- Email: [email protected]
Calcium-dependent component of massive increase in extracellular potassium during cerebral ischemia as demonstrated by microdialysis in vivo
Brain Research, 567 (1991) 57-63 © 1991 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/91/$03.50
57
BRES 17241
Calcium-dependent component of massive increase in extracellular potassium during cerebral ischemia as demonstrated by microdialysis in vivo Yoichi Katayama 1'2., Toru Tamura 1, Donald P. Becker I and Takashi Tsubokawa 2 1Division of Neurosurgery, UCLA School of Medicine, University of California at Los Angeles, CA 90024 (U.S.A.) and 2Department of Neurological Surgery, Nihon University School of Medicine, Tokyo (Japan) (Accepted 23 July 1991) Key words: Cerebral ischemia; Calcium ion; Potassium ion; Microdialysis; Spreading depression This study characterizes the physiological features and limitations of K+-free dialysis to detect changes in extracellular concentration of K ÷ ([K+]c) in the rat hippoeampus in vivo. It also demonstrates the effects of Ca2+-free perfusate containing Co2+ or Mg2+, which blocks Ca2+ entry into the presynaptic nerve terminal, on the abrupt increase in [K+]c detected by this technique during cerebral ischemia. K+-free dialysis for 40 rain caused no significant changes in the baseline [K+]c. In contrast, Ca2+-free dialysis for 40 min significantly reduced the extracellular Ca2+ concentration. Under this condition, together with addition of Co2+ or Mg2+ to the perfusate, the increase in [K+]c was delayed, and a delay in reaching the maximum level was observed in a dose-dependent manner. These results are consistent with the hypothesis that the initial increase in [K+]. during cerebral isehemia is related to the Ca2+-dependent exocytotic release of neurotransmitters from depolarized nerve terminals.
INTRODUCTION During cerebral ischemia, an abrupt increase in extracellular concentration of K + ([K+].) to 50-60 mM is observed following a slow increase 3'5't1'15'33. Since this rapid and dramatic increase in [K+], invariably begins when [K+]¢ reaches a level of 6-10 mM 3'5'11'15'33, neurotransmitter release and a sudden K + flux through transmitter-gated ion channels of neuronal cells have been postulated to be the cause of this phenomenon 22'25'32. ThUS, when [K+]e reaches this level, the nerve terminals may be depolarized, causing Ca 2÷ entry into the presynaptic nerve terminal, and neurotransmitters are thereby released. Among the various neurotransmitters, excitatory amino acids (EAAs) seem the most likely substances which could produce such a remarkable K + flux 7"21'27'28. Ca2+-dependent exocytotic release of E A A s in response to an elevated [K+]e has been demonstrated by in vitro studies, employing synaptosome preparations or brain slices 2'6'2°'23'24. It has also been found by microdialysis in vivo that EAAs are in fact released during cerebral ischemia 4'1°'1sA9, and this release is completely abolished when Ca2÷-free perfusate containing Co 2÷, which blocks Ca 2÷ entry into the nerve terminal, is used 1°,t9.
If the Ca2+-dependent exocytotic release of neurotransmitters, especiaUy EAAs, is responsible for the abrupt increase in [K+]e, inhibition of C a 2+ entry into the nerve terminal would tend to alter the time course of the changes in [K+]e during cerebral ischemia. The present study examined this hypothesis by testing the effect of Ca2+-free perfusate containing C o 2+ o r Mg 2+, which blocks C a 2+ entry, on the changes in [K+]¢ detected by microdialysis2°'21 during cerebral ischemia. While the detection of changes in [K+], with microdialysis has not been described previously, such a technique would provide a unique opportunity to examine the mechanism of the changes in [K+], during ischemia, by monitoring other neurochemical changes simultaneously and, more importantly, by manipulating neurochemical processes through the administration of various agents in situ via the dialysis probe. In order to facilitate the detection of [K+], changes by microdialysis, however, the use of K+-free perfusate for the dialysis is necessary. This would inevitably cause a slow but continuous removal of K ÷ from the extracellular space (ECS) and could potentially produce an unphysiologically lowered [K+L level. The present study, therefore, first characterized the physiological features and limitations of this technique for the detection of changes in [K+], and then investigated the effects of Ca2+-free perfusate containing C o 2+ o r Mg 2+.
*On leave from Department of Neurological Surgery, Nihon University School of Medicine, Tokyo, Japan. Correspondence: Y. Katayama, Department of Neurological Surgery, Nihon University School of Medicine, Itabashi-ku, Tokyo 173, Japan. Fax: (81) 3 3554 0425.
58 MATERIALS AND METHODS
Procedures for animal preparation Young adult Sprague-Dawley rats (n = 36) weighing 180-240 g were used. The animals were maintained in an environmentally controlled room with a 12 h light/dark cycle and were allowed free access to food and water. They were anesthetized with a mixture of nitrous oxide (66%), oxygen (33%) and enflurane (1%), and placed in a stereotaxic frame with the nosebar setting at 2.5 mm below the interaural level. Skin wounds and pressure points were infiltrated with 1% xylocaine. The rectal temperature was maintained at 3738 °C using a heating pad. A pair of dialysis probes (CMA/10, Bioanalytical System Inc.; o.d., 500 #; effective length, 3 mm; cut off, 20,000 MW) were lowered vertically through small skull holes into the brain, placing the tip in the hippocampus (3.8-4.8 mm caudal to the bregma, 2.0 mm lateral to the midline and 3.5 mm below the surface of the dura). Thus, approximately 50% of the effective length of the probe was located within the hippocampus. The probes were initially perfused with Ringer solution (adjusted to pH 7.4) at a rate of 5,0 #l/min.
Procedures for ischemia induction and microdialysis Global cerebral ischemia was induced by decapitation. The temperature of the perfusate was prepared to be at 38 °C, and the temperature was maintained with a chamber filled with water at 38 °C in which the whole length of the inlet tubing of the dialysis system was placed. Dialysate fractions were collected at 1 min intervals. The length of the outlet tube was adjusted to a length that resulted in the dead space of the probe and tubing being 5.0 M1. Thus, there would be a 1 min delay between the changes observed in the dialysate and the actual changes occurring in the brain. At the end of the experiment, the position of the probe was confirmed anatomically. In order to facilitate the detection of [K+]~ changes by microdialysis, K+-free perfusate was used for the dialysis. The K ÷ concentration of each dialysate ([K+]d) was measured with a K+-sensitive electrode in vitro. The recovery rate of the dialysis system for K + in vitro was determined using a calibration solution ([K+]d/ [K+L,lib). The K ÷ concentration of the calibration solution
([K+]calib) was varied from 1 to 100 mM, replacing equimolar sodium chloride in the Ringer solution. The recovery rate obtained in vitro ranged from 6 to 13% at the flow rate of the present study. A reasonably linear relationship was observed between the [K+]o and the [K+]¢anb within the physiological range (Fig. 1A). The [K+]d changed rapidly in response to sudden changes in [K+]calibwith a 1 min delay for the dead space of the dialysis system (Fig. 1B). In order to confirm Ca z+ depletion with the Ca2÷-free dialysis (see below), the Ca z+ concentration ([Ca2+]o) was similarly measured with a CaZ÷-sensitive electrode in vitro. Evaluation of the effects of K+-free dialysis In order to evaluate the effect of K+-free dialysis on [K÷]~, two sets of experimental data were analyzed: (1) the sequential changes in [K+]a obtained from the same probe during the continued K +free dialysis; and (2) the differences in [K÷]o obtained from two probes in the same animal each perfused with K÷-free perfusate for a different duration. Thus, the dialysis with the K÷-free Ringer solution was first initiated with one of the two probes (test dialysis) while the other probe, placed in the contralateral hippocampus, was perfused with normal Ringer solution. The K+-free dialysis with the second probe was then initiated at 20, 40 or 60 min after the onset of the test dialysis (control dialysis). In 3 experiments, the changes in [K+]e were recorded with a K+-sensitive electrode in vivo in the hippocampus near the site of the test dialysis (2 mm lateral from the tip of the dialysis probe and 3.0 mm below the surface of the cortex). The K+-sensitive electrode was made by glass micropipettes having a common tip diameter of 3 M. Cerebral ischemia was induced 5 min following the initiation of the control dialysis. The time of ischemia induction therefore corresponded to 25 (n -- 6), 45 (n = 5) or 65 (n = 70 min after the initiation of the test dialysis.
Evaluation of the effects of Cae+-free dialysis In one of the two probes chosen at random, the perfusate was switched to CaZ+-free Ringer solution 5 min following the initiation of perfusion (test dialysis). The other probe was perfused with a solution containing Ca 2÷ throughout the experiment (control dialysis). The perfusates for the control and test dialysis were then switched to K+-free, or K+4Ca2+-free Ringer solutions, respectively, at 40 min after the initiation of the perfusion, and measurements of [K+]d were started. In most animals, cobalt chloride (1-10 mM; n = 26) or magnesium chloride (10 mM; n = 6) was added to
6.0
z
A
0
5.0-
~g
4o-
0.3
A POTASSIUM-FREEDIALYSIS z~ ©m
O< LUI-
3.0
m u.
