- Email: [email protected]
Astrocytes contribute to regulation of extracellular calcium and potassium in the rat cerebral cortex during spreading depression
Brain Research 1012 (2004) 177 – 184 www.elsevier.com/locate/brainres
Research report
Astrocytes contribute to regulation of extracellular calcium and potassium in the rat cerebral cortex during spreading depression Xiao-Yuan Lian, Janet L. Stringer * Department of Pharmacology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA Accepted 3 April 2004 Available online 10 May 2004
Abstract This study used spreading depression (SD), which is characterized by redistribution of ions, to examine the role of astrocytes in the regulation of extracellular potassium ([K+]o) and calcium ([Ca2 +]o) levels. Recurrent spreading depression episodes were induced by application of 3 M potassium chloride to the cortex of adult anesthetized rats while monitoring the extracellular direct current (DC) potential shifts and changes in [K+]o or [Ca2 +]o 6 – 7 mm away. The reversible glial toxins, fluorocitrate (FC) and fluoroacetate (FA), were injected locally into the cortex at doses that are selective for reducing glial function. The peak changes and area under the curve for [K+]o and [Ca2 +]o, recovery rate for [K+]o, and interval between spreading depression episodes were measured before and at various times after administration of the toxins. Both fluorocitrate and fluroacetate slowed the recovery of the [K+]o and altered the recovery of the [Ca2 +]o. Local injection of glutamate uptake inhibitors or barium had no effect on the peak changes in [K+]o or the rate of recovery of the [K+]o. The slowing of the recovery rate is consistent with the hypothesis that glial cells play a role in the return of [K+]o to baseline after spreading depression in the cortex in vivo. The change in movement of calcium after administration of FC suggests that astrocytes normally extrude calcium during spreading depression, resulting in rapid recovery of the levels of [Ca2 +]o with an overshoot. These findings demonstrate that astrocytes contribute to the regulation of both potassium and calcium during and after a stress to the ionic homeostatic mechanisms. D 2004 Elsevier B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Epilepsy: basic mechanisms Keywords: Fluorocitrate; Fluoroacetate; Ion homeostasis
1. Introduction Astrocytes have been postulated to play a role in maintaining, or restoring, extracellular ion concentrations, particularly potassium, during and after intense neuronal activity. During stimulus trains or seizure activity in the hippocampus, the extracellular potassium concentration ([K+]o) will reach a plateau level that is sustained until the end of the neuronal activity [29]. Previously, it has been shown that astrocytes participate in the return of the [K+]o to normal levels after seizures in the dentate gyrus, but not in the determination of the plateau level reached during seizures [29]. The plateau level appears to be determined more by active uptake into neurons by the Na+ – K+ ATPase, than * Corresponding author. Tel.: +1-713-798-7294; fax: +1-713-7983145. E-mail address: [email protected] (J.L. Stringer). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2004.04.011
by uptake into glial cells [30]. In the absence of neuronal activity, uptake of potassium (after iontophoretic application) appears to be mediated largely through barium-sensitive uptake mechanisms in glia [12]. Thus, it appears that astrocytes can participate in the regulation of [K+]o, but that the mechanisms responsible are variable depending on the conditions. Much less is known about the role of astrocytes in the maintenance of the extracellular calcium concentration ([Ca2 +]o). [Ca2 +]o decreases during intense neuronal activity, including seizures [15,25]. This is thought to reflect the entry of calcium into neurons through voltage-gated calcium channels. Although most populations of astrocytes express voltage-gated calcium channels, it is not known whether calcium also moves into glial cells during neuronal activity. Intracellular calcium in astrocytes increases in response to membrane depolarization, which will also occur during neuronal activity [13]. However, much of the increase in
178
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
internal calcium is thought to come from internal stores. Whether astrocytes contribute to recovery of [Ca2 +]o during neuronal activity, which would require active extrusion of calcium, is not known. Spreading depression (SD) is characterized by rapid depolarization of both neurons and glia with redistribution of ionic gradients between the intracellular and extracellular spaces. Spreading depression evolves as a regenerative process and propagates as a wave in brain tissues [23]. The major extracellular ionic shifts that occur during SD include an increase in potassium [28], and decreases in sodium, chloride, and calcium [6,9,14]. During SD, changes in the ionic concentrations are greater than those recorded during seizure activity, making SD a good model for studying the regulation of the extracellular ionic environment in vivo under conditions of extreme stress, but without the neuronal firing characteristic of seizure activity. In the present experiments, we examined the regulation of [K+]o and [Ca2 +]o during cortical SD after local injections of the reversible glial toxins fluorocitrate (FC) or fluoroacetate (FA). FC and FA are selectively taken up by astrocytes and inhibit aconitase, thus inhibiting the generation of ATP through the Krebs cycle [10,26].
