72
%cur..~iem'c Lcl~', i~. i~6 t i g92~ ~2 , 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved (t~04-3940'~2'$ i):, c)~
NSL 08408
Relationships between electrically induced slow negative potentials and changes in extracellular potassium concentrations in cerebral cortex of the cat Elijah Ocherashvili a n d A l e x a n d e r R o i t b a k Beritashvili Institute of Physiology, Georgian Academy (~fSeienees, Tbilisi {Georgia J (Received 15 April 199 l: Revised version received 19 November 1991: Accepted 27 November 1991 )
Key words: Cerebral cortex: Slow negative potential: Extracellular potassium concentration Experiments were carried out on cats under deep nembutal anaesthesia. Local cortical stimulation evoked both a slow negative potential (SNP) and an increase in extracellular potassium ([K+]o) which were maximal in the upper cortical layers. The decline in the K*-potential in deeper layers was slower than in the upper ones so that with time [K-]o became equal in upper and deeper layers. No SNP was recorded at that time, perhaps explaining why K--potential decay is of longer duration than the SNE Transcallosal stimulation also evoked an SNP but gave a lesser difference in A[K+],~in upper and deeper layers. The lower amplitude of the SNP evoked by transcallosal stimulation may be due to these circumstances. These data could clarify the mechanism of correlation disturbance between SNPs and K ~-potentials in the cerebral cortex.
It has been established that slow negative potentials (SNP) in different regions of the nervous system are associated with an increase in extracellular potassium concentration ([K+]o) and glial depolarization [2, 4, 6, 10]. Triple correlations between these measures have confirmed the supposition of the glial origin of the SNP [5, 10]. Mathematical analysis has shown that the contribution of neuronal elements to SNP generation evoked by surface cortical stimulation is insignificant [7]. However, it is also known that the correlation between SNE K ~potential and glial depolarization can be disturbed: SNP normally decays more rapidly than glial depolarization and the K+-potential. The positive correlation between the amplitude of the SNP and A[K+]o is changed at supramaximal intensities or high frequencies of stimulation [8]. The present investigation was aimed at elucidating the mechanisms of these discrepancies. Experiments were performed on 11 cats under nembutal anaesthesia (60-70 mg/kg). The sites of stimulation and recording were on the suprasylvian gyrus. Bipolar stimulating electrodes were used for local stimulation of the cortical surface (local stimulation) and transcallosal stimulation with 200/lm and 1.5 mm interpolar distance respectively. Stimulating electrodes were located on the Correspondence: E. Ocherashvili, Beritashvili Institute of Physiology. Georgian Academy of Sciences, Tbilisi. Georgia.
pial surface of the cortex. Local stimulation was performed by single stimuli and transcallosal stimulation by brief rhythmic trains, because a single stimulus evoked very weak A[K~]o and SNP or failed to evoke these. At a distance of 1 mm from the electrodes which were used for local stimulation, an Ag-AgC1 macroelectrode and a double barreled microelectrode were placed, the latter with one K*-selective capillary (for measuring the A[K+]o) and one NaC1 reference capillary (used also for recording the electrical field potential). After the arrangement of electrodes, the cortex was covered with a mixture of dental wax and vaseline oil, to prevent brain pulsations. The temperature of the cortex was 29-30 ° C. Anaesthesia and low cortex temperature abolished spontaneous activity and allowed stable recording over 4-5 h. K+-potentials and SNPs were recorded with use of high impedance negative capacity input preamplifiers and displayed on ink recorders. Fig. 1 shows that on local stimulation a SNP was recorded from the surface while a positive potential was recorded from the middle layers of the cortex. The mean + S.E.M. of the SNP amplitude was 0.95 _+ 0.18 mV for the upper layers. (It is known [1] that the SNP inverts deep within the cortex.) In the example shown in Fig. l A[K+]o was ca. 1.3 mmol/1 in the upper layers (100/lm) and ca. 0.74 mmol/l in the middle ones (500 jim), where the positive potential was recorded (Fig. I B). Com-
73
A
!
