian gyrus of the cat.’ In this experimental paradigm, two stimuli were given belonging to two different modalities, with various time intervals between them, and the action of the first stimulus upon the potential evoked by the second one was studied. A quotient of the calculated and measured amplitude of this potential expressed the extent of interaction. This might be more or less than one and if it equalled one, it was taken as a lack of interaction. Unfortunately the potential usefulness of this paradigm was not recognized for a long time. In our earlier experiments” Berman’s paradigm was applied with several modifications. The time sequence of two stimuli of different modalities was adjusted so as to achieve a complete synchronism of the two potentials evoked by them, what meant an overlapping of the positive and negative phases, respectively. The experiments were performed on the anterior suprasylvian gyrus of the cat anaesthetized with barbiturate and in some cases with chloralose. The comparison of amplitudes obtained with separate and compound stimulations gave an opportunity to judge the type and extent of interaction (occlusion and facilitation). In this paper we give an account of the influences which are able to modify these interASG, anterior suprasylvian gyrus; MFR, mesencephalic reticular formation.
Abbreviations:
actions between acoustic and somatosensory evoked potentials in the anterior suprasylvian gyrus (ASG) of the cat, and an attempt will be made to draw conclusions as to their mechanism and the origin of their dynamic changes. EXPERIMENTALPROCEDURES The experiments were performed on adult cats of both sexes, anaesthetized with pentobarbital (Nembutal; 40mg/kg, i.p.). After venous and tracheal cannulation, the left cerebral hemisphere was exposed and covered with mineral oil. The experiment began after 2 h rest. Acoustic click stimuli were applied to the right ear via a miniature earphone, connected to a square wave stimulator. The stimuli had a duration of 1 ms and were always supramaximal. Somatosensory evoked potentials were obtained by stimulating the right forepaw with electric pulses of 3-8 V and 0.3-0.5 ms in duration, which also proved to be supramaximal. Mesencephalic reticular formation (MFR) was stimulated with bipolar needle electrodes according to the stereotaxic co-ordinates given by Jasper and Ajmone Marsan’ (FR: 3; L: 3.5; V: - 1). At the end of the experiments electrolytic lesions were placed at the tips of the bipolar electrodes and localized by standard histological procedures. The evoked potentials were recorded from the surface of cortex with ball-tipped silver wire electrodes attached to a two channel amplifier system with a time constant of 0.5 s. Blocks of 20 or 50 potentials were averaged with a Motorola MC-6800 computer and drawn using an X-Y plotter. Comparison of results taken from different series of stimulation was performed by using Student’s double 645
.I. Toldi et al.
646
t test and in some cases correlation were calculated.
of stimulus pairs as with separate stimulation. The extent of interaction (I) was calculated according to the formula presented in Fig. 1. The amplitude obtained with compound stimulation (given by arbitrary units) is divided by the sum of the separately recorded potentials. From this quotient a percentage value was calculated which expressed the extent of interaction independently of voltage values. In the experiments three types of interaction occurred: (1) algebraic summation, which meant a lack of interaction with I = 0; (2) occlusion with I < 0 and (3) facilitation at I > 0. First we examined whether the basic levels of interactions show any correlation with one another in different polysensory areas. One example of a comparison between the posterior middle suprasylvian association area and the anterior lateral association area is presented in Fig. 2. The records taken for as long as 30 min show that the level of interaction varied around zero with low occlusion and facilitation in the anterior lateral association area. In the posterior middle suprasylvian association area a rather high level of occlusion was present, at around 30”/,. Although the variability was somewhat greater than in the anterior lateral association area, it held the overall tendency throughout the recording time. Among minute variations of occlusion no correlation could be found between the two recorded sites.
coefficients (r)
Materials
Drugs were obtained from Hungarian and foreign pharmaceutical companies; amphetamine and y-glutamyl-taurine (LitoralonR) from Chinoin Ltd; haloperidol and reserpine from Kiibanyai Gyogyszerartigyar; and Propranolol from Coch-Light. RESULTS
Types of interactions of acoustic and somatosensory evoked potentials
Acoustic and somatosensory stimulation elicited potentials of 200-5OOpV amplitude in the ASG. Their latency according to the earlier findings’,” was near to those obtainable in the primary sensory areas. With acoustic evoked potentials it ranged from 9 to 11 ms and was 8-10 ms with somatosensory potentials. Actual differences in latency were taken into account and compound stimulus pairs were given with an interval which ensured complete overlapping of the respective positive and negative phases. First, a block of acoustic potentials was recorded with 1 Hz frequency, then the same was done with somatosensory potentials. Afterwards compound stimulation followed with the same frequency and number
A
AC
ss
/2 ‘1. ;
z
:L 1 9 r_
0
i
I
42
B
AC
37
62 I= 1--l).lOO% 42*37 I
=-21,51% (0ccIUsron)
2oiiisec
62
ss
I* (7343*27 I
43 Fig. 1. Types of interactton
27
-,).l~“ib
* 4,28”k (facthtatton)
73
between acoustic (AC) and somatosensory (Ss) evoked potentials. (A) Occlusion; amplitude of the compound response (62) is lower than the algebraic sum of the separately recorded potentials (42 + 37). According to the formula this means an occlusion of 2 I .5’% (B) Facilitation; the compound response is higher than the sum of separately recorded potentials. Calculation is the same as in case of occlusion.