2. Materials and methods All animal experiments were carried out in accordance with the NIH guide for the care and use of laboratory animals (NIH publication No. 8023, revised 1996) and with the approval of the local Animal Use Committee. All efforts were made to minimize the number of animals used and any suffering that might occur. Adult male Sprague – Dawley rats (170 –240 g) were anesthesized with urethane (1.2 – 1.5 g/kg, i.p.) and placed into a stereotaxic frame. Body temperature was maintained at 37 F 0.1 jC with a heating blanket. The skull was exposed and two burr holes were placed on the right side of the skull. The burr hole used for 3 M KCl application was made 2 mm anterior and 2 mm lateral to bregma, and the other burr hole for recordings of
direct current (DC) potential shifts and [K+]o (or [Ca2 +]o) was made 2 –3 mm lateral and 4– 5 mm posterior to bregma. The animals were grounded through a subcutaneous Ag/ AgCl wire in the scapular region. Recordings of extracellular DC potential shift and [K+]o (or [Ca2 +]o) were made with double-barreled ion-selective electrodes manufactured as described previously [21,25]. One barrel of the double-barreled electrodes was silanized with 15% tri-N-butylchlorosilane (Alfrebro, Monroe, OH, USA) in chloroform. For potassium-sensitive electrodes, the tip was filled with Corning 477 317 potassium (Corning Medical, Medfield, MA, USA) or WPI IE190 (World Precision Instruments, Sarasota, FL, USA) and then backfilled with 1 M potassium acetate. No difference was found in the results obtained with the two different exchangers. For calcium electrodes, the tip was filled with WPI IE200 and backfilled with 0.1 M CaCl2. The reference barrels were filled with 2 M NaCl. The reference and ion signals were amplified (Axoprobe 1A; Axon Instruments, USA) and displayed on a chart recorder. Electrodes were calibrated with a series of concentrations of standard solutions before and after each experiment and, on rare occasions when the two calibrations differed by more than 10%, the results from that experiment were not utilized in the analysis. The electrodes had a 45- to 59-mV change in potential for every log increase in ion concentration, although most electrodes were not perfectly linear on the log scale. The recording electrodes were placed in the cortex (down 1– 1.5 mm from dura). Cortical spreading depression was induced by application of 3 M KCl to the surface of the cortex in the anterior burr hole after removal of the dura. During each episode of SD, the [K+]o increased to a peak level and then recovered to baseline (Fig. 1). The peak change in [K+]o was defined as the maximum change in [K+]o (in millimolar) during each SD event. The recovery rate of the [K+]o was defined as the change in [K+]o (in millimolar) from the peak level to 80% of the return to baseline divided by the time for this change to occur, giving a change in [K+]o per second (in millimolar per second). The peak changes in calcium were measured as the maximal decrease or increase in [Ca2 +]o during the spread-
Fig. 1. Changes in DC and extracellular potassium (K+) and extracellular calcium (Ca2 +) during high potassium-induced spreading depression in normal rat cortex. Typical recordings of DC, K+ (A), and Ca2 + (B) are presented and the measurements of each of the parameters are indicated (see Materials and methods for details).