mV
o_
B
0.1
4,
K+
3 Fig. 1. Connection between the duration of the stimulus-induced SNP and changes in [K*],, in the upper and middle layers of the cortex. Electrode arrangement shown in diagram: 1, macroelectrode: 2. K ~microelectrode: 3, stimulating electrodes. Single stimulus (40 V, 0.1 ms) was applied to the cortical surface. Alter recording the SNP from the cortical surface and K~-potential from the upper layers (0.1 mm) the K+-microelectrode was inserted to a depth of 0.5 mm. A: SNP (upper trace) recorded from the cortical surface by means of microelectrode 1 and positive potential (lower trace) recorded from the depth of the cortex (0.5 ram) by means of the K'-microelectrode reference channel, B: superposition of K-potentials recorded from upper (0.1 ram) and middle (0.5 mm) layers. Vertical dotted lines in A show the time of the decay of the SNP to the baseline and in B the time when [K-]o in upper and middle layers became equal.
parable figures for A[K+]~,for upper and middle layers for six animals were 1.1 + 0.09 and 0.38 + 0.08 mmol/1 respectively (t,5 = 8.0: P < 0.001 ). It is established that the half time for decay (to.5) of the K+-potential is inversely related to its amplitude: the larger the amplitude the shorter is the t..5 [2, 9]. In our experiments t0.s of the K +potential recorded from the upper layers was smaller than the t..5 of the K+-potential recorded from the middle layers (mean _+ S.E.M. = 3.7 +_ 0.31 s, 4.5 _+0.29 s respectively; t,5 = 2.7: P < 0.05}. These differences in t,,5 and differences shown from superposition of the two curves (Fig. 1B) illustrate that [K+],, becomes equal in upper and middle layers at ca. 13 s (marked by the vertical dotted
line). Comparisons of recordings from three animals like those shown in Fig. 1A and B allow us to conclude that the SNP to local stimulation of the cortex decays to baseline when [K+],, in the upper and middle layers become equal. SNPs and K+-potentials recorded at similar cortical depths to transcallosal stimulation were similar to each other in contrast to those evoked by local stimulation. Amplitudes of the SNP in upper and middle layers to transcallosal stimulation were 0.68 +_ 0.06 and 0.56 _+ 0.05 mV (t,5 = 2.0: P = n.s.) respectively corresponding to A[K']o of 1.5 + 0.11 and 1.4 + 0.08 mmol/1 (I,5 - 1.0: P = n.s.). Fig. 2 shows an example of the depth profiles of K +potentials and the SNP evoked by local and transcallosal stimulations in one animal. On local stimulation, at a depth of 50/am A[K']o was 1.1 mmol/1 and at a depth of 300/am 0.4 mmol/l (Fig. 2A). At the same depths A[K+],,, evoked by transcallosal stimulation, was 1.5 retool/1 and 1.2 retool/l, respectively (Fig. 2B). It is important to stress that with local stimulation the attenuation of the A[K+]o from upper to middle layers was much more prominent than with transcallosal stimulation. The SNPs evoked by local and transcallosal stimulation also differed markedly from each other. The amplitude of the SNP recorded from a depth of 50/am with local stimulation was 1.3 mV and the amplitude of the SNP at the same depth, evoked by transcallosal stimulation, was 0.6 mV (in spite of greater A[K-],,). The SNP on local stimulation was attenuated sharply in the cortical depth and at 300~,00/am began to invert (Fig. 2A). The SNP evoked by transcallosal stimulation was insignificantly attenuated and did not show an}' tendency to invert at the above mentioned depths in the cortex (Fig. 2B). According to previously advanced hypothesis [6], short duration of the SNP in comparison with the K'potential and glial depolarization, is due to "smoothing" of the A[K ~]o. This results in attenuation and cessation of the extracellular electrical current. These processes proceed against a background of increased [K ~],, and continuous glial depolarization. The present results confirm this hypothesis and as illustrated in Fig. 1 show that the decay of the SNP to the baseline occurs when the [K ],~ in the upper and middle layers become equal but still remains higher than normal. A positive correlation between A[K-]o and the amplitude of SNP has been described in several publications [2, 4, 6], but, as mentioned, this correlation can be disturbed [8]. For instance (Fig. 2) a more pronounced increase in [K ~],, in the superficial layers produced by transcallosal stimulation was associated with a SNP of smaller amplitude than that elicited by local stimulation when A[K'],, was smaller. Recordings of K+-potentials and SNP depth profiles
74
B
A J,m " -
mV 0
"
0 .?