647
Dynamic interactions of evoked potentials
Fig. 2. Interaction of acoustic and somatosensory evoked potentials in the anterior lateral association area (ALA) and posterior middle suprasylvian association area (PMSA) as recorded simultaneously over a 30min period. In ALA interactions are very weak, while in PMSA occlusion is dominant. The same holds for another pair of loci in the ASG under resting conditions (Fig. 3). The correlation coefficient between their occlusion values was 0.075. However, as MFR was stimulated with 200 Hz frequency for 33 min, the occlusion of acoustic and somatosensory potentials suddenly increased at both sites. After the stimulation period the occlusion remained high at point 2, while at point 1 it resumed its initial level. The parallel of the MFR action at these loci is obvious and the correlation coefficient is
M-7,6*5.9
c*
orxhJsmol
MS 6,8'21
c,c,
r- 0,075
rather high (r = 0.744). Thus it appears that each point of ASG exhibits a particular level of occlusion, apparently independent of other ones, but factors modifying the level of interactions manifest themselves uniformly. Modljkation
M-I43’6,O
VW
p
pea,,
stimulus trains
After having established the level of interaction (mostly occlusion) at particular points of ASG, 2 Hz compound stimulation was given for 2 min (Fig. 4C).
M-32.6*7,9
9
of interactions following
MFR2(zoo~)
b
G
M-2X0’ 6.0 NFR, m,
r*o,#4
@ 3min
Fig. 3. The effect of MFR stimulation on the occlusion level at two (1,2) points of ASG, denoted in insert top right. In the control period (C) no correlation was present between the two sites in view of interactions. Stimulation of MFR, however, acted upon them in the same manner, reflected also by the correlation coefficient, r = 0.744. Enhanced occlusion persisted in point 2. Stimulation sites in the MFR are denoted in the scheme bottom right. g, gyrus; M, mean.
J. Toldi et al.
648
i b
mlLficx1~.
I
AC
n.4 n-3
'th1113.53 M-3.33t2.35
h+QPO
M-8.7%X3
M.lO.ma47M=~istOa2
M-18.66a94M-lwa5 p
i=o.OOl
Fig. 4. The effects of 2 Hz conditioning (c) stimulation on the interactions of evoked potentials. (A) Occlusion levels before and after c as measured in 2 min periods. (B) Mean occlusion level with SD before
and after c. The difference is highly significant (P < 0.001). (C) Diagrams showing the relation between evoked response amplitudes (abscissa) and occlusion levels (ordinates) in all the 2 min period presented in (A). (D) Acoustic (AC) and somatosensory (Ss) potentials remained practically unchanged throughout the recording period, while occlusion levels showed a rather wide variance. The compound responses decreased significantly after c and this led to elevation of occlusion level presented in (B). M, mean.
The changes subsequent to it were characterized by temporary elevation of occlusion as measured during the next 1 Hz stimulation period. After the 2 Hz stimulus train a 2min rest was interposed. The maximal duration of such reversible changes was 16min. The course of experiments of this type are illustrated in Fig. 4 and an insight into the interrelations between amplitude of evoked potentials (at separate and compound stimulations) and occlusion values is given. Occlusion values are diagrammed as recorded in 6min control period at 1 Hz stimulus frequency (Fig. 4A). In the first 2 min sampling period there was 0% occlusion; in the two subsequent ones 5% occlusion was obtained. The average value of this control period is presented in Fig. 4B (first bar) with SD (3.33 f 2.35). After the conditioning stimulus series (c) occlusion became much more intensive over an 8 min period and although it showed large fluctuations, its average differed from the control values significantly (mean: 11 f 3.53, P < 0.001) (Fig. 4B, second bar). In Fig. 4C variations of acoustic and somatosensory evoked potential amplitudes are presented as recorded with separate and compound stimulations and it can be judged how they were related to the actual level of occlusion. It is obvious that, at separate stimulation, acoustic and somatosensory potentials show minute changes even after 2 min of 2 Hz conditioning stimulation. The compound response, however, decreased significantly from 18.66 f 0.5 and this led to elevation of the occlusion level (Fig. 4C and D, last diagrams). One is led to conclude that the amplitude
of evoked potentials and the extent of their occlusion are determined by different factors and show independence from each other within very wide limits. In a few cases conditioning stimulation (also 2 Hz) resulted in transitory facilitation but its extent also proved to be independent of the evoked potential amplitude (Fig. 5). In another series of experiments conditioning stimulations of 5 and 10 Hz were also applied but they failed to exert any action upon occlusion level. The dependence frequency
of occlusion on the stimulution
Another way of exploring the dependence of interaction on the stimulation rate was chosen, when separate and compound stimulation was made with 3 Hz frequency (Fig. 6). First the occlusion level was determined with 1 Hz stimulation over a period of 45 min. Then the frequency was raised to 3 HZ at both separate and compound stimulations and the occlusion level was measured for 30 min. As shown in Fig. 6, this led to an abrupt elevation of occlusion all over the period of a higher stimulation rate. After 30min break the occlusion levels returned to the original one with far less variability than before. Pharmacological injuences upon interactions acoustic and somatosensory potentials
of
Since interactions of evoked potentials seemed to have something to do with equilibrium of excitatory and inhibitory processes, the effects of some drugs,
649
Dynamic interactions of evoked potentials
C
Table 1. Effect of amphetamine on the occlusion of acoustic and somatosensory evoked potentials
IhC AC -P
SSA
r ?r 3
D
min break
l/WC
ss* 5
Amphetamine (mg/kg, i.v.)