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
ing depression event. The area under the curve was measured from digital tracings using ImageJ (NIH) and the baseline measurement was set at 100%. The time interval (in minutes) between two contiguous SD events was measured as the time between the peaks of the DC traces. When neuronal damage is present, it has been reported that SD waves become smaller and propagate slower [17], which would result in an increase in the interval between SD events. Therefore, this measurement was used as an indication of possible neuronal damage with the local injections. If there was a lengthening of the interval between SD events, it was assumed that there was neuronal damage present and the data from that animal were not used in the analysis of possible glial functions. Two selective, and reversible, toxins were used to reduce astrocytic function—FC and FA. FC is selectively taken up by glial cells and is a ‘suicide’ substrate for the enzyme aconitase [5], temporarily depressing glial function [4,7, 10,22]. FA is taken up into cells and converted into FC. The FC solution was prepared as described by Paulsen et al [22]. Briefly, 8 mg of the barium salt of DL-fluorocitric acid (Sigma, St. Louis, MO, USA) was dissolved in 0.1 M HCl, precipitated by addition of 0.1 M Na2SO4, then buffered with 0.1 M Na2HPO4 and centrifuged at 1000 g for 5 min. The supernatant containing FC was diluted with phosphate-buffered saline to the final concentration (pH 7.3 –7.4). Animals were randomly assigned to receive one of four concentrations of FC (0.01, 0.02, 0.05, or 0.5 mM). Baseline recordings were obtained and then 2 Al of one concentration of FC was injected into the right cortex (down 1 –1.5 mm) via a needle placed into the groove between the two halves of the doublebarreled electrode. FA (pH 7.4) was made by diluting a 1-M aqueous stock solution of the sodium salt of monofluoroacetic acid (Sigma) immediately before use to a final concentration of 10 or 20 mM with phosphate-buffered saline [18]. Animals were randomly assigned to receive one of the two concentrations of FA. Baseline recordings were obtained and then 2 Al of FA was injected into the right cortex. The interval, peak changes in [K+]o and [Ca2 +]o, and recovery rate of [K+]o were monitored for at least 2 h after injection of either FC or FA. Doses of FC and FA were based on reports in the literature [7,18,22,27] and our preliminary experiments. The action of FC appears to be maximal in the first 2– 4 h after injection [22]. Therefore, in the present study, the effects of FC and FA were measured within 3 h after toxin application. Controls consisted of animals injected with matched volumes of the vehicle. After completion of recording, all animals were perfused through the heart with 4% neutral-buffered paraformaldehyde. Brains were postfixed overnight. Sections 35 Am thick were cut horizontally using a Vibratome (Technical Products) and then processed for glial fibrillary acidic protein (GFAP) immunohistochemistry (1:500; Dako, Carpinteria, CA, USA) with diaminobenzidine for visualization. This verified that the local injections of FC or FA had no effect on the morphology of the astrocytes.
179
Astrocytes are known to take up glutamate that has been released from neurons [8,18,28]. This glutamate is converted into glutamine, for conversion back into glutamate in neurons. Therefore, a loss of astrocytic function may result in a disturbance in glutamate homeostasis. In fact, dialysis of the hippocampus with FC has been reported to cause, first, a decrease and then an increase in the extracellular levels of glutamate during spreading depression [16]. To control for this possible effect of the glial toxins, additional animals were locally injected with 4 Al of 3 mM glutamate (Sigma), or 4 Al of 2 mM DL-threo-h-benzyloxyaspartate (TBOA; Tocris Cookson, Ellisville, MO, USA), a nonspecific glutamate uptake inhibitor, or 4 Al of the combination of 2 mM TBOA and 0.05 mM FC. Solutions of TBOA and glutamate were made in phosphate-buffered saline, pH 7.3 – 7.4. DC and [K+]o recordings were obtained before and then up to 4 h after local injections of these solutions. Because of the variability in the SD episodes, the parameters were measured and averaged for four to six SD episodes. SD events more than 10 min before or after the time point were not included in the analysis. The peak [K+]o and [Ca2 +]o and recovery rate for [K+]o before and after FC/ FA or vehicle were determined in each animal and, within an experimental group, compared with a paired t test. The means for these numbers are reported in the Results section. The change in each parameter at each time point after injection of FC or FA was then calculated for each animal. The results after injection of vehicle or the various doses of FC or FA were compared using one-way analysis of variance (ANOVA) followed by the Dunnett’s test for multiple comparisons to control. Results for FC and FA were analyzed separately. There was no difference between the statistical significance determined by the paired t test and the ANOVA. For example, if an FC injection produced a statistically significant change in the recovery rate of potassium compared to baseline with the paired t test, then the results from this drug group were also statistically different from the injection of vehicle as determined with the ANOVA.