OO
~ 10 s
~
Oo
[i+ 100. •
200.
o 15o
• o
~
3 0 ~
_2",.
o
100
.8
0.8 ' ;'
1;6
0
•
zoo
0
1:2 K+ rnf'l '
11 mV
0
•
'300
o
4°e°J 00.4 , Ji,'O,
•
'
•
lS£ 300t •
/3"I
0
•
0
• 400
e0.4 i
•
i I
0.8
~8
0
•
• :
1.2 K÷ n ~ •
lll]
J
mV
Fig. 2. Depth profile of the SNP and of the K+-potentials evoked by local and transcallosal stimulation. Electrode arrangement shown in diagram: 1, electrodes for transcallosal stimulation; 2, electrodes for ipsilateral local stimulation; 3, K+-microelectrode. A,B: numbers on inc records indicate the depth from the cortical surface in/~m; upper traces, electrical field potentials; lower traces, K÷-potentials; graphs show the amplitude of SNPs (0) and K+-potentials (()) in mV and mmol/1, respectively, plotted as a function of the cortical depth in/lm. A: SNPs and K+-potentials elicited by a single stimulus (50 V, 0.1 ms) applied to the cortical surface through the electrode 2. B: the same, evoked by transcallosal stimulation (20 stimuli, 20 Hz, 80 V, 0.3 ms) through electrode l.
show that on local stimulation the difference in values of A[K+]o in the upper and middle layers was much more marked than on transcallosal stimulation. It is suggested that the difference in the degree of depolarization of glial cells located in the upper and middle layers is also much more prominent on local stimulation of the cortical surface. Evidently, there might appear a more intensive electrical current in the extracellular space with local stimulation than with transcallosal stimulation. It is suggested that local stimulation of the cortical surface evokes activation of neuronal elements and a local increase in [K+]o predominantly in the superficial layers, whereas with transcallosal stimulation the sources of the K ÷ are distributed more diffusely in the cortical depth. This corresponds with our electrophysiological observations, described here (Fig. 2B) and elsewhere [3]. The data presented here demonstrate that measures of the SNP are dependent not only on magnitude and duration of changes in [K+]o but also on the depth distribution of K ÷ in the cortex.
We thank R Stys and R Laming for reading and correcting the manuscript and also R. Kapel and K. Kikabidze for technical assistance. 1 Bobrov, A.V., Cortical electrical field during slow component of the cortical direct response, Fiziol. Zh. SSSR, 57 (1971) 656-663. 2 Heinemann, U., Lux, H.D., Marciani, M,G. and Hofmeier, G.,
Slow potentials in relation to changes in extracellular potassium activity in the cortex of cats. In E.-J. Speckman and H. Caspers (Eds.), Origin of Cerebral Field Potentials, Georg Yhieme, Stuttgart, 1979, pp. 33-48. 3 Kikabidze, K.G. and Ocherashvili, I.V., Slow negativity evoked with transcallosal stimulation in the cat cortex, Fiziol. Zh. SSSR, 73 (1987) 1032 1038. 40cherashvili, I.V., Slow electrical field potentials arising after the primary response in the somatosensory cortex at the stimulation of the thalamic ventroposterolateral nucleus in cat, Neurofiziologia, 17 (1985) 435-441. 5 Roitbak, A.I., Slow negative potentials of the cortex and neuroglia~ In V.V. Parin (Ed.), Current Problems in the Physiology and Pathology of the Nervous System, Meditsina, Moscow, 1965, pp. 68-92. 6 Roitbak, A.I., Neuroglia: properties, functions and significance in nervous activity. In Soy. Sci. Rev. F. Physiol. Gen. Biol. Vol. 2, Harwood Academic Publishers GmbH, UK, 1988, pp. 355-402. 7 Roitbak, A.I., Fanardjian, V.V., Melkonyan, D.S. and Melkonyan, A.A., Contribution of glia and neurons to the surface-negative potentials of the cerebral cortex during its electrical stimulation, Neuroscience, 20 (1987) 1057-1067. 8 Roitbak, A.I., Ocherashvili, 1.V. and Kapel, R.G., Prolonged surface-negative potentials and the changes of the extracetlutar potassium concentration in the cerebral cortex at different parameters of its electrical stimulation, Fiziol. Zh. SSSR, 73 (1987) 227 283. 9 Roitbak A.I. and Ocherashvili, I.V., Changes in extracellular potassium concentration in the cerebral cortex at different para, meters of its electrical stimulation, Fiziol. Zh. SSSR, 73 (1987) 277: 283. 10 Somjen, G.G.,Electrophysiology of neuroglia, Annu. Rev. Physiol., 37 (1975) 163-190.