Decrease of occlusion (“/,)
Correlation coefficient: r
0.13 0.36 0.42 0.50
3.1 6.3 6.2 9.8
0.93
capable of influencing the arousal level of the brain, were examined. Amphetamine was applied iv. at four different doses, from 0.13 to 0.50 mg/kg (Table 1). All doses caused a lowering of occlusion, i.e. a shift towards facilitation (Fig. 7). The decrease of occlusion, as expressed in per cent, correlated well with the doses of amphetamine (r = 0.93), indicating causal relationship between them. The effects of haloperidol and reserpine (antagonists of amphetamine) were now looked at. Haloperidol was administered i.v. in a dose of 0.5 mg/kg, 3:h prior to the recording, preventing the decrease of occlusion by amphetamine (Fig. 8). Reserpine (0.42 mg/kg) behaved similarly (not presented here). None of these drugs seemed to influence the level of occlusion itself, but blocked changes of occlusion induced by amphetamine. A newly discovered dipeptide (7 -glutamyltaurine) proved to have effects on central nervous functions analagous with those of amphetamine (Fig. 7). In the present experiments it was tested whether this parallel also holds for the interactions of evoked potentials. On administering y-glutamyltaurine in a wide range of doses, a dose-response relationship could be observed, if not as strict as in the case of amphetamine (Table 2), with a correlation coefficient of 0.53. The
-P > z
20%
1
8
Fig. 5. The facilitatory effect of 2Hz conditioning stimulation on the interaction of acoustic (AC)and somatosensory (Ss) responses. (A) Amplitudes of separate and compound responses at 1 Hz frequency. (B) Compound responses recorded during conditioning stimulation. (C) Separate and compound evoked potentials during 1 Hz test period. (D) The same as in (C) after 3 min rest. Facilitation of AC and compound responses is obvious. This led to a facilitatory shift in the interaction too.
facilitation%1
3lsec
vsec P
30min break
-40 occlusion ‘!/(j --I
ll=15
Il=lO
M-9.73? 564
M-2545 p
54
n-10 M=9’1,46 P
Fig. 6. Effect of elevating the stimulus frequency on the level of occlusion. During 3 Hz stimulation the mean level rose to more than twice that of the control period (C) and returned to the initial one after 30min intermission before returning to 1 Hz. All changes are statistically significant. M, mean.
J Toldi el al.
650
B
Litoralon rwkg
facilitatii%l
in M-25.14+288
M.25J6’4.02
M-15,1*3,71
M-16.43’413
p
p
kg
028mg/kg
facthtatlon% 10
2.6-Wkg LIpIon
am hetamme P
ha’Ywr’do’
on
/ Il
-I
M-13.25
-30
22 55
M- 10.67*3
8
M-1368+457
control 1 occluslon~/OI Fig.
8. The
effect
of
haloperidol (O.SOmg/kg) on the facilitatory y-glutamyltaurine (Litoralon). M, mean.
action of y -glutamyltaurine was similar to that of amphetamine and was also blocked by haloperidol (Fig. 8). Propranolol, a well known blocker of fl adrenergic receptors, not only antagonized the facilitatory action of amphetamine and y-glutamyltaurine, but also strongly enhanced the occlusive interaction of acoustic and somatosensory potentials (Fig. 9). DISCUSSION
The existence of interaction between evoked potentials representing different modalities has been proven in the experiments presented in this and earlierlo papers. Although it seems likely that the two kinds of stimuli excite separate populations of neurons, our experiments indicate that they are bound with each other by dynamic interrelations. The folTable
2. Effect of y-glutamyltaurine on the occlusion acoustic and somatosensory evoked potentials
y-Glutamyltaurine (mg/kg, i.v.) 1.81 2.10 3.17 3.17 3.60 3.77 3.61 5.00
Decrease of occlusion (%) 5.00 7.40 4.04 3.37 8.50 3.37 6.20 14.60
of
action
of
amphetamine
and
lowing observations provide information as to the nature of the interaction: (1) the most dominant form of interactions is occlusion, which may be the resultant of occlusive and facilitatory mechanisms. Thus, changes in the level of occlusion may be due to changes in facilitatory mechanisms, too. (2) Elevation of occlusion level can be attained in more than one way: (a) with 2 Hz conditioning series; (b) with 3 Hz stimulation frequency at both separate and compound stimulations; (c) with MFR stimulation at
Propranolol : 0,37 mgl kg Litoralon:3,7 @kg
fachtatlon% I
-30 -40.
_._
Correlation coefficient: r
-50 occtus0n %1
3 min n-10 M=39,8’ 5,42
n=9 M=15,0*7;23 P
0.53
Fig. 9. The effect of propranolol (0.37 mg/kg) on the occlusion level. This drug does not only antagonize y-glutamyltaurine (Litoralon), but causes a great enhancement of occlusion. M. mean
651
Dynamic interactions of evoked potentials
Mod 1
Mod2
\
Mod1
MFR(+) 3c/s(+)
(
/
Propr~“olol t-1
Litoralon (+I
\
Mod2
Mod1
Mod 2
Arndhetamine (+I
Fig. 10. Schemes to explain the supposed localization of inhibitory (A) and excitatory (B) interneurons, and overlapping of different modalities (C) within simple neuronal nets producing evoked potentials of both modalities. (A) Compound stimulation (Mod:1 and Mod:2) evokes smaller evoked potential than the algebraic sum of separately evoked responses (Mod:1 and Mod:2 respectively), because of the activated inhibitory interneuron by the converging afferents from Mod: 1 and Mod:2. In case of MFR stimulation or stimulation with higher frequencies the inhibitory effect is larger and the occlusion is enhanced. (B) Impulses from Mod:1 and Mod:2 converge on an excitatory interneuron which extends excitation to pyramidal cells being at rest at separate stimulation. Amphetamine and y-glutamyltaurine (Litoralon) enhance, and propranolol depresses transmission at this synapse. (C) Afferent fibres (Mod: 1 and Mod:2) terminate on pyramidal cells with synapses of different strength. Stimulation of Mod: 1 excites strongly only neuron I while its effect on neuron II is weak (dotted line). Mod:2 excites both neurons strongly. In course of 2Hz compound stimulation the strongly excitatory synapse may (heterosynaptically?) facilitate the weaker one (arrow). This results in reinforcement of Mod: 1 synapse on neuron II. Mod: 1 evoked potential will be augmented but the effect of compound stimulation unchanged. All this leads to enhancement of occlusion. (See facilitation of acoustic evoked potential in Fig. 5).