20
1
-- 0.2
~2 zr~
8~ 0.1 -'0 00
• 20
5 10 15 20 25 30 35 TIME (MIN)
40 60 80 POTASSIUMCONCENTRATION OF CALIBRATIONSOLUTION (mM)
100
DEAD S 120
Fig. 1. A: representative example of the relationship between the K ÷ concentration of the calibration solution ([K÷]¢aiib) and the dialysate ([K÷]d). Results from 4 measurements with the same probe are plotted (closed circles). The curve represents a polynomial approximation of the relationship. The slope of the curve ([K+]J[K+L.Iib) represents the recovery rate in vitro. B: changes in [K÷ld (in vitro) in response to sudden changes in [K÷]~alib (open circles). The changes in [K+]~.lib (from 1 mM to 60 raM) are illustrated schematically by a line (no relation to concentration scale). The responses are delayed for 1 rain due to the dead space in the outlet tubing of the dialysis system (5 MI).
--~l
o
P
B
~
CALCIUM-FREEDIALYSIS
0.0 0
5 10 15 DURATION OF POTASSIUM-FREE(A) OR CALCIUM-FREE(B) DIALYSIS(MIN)
20
Fig. 2. Changes in baseline level of [K+]d during K+-free dialysis (A, n = 18) and baseline level of [Ca2+]d during Ca2+-free dialysis (B, n = 7). The perfusate was changed to K+-free or Ca2÷-free Ringer solution at time zero. [K+]d and [Ca2+]d during the initial 3 fractions are not shown because of the levels of K ÷ (3 raM) and Ca 2+ (1 mM) contained in the previous perfusate. The responses are delayed due to the dead space in the inlet as well as outlet tubing of the dialysis system. Each curve represents a logarithmic approximation of the changes.
59 TABLE I
Effects of K+-free dialysis K+-free dialysis (duration)
n
Pre-ischemia baseline [K+]d (raM)
Latency (rain)* Onset (L1)
Maximum (L2)
Post-ischemia maximum [K+]d (raM)
Control (5 min) Test 25 min)
6
0.22 -+ 0.02 0.21 --- 0.02
2.5 --- 0.4 2.7 --+ 0.3
3.5 --- 0.4 3.7 --- 0.3
1.33 - 0.12 1.29 - 0.10
Control (5 rain) Test (45 min)
5
0.21 -+ 0.03 0.19 -+ 0.03
2.4 - 0.4 2.6 +-- 0.5
3.8 -+ 0.4 3.8 --- 0.5
1.27 - 0.12 1.19 --+ 0.10
Control (5 rain) Test (65 min)
7
0.20 -+ 0.01 0.17 -+ 0.01"*
2.3 -+ 0.4 2.7 --- 0.3
3.3 --- 0.4 4.0 + 0.3
1.30 - 0.11 1.10 +-- 0.10
Paired data (mean --- S.E.M.); *a fraction for dead space (1 min) is excluded. **P < 0.05, as compared to the value at 5 min obtained from the same dialysis site with Student paired t test. No significant difference detected in comparison to the control dialysis site.
the perfusate for the test dialysis. The osmolarity of each perfusate was maintained constant by changing the concentration of sodium chloride. Cerebral ischemia was induced 5 rain following the initiation of the [K+]a measurements.
Statistical analysis The changes in [K+]a values obtained in each animal were analyzed by Student paired t test. The latencies of defined events were expressed as the numbers of the fraction sampled after isehemia and compared between the test and control probes by Wilcoxon matched-pair signed rank test.
RESULTS
Detection of changes in [K+]e with K+-free dialysis +
Following the initiation of the K -free dialysis in vivo,
1.2
TIME AFTER ISCHEMIA INDUCTION (MIN) -2 0 2 4 6 8 10 12 14 i I I i i i i i i ~
-4
1.o-
H
16 i
A
DEAD SPACE
~ 0.8-
[K+]o decreased rapidly from the higher values due to the previous K+-containing perfusate, and a stationary level of [K+]a ranging from 0.14 to 0.29 mM was obtained within 5 rain (0.22 - 0.01 mM, n = 18; Fig. 2A). Although a very slow, progressive decrease did occur thereafter, the [K+]o appeared to remain at approximately the same level for a considerable period of time (Fig. 2A and Table I). The presence of a slight but significant decrease in the baseline [K+]d became apparent only when the K+-free dialysis was continued for 60 min (5 min vs 65 min at the same site, P < 0.05, n = 7, Table I). The recording of [K+]~ with a K+-sensitive electrode in the proximity of the site of dialysis (distance, 2 mm) failed to reveal a decrease in baseline [K+]¢ during the prolonged K+-free dialysis.
/
/
1.8-
~ i / 1 ~
z
O
DA I LYSS I
oct
A
1.5" _z o
i 0.6-
o
20
0
w
20
40
60
80
100
•
0.9-
!
t o.o-
o01(~ 0.2 ~
--I
10
B
--
0.3-
0.0 v.v
1
-5 -'3 -1 ; '3 5 7 9 11 13 15 ACTUAL TIME AFTER ISCHEMIA INDUCTION (MIN)
Fig. 3. Comparison of time courses of [K+]d measured by microdialysis (A, closed circles) and [K+]e measured with a K+-sensitive electrode (B) during cerebral ischemia. Each symbol representing [K+]a of the 1 min fraction is located at the midpoint of the period in which the fraction was collected. Perfusion rate, 5.0 #l/rain. The difference in time scale for A ([K+]d), and B ([K+]¢) accounts for the delay (1 rain) of the [K+]d response due to the dead space (5 #1) of the outlet side of the dialysis system. The ischemia was induced 20 min after the onset of K+-free dialysis.
0
20 40 60 80 100 EXTRACELLULAR CONCENTRATION OF POTASSIUM (mM)
Fig. 4. A: representative example of relationships between [K+]e measured with a microelectrode (average during 1 min period for dialysate collection) and [K+]d detected microdialysis in the same animal. Date for varying [K+]e were from the period during ischemia. The curve represents an exponential approximation of the relationship. The slope of the curve ([K+]J[K+]c) represents the recovery rate in vivo (of. Fig. 1A). B: representative example of the changes in ratio of the recovery rates in vivo vs in vitro (in vivo/in vitro) at various [K+]e during ischemia. The curve represents a polynomial approximation of the changes.
60 TABLE IT
Effects of Ca2+-free dialysis with Co2+ of Mg2+ on baseline and maximum [K+]d
[K+ld (raM)
Control dialysis Ca2+-free dialysis (n -- 7) Control dialysis Ca2+-free dialysis with
Co 2+
(10 mM) (n = 13)
Control dialysis Ca2+-free dialysis with Mg2+ (10 mM) (n = 6)
Pre-ischemic baseline level
Post-ischemic maximum level
0.21 --- 0,01 0.20 + 0,01
1.28 --- 0.10 1.25 --- 0.11
0.19 --- 0.01 0.18 -+ 0.01
1.18 --- 0.09 1.13 --- 0.11
0.21 --- 0,02 0.21 --- 0,02
1.50 --+ 0.10 1.45 -+ 0.07
Paired data (mean -+ S.E.M.); no significant differences were seen between the two dialysis sites with Student paired t-test.
•
A
0.8" 1.0"
~
N
0.6" 0.40.2" ~
o'12-B 0.0
v
-4 1.4-
TROL CO 0mM
---o--2
0
2
4
, 8
6
10
12
, 14
16
1.0-
O z 0
0.8-
<
0.6-
z 0.4 ,,, O z 0.2 O O 19
0.o
< -4 >- 1.6
f,/)
---o-.
-2
.
0
.
.
2
.
4
.
.
6
.
8
.
10
Induction of the global ischemia produced a slight increase in [K+]d in one or more fractions, in addition to the fraction for the dead space, and a marked increase in subsequent fractions (Fig. 3A). The onset of the sudden, large increase corresponded to an abrupt increase in [K+]e as measured by the K+-sensitive electrode in the proximity of the site of dialysis, when a 1 min delay for the response in [K+]d due to the dead space was taken into account (cf. Fig. 3A and B). The [K+L measured by the electrode showed an initial slow increase from a baseline value of approximately 3 mM to a level of 6 mM, and subsequently a rapid increase up to values ranging from 60 to 80 mM (Fig. 3B). The [K+]d m e a s u r e d from the dialysate increased invariably to a max-
ROL Co 5ram 12
14
2.0
16
~==
~ 10
uJ ~
1.0 0.8"
0.6" 0.4" 0.2-
CO
m Z
1.21
+
CONTROL Co lOmM
~ IN ~
,,<,
L2 (TESF-CONT)
,
. . . . . .
l
TIME
Oa. WZ
8o
0.5
z
o0
Z
TEST
a -0.5 CONTROL
0.0 -4
[]L'(rTEST-CONT) []
-2 0 2 4 6 8 10 12 14 TIME AFTER ISCHEMIA INDUCTION (MIN)
16
Fig. 5. Representative examples of the time course of changes in [K+]d during isehemia with CaZ+-free dialysis employing various concentrations of Co2+ (test dialysis, open circles). The time course of the changes in [K+]a detected in the same animal with the probe perfused with Ca 2+ and without Co 2+ is also shown (control dialysis, filled circles). A: 0 mM Co2+; B: 5 mM CoZ+; C: 10 mM Co 2÷. Each symbol representing [K+]a of the 1 min fraction is located at the midpoint of the period in which the fraction was collected. Perfusion rate, 5.0/tl/min. The response of [K+]d to the ischemia is delayed 1 min due to the dead space (5 #1) of the outlet side of the dialysis system.