%cur..~iem'c Lcl~', i~. i~6 t i g92~ ~2 , 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved (t~04-3940'~2'$ i):, c)~
NSL 08408
Relationships between electrically induced slow negative potentials and changes in extracellular potassium concentrations in cerebral cortex of the cat Elijah Ocherashvili a n d A l e x a n d e r R o i t b a k Beritashvili Institute of Physiology, Georgian Academy (~fSeienees, Tbilisi {Georgia J (Received 15 April 199 l: Revised version received 19 November 1991: Accepted 27 November 1991 )
Key words: Cerebral cortex: Slow negative potential: Extracellular potassium concentration Experiments were carried out on cats under deep nembutal anaesthesia. Local cortical stimulation evoked both a slow negative potential (SNP) and an increase in extracellular potassium ([K+]o) which were maximal in the upper cortical layers. The decline in the K*-potential in deeper layers was slower than in the upper ones so that with time [K-]o became equal in upper and deeper layers. No SNP was recorded at that time, perhaps explaining why K--potential decay is of longer duration than the SNE Transcallosal stimulation also evoked an SNP but gave a lesser difference in A[K+],~in upper and deeper layers. The lower amplitude of the SNP evoked by transcallosal stimulation may be due to these circumstances. These data could clarify the mechanism of correlation disturbance between SNPs and K ~-potentials in the cerebral cortex.
It has been established that slow negative potentials (SNP) in different regions of the nervous system are associated with an increase in extracellular potassium concentration ([K+]o) and glial depolarization [2, 4, 6, 10]. Triple correlations between these measures have confirmed the supposition of the glial origin of the SNP [5, 10]. Mathematical analysis has shown that the contribution of neuronal elements to SNP generation evoked by surface cortical stimulation is insignificant [7]. However, it is also known that the correlation between SNE K ~potential and glial depolarization can be disturbed: SNP normally decays more rapidly than glial depolarization and the K+-potential. The positive correlation between the amplitude of the SNP and A[K+]o is changed at supramaximal intensities or high frequencies of stimulation [8]. The present investigation was aimed at elucidating the mechanisms of these discrepancies. Experiments were performed on 11 cats under nembutal anaesthesia (60-70 mg/kg). The sites of stimulation and recording were on the suprasylvian gyrus. Bipolar stimulating electrodes were used for local stimulation of the cortical surface (local stimulation) and transcallosal stimulation with 200/lm and 1.5 mm interpolar distance respectively. Stimulating electrodes were located on the Correspondence: E. Ocherashvili, Beritashvili Institute of Physiology. Georgian Academy of Sciences, Tbilisi. Georgia.
pial surface of the cortex. Local stimulation was performed by single stimuli and transcallosal stimulation by brief rhythmic trains, because a single stimulus evoked very weak A[K~]o and SNP or failed to evoke these. At a distance of 1 mm from the electrodes which were used for local stimulation, an Ag-AgC1 macroelectrode and a double barreled microelectrode were placed, the latter with one K*-selective capillary (for measuring the A[K+]o) and one NaC1 reference capillary (used also for recording the electrical field potential). After the arrangement of electrodes, the cortex was covered with a mixture of dental wax and vaseline oil, to prevent brain pulsations. The temperature of the cortex was 29-30 ° C. Anaesthesia and low cortex temperature abolished spontaneous activity and allowed stable recording over 4-5 h. K+-potentials and SNPs were recorded with use of high impedance negative capacity input preamplifiers and displayed on ink recorders. Fig. 1 shows that on local stimulation a SNP was recorded from the surface while a positive potential was recorded from the middle layers of the cortex. The mean + S.E.M. of the SNP amplitude was 0.95 _+ 0.18 mV for the upper layers. (It is known [1] that the SNP inverts deep within the cortex.) In the example shown in Fig. l A[K+]o was ca. 1.3 mmol/1 in the upper layers (100/lm) and ca. 0.74 mmol/l in the middle ones (500 jim), where the positive potential was recorded (Fig. I B). Com-
73
A
!