200 Hz frequency and (d) by blocking central adrenergic receptors with propranolol. A decrease of occlusion (which may be equivalent with enhancement of facilitation) was seen after amphetamine and y-glutamyltaurine, while haloperidol and reserpine prevented these effects. Since the most frequent interaction between evoked potentials is occlusion, changes of interactions will be dealt with in terms of occlusion. Elevation of the occlusion level can be substantiated in two ways: (a) increase of one or both evoked potentials can be followed by unchanged amplitude at compound stimulation and (b) evoked potentials obtained separately remain unchanged, but the compound response decreases. The neuron population excited by compound stimulation remains the same in case (a), and becomes smaller in case (b) (see Fig. 10). One is led to conclude, that the latter is caused by inhibition, activated by impulses converging to some interneurons. If these inhibitory interneurons are blocked by chloralose, no occlusion is present. This might have been the case in the experiments of Thompson et ~1.;~confirmed by Toldi and Feher.“’ At the same time, a partial overlapping between the pyramidal neuron populations innervated by the two different afferent systems in the ASG is indicated, but their interactions are mediated partly by interneurons of inhibitory and, probably, excitatory character. In barbiturate anaesthesia inhibitory dominance seems
to prevail. However, particular stimulus situations and drugs are able to modify the equilibrium between excitatory and inhibitory interneuron populations: amphetamine and y -glutamyltaurine seem to favour facilitation, 3 Hz stimulation and blocking of ,8 adrenergic receptors enhance occlusion. Shifts in either direction can be brought about by enhanced activity of the one, or depression of the other population, or by both. The two schemes presented below visualize the supposed localization of inhibitory (Fig. 10A) and excitatory (Fig. 10B) interneurons, and overlapping of modalities (Fig. 1OC) within simple neuronal nets producing evoked potentials in the ASG. Schemes can be built up, of course, in a mirror image, in view of modalities. The organization principles reflected by them are supposed to be present as co-existing subsystems of the cortical network. The modifications of occlusion level are reversible and commensurable in duration with those observed by Rutledge,’ Chalupa et aL4 and Morrel et ~1.~Their dynamics are probably dependent on the anaesthesia and type of anaesthetic, therefore experiments on waking animals are required. The fact that modifications of occlusion level are largely independent of the actual amplitude of separately evoked potentials, indicates that the subsystems responsible for interactions are not identical with those generating sensory evoked potentials. This
652
J. Toldi er al.
has been indicated by the differential effect of some stimulation paradigms, MFR stimulation and the drugs examined. The sites of action of the drugs applied in the present experiments require further study. The most probable targets of their actions are cortical icterneurons, reticulocortical afferents and terminals of the catecholaminergic system. This can be cleared up only by further pharmacological investigations.
Summary
Our experiments have revealed a new dimension of cerebral cortical function in which cortical organization, co-operation of thalamocortical afferent systems and reticular formation can be examined from new aspects. Acknowledgements-The authors are deeply indebted Mrs. A. Vecsernyes, Mrs. A. Sziics and Mrs. A. Jo&
to
REFERENCES 1. Berman
2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
A. L. (1961) Interactions of cortical responses to somatic and auditory stimuli in anterior ectosylvian gyrus of cat. J. Neurophysiol. 24, 608-620. Bremer F. (1952) Analyse oscillographique des reponses sensorielles des ecorces cerebrale et ctrtbelleuse. Revue Neural. 87, 65-92. Bremer F., Bonnet V. and Terzuolo C. (1954) Etude Clectrophysiologique des aires auditives corticales du chat. Archs int. Physiol. 62, 39&428. Chalupa L. M., Macadar A. W. and Lindsley D. B. (1975) Response plasticity of lateral geniculate neurons during and after pairing of auditory and visual stimuli. Science, N. Y. 190, 29(t292. Jasper H. H. and Ajmone-Marsan C. (1954) A Stereotaxic Atlas of the Diencephalon of the Cat. NRC of Canada, Ottawa. University of Toronto Press, Toronto. Morrel F., Hoeppner T. J. and De Toledo-Morrel L. (1983) Conditioning of single units in visual association cortex: cell-specific behavior within a small population. Expl Neural. 80, 111-146. Poliakova A. G. (1972) Origin of the early component of the evoked response in the association cortex of the cat. Electroenceph. clin. Neurophysiol. 32, 129-139. Rutledge L. T. (1965) Facilitation: electrical response enhanced by conditional excitation of cerebral cortex. Science, N.Y. 148, 12461248. Thompson R. F., Smith H. E. and Bliss D. (1963) Auditory, somatic sensory and visual response interactions and interrelations in association and primary cortical fields of the cat. J. Neurophysiol. 26, 365-378. Toldi J. and Feher 0. (1980) Dynamic interactions of evoked potentials in the association cortex of the cat. Actaphysiol. hung. 55, 19-30. Toldi J., Rojik I. and Feher 0. (1981) Two different polysensory systems in the suprasylvian gyrus of the cat. An electrophysiological and autoradiographic study. Neuroscience 6, 2539-2545. (Accepted 27 June 1984)
NeuroscienceVol. 13, No. 3, PP. 645-652, 1984 Printed in Great Britain
IBRO
DYNAMIC INTERACTIONS OF EVOKED POTENTIALS IN A POLYSENSORY CORTEX OF THE CAT J. TOLDI, 0. FE&R and L. FEUER* Department of Comparative Physiology, Jbzsef Attila University, Szeged, Hungary, POB 533, H-6701. *Chinoin Pharmaceutical and Chemical Works Ltd., Budapest, T6 u. l., Hungary, H-1325 Abstract-Interactions of acoustic and somatosensory evoked potentials were studied in the anterior suprasylvian gyrus of the cat. Two kinds of interaction could be observed: occlusion or facilitation. In most cases occlusion was observed. The interactions showed dynamic changes and were susceptible to different kinds of influences. After having determined the control values of interaction over a period of several minutes, reversible enhancement of occlusion was observed after synchronous activation of the acoustic and somatosensory inputs with 2Hz frequency. The same effect could be observed during
stimulation of the mesencephalic reticular formation with 200 Hz frequency. The interactions could also be influenced by amphetamine and y-glutamyltaurine, known as drugs capable of influencing the arousal level of the brain. During treatment with amphetamine the interaction was shifted in the facilitatory direction. The antagonists of amphetamine (haloperidol and reserpine) prevented this effect. The authors suggest that the interactions of acoustic and somatosensory responses are mediated by interneurons (inhibitory and/or excitatory) and particular stimulus situations and drugs are able to modify the equilibrium between excitatory and inhibitory subsystems.