CO 0mM
Co 1raM
Co 5raM
Col0mM
CALCIUM-FREE DIALYSIS
Fig. 6. Effects of Ca2+-free dialysis employing various concentrations of Co2+ (test dialysis) on the latencies of onset of the increase in [K+]a (L 0 and the time at which the maximum level of increased [K+]d was reached (Lz). The effects are demonstrated in comparison with the latencies observed in the control dialysis in the same animal (test-control). The latencies were measured as the number of the 1 min dialysis fractions. The onset (L1) was defined as the fraction which demonstrated a sudden increase in [K+]d reaching more than 2-fold baseline, and the time at which the maximum level was attained (L2) was defined as the fraction reaching a level within 2 S.D. of the higher [K+]d level after the onset. *P < 0.01, n = 13, Wilcoxon matched-pair signed rank test.
61 TABLE III Effects of Ca2+-free dialysis with Co2+ or Mg2+ on time courses of the increase in [K+]e during ischemia Latency (rain)* Onset (L1)
Maximum (L2)
Duration (min) .from the onset to the maximum (L2-L~)
Control dialysis Ca2+-free dialysis with 10 mM Co2÷ (n = 13)
2.9 --- 0.3 4.0 --- 0.3***
4.5 --- 0.4 6.2 +--0.3***
1.6 - 0.2 2.2 -+ 0.2**
Control dialysis Ca2+-free dialysis with 10 m M Mg 2+ (n -- 6)
2.2 --- 0.2 2.8 -+ 0.3**
3.5 --- 0.3 4.3 --- 0.4**
1.3 +- 0.2 1.7 --- 0.2
Paired date (mean +- S.E.M.); *the fraction for dead space is excluded. **P < 0.05, ***P < 0.01 Wilcoxon matched-pair signed rank test.
imum level ranging from 0.8 to 2.0 mM (1.30 - 0.11 raM, n = 18; Fig. 3A and Table I). There were no significant differences in the maximum level of [K+]d attained eventually during the ischemia (Table I). The onset of the large increase in [K÷]d was arbitrarily defined as the fraction that demonstrated a sudden increase in [K+]d reaching more than 2-fold baseline, and the latency of onset was expressed as the number of the fraction after ischemia induction excluding the dead space (L1). The time that the maximum level of [K+]d was attained was defined as the fraction reaching a [K+]d level within 2 S.D. of the higher [K+]d level, and its latency was similarly expressed as the number of the fraction (1-2). The difference in duration of K+-free dialysis between the test and control dialysis, which varied from 20 to 60 min, resulted in no significant difference in these latencies (Table I). The results of direct recording of the changes in [K+]~ with the K+-sensitive electrode were used for computation of the recovery rate of the dialysis system for K ÷ in vivo ([K+]d/[K+]e). The obtained recovery rate in vivo was always smaller than the recovery rate determined in vitro (Fig. 4A, cf. Fig. 1A). It was also found that the ratio of the recovery rate in vivo to the recovery rate in vitro dropped further following the ischemia induction in association with the elevation of [K+]e during ischemia (Fig. 4B). Modification o f changes in [K+]e with Ca2+-free dialysis
Following the initiation of the Ca2+-free dialysis, [Ca2+]d decreased rapidly, and a relatively stationary level of [Ca2+]d, ranging from 0.08 to 0.12 mM, was obtained within 5 min (Fig. 2B, filled circles). The higher values for the initial period were due to Ca 2÷ contained in the perfusate previously. A gradual decrease in [Ca2+]d, however, continued thereafter (Fig. 2B). This effect of the Ca2+-free dialysis on [Ca2+]d was in clear contrast to the effect of the K+-free dialysis on [K+]d (cf. Fig. 2A and B). The baseline [Ca2+]d often decreased to
less than half within 40 min following the initiation of Ca2+-free dialysis. The Ca2+-free dialysis, regardless of whether or not Co 2+ was added to the perfusate, produced no significant changes in baseline [K+]d (Table II). The development of the large increase in [K+]d following the induction of ischemia was significantly delayed by the dialysis with Ca2+-free perfusate containing Co 2÷ (Fig. 5B,C). No significant effect was detected with Ca2+-free dialysis alone (Fig. 5A). The effects of Co 2+ were demonstrated to be dose-dependent (Fig. 6A,B) as a prolonged latency of o n s e t (L1, P < 0.01, n = 13, Table III), an additional delay in reaching the maximum level (L 2, P < 0.01, n = 13; Table III) and a prolonged duration from the onset to the maximum level (L1-L2, P < 0.01, n = 13; Table III). Similar effects were also observed with Ca2+-free, Mg2+-containing dialysis (Table III). DISCUSSION If K+-free dialysis removed K ÷ only from the ECS, the removal of 0.2 mM at 5/A/min as dialysate would result in a loss of 1.0 nmol/min from the ECS which could deplete all the K ÷ in 3.3/zl ECS after 10 min. No remarkable changes in baseline [K+]d were observed in the present study, however, suggesting that K ÷ is replenished continuously by transcellular K + movement during K+-free dialysis 25. The [K+]d changed rapidly in response to sudden changes in the K ÷ concentration of the calibration solution and sensitively responded to the rapid increase in [K+]e during cerebral ischemia. It appears therefore that K+-free dialysis provides reasonable information regarding the timing of large changes of [K+]e, such as that observed during cerebral ischemia, in shortterm experiments. The recovery rate in vivo was always smaller than the recovery rate determined in vitro. This may reflect the lower effective surface area of the dialysis system in vivo which has recently been demonstrated at flow rates rang-
62 ing from 2 to 10/A/min 1. The further decrease in recovery rate in vivo occurring in association with the elevation of [K+]~ during ischemia may be accounted for by a decrease in the effective surface area of the dialysis system in vivo due presumably to shrinkage of the ECS, which has been demonstrated to take place concomitantly with the abrupt increase in [K+]e 14. Without knowing the changes in effective surface area of the dialysis system in vivo, microdialysis can provide only a qualitative measure of the changes in the concentration of any substance in the ECS under pathological conditions such as ischemia. K+-free dialysis is still useful, however, for identifying the timing of the abrupt K ÷ flux and for demonstrating other neurochemical changes simultaneously using the same dialysate fractions. Further, neurochemical processes can be manipulated experimentally through the administration of various agents in situ via the same dialysis probe. Thus, K+-free dialysis provides a less complicated and less invasive procedure for investigating various neurochemical changes associated with the sudden K + flux induced by cerebral ischemia. The present data showed that the abrupt and massive increase in [K+]~ occurring during cerebral ischemia was significantly delayed by dialysis with Ca2+-free peffusate containing Co 2÷ or Mg 2+. The rapid component of the massive increase in [K+]e, which normally lasts only for less than a minute, comprises approximately 75% of the maximum level of [K+]¢ attained during ischemia 11'12'2°. The observed effects therefore appear to involve inhibition of this rapid component. In view of the non-linear relationship between [K+]e and [K+]d, however, there is a possibility that inhibition of the subsequent slow component of [K+]~ is also involved. CaZ+-free dialysis produces a marked decrease in the baseline [Ca2+]~, indicating that the CaZ+-free dialysis removes Ca 2+ from the ECS rapidly 8'1°. Employing 45Ca autoradiography, we have confirmed Ca 2+ depletion within a brain area of approximately 2-3 mm in diameter by the Ca2+-free dialysis when continued for more than 30 min (unpublished observations). In addition, the presence of Co 2÷ or Mg 2÷ in the ECS inhibits Ca 2÷ entry into the presynaptic nerve terminal 1°As'19'33 or postsynaptic nerve cells. Thus, the results of the present study indicate that a major component of the initial increase in [K+]e during cerebral ischemia is dependent upon Ca 2÷ flux from the ECS into the cells. Ca 2+ has a number of actions when it enters the postsynaptic nerve cells, including the activation of numerous enzyme systems, and some of these changes result in an increase in neuronal cell membrane permeability29. The effects observed in the present study may, in part, be attributable to an inhibition of these processes. When Ca 2+ enters the presynaptic nerve terminal, it
mediates
the
exocytotic
release
of
neurotransmit-
ters 2'6A9'22'23. As mentioned earlier, it has been postulated that the abrupt increase in [K+]e occurring during ischemia is caused by neurotransmitter release from the nerve terminal depolarized by a [K+]e elevated to 6-10 mM 21'24'31. The presence of the CaZ+-dependent component demonstrated in the present study is consistent with this hypothesis, since exocytotic release of neurotransmitter is dependent upon Ca 2÷ entry into the presynaptic nerve terminal. It is possible that the induced ionic environment depresses metabolism and increases the glucose stores within the brain tissue 14. Thus, the observed delay in the increase in [K÷]~ during ischemia may be a consequence of a reduced metabolism and increased glucose content. We have, however, observed no changes in the local glucose utilization rate, as evaluated by [14C]deoxyglucose autoradiography, in the brain area peffused with Ca2+-free perfusate containing Co 2+ (unpublished observations). A more likely mechanism for the delay in the increase in [K+]~ is an inhibition of Ca2+-dependent exocytotic release of neurotransmitters. EAAs produce marked ionic movements across neuronal cell membranes, including K ÷ flux from the cells 7'11'14'20. Such an ionic event would facilitate activity of the energy-dependent ion pump and energy depletion. The extracellular concentration of EAAs measured by microdialysis in vivo has been shown to increase during ischemia 1°'17'18, beginning concomitantly with the onset of the abrupt increase in [K+L during ischemia TM. EAAs are thus considered the most likely neurotransmitter candidate to mediate such an ionic event. In vitro studies employing synaptosomal preparations or brain slices have shown that an elevated [K+]~ causes a rapid Ca2+-dependent exocytotic release from the vesicular pool and a slow Ca2+-independent release from the cytoplasmic pool of EAAs 4°'41. Elevation of the E A A concentration during 10 min ischemia has been reported to be completely abolished when Ca2+-free perfusate containing Co 2+ is used, indicating involvement of Ca2+-dependent exocytotic release 1°. We have recently observed that rapid E A A release concomitant with the abrupt increase in [K+]e is primarily CaZ+-dependent. The Ca 2÷dependency of the abrupt increase in [K+]¢ supports the hypothesis that the Ca2+-dependent E A A release is a major cause of the abrupt increase in [K+]~.