mV
o_
B
0.1
4,
K+
3 Fig. 1. Connection between the duration of the stimulus-induced SNP and changes in [K*],, in the upper and middle layers of the cortex. Electrode arrangement shown in diagram: 1, macroelectrode: 2. K ~microelectrode: 3, stimulating electrodes. Single stimulus (40 V, 0.1 ms) was applied to the cortical surface. Alter recording the SNP from the cortical surface and K~-potential from the upper layers (0.1 mm) the K+-microelectrode was inserted to a depth of 0.5 mm. A: SNP (upper trace) recorded from the cortical surface by means of microelectrode 1 and positive potential (lower trace) recorded from the depth of the cortex (0.5 ram) by means of the K'-microelectrode reference channel, B: superposition of K-potentials recorded from upper (0.1 ram) and middle (0.5 mm) layers. Vertical dotted lines in A show the time of the decay of the SNP to the baseline and in B the time when [K-]o in upper and middle layers became equal.
parable figures for A[K+]~,for upper and middle layers for six animals were 1.1 + 0.09 and 0.38 + 0.08 mmol/1 respectively (t,5 = 8.0: P < 0.001 ). It is established that the half time for decay (to.5) of the K+-potential is inversely related to its amplitude: the larger the amplitude the shorter is the t..5 [2, 9]. In our experiments t0.s of the K +potential recorded from the upper layers was smaller than the t..5 of the K+-potential recorded from the middle layers (mean _+ S.E.M. = 3.7 +_ 0.31 s, 4.5 _+0.29 s respectively; t,5 = 2.7: P < 0.05}. These differences in t,,5 and differences shown from superposition of the two curves (Fig. 1B) illustrate that [K+],, becomes equal in upper and middle layers at ca. 13 s (marked by the vertical dotted
line). Comparisons of recordings from three animals like those shown in Fig. 1A and B allow us to conclude that the SNP to local stimulation of the cortex decays to baseline when [K+],, in the upper and middle layers become equal. SNPs and K+-potentials recorded at similar cortical depths to transcallosal stimulation were similar to each other in contrast to those evoked by local stimulation. Amplitudes of the SNP in upper and middle layers to transcallosal stimulation were 0.68 +_ 0.06 and 0.56 _+ 0.05 mV (t,5 = 2.0: P = n.s.) respectively corresponding to A[K']o of 1.5 + 0.11 and 1.4 + 0.08 mmol/1 (I,5 - 1.0: P = n.s.). Fig. 2 shows an example of the depth profiles of K +potentials and the SNP evoked by local and transcallosal stimulations in one animal. On local stimulation, at a depth of 50/am A[K']o was 1.1 mmol/1 and at a depth of 300/am 0.4 mmol/l (Fig. 2A). At the same depths A[K+],,, evoked by transcallosal stimulation, was 1.5 retool/1 and 1.2 retool/l, respectively (Fig. 2B). It is important to stress that with local stimulation the attenuation of the A[K+]o from upper to middle layers was much more prominent than with transcallosal stimulation. The SNPs evoked by local and transcallosal stimulation also differed markedly from each other. The amplitude of the SNP recorded from a depth of 50/am with local stimulation was 1.3 mV and the amplitude of the SNP at the same depth, evoked by transcallosal stimulation, was 0.6 mV (in spite of greater A[K-],,). The SNP on local stimulation was attenuated sharply in the cortical depth and at 300~,00/am began to invert (Fig. 2A). The SNP evoked by transcallosal stimulation was insignificantly attenuated and did not show an}' tendency to invert at the above mentioned depths in the cortex (Fig. 2B). According to previously advanced hypothesis [6], short duration of the SNP in comparison with the K'potential and glial depolarization, is due to "smoothing" of the A[K ~]o. This results in attenuation and cessation of the extracellular electrical current. These processes proceed against a background of increased [K ~],, and continuous glial depolarization. The present results confirm this hypothesis and as illustrated in Fig. 1 show that the decay of the SNP to the baseline occurs when the [K ],~ in the upper and middle layers become equal but still remains higher than normal. A positive correlation between A[K-]o and the amplitude of SNP has been described in several publications [2, 4, 6], but, as mentioned, this correlation can be disturbed [8]. For instance (Fig. 2) a more pronounced increase in [K ~],, in the superficial layers produced by transcallosal stimulation was associated with a SNP of smaller amplitude than that elicited by local stimulation when A[K'],, was smaller. Recordings of K+-potentials and SNP depth profiles
74
B
A J,m " -
mV 0
"
0 .?