The examination of the interactions of evoked potentials began with the experiments of Bremer,2,3 some 30 years ago. This field of research proved to be very fruitful because it gave insight into problems such as the organization of polysensory structures. In studying the interactions of acoustic and somatosensory evoked potentials Berman has shown that there exists a considerable overlapping of these areas in the anterior ectosylvian gyrus of the cat.’ In this experimental paradigm, two stimuli were given belonging to two different modalities, with various time intervals between them, and the action of the first stimulus upon the potential evoked by the second one was studied. A quotient of the calculated and measured amplitude of this potential expressed the extent of interaction. This might be more or less than one and if it equalled one, it was taken as a lack of interaction. Unfortunately the potential usefulness of this paradigm was not recognized for a long time. In our earlier experiments” Berman’s paradigm was applied with several modifications. The time sequence of two stimuli of different modalities was adjusted so as to achieve a complete synchronism of the two potentials evoked by them, what meant an overlapping of the positive and negative phases, respectively. The experiments were performed on the anterior suprasylvian gyrus of the cat anaesthetized with barbiturate and in some cases with chloralose. The comparison of amplitudes obtained with separate and compound stimulations gave an opportunity to judge the type and extent of interaction (occlusion and facilitation). In this paper we give an account of the influences which are able to modify these interASG, anterior suprasylvian gyrus; MFR, mesencephalic reticular formation.
Abbreviations:
actions between acoustic and somatosensory evoked potentials in the anterior suprasylvian gyrus (ASG) of the cat, and an attempt will be made to draw conclusions as to their mechanism and the origin of their dynamic changes. EXPERIMENTALPROCEDURES The experiments were performed on adult cats of both sexes, anaesthetized with pentobarbital (Nembutal; 40mg/kg, i.p.). After venous and tracheal cannulation, the left cerebral hemisphere was exposed and covered with mineral oil. The experiment began after 2 h rest. Acoustic click stimuli were applied to the right ear via a miniature earphone, connected to a square wave stimulator. The stimuli had a duration of 1 ms and were always supramaximal. Somatosensory evoked potentials were obtained by stimulating the right forepaw with electric pulses of 3-8 V and 0.3-0.5 ms in duration, which also proved to be supramaximal. Mesencephalic reticular formation (MFR) was stimulated with bipolar needle electrodes according to the stereotaxic co-ordinates given by Jasper and Ajmone Marsan’ (FR: 3; L: 3.5; V: - 1). At the end of the experiments electrolytic lesions were placed at the tips of the bipolar electrodes and localized by standard histological procedures. The evoked potentials were recorded from the surface of cortex with ball-tipped silver wire electrodes attached to a two channel amplifier system with a time constant of 0.5 s. Blocks of 20 or 50 potentials were averaged with a Motorola MC-6800 computer and drawn using an X-Y plotter. Comparison of results taken from different series of stimulation was performed by using Student’s double 645
.I. Toldi et al.
646
t test and in some cases correlation were calculated.
of stimulus pairs as with separate stimulation. The extent of interaction (I) was calculated according to the formula presented in Fig. 1. The amplitude obtained with compound stimulation (given by arbitrary units) is divided by the sum of the separately recorded potentials. From this quotient a percentage value was calculated which expressed the extent of interaction independently of voltage values. In the experiments three types of interaction occurred: (1) algebraic summation, which meant a lack of interaction with I = 0; (2) occlusion with I < 0 and (3) facilitation at I > 0. First we examined whether the basic levels of interactions show any correlation with one another in different polysensory areas. One example of a comparison between the posterior middle suprasylvian association area and the anterior lateral association area is presented in Fig. 2. The records taken for as long as 30 min show that the level of interaction varied around zero with low occlusion and facilitation in the anterior lateral association area. In the posterior middle suprasylvian association area a rather high level of occlusion was present, at around 30”/,. Although the variability was somewhat greater than in the anterior lateral association area, it held the overall tendency throughout the recording time. Among minute variations of occlusion no correlation could be found between the two recorded sites.
coefficients (r)
Materials
Drugs were obtained from Hungarian and foreign pharmaceutical companies; amphetamine and y-glutamyl-taurine (LitoralonR) from Chinoin Ltd; haloperidol and reserpine from Kiibanyai Gyogyszerartigyar; and Propranolol from Coch-Light. RESULTS
Types of interactions of acoustic and somatosensory evoked potentials
Acoustic and somatosensory stimulation elicited potentials of 200-5OOpV amplitude in the ASG. Their latency according to the earlier findings’,” was near to those obtainable in the primary sensory areas. With acoustic evoked potentials it ranged from 9 to 11 ms and was 8-10 ms with somatosensory potentials. Actual differences in latency were taken into account and compound stimulus pairs were given with an interval which ensured complete overlapping of the respective positive and negative phases. First, a block of acoustic potentials was recorded with 1 Hz frequency, then the same was done with somatosensory potentials. Afterwards compound stimulation followed with the same frequency and number
A
AC
ss
/2 ‘1. ;
z
:L 1 9 r_
0
i
I
42
B
AC
37
62 I= 1--l).lOO% 42*37 I
=-21,51% (0ccIUsron)
2oiiisec
62
ss
I* (7343*27 I
43 Fig. 1. Types of interactton
27
-,).l~“ib
* 4,28”k (facthtatton)
73
between acoustic (AC) and somatosensory (Ss) evoked potentials. (A) Occlusion; amplitude of the compound response (62) is lower than the algebraic sum of the separately recorded potentials (42 + 37). According to the formula this means an occlusion of 2 I .5’% (B) Facilitation; the compound response is higher than the sum of separately recorded potentials. Calculation is the same as in case of occlusion.