Acknowledgement. This work was supported by Grants from the NIH (NS27544), Lind Lawrence Foundation, Annie Laurie Aitken Charitable Trust and World Boxing Council, and a research grant for Cardiovascular Disease (2A-2) from the Ministry of Health and Welfare of Japan.
63 REFERENCES 1 Alexander, G.M., Grothusen, J.R. and Schwartzman, R.J., Flow dependent changes in the effective surface area of microdialysis probes, Life Sci., 43 (1988) 595-601. 2 Arnfred, T. and Hertz, L., Effects of potassium and glutamate on brain cortex slices: uptake and release of glutamic and other amino acids., J. Neurochem., 18 (1971) 259-265. 3 Astrnp, J., Symon, L., Branston, N.M. and Lassen, N.A., Cortical evoked potential and extraceUular K + and H + at critical levels of brain ischemia, Stroke, 8 (1977) 51-57. 4 Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H., Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis, J. Neurochem., 43 (1984) 1369-1374. 5 Blank Jr., W.E and Kirshner, H.S., The kinetics of extracellular potassium changes during hypoxia and anoxia in the cat cerebral cortex, Brain Research, 123 (1977) 113-124. 6 Bosley, T.M., Woodhams, EL., Gordon, R.D. and Bal~s, R., Effects of anoxia on the stimulated release of amino acid neurotransmitters in the cerebellum in vitro, J. Neurochem., 40 (1983) 188-210. 7 Choi, D.W., Ionic dependence of glutamate neurotoxicity, J. Neurosci., (1987) 369-379. 8 Damsma, G., Westernik, B.H.C., Imperato, A., Rollema, H., DeVries, J.B. and Horn, A.S., Automated brain dialysis of acetylcholine in freely moving rats: detection of basal acetylcholine, Life Sci., 41 (1987) 873-876. 9 Delgado, J.M.R., Defeudis, EV., Roth, R.H., Ryngo, D.K. and Mitruka, B.M., Dialytrode for long term intercerebral perfusion in awake monkeys, Arch. Int. Pharmacodyn. Ther., 198 (1971) 9-21. 10 Drejer, J., Benveniste, H., Diemer, N.H. and Schousboe, A., Cellular origin of ischemia-induced glutamate release from brain tissue in vivo and in vitro, J. Neurochem., 45 (1985) 145-151. 11 Hansen, A.J., Extracellular potassium concentration in juvenile and adult rat brain cortex during anoxia, Acta Physiol. Scand., 99 (1977) 412-420. 12 Hansen, A.J. and Nordstrem, C.-H., Brain extracellular potassium and energy metabolism during ischemia in juvenile rats after exposure to hypoxia for 24 h, J. Neurochem., 32 (1979) 915-920. 13 Hansen, A.J. and Zeuthan, T., Extracellular ion concentrations during spreading depression and ischemia in the rat brain cortex, Acta Physiol. Scand., 113 (1981) 437-445. 14 Hansen, A.J., Effect of anoxia on ion distribution in the brain, Physiol. Rev., 65 (1985) 101-148. 15 Hossmann, K.-A., Sakaki, S. and Zimmermann, V., Cation activities in reversible ischemia of the cat brain, Stroke, 8 (1977) 77-81. 16 Imperato, A. and DiChiara, G., Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the in vivo release of endogenous dopamine and metabolites, J. Neurosci., 4 (1984) 966-977. 17 Katayama, Y., Becker, D.P., Tamura, T., Martin, N. and Tsubokawa, T., Inhibition of massive ionic fluxes during cerebral ischemia (Terminal depolarization) with excitatory amino acid
18
19
20 21 22
23 24 25
26
27 28 29 30
31
32 33
34
antagonist as demonstrated by microdialysis, J. Cerebr. Blood Flow Metabol., 9 (1989) $57. Katayama, Y., Tamura, T., Kawamata, T. and Becket, D.P., Concomitance and dependence of massive potassium flux to early glutamate release during cerebral ischemia in vivo, Neurosci. Abstr., 15 (1989) 358. Katayama, Y., Kawamata, T., Tamura, T., Hovda, D.A., Becker, D.P. and Tsubokawa, T., Calcium-dependent glutamate release concomitant with massive potassium flux during cerebral ischemia in vivo, Brain Research, 558 (1991) 136-140. Martins-Ferreira, H., DeOliveira Castro, G., Struchiner, C.J. and Rodrigues, ES., Circling spreading depression in isolated chick retina, J. Neurophysiol., 37 (1974) 773-784. Mayer, M.L. and Westbrook, G.L., Cellular mechanisms underlying excitotoxicity, Trends Neurosci., 10 (1987) 59-61. Moghaddam, B., Schenk, J.O., Stewart, W.B. and Hansen, A.J., Temporal relationship between neurotransmitter release and ion flux during spreading depression and anoxia, Can. J. Physiol. Pharmacol., 54 (1987) 1105-1110. Nadler, J.V., Vaca, K.W., White, W.E, Lynch, G.S. and Cotman, C.W., Aspartate and glutamate as possible transmitters of excitatory hippocampal afferents, Nature, 260 (1976) 538-540. Nicholls, D.G. and Sihra, T.S., Synaptosomes possess an exocytotic pool of glutamate, Nature, 321 (1986) 772-773. Nicholson, C. and Kraig, R.P., The behavior of extracellular ions during spreading depression. In T. Zeuthen (Ed.), The Application of Ion Electrodes, Elsevier, Amsterdam, 1981, pp. 217-238. Nicholson, C., Phillips, J.M. and Gardner-Medwin, A.R., Diffusion from an iontophoretic point source in the brain. Role of tortuosity and volume fraction, Brain Research, 169 (1979) 580584. Olney, J.W., Price, M.T., Samson, L. and Labuyere, J., The role of specific ions in glutamate neurotoxicity, Neurosci. Lett., 65 (1986) 65-71. Rothman, S.M. and Olney, J.W., Excitotoxicity and the NMDA receptor, Trends Neurosci., 10 (1987) 299-302. Sanchez-Prieto, J., Sihra, T.S. and NichoUs, D.G., Characterization of the exocytotic release of glutamate from guinea-pig cerebral cortical synaptosomes, J. Neurochem., 49 (1987) 58-64. Siesjo, B.K. and Wieloch, T., Molecular mechanisms of ischemic brain damage: calcium related events. In E Plum and W. Pulsinelli (Eds.), Cerebrovascular Diseases, Raven, New York, 1985, pp. 187-197. Ungerstedt, U., Herrera-Manschitz, M., Jungnelius, U., Stahle, L., Tossman, U. and Zetterstrom, T., Dopamine synaptic mechanisms reflected in studies combining behavioral recordings and brain dialysis, Adv. Dopamine Res., 17 (1982) 219-231. Van Harreveld, A., Two mechanisms for spreading depression in chicken retina, J. Neurobiol., 9 (1979) 419-431. Vyscocil, E, Kriz, N. and Bures, J., Potassium selective microelectrodes used for measuring the extracellular brain potassium during spreading depression and anoxic depolarization in rats, Brain Research, 39 (1973) 255-259. Westerink, B.H.C., Damsma, G., Rollema, H., De Vries, J.B. and Horn, A.S., Scope and limitations of in vivo brain dialysis: a comparison of its application to various neurotransmitter systems, Life Sci., 41 (1987) 1763-1776.