OO
~ 10 s
~
Oo
[i+ 100. •
200.
o 15o
• o
~
3 0 ~
_2",.
o
100
.8
0.8 ' ;'
1;6
0
•
zoo
0
1:2 K+ rnf'l '
11 mV
0
•
'300
o
4°e°J 00.4 , Ji,'O,
•
'
•
lS£ 300t •
/3"I
0
•
0
• 400
e0.4 i
•
i I
0.8
~8
0
•
• :
1.2 K÷ n ~ •
lll]
J
mV
Fig. 2. Depth profile of the SNP and of the K+-potentials evoked by local and transcallosal stimulation. Electrode arrangement shown in diagram: 1, electrodes for transcallosal stimulation; 2, electrodes for ipsilateral local stimulation; 3, K+-microelectrode. A,B: numbers on inc records indicate the depth from the cortical surface in/~m; upper traces, electrical field potentials; lower traces, K÷-potentials; graphs show the amplitude of SNPs (0) and K+-potentials (()) in mV and mmol/1, respectively, plotted as a function of the cortical depth in/lm. A: SNPs and K+-potentials elicited by a single stimulus (50 V, 0.1 ms) applied to the cortical surface through the electrode 2. B: the same, evoked by transcallosal stimulation (20 stimuli, 20 Hz, 80 V, 0.3 ms) through electrode l.
show that on local stimulation the difference in values of A[K+]o in the upper and middle layers was much more marked than on transcallosal stimulation. It is suggested that the difference in the degree of depolarization of glial cells located in the upper and middle layers is also much more prominent on local stimulation of the cortical surface. Evidently, there might appear a more intensive electrical current in the extracellular space with local stimulation than with transcallosal stimulation. It is suggested that local stimulation of the cortical surface evokes activation of neuronal elements and a local increase in [K+]o predominantly in the superficial layers, whereas with transcallosal stimulation the sources of the K ÷ are distributed more diffusely in the cortical depth. This corresponds with our electrophysiological observations, described here (Fig. 2B) and elsewhere [3]. The data presented here demonstrate that measures of the SNP are dependent not only on magnitude and duration of changes in [K+]o but also on the depth distribution of K ÷ in the cortex.
We thank R Stys and R Laming for reading and correcting the manuscript and also R. Kapel and K. Kikabidze for technical assistance. 1 Bobrov, A.V., Cortical electrical field during slow component of the cortical direct response, Fiziol. Zh. SSSR, 57 (1971) 656-663. 2 Heinemann, U., Lux, H.D., Marciani, M,G. and Hofmeier, G.,
Slow potentials in relation to changes in extracellular potassium activity in the cortex of cats. In E.-J. Speckman and H. Caspers (Eds.), Origin of Cerebral Field Potentials, Georg Yhieme, Stuttgart, 1979, pp. 33-48. 3 Kikabidze, K.G. and Ocherashvili, I.V., Slow negativity evoked with transcallosal stimulation in the cat cortex, Fiziol. Zh. SSSR, 73 (1987) 1032 1038. 40cherashvili, I.V., Slow electrical field potentials arising after the primary response in the somatosensory cortex at the stimulation of the thalamic ventroposterolateral nucleus in cat, Neurofiziologia, 17 (1985) 435-441. 5 Roitbak, A.I., Slow negative potentials of the cortex and neuroglia~ In V.V. Parin (Ed.), Current Problems in the Physiology and Pathology of the Nervous System, Meditsina, Moscow, 1965, pp. 68-92. 6 Roitbak, A.I., Neuroglia: properties, functions and significance in nervous activity. In Soy. Sci. Rev. F. Physiol. Gen. Biol. Vol. 2, Harwood Academic Publishers GmbH, UK, 1988, pp. 355-402. 7 Roitbak, A.I., Fanardjian, V.V., Melkonyan, D.S. and Melkonyan, A.A., Contribution of glia and neurons to the surface-negative potentials of the cerebral cortex during its electrical stimulation, Neuroscience, 20 (1987) 1057-1067. 8 Roitbak, A.I., Ocherashvili, 1.V. and Kapel, R.G., Prolonged surface-negative potentials and the changes of the extracetlutar potassium concentration in the cerebral cortex at different parameters of its electrical stimulation, Fiziol. Zh. SSSR, 73 (1987) 227 283. 9 Roitbak A.I. and Ocherashvili, I.V., Changes in extracellular potassium concentration in the cerebral cortex at different para, meters of its electrical stimulation, Fiziol. Zh. SSSR, 73 (1987) 277: 283. 10 Somjen, G.G.,Electrophysiology of neuroglia, Annu. Rev. Physiol., 37 (1975) 163-190.