647
Dynamic interactions of evoked potentials
Fig. 2. Interaction of acoustic and somatosensory evoked potentials in the anterior lateral association area (ALA) and posterior middle suprasylvian association area (PMSA) as recorded simultaneously over a 30min period. In ALA interactions are very weak, while in PMSA occlusion is dominant. The same holds for another pair of loci in the ASG under resting conditions (Fig. 3). The correlation coefficient between their occlusion values was 0.075. However, as MFR was stimulated with 200 Hz frequency for 33 min, the occlusion of acoustic and somatosensory potentials suddenly increased at both sites. After the stimulation period the occlusion remained high at point 2, while at point 1 it resumed its initial level. The parallel of the MFR action at these loci is obvious and the correlation coefficient is
M-7,6*5.9
c*
orxhJsmol
MS 6,8'21
c,c,
r- 0,075
rather high (r = 0.744). Thus it appears that each point of ASG exhibits a particular level of occlusion, apparently independent of other ones, but factors modifying the level of interactions manifest themselves uniformly. Modljkation
M-I43’6,O
VW
p
pea,,
stimulus trains
After having established the level of interaction (mostly occlusion) at particular points of ASG, 2 Hz compound stimulation was given for 2 min (Fig. 4C).
M-32.6*7,9
9
of interactions following
MFR2(zoo~)
b
G
M-2X0’ 6.0 NFR, m,
r*o,#4
@ 3min
Fig. 3. The effect of MFR stimulation on the occlusion level at two (1,2) points of ASG, denoted in insert top right. In the control period (C) no correlation was present between the two sites in view of interactions. Stimulation of MFR, however, acted upon them in the same manner, reflected also by the correlation coefficient, r = 0.744. Enhanced occlusion persisted in point 2. Stimulation sites in the MFR are denoted in the scheme bottom right. g, gyrus; M, mean.
J. Toldi et al.
648
i b
mlLficx1~.
I
AC
n.4 n-3
'th1113.53 M-3.33t2.35
h+QPO
M-8.7%X3
M.lO.ma47M=~istOa2
M-18.66a94M-lwa5 p
i=o.OOl
Fig. 4. The effects of 2 Hz conditioning (c) stimulation on the interactions of evoked potentials. (A) Occlusion levels before and after c as measured in 2 min periods. (B) Mean occlusion level with SD before
and after c. The difference is highly significant (P < 0.001). (C) Diagrams showing the relation between evoked response amplitudes (abscissa) and occlusion levels (ordinates) in all the 2 min period presented in (A). (D) Acoustic (AC) and somatosensory (Ss) potentials remained practically unchanged throughout the recording period, while occlusion levels showed a rather wide variance. The compound responses decreased significantly after c and this led to elevation of occlusion level presented in (B). M, mean.
The changes subsequent to it were characterized by temporary elevation of occlusion as measured during the next 1 Hz stimulation period. After the 2 Hz stimulus train a 2min rest was interposed. The maximal duration of such reversible changes was 16min. The course of experiments of this type are illustrated in Fig. 4 and an insight into the interrelations between amplitude of evoked potentials (at separate and compound stimulations) and occlusion values is given. Occlusion values are diagrammed as recorded in 6min control period at 1 Hz stimulus frequency (Fig. 4A). In the first 2 min sampling period there was 0% occlusion; in the two subsequent ones 5% occlusion was obtained. The average value of this control period is presented in Fig. 4B (first bar) with SD (3.33 f 2.35). After the conditioning stimulus series (c) occlusion became much more intensive over an 8 min period and although it showed large fluctuations, its average differed from the control values significantly (mean: 11 f 3.53, P < 0.001) (Fig. 4B, second bar). In Fig. 4C variations of acoustic and somatosensory evoked potential amplitudes are presented as recorded with separate and compound stimulations and it can be judged how they were related to the actual level of occlusion. It is obvious that, at separate stimulation, acoustic and somatosensory potentials show minute changes even after 2 min of 2 Hz conditioning stimulation. The compound response, however, decreased significantly from 18.66 f 0.5 and this led to elevation of the occlusion level (Fig. 4C and D, last diagrams). One is led to conclude that the amplitude
of evoked potentials and the extent of their occlusion are determined by different factors and show independence from each other within very wide limits. In a few cases conditioning stimulation (also 2 Hz) resulted in transitory facilitation but its extent also proved to be independent of the evoked potential amplitude (Fig. 5). In another series of experiments conditioning stimulations of 5 and 10 Hz were also applied but they failed to exert any action upon occlusion level. The dependence frequency
of occlusion on the stimulution
Another way of exploring the dependence of interaction on the stimulation rate was chosen, when separate and compound stimulation was made with 3 Hz frequency (Fig. 6). First the occlusion level was determined with 1 Hz stimulation over a period of 45 min. Then the frequency was raised to 3 HZ at both separate and compound stimulations and the occlusion level was measured for 30 min. As shown in Fig. 6, this led to an abrupt elevation of occlusion all over the period of a higher stimulation rate. After 30min break the occlusion levels returned to the original one with far less variability than before. Pharmacological injuences upon interactions acoustic and somatosensory potentials
of
Since interactions of evoked potentials seemed to have something to do with equilibrium of excitatory and inhibitory processes, the effects of some drugs,
649
Dynamic interactions of evoked potentials
C
Table 1. Effect of amphetamine on the occlusion of acoustic and somatosensory evoked potentials
IhC AC -P
SSA
r ?r 3
D
min break
l/WC
ss* 5
Amphetamine (mg/kg, i.v.)