57
BRES 17241
Calcium-dependent component of massive increase in extracellular potassium during cerebral ischemia as demonstrated by microdialysis in vivo Yoichi Katayama 1'2., Toru Tamura 1, Donald P. Becker I and Takashi Tsubokawa 2 1Division of Neurosurgery, UCLA School of Medicine, University of California at Los Angeles, CA 90024 (U.S.A.) and 2Department of Neurological Surgery, Nihon University School of Medicine, Tokyo (Japan) (Accepted 23 July 1991) Key words: Cerebral ischemia; Calcium ion; Potassium ion; Microdialysis; Spreading depression This study characterizes the physiological features and limitations of K+-free dialysis to detect changes in extracellular concentration of K ÷ ([K+]c) in the rat hippoeampus in vivo. It also demonstrates the effects of Ca2+-free perfusate containing Co2+ or Mg2+, which blocks Ca2+ entry into the presynaptic nerve terminal, on the abrupt increase in [K+]c detected by this technique during cerebral ischemia. K+-free dialysis for 40 rain caused no significant changes in the baseline [K+]c. In contrast, Ca2+-free dialysis for 40 min significantly reduced the extracellular Ca2+ concentration. Under this condition, together with addition of Co2+ or Mg2+ to the perfusate, the increase in [K+]c was delayed, and a delay in reaching the maximum level was observed in a dose-dependent manner. These results are consistent with the hypothesis that the initial increase in [K+]. during cerebral isehemia is related to the Ca2+-dependent exocytotic release of neurotransmitters from depolarized nerve terminals.
INTRODUCTION During cerebral ischemia, an abrupt increase in extracellular concentration of K + ([K+].) to 50-60 mM is observed following a slow increase 3'5't1'15'33. Since this rapid and dramatic increase in [K+], invariably begins when [K+]¢ reaches a level of 6-10 mM 3'5'11'15'33, neurotransmitter release and a sudden K + flux through transmitter-gated ion channels of neuronal cells have been postulated to be the cause of this phenomenon 22'25'32. ThUS, when [K+]e reaches this level, the nerve terminals may be depolarized, causing Ca 2÷ entry into the presynaptic nerve terminal, and neurotransmitters are thereby released. Among the various neurotransmitters, excitatory amino acids (EAAs) seem the most likely substances which could produce such a remarkable K + flux 7"21'27'28. Ca2+-dependent exocytotic release of E A A s in response to an elevated [K+]e has been demonstrated by in vitro studies, employing synaptosome preparations or brain slices 2'6'2°'23'24. It has also been found by microdialysis in vivo that EAAs are in fact released during cerebral ischemia 4'1°'1sA9, and this release is completely abolished when Ca2÷-free perfusate containing Co 2÷, which blocks Ca 2÷ entry into the nerve terminal, is used 1°,t9.
If the Ca2+-dependent exocytotic release of neurotransmitters, especiaUy EAAs, is responsible for the abrupt increase in [K+]e, inhibition of C a 2+ entry into the nerve terminal would tend to alter the time course of the changes in [K+]e during cerebral ischemia. The present study examined this hypothesis by testing the effect of Ca2+-free perfusate containing C o 2+ o r Mg 2+, which blocks C a 2+ entry, on the changes in [K+]¢ detected by microdialysis2°'21 during cerebral ischemia. While the detection of changes in [K+], with microdialysis has not been described previously, such a technique would provide a unique opportunity to examine the mechanism of the changes in [K+], during ischemia, by monitoring other neurochemical changes simultaneously and, more importantly, by manipulating neurochemical processes through the administration of various agents in situ via the dialysis probe. In order to facilitate the detection of [K+], changes by microdialysis, however, the use of K+-free perfusate for the dialysis is necessary. This would inevitably cause a slow but continuous removal of K ÷ from the extracellular space (ECS) and could potentially produce an unphysiologically lowered [K+L level. The present study, therefore, first characterized the physiological features and limitations of this technique for the detection of changes in [K+], and then investigated the effects of Ca2+-free perfusate containing C o 2+ o r Mg 2+.
*On leave from Department of Neurological Surgery, Nihon University School of Medicine, Tokyo, Japan. Correspondence: Y. Katayama, Department of Neurological Surgery, Nihon University School of Medicine, Itabashi-ku, Tokyo 173, Japan. Fax: (81) 3 3554 0425.
58 MATERIALS AND METHODS
Procedures for animal preparation Young adult Sprague-Dawley rats (n = 36) weighing 180-240 g were used. The animals were maintained in an environmentally controlled room with a 12 h light/dark cycle and were allowed free access to food and water. They were anesthetized with a mixture of nitrous oxide (66%), oxygen (33%) and enflurane (1%), and placed in a stereotaxic frame with the nosebar setting at 2.5 mm below the interaural level. Skin wounds and pressure points were infiltrated with 1% xylocaine. The rectal temperature was maintained at 3738 °C using a heating pad. A pair of dialysis probes (CMA/10, Bioanalytical System Inc.; o.d., 500 #; effective length, 3 mm; cut off, 20,000 MW) were lowered vertically through small skull holes into the brain, placing the tip in the hippocampus (3.8-4.8 mm caudal to the bregma, 2.0 mm lateral to the midline and 3.5 mm below the surface of the dura). Thus, approximately 50% of the effective length of the probe was located within the hippocampus. The probes were initially perfused with Ringer solution (adjusted to pH 7.4) at a rate of 5,0 #l/min.
Procedures for ischemia induction and microdialysis Global cerebral ischemia was induced by decapitation. The temperature of the perfusate was prepared to be at 38 °C, and the temperature was maintained with a chamber filled with water at 38 °C in which the whole length of the inlet tubing of the dialysis system was placed. Dialysate fractions were collected at 1 min intervals. The length of the outlet tube was adjusted to a length that resulted in the dead space of the probe and tubing being 5.0 M1. Thus, there would be a 1 min delay between the changes observed in the dialysate and the actual changes occurring in the brain. At the end of the experiment, the position of the probe was confirmed anatomically. In order to facilitate the detection of [K+]~ changes by microdialysis, K+-free perfusate was used for the dialysis. The K ÷ concentration of each dialysate ([K+]d) was measured with a K+-sensitive electrode in vitro. The recovery rate of the dialysis system for K + in vitro was determined using a calibration solution ([K+]d/ [K+L,lib). The K ÷ concentration of the calibration solution
([K+]calib) was varied from 1 to 100 mM, replacing equimolar sodium chloride in the Ringer solution. The recovery rate obtained in vitro ranged from 6 to 13% at the flow rate of the present study. A reasonably linear relationship was observed between the [K+]o and the [K+]¢anb within the physiological range (Fig. 1A). The [K+]d changed rapidly in response to sudden changes in [K+]calibwith a 1 min delay for the dead space of the dialysis system (Fig. 1B). In order to confirm Ca z+ depletion with the Ca2÷-free dialysis (see below), the Ca z+ concentration ([Ca2+]o) was similarly measured with a CaZ÷-sensitive electrode in vitro. Evaluation of the effects of K+-free dialysis In order to evaluate the effect of K+-free dialysis on [K÷]~, two sets of experimental data were analyzed: (1) the sequential changes in [K+]a obtained from the same probe during the continued K +free dialysis; and (2) the differences in [K÷]o obtained from two probes in the same animal each perfused with K÷-free perfusate for a different duration. Thus, the dialysis with the K÷-free Ringer solution was first initiated with one of the two probes (test dialysis) while the other probe, placed in the contralateral hippocampus, was perfused with normal Ringer solution. The K+-free dialysis with the second probe was then initiated at 20, 40 or 60 min after the onset of the test dialysis (control dialysis). In 3 experiments, the changes in [K+]e were recorded with a K+-sensitive electrode in vivo in the hippocampus near the site of the test dialysis (2 mm lateral from the tip of the dialysis probe and 3.0 mm below the surface of the cortex). The K+-sensitive electrode was made by glass micropipettes having a common tip diameter of 3 M. Cerebral ischemia was induced 5 min following the initiation of the control dialysis. The time of ischemia induction therefore corresponded to 25 (n -- 6), 45 (n = 5) or 65 (n = 70 min after the initiation of the test dialysis.
Evaluation of the effects of Cae+-free dialysis In one of the two probes chosen at random, the perfusate was switched to CaZ+-free Ringer solution 5 min following the initiation of perfusion (test dialysis). The other probe was perfused with a solution containing Ca 2÷ throughout the experiment (control dialysis). The perfusates for the control and test dialysis were then switched to K+-free, or K+4Ca2+-free Ringer solutions, respectively, at 40 min after the initiation of the perfusion, and measurements of [K+]d were started. In most animals, cobalt chloride (1-10 mM; n = 26) or magnesium chloride (10 mM; n = 6) was added to
6.0
z
A
0
5.0-
~g
4o-
0.3
A POTASSIUM-FREEDIALYSIS z~ ©m
O< LUI-
3.0
m u.
20
1
-- 0.2
~2 zr~
8~ 0.1 -'0 00
• 20
5 10 15 20 25 30 35 TIME (MIN)
40 60 80 POTASSIUMCONCENTRATION OF CALIBRATIONSOLUTION (mM)
100
DEAD S 120
Fig. 1. A: representative example of the relationship between the K ÷ concentration of the calibration solution ([K÷]¢aiib) and the dialysate ([K÷]d). Results from 4 measurements with the same probe are plotted (closed circles). The curve represents a polynomial approximation of the relationship. The slope of the curve ([K+]J[K+L.Iib) represents the recovery rate in vitro. B: changes in [K÷ld (in vitro) in response to sudden changes in [K÷]~alib (open circles). The changes in [K+]~.lib (from 1 mM to 60 raM) are illustrated schematically by a line (no relation to concentration scale). The responses are delayed for 1 rain due to the dead space in the outlet tubing of the dialysis system (5 MI).
--~l
o
P
B
~
CALCIUM-FREEDIALYSIS
0.0 0
5 10 15 DURATION OF POTASSIUM-FREE(A) OR CALCIUM-FREE(B) DIALYSIS(MIN)
20
Fig. 2. Changes in baseline level of [K+]d during K+-free dialysis (A, n = 18) and baseline level of [Ca2+]d during Ca2+-free dialysis (B, n = 7). The perfusate was changed to K+-free or Ca2÷-free Ringer solution at time zero. [K+]d and [Ca2+]d during the initial 3 fractions are not shown because of the levels of K ÷ (3 raM) and Ca 2+ (1 mM) contained in the previous perfusate. The responses are delayed due to the dead space in the inlet as well as outlet tubing of the dialysis system. Each curve represents a logarithmic approximation of the changes.
59 TABLE I
Effects of K+-free dialysis K+-free dialysis (duration)
n
Pre-ischemia baseline [K+]d (raM)
Latency (rain)* Onset (L1)
Maximum (L2)
Post-ischemia maximum [K+]d (raM)
Control (5 min) Test 25 min)
6
0.22 -+ 0.02 0.21 --- 0.02
2.5 --- 0.4 2.7 --+ 0.3
3.5 --- 0.4 3.7 --- 0.3
1.33 - 0.12 1.29 - 0.10
Control (5 rain) Test (45 min)
5
0.21 -+ 0.03 0.19 -+ 0.03
2.4 - 0.4 2.6 +-- 0.5
3.8 -+ 0.4 3.8 --- 0.5
1.27 - 0.12 1.19 --+ 0.10
Control (5 rain) Test (65 min)
7
0.20 -+ 0.01 0.17 -+ 0.01"*
2.3 -+ 0.4 2.7 --- 0.3
3.3 --- 0.4 4.0 + 0.3
1.30 - 0.11 1.10 +-- 0.10
Paired data (mean --- S.E.M.); *a fraction for dead space (1 min) is excluded. **P < 0.05, as compared to the value at 5 min obtained from the same dialysis site with Student paired t test. No significant difference detected in comparison to the control dialysis site.
the perfusate for the test dialysis. The osmolarity of each perfusate was maintained constant by changing the concentration of sodium chloride. Cerebral ischemia was induced 5 rain following the initiation of the [K+]a measurements.
Statistical analysis The changes in [K+]a values obtained in each animal were analyzed by Student paired t test. The latencies of defined events were expressed as the numbers of the fraction sampled after isehemia and compared between the test and control probes by Wilcoxon matched-pair signed rank test.
RESULTS
Detection of changes in [K+]e with K+-free dialysis +
Following the initiation of the K -free dialysis in vivo,
1.2
TIME AFTER ISCHEMIA INDUCTION (MIN) -2 0 2 4 6 8 10 12 14 i I I i i i i i i ~
-4
1.o-
H
16 i
A
DEAD SPACE
~ 0.8-
[K+]o decreased rapidly from the higher values due to the previous K+-containing perfusate, and a stationary level of [K+]a ranging from 0.14 to 0.29 mM was obtained within 5 rain (0.22 - 0.01 mM, n = 18; Fig. 2A). Although a very slow, progressive decrease did occur thereafter, the [K+]o appeared to remain at approximately the same level for a considerable period of time (Fig. 2A and Table I). The presence of a slight but significant decrease in the baseline [K+]d became apparent only when the K+-free dialysis was continued for 60 min (5 min vs 65 min at the same site, P < 0.05, n = 7, Table I). The recording of [K+]~ with a K+-sensitive electrode in the proximity of the site of dialysis (distance, 2 mm) failed to reveal a decrease in baseline [K+]¢ during the prolonged K+-free dialysis.
/
/
1.8-
~ i / 1 ~
z
O
DA I LYSS I
oct
A
1.5" _z o
i 0.6-
o
20
0
w
20
40
60
80
100
•
0.9-
!
t o.o-
o01(~ 0.2 ~
--I
10
B
--
0.3-
0.0 v.v
1
-5 -'3 -1 ; '3 5 7 9 11 13 15 ACTUAL TIME AFTER ISCHEMIA INDUCTION (MIN)
Fig. 3. Comparison of time courses of [K+]d measured by microdialysis (A, closed circles) and [K+]e measured with a K+-sensitive electrode (B) during cerebral ischemia. Each symbol representing [K+]a of the 1 min fraction is located at the midpoint of the period in which the fraction was collected. Perfusion rate, 5.0 #l/rain. The difference in time scale for A ([K+]d), and B ([K+]¢) accounts for the delay (1 rain) of the [K+]d response due to the dead space (5 #1) of the outlet side of the dialysis system. The ischemia was induced 20 min after the onset of K+-free dialysis.
0
20 40 60 80 100 EXTRACELLULAR CONCENTRATION OF POTASSIUM (mM)
Fig. 4. A: representative example of relationships between [K+]e measured with a microelectrode (average during 1 min period for dialysate collection) and [K+]d detected microdialysis in the same animal. Date for varying [K+]e were from the period during ischemia. The curve represents an exponential approximation of the relationship. The slope of the curve ([K+]J[K+]c) represents the recovery rate in vivo (of. Fig. 1A). B: representative example of the changes in ratio of the recovery rates in vivo vs in vitro (in vivo/in vitro) at various [K+]e during ischemia. The curve represents a polynomial approximation of the changes.
60 TABLE IT
Effects of Ca2+-free dialysis with Co2+ of Mg2+ on baseline and maximum [K+]d
[K+ld (raM)
Control dialysis Ca2+-free dialysis (n -- 7) Control dialysis Ca2+-free dialysis with
Co 2+
(10 mM) (n = 13)
Control dialysis Ca2+-free dialysis with Mg2+ (10 mM) (n = 6)
Pre-ischemic baseline level
Post-ischemic maximum level
0.21 --- 0,01 0.20 + 0,01
1.28 --- 0.10 1.25 --- 0.11
0.19 --- 0.01 0.18 -+ 0.01
1.18 --- 0.09 1.13 --- 0.11
0.21 --- 0,02 0.21 --- 0,02
1.50 --+ 0.10 1.45 -+ 0.07
Paired data (mean -+ S.E.M.); no significant differences were seen between the two dialysis sites with Student paired t-test.
•
A
0.8" 1.0"
~
N
0.6" 0.40.2" ~
o'12-B 0.0
v
-4 1.4-
TROL CO 0mM
---o--2
0
2
4
, 8
6
10
12
, 14
16
1.0-
O z 0
0.8-
<
0.6-
z 0.4 ,,, O z 0.2 O O 19
0.o
< -4 >- 1.6
f,/)
---o-.
-2
.
0
.
.
2
.
4
.
.
6
.
8
.
10
Induction of the global ischemia produced a slight increase in [K+]d in one or more fractions, in addition to the fraction for the dead space, and a marked increase in subsequent fractions (Fig. 3A). The onset of the sudden, large increase corresponded to an abrupt increase in [K+]e as measured by the K+-sensitive electrode in the proximity of the site of dialysis, when a 1 min delay for the response in [K+]d due to the dead space was taken into account (cf. Fig. 3A and B). The [K+L measured by the electrode showed an initial slow increase from a baseline value of approximately 3 mM to a level of 6 mM, and subsequently a rapid increase up to values ranging from 60 to 80 mM (Fig. 3B). The [K+]d m e a s u r e d from the dialysate increased invariably to a max-
ROL Co 5ram 12
14
2.0
16
~==
~ 10
uJ ~
1.0 0.8"
0.6" 0.4" 0.2-
CO
m Z
1.21
+
CONTROL Co lOmM
~ IN ~
,,<,
L2 (TESF-CONT)
,
. . . . . .
l
TIME
Oa. WZ
8o
0.5
z
o0
Z
TEST
a -0.5 CONTROL
0.0 -4
[]L'(rTEST-CONT) []
-2 0 2 4 6 8 10 12 14 TIME AFTER ISCHEMIA INDUCTION (MIN)
16
Fig. 5. Representative examples of the time course of changes in [K+]d during isehemia with CaZ+-free dialysis employing various concentrations of Co2+ (test dialysis, open circles). The time course of the changes in [K+]a detected in the same animal with the probe perfused with Ca 2+ and without Co 2+ is also shown (control dialysis, filled circles). A: 0 mM Co2+; B: 5 mM CoZ+; C: 10 mM Co 2÷. Each symbol representing [K+]a of the 1 min fraction is located at the midpoint of the period in which the fraction was collected. Perfusion rate, 5.0/tl/min. The response of [K+]d to the ischemia is delayed 1 min due to the dead space (5 #1) of the outlet side of the dialysis system.