Decrease of occlusion (“/,)
Correlation coefficient: r
0.13 0.36 0.42 0.50
3.1 6.3 6.2 9.8
0.93
capable of influencing the arousal level of the brain, were examined. Amphetamine was applied iv. at four different doses, from 0.13 to 0.50 mg/kg (Table 1). All doses caused a lowering of occlusion, i.e. a shift towards facilitation (Fig. 7). The decrease of occlusion, as expressed in per cent, correlated well with the doses of amphetamine (r = 0.93), indicating causal relationship between them. The effects of haloperidol and reserpine (antagonists of amphetamine) were now looked at. Haloperidol was administered i.v. in a dose of 0.5 mg/kg, 3:h prior to the recording, preventing the decrease of occlusion by amphetamine (Fig. 8). Reserpine (0.42 mg/kg) behaved similarly (not presented here). None of these drugs seemed to influence the level of occlusion itself, but blocked changes of occlusion induced by amphetamine. A newly discovered dipeptide (7 -glutamyltaurine) proved to have effects on central nervous functions analagous with those of amphetamine (Fig. 7). In the present experiments it was tested whether this parallel also holds for the interactions of evoked potentials. On administering y-glutamyltaurine in a wide range of doses, a dose-response relationship could be observed, if not as strict as in the case of amphetamine (Table 2), with a correlation coefficient of 0.53. The
-P > z
20%
1
8
Fig. 5. The facilitatory effect of 2Hz conditioning stimulation on the interaction of acoustic (AC)and somatosensory (Ss) responses. (A) Amplitudes of separate and compound responses at 1 Hz frequency. (B) Compound responses recorded during conditioning stimulation. (C) Separate and compound evoked potentials during 1 Hz test period. (D) The same as in (C) after 3 min rest. Facilitation of AC and compound responses is obvious. This led to a facilitatory shift in the interaction too.
facilitation%1
3lsec
vsec P
30min break
-40 occlusion ‘!/(j --I
ll=15
Il=lO
M-9.73? 564
M-2545 p
54
n-10 M=9’1,46 P
Fig. 6. Effect of elevating the stimulus frequency on the level of occlusion. During 3 Hz stimulation the mean level rose to more than twice that of the control period (C) and returned to the initial one after 30min intermission before returning to 1 Hz. All changes are statistically significant. M, mean.
J Toldi el al.
650
B
Litoralon rwkg
facilitatii%l
in M-25.14+288
M.25J6’4.02
M-15,1*3,71
M-16.43’413
p
p
kg
028mg/kg
facthtatlon% 10
2.6-Wkg LIpIon
am hetamme P
ha’Ywr’do’
on
/ Il
-I
M-13.25
-30
22 55
M- 10.67*3
8
M-1368+457
control 1 occluslon~/OI Fig.
8. The
effect
of
haloperidol (O.SOmg/kg) on the facilitatory y-glutamyltaurine (Litoralon). M, mean.
action of y -glutamyltaurine was similar to that of amphetamine and was also blocked by haloperidol (Fig. 8). Propranolol, a well known blocker of fl adrenergic receptors, not only antagonized the facilitatory action of amphetamine and y-glutamyltaurine, but also strongly enhanced the occlusive interaction of acoustic and somatosensory potentials (Fig. 9). DISCUSSION
The existence of interaction between evoked potentials representing different modalities has been proven in the experiments presented in this and earlierlo papers. Although it seems likely that the two kinds of stimuli excite separate populations of neurons, our experiments indicate that they are bound with each other by dynamic interrelations. The folTable
2. Effect of y-glutamyltaurine on the occlusion acoustic and somatosensory evoked potentials
y-Glutamyltaurine (mg/kg, i.v.) 1.81 2.10 3.17 3.17 3.60 3.77 3.61 5.00
Decrease of occlusion (%) 5.00 7.40 4.04 3.37 8.50 3.37 6.20 14.60
of
action
of
amphetamine
and
lowing observations provide information as to the nature of the interaction: (1) the most dominant form of interactions is occlusion, which may be the resultant of occlusive and facilitatory mechanisms. Thus, changes in the level of occlusion may be due to changes in facilitatory mechanisms, too. (2) Elevation of occlusion level can be attained in more than one way: (a) with 2 Hz conditioning series; (b) with 3 Hz stimulation frequency at both separate and compound stimulations; (c) with MFR stimulation at
Propranolol : 0,37 mgl kg Litoralon:3,7 @kg
fachtatlon% I
-30 -40.
_._
Correlation coefficient: r
-50 occtus0n %1
3 min n-10 M=39,8’ 5,42
n=9 M=15,0*7;23 P
0.53
Fig. 9. The effect of propranolol (0.37 mg/kg) on the occlusion level. This drug does not only antagonize y-glutamyltaurine (Litoralon), but causes a great enhancement of occlusion. M. mean
651
Dynamic interactions of evoked potentials
Mod 1
Mod2
\
Mod1
MFR(+) 3c/s(+)
(
/
Propr~“olol t-1
Litoralon (+I
\
Mod2
Mod1
Mod 2
Arndhetamine (+I
Fig. 10. Schemes to explain the supposed localization of inhibitory (A) and excitatory (B) interneurons, and overlapping of different modalities (C) within simple neuronal nets producing evoked potentials of both modalities. (A) Compound stimulation (Mod:1 and Mod:2) evokes smaller evoked potential than the algebraic sum of separately evoked responses (Mod:1 and Mod:2 respectively), because of the activated inhibitory interneuron by the converging afferents from Mod: 1 and Mod:2. In case of MFR stimulation or stimulation with higher frequencies the inhibitory effect is larger and the occlusion is enhanced. (B) Impulses from Mod:1 and Mod:2 converge on an excitatory interneuron which extends excitation to pyramidal cells being at rest at separate stimulation. Amphetamine and y-glutamyltaurine (Litoralon) enhance, and propranolol depresses transmission at this synapse. (C) Afferent fibres (Mod: 1 and Mod:2) terminate on pyramidal cells with synapses of different strength. Stimulation of Mod: 1 excites strongly only neuron I while its effect on neuron II is weak (dotted line). Mod:2 excites both neurons strongly. In course of 2Hz compound stimulation the strongly excitatory synapse may (heterosynaptically?) facilitate the weaker one (arrow). This results in reinforcement of Mod: 1 synapse on neuron II. Mod: 1 evoked potential will be augmented but the effect of compound stimulation unchanged. All this leads to enhancement of occlusion. (See facilitation of acoustic evoked potential in Fig. 5).
200 Hz frequency and (d) by blocking central adrenergic receptors with propranolol. A decrease of occlusion (which may be equivalent with enhancement of facilitation) was seen after amphetamine and y-glutamyltaurine, while haloperidol and reserpine prevented these effects. Since the most frequent interaction between evoked potentials is occlusion, changes of interactions will be dealt with in terms of occlusion. Elevation of the occlusion level can be substantiated in two ways: (a) increase of one or both evoked potentials can be followed by unchanged amplitude at compound stimulation and (b) evoked potentials obtained separately remain unchanged, but the compound response decreases. The neuron population excited by compound stimulation remains the same in case (a), and becomes smaller in case (b) (see Fig. 10). One is led to conclude, that the latter is caused by inhibition, activated by impulses converging to some interneurons. If these inhibitory interneurons are blocked by chloralose, no occlusion is present. This might have been the case in the experiments of Thompson et ~1.;~confirmed by Toldi and Feher.“’ At the same time, a partial overlapping between the pyramidal neuron populations innervated by the two different afferent systems in the ASG is indicated, but their interactions are mediated partly by interneurons of inhibitory and, probably, excitatory character. In barbiturate anaesthesia inhibitory dominance seems
to prevail. However, particular stimulus situations and drugs are able to modify the equilibrium between excitatory and inhibitory interneuron populations: amphetamine and y -glutamyltaurine seem to favour facilitation, 3 Hz stimulation and blocking of ,8 adrenergic receptors enhance occlusion. Shifts in either direction can be brought about by enhanced activity of the one, or depression of the other population, or by both. The two schemes presented below visualize the supposed localization of inhibitory (Fig. 10A) and excitatory (Fig. 10B) interneurons, and overlapping of modalities (Fig. 1OC) within simple neuronal nets producing evoked potentials in the ASG. Schemes can be built up, of course, in a mirror image, in view of modalities. The organization principles reflected by them are supposed to be present as co-existing subsystems of the cortical network. The modifications of occlusion level are reversible and commensurable in duration with those observed by Rutledge,’ Chalupa et aL4 and Morrel et ~1.~Their dynamics are probably dependent on the anaesthesia and type of anaesthetic, therefore experiments on waking animals are required. The fact that modifications of occlusion level are largely independent of the actual amplitude of separately evoked potentials, indicates that the subsystems responsible for interactions are not identical with those generating sensory evoked potentials. This
652
J. Toldi er al.
has been indicated by the differential effect of some stimulation paradigms, MFR stimulation and the drugs examined. The sites of action of the drugs applied in the present experiments require further study. The most probable targets of their actions are cortical icterneurons, reticulocortical afferents and terminals of the catecholaminergic system. This can be cleared up only by further pharmacological investigations.
Summary
Our experiments have revealed a new dimension of cerebral cortical function in which cortical organization, co-operation of thalamocortical afferent systems and reticular formation can be examined from new aspects. Acknowledgements-The authors are deeply indebted Mrs. A. Vecsernyes, Mrs. A. Sziics and Mrs. A. Jo&
to
REFERENCES 1. Berman
2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
A. L. (1961) Interactions of cortical responses to somatic and auditory stimuli in anterior ectosylvian gyrus of cat. J. Neurophysiol. 24, 608-620. Bremer F. (1952) Analyse oscillographique des reponses sensorielles des ecorces cerebrale et ctrtbelleuse. Revue Neural. 87, 65-92. Bremer F., Bonnet V. and Terzuolo C. (1954) Etude Clectrophysiologique des aires auditives corticales du chat. Archs int. Physiol. 62, 39&428. Chalupa L. M., Macadar A. W. and Lindsley D. B. (1975) Response plasticity of lateral geniculate neurons during and after pairing of auditory and visual stimuli. Science, N. Y. 190, 29(t292. Jasper H. H. and Ajmone-Marsan C. (1954) A Stereotaxic Atlas of the Diencephalon of the Cat. NRC of Canada, Ottawa. University of Toronto Press, Toronto. Morrel F., Hoeppner T. J. and De Toledo-Morrel L. (1983) Conditioning of single units in visual association cortex: cell-specific behavior within a small population. Expl Neural. 80, 111-146. Poliakova A. G. (1972) Origin of the early component of the evoked response in the association cortex of the cat. Electroenceph. clin. Neurophysiol. 32, 129-139. Rutledge L. T. (1965) Facilitation: electrical response enhanced by conditional excitation of cerebral cortex. Science, N.Y. 148, 12461248. Thompson R. F., Smith H. E. and Bliss D. (1963) Auditory, somatic sensory and visual response interactions and interrelations in association and primary cortical fields of the cat. J. Neurophysiol. 26, 365-378. Toldi J. and Feher 0. (1980) Dynamic interactions of evoked potentials in the association cortex of the cat. Actaphysiol. hung. 55, 19-30. Toldi J., Rojik I. and Feher 0. (1981) Two different polysensory systems in the suprasylvian gyrus of the cat. An electrophysiological and autoradiographic study. Neuroscience 6, 2539-2545. (Accepted 27 June 1984)