CO 0mM
Co 1raM
Co 5raM
Col0mM
CALCIUM-FREE DIALYSIS
Fig. 6. Effects of Ca2+-free dialysis employing various concentrations of Co2+ (test dialysis) on the latencies of onset of the increase in [K+]a (L 0 and the time at which the maximum level of increased [K+]d was reached (Lz). The effects are demonstrated in comparison with the latencies observed in the control dialysis in the same animal (test-control). The latencies were measured as the number of the 1 min dialysis fractions. The onset (L1) was defined as the fraction which demonstrated a sudden increase in [K+]d reaching more than 2-fold baseline, and the time at which the maximum level was attained (L2) was defined as the fraction reaching a level within 2 S.D. of the higher [K+]d level after the onset. *P < 0.01, n = 13, Wilcoxon matched-pair signed rank test.
61 TABLE III Effects of Ca2+-free dialysis with Co2+ or Mg2+ on time courses of the increase in [K+]e during ischemia Latency (rain)* Onset (L1)
Maximum (L2)
Duration (min) .from the onset to the maximum (L2-L~)
Control dialysis Ca2+-free dialysis with 10 mM Co2÷ (n = 13)
2.9 --- 0.3 4.0 --- 0.3***
4.5 --- 0.4 6.2 +--0.3***
1.6 - 0.2 2.2 -+ 0.2**
Control dialysis Ca2+-free dialysis with 10 m M Mg 2+ (n -- 6)
2.2 --- 0.2 2.8 -+ 0.3**
3.5 --- 0.3 4.3 --- 0.4**
1.3 +- 0.2 1.7 --- 0.2
Paired date (mean +- S.E.M.); *the fraction for dead space is excluded. **P < 0.05, ***P < 0.01 Wilcoxon matched-pair signed rank test.
imum level ranging from 0.8 to 2.0 mM (1.30 - 0.11 raM, n = 18; Fig. 3A and Table I). There were no significant differences in the maximum level of [K+]d attained eventually during the ischemia (Table I). The onset of the large increase in [K÷]d was arbitrarily defined as the fraction that demonstrated a sudden increase in [K+]d reaching more than 2-fold baseline, and the latency of onset was expressed as the number of the fraction after ischemia induction excluding the dead space (L1). The time that the maximum level of [K+]d was attained was defined as the fraction reaching a [K+]d level within 2 S.D. of the higher [K+]d level, and its latency was similarly expressed as the number of the fraction (1-2). The difference in duration of K+-free dialysis between the test and control dialysis, which varied from 20 to 60 min, resulted in no significant difference in these latencies (Table I). The results of direct recording of the changes in [K+]~ with the K+-sensitive electrode were used for computation of the recovery rate of the dialysis system for K ÷ in vivo ([K+]d/[K+]e). The obtained recovery rate in vivo was always smaller than the recovery rate determined in vitro (Fig. 4A, cf. Fig. 1A). It was also found that the ratio of the recovery rate in vivo to the recovery rate in vitro dropped further following the ischemia induction in association with the elevation of [K+]e during ischemia (Fig. 4B). Modification o f changes in [K+]e with Ca2+-free dialysis
Following the initiation of the Ca2+-free dialysis, [Ca2+]d decreased rapidly, and a relatively stationary level of [Ca2+]d, ranging from 0.08 to 0.12 mM, was obtained within 5 min (Fig. 2B, filled circles). The higher values for the initial period were due to Ca 2÷ contained in the perfusate previously. A gradual decrease in [Ca2+]d, however, continued thereafter (Fig. 2B). This effect of the Ca2+-free dialysis on [Ca2+]d was in clear contrast to the effect of the K+-free dialysis on [K+]d (cf. Fig. 2A and B). The baseline [Ca2+]d often decreased to
less than half within 40 min following the initiation of Ca2+-free dialysis. The Ca2+-free dialysis, regardless of whether or not Co 2+ was added to the perfusate, produced no significant changes in baseline [K+]d (Table II). The development of the large increase in [K+]d following the induction of ischemia was significantly delayed by the dialysis with Ca2+-free perfusate containing Co 2÷ (Fig. 5B,C). No significant effect was detected with Ca2+-free dialysis alone (Fig. 5A). The effects of Co 2+ were demonstrated to be dose-dependent (Fig. 6A,B) as a prolonged latency of o n s e t (L1, P < 0.01, n = 13, Table III), an additional delay in reaching the maximum level (L 2, P < 0.01, n = 13; Table III) and a prolonged duration from the onset to the maximum level (L1-L2, P < 0.01, n = 13; Table III). Similar effects were also observed with Ca2+-free, Mg2+-containing dialysis (Table III). DISCUSSION If K+-free dialysis removed K ÷ only from the ECS, the removal of 0.2 mM at 5/A/min as dialysate would result in a loss of 1.0 nmol/min from the ECS which could deplete all the K ÷ in 3.3/zl ECS after 10 min. No remarkable changes in baseline [K+]d were observed in the present study, however, suggesting that K ÷ is replenished continuously by transcellular K + movement during K+-free dialysis 25. The [K+]d changed rapidly in response to sudden changes in the K ÷ concentration of the calibration solution and sensitively responded to the rapid increase in [K+]e during cerebral ischemia. It appears therefore that K+-free dialysis provides reasonable information regarding the timing of large changes of [K+]e, such as that observed during cerebral ischemia, in shortterm experiments. The recovery rate in vivo was always smaller than the recovery rate determined in vitro. This may reflect the lower effective surface area of the dialysis system in vivo which has recently been demonstrated at flow rates rang-
62 ing from 2 to 10/A/min 1. The further decrease in recovery rate in vivo occurring in association with the elevation of [K+]~ during ischemia may be accounted for by a decrease in the effective surface area of the dialysis system in vivo due presumably to shrinkage of the ECS, which has been demonstrated to take place concomitantly with the abrupt increase in [K+]e 14. Without knowing the changes in effective surface area of the dialysis system in vivo, microdialysis can provide only a qualitative measure of the changes in the concentration of any substance in the ECS under pathological conditions such as ischemia. K+-free dialysis is still useful, however, for identifying the timing of the abrupt K ÷ flux and for demonstrating other neurochemical changes simultaneously using the same dialysate fractions. Further, neurochemical processes can be manipulated experimentally through the administration of various agents in situ via the same dialysis probe. Thus, K+-free dialysis provides a less complicated and less invasive procedure for investigating various neurochemical changes associated with the sudden K + flux induced by cerebral ischemia. The present data showed that the abrupt and massive increase in [K+]~ occurring during cerebral ischemia was significantly delayed by dialysis with Ca2+-free peffusate containing Co 2÷ or Mg 2+. The rapid component of the massive increase in [K+]e, which normally lasts only for less than a minute, comprises approximately 75% of the maximum level of [K+]¢ attained during ischemia 11'12'2°. The observed effects therefore appear to involve inhibition of this rapid component. In view of the non-linear relationship between [K+]e and [K+]d, however, there is a possibility that inhibition of the subsequent slow component of [K+]~ is also involved. CaZ+-free dialysis produces a marked decrease in the baseline [Ca2+]~, indicating that the CaZ+-free dialysis removes Ca 2+ from the ECS rapidly 8'1°. Employing 45Ca autoradiography, we have confirmed Ca 2+ depletion within a brain area of approximately 2-3 mm in diameter by the Ca2+-free dialysis when continued for more than 30 min (unpublished observations). In addition, the presence of Co 2÷ or Mg 2÷ in the ECS inhibits Ca 2÷ entry into the presynaptic nerve terminal 1°As'19'33 or postsynaptic nerve cells. Thus, the results of the present study indicate that a major component of the initial increase in [K+]e during cerebral ischemia is dependent upon Ca 2÷ flux from the ECS into the cells. Ca 2+ has a number of actions when it enters the postsynaptic nerve cells, including the activation of numerous enzyme systems, and some of these changes result in an increase in neuronal cell membrane permeability29. The effects observed in the present study may, in part, be attributable to an inhibition of these processes. When Ca 2+ enters the presynaptic nerve terminal, it
mediates
the
exocytotic
release
of
neurotransmit-
ters 2'6A9'22'23. As mentioned earlier, it has been postulated that the abrupt increase in [K+]e occurring during ischemia is caused by neurotransmitter release from the nerve terminal depolarized by a [K+]e elevated to 6-10 mM 21'24'31. The presence of the CaZ+-dependent component demonstrated in the present study is consistent with this hypothesis, since exocytotic release of neurotransmitter is dependent upon Ca 2÷ entry into the presynaptic nerve terminal. It is possible that the induced ionic environment depresses metabolism and increases the glucose stores within the brain tissue 14. Thus, the observed delay in the increase in [K÷]~ during ischemia may be a consequence of a reduced metabolism and increased glucose content. We have, however, observed no changes in the local glucose utilization rate, as evaluated by [14C]deoxyglucose autoradiography, in the brain area peffused with Ca2+-free perfusate containing Co 2+ (unpublished observations). A more likely mechanism for the delay in the increase in [K+]~ is an inhibition of Ca2+-dependent exocytotic release of neurotransmitters. EAAs produce marked ionic movements across neuronal cell membranes, including K ÷ flux from the cells 7'11'14'20. Such an ionic event would facilitate activity of the energy-dependent ion pump and energy depletion. The extracellular concentration of EAAs measured by microdialysis in vivo has been shown to increase during ischemia 1°'17'18, beginning concomitantly with the onset of the abrupt increase in [K+L during ischemia TM. EAAs are thus considered the most likely neurotransmitter candidate to mediate such an ionic event. In vitro studies employing synaptosomal preparations or brain slices have shown that an elevated [K+]~ causes a rapid Ca2+-dependent exocytotic release from the vesicular pool and a slow Ca2+-independent release from the cytoplasmic pool of EAAs 4°'41. Elevation of the E A A concentration during 10 min ischemia has been reported to be completely abolished when Ca2+-free perfusate containing Co 2+ is used, indicating involvement of Ca2+-dependent exocytotic release 1°. We have recently observed that rapid E A A release concomitant with the abrupt increase in [K+]e is primarily CaZ+-dependent. The Ca 2÷dependency of the abrupt increase in [K+]¢ supports the hypothesis that the Ca2+-dependent E A A release is a major cause of the abrupt increase in [K+]~.
Acknowledgement. This work was supported by Grants from the NIH (NS27544), Lind Lawrence Foundation, Annie Laurie Aitken Charitable Trust and World Boxing Council, and a research grant for Cardiovascular Disease (2A-2) from the Ministry of Health and Welfare of Japan.
63 REFERENCES 1 Alexander, G.M., Grothusen, J.R. and Schwartzman, R.J., Flow dependent changes in the effective surface area of microdialysis probes, Life Sci., 43 (1988) 595-601. 2 Arnfred, T. and Hertz, L., Effects of potassium and glutamate on brain cortex slices: uptake and release of glutamic and other amino acids., J. Neurochem., 18 (1971) 259-265. 3 Astrnp, J., Symon, L., Branston, N.M. and Lassen, N.A., Cortical evoked potential and extraceUular K + and H + at critical levels of brain ischemia, Stroke, 8 (1977) 51-57. 4 Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H., Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis, J. Neurochem., 43 (1984) 1369-1374. 5 Blank Jr., W.E and Kirshner, H.S., The kinetics of extracellular potassium changes during hypoxia and anoxia in the cat cerebral cortex, Brain Research, 123 (1977) 113-124. 6 Bosley, T.M., Woodhams, EL., Gordon, R.D. and Bal~s, R., Effects of anoxia on the stimulated release of amino acid neurotransmitters in the cerebellum in vitro, J. Neurochem., 40 (1983) 188-210. 7 Choi, D.W., Ionic dependence of glutamate neurotoxicity, J. Neurosci., (1987) 369-379. 8 Damsma, G., Westernik, B.H.C., Imperato, A., Rollema, H., DeVries, J.B. and Horn, A.S., Automated brain dialysis of acetylcholine in freely moving rats: detection of basal acetylcholine, Life Sci., 41 (1987) 873-876. 9 Delgado, J.M.R., Defeudis, EV., Roth, R.H., Ryngo, D.K. and Mitruka, B.M., Dialytrode for long term intercerebral perfusion in awake monkeys, Arch. Int. Pharmacodyn. Ther., 198 (1971) 9-21. 10 Drejer, J., Benveniste, H., Diemer, N.H. and Schousboe, A., Cellular origin of ischemia-induced glutamate release from brain tissue in vivo and in vitro, J. Neurochem., 45 (1985) 145-151. 11 Hansen, A.J., Extracellular potassium concentration in juvenile and adult rat brain cortex during anoxia, Acta Physiol. Scand., 99 (1977) 412-420. 12 Hansen, A.J. and Nordstrem, C.-H., Brain extracellular potassium and energy metabolism during ischemia in juvenile rats after exposure to hypoxia for 24 h, J. Neurochem., 32 (1979) 915-920. 13 Hansen, A.J. and Zeuthan, T., Extracellular ion concentrations during spreading depression and ischemia in the rat brain cortex, Acta Physiol. Scand., 113 (1981) 437-445. 14 Hansen, A.J., Effect of anoxia on ion distribution in the brain, Physiol. Rev., 65 (1985) 101-148. 15 Hossmann, K.-A., Sakaki, S. and Zimmermann, V., Cation activities in reversible ischemia of the cat brain, Stroke, 8 (1977) 77-81. 16 Imperato, A. and DiChiara, G., Trans-striatal dialysis coupled to reverse phase high performance liquid chromatography with electrochemical detection: a new method for the study of the in vivo release of endogenous dopamine and metabolites, J. Neurosci., 4 (1984) 966-977. 17 Katayama, Y., Becker, D.P., Tamura, T., Martin, N. and Tsubokawa, T., Inhibition of massive ionic fluxes during cerebral ischemia (Terminal depolarization) with excitatory amino acid
18
19
20 21 22
23 24 25
26
27 28 29 30
31
32 33
34
antagonist as demonstrated by microdialysis, J. Cerebr. Blood Flow Metabol., 9 (1989) $57. Katayama, Y., Tamura, T., Kawamata, T. and Becket, D.P., Concomitance and dependence of massive potassium flux to early glutamate release during cerebral ischemia in vivo, Neurosci. Abstr., 15 (1989) 358. Katayama, Y., Kawamata, T., Tamura, T., Hovda, D.A., Becker, D.P. and Tsubokawa, T., Calcium-dependent glutamate release concomitant with massive potassium flux during cerebral ischemia in vivo, Brain Research, 558 (1991) 136-140. Martins-Ferreira, H., DeOliveira Castro, G., Struchiner, C.J. and Rodrigues, ES., Circling spreading depression in isolated chick retina, J. Neurophysiol., 37 (1974) 773-784. Mayer, M.L. and Westbrook, G.L., Cellular mechanisms underlying excitotoxicity, Trends Neurosci., 10 (1987) 59-61. Moghaddam, B., Schenk, J.O., Stewart, W.B. and Hansen, A.J., Temporal relationship between neurotransmitter release and ion flux during spreading depression and anoxia, Can. J. Physiol. Pharmacol., 54 (1987) 1105-1110. Nadler, J.V., Vaca, K.W., White, W.E, Lynch, G.S. and Cotman, C.W., Aspartate and glutamate as possible transmitters of excitatory hippocampal afferents, Nature, 260 (1976) 538-540. Nicholls, D.G. and Sihra, T.S., Synaptosomes possess an exocytotic pool of glutamate, Nature, 321 (1986) 772-773. Nicholson, C. and Kraig, R.P., The behavior of extracellular ions during spreading depression. In T. Zeuthen (Ed.), The Application of Ion Electrodes, Elsevier, Amsterdam, 1981, pp. 217-238. Nicholson, C., Phillips, J.M. and Gardner-Medwin, A.R., Diffusion from an iontophoretic point source in the brain. Role of tortuosity and volume fraction, Brain Research, 169 (1979) 580584. Olney, J.W., Price, M.T., Samson, L. and Labuyere, J., The role of specific ions in glutamate neurotoxicity, Neurosci. Lett., 65 (1986) 65-71. Rothman, S.M. and Olney, J.W., Excitotoxicity and the NMDA receptor, Trends Neurosci., 10 (1987) 299-302. Sanchez-Prieto, J., Sihra, T.S. and NichoUs, D.G., Characterization of the exocytotic release of glutamate from guinea-pig cerebral cortical synaptosomes, J. Neurochem., 49 (1987) 58-64. Siesjo, B.K. and Wieloch, T., Molecular mechanisms of ischemic brain damage: calcium related events. In E Plum and W. Pulsinelli (Eds.), Cerebrovascular Diseases, Raven, New York, 1985, pp. 187-197. Ungerstedt, U., Herrera-Manschitz, M., Jungnelius, U., Stahle, L., Tossman, U. and Zetterstrom, T., Dopamine synaptic mechanisms reflected in studies combining behavioral recordings and brain dialysis, Adv. Dopamine Res., 17 (1982) 219-231. Van Harreveld, A., Two mechanisms for spreading depression in chicken retina, J. Neurobiol., 9 (1979) 419-431. Vyscocil, E, Kriz, N. and Bures, J., Potassium selective microelectrodes used for measuring the extracellular brain potassium during spreading depression and anoxic depolarization in rats, Brain Research, 39 (1973) 255-259. Westerink, B.H.C., Damsma, G., Rollema, H., De Vries, J.B. and Horn, A.S., Scope and limitations of in vivo brain dialysis: a comparison of its application to various neurotransmitter systems, Life Sci., 41 (1987) 1763-1776.