3. Results To determine the reproducibility and stability of the SD events and the corresponding shifts in extracellular potassium concentration ([K+]o), spreading depression events were monitored continuously for 8 h in one animal. The recovery rate of the [K+]o and peak [K+]o were stable for 4 h. As controls for the injection and to further test the stability and reproducibility of the SD, animals (n = 5) were followed before and after injection of vehicle. After 90 min of recording of the DC and [K+]o (or [Ca2 +]o) changes during SD, 4 Al of phosphate-buffered saline was injected into the cortex. The baseline measurements were quite stable with a variability of less than 15% between consecutive SD events. The measurements of [K+]o and [Ca2 +]o changes were also stable for up to 3 h after injection of PBS (n = 6, data not
180
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
shown). There was no statistically significant change in the peak [K+]o, recovery rate of potassium, area under the curve for both [K+]o and [Ca2 +]o, or peak changes in [Ca2 +]o after injection of vehicle ( p>0.05). Beyond 3 h, there was a gradual decline in the peak [K+]o and an increase in the time between SD episodes. The physiological basis for these changes is not known. Concentrations of FC lower than 0.01 mM had no effect on any of the measured parameters for [K+]o (data not shown). Injection of 0.01 (n = 6) or 0.02 mM FC (n = 6) significantly increased the peak [K+]o measured during SD (47 F 2.7 mM, 1 h after FC injection compared to 35 F 2.0 mM before FC injection, p < 0.05), while having no effect on the other measured parameters, including the interval between SD events. Injection of 0.05 mM FC (Fig. 2; n = 8) had no significant effect on the peak change in [K+]o (35.3 F 1.5, 1 h after injection of FC compared to 36.7 F 1.7 mM baseline, p>0.05), but the recovery of [K+]o was significantly slowed (61.0 F 7.0% of the baseline recovery rate, p < 0.01) and the area under the curve was significantly increased ( p < 0.005; 1 h after FC, the area was 214 F 22% compared to the baseline of 100%) . With this dose of FC, there was a trend towards a decrease in the interval between SD events, but this did not reach statistical significance (baseline: 4.6 F 0.5; 1 h after FC: 3.9 F 0.5; 2 h: 3.3 F 0.7). Injection of 0.5 mM FC (n = 6) significantly decreased the peak change in [K+]o (22.1 F 1.1 mM, 1 h after injection of FC compared to 40.8 F 1.0 mM baseline, p < 0.01) and significantly slowed the recovery rate of [K+]o (21.5 F 2.9% of the baseline recovery rate, p < 0.01). At doses < 0.5 mM, there was no change in the baseline level of [K+]o between episodes of SD. At doses above 0.5 mM FC, the interval was significantly increased 2 h after FC (baseline: 4.8 F 0.9 min; 1 h after FC: 3.4 F 0.3; 2 h: 3.0 F 0.0.3) and baseline levels of [K+]o rose gradually. Therefore, doses above 0.5 mM were not used for the current analysis. There were some changes in the shape of the [K+]o tracing after injection of FC that are not represented by the measured parameters. Before injection of FC, the shape of the [K+]o curve was symmetric with a smooth increase and then decreased (Fig. 1; Fig. 2, ‘‘before’’). Early after administration of FC, there was a broadening at the peak of the [K+]o recording (Fig. 2). At higher doses of FC, or after more time had passed, there appeared to be two phases to the increase in [K+]o. An early fast phase, which appeared unchanged from baseline, was followed by a slower phase. The results with FC were confirmed with local injections of FA, which is not a barium salt. Injection of 2 Al of 10 mM FA (n = 6) significantly increased the peak change in [K+]o at 0.5, 1, and 2 h after injection (49.4 F 2.3 mM, 1 h after injection compared to 28.5 F 1.22 mM baseline, p < 0.01). Injection of 2 Al of 20 mM FA (n = 6) had no significant effect on the peak [K+]o reached during SD (25.6 F 1.5 mM, 1 h after injection compared to 28.2 F 2.3 baseline, p>0.05). The recovery rate was significantly decreased at 1 h (60.3 F 7.1% of the baseline recovery
Fig. 2. The DC and K+ recordings obtained before and after injection of FC or FA during high K+-induced cortical SD. Results from three different animals are presented. In (A), 2 h after administration of FC (0.05 mM), there was an increase in the width of the [K+]o trace and a decrease in the interval between SD events. In (B), the change in the shape of the potassium transient is illustrated at 1 and 2 h after injection of FC. In (C), the shape of the [K+]o curve before and at three time points after injection of 20 mM FA is shown. Calibrations are indicated on the figure and are the same in all panels.
rate, p < 0.01). Injections of FA were associated with the same changes in shape (Fig. 2C) that were described after injection of FC. There was broadening of the [K+]o tracing with the appearance of a longer latency peak. To further test that the residual barium was not the source of the changes after administration of FC, 100 AM barium chloride solution was injected into the cortex using identical methods. Local injection of BaCl2 had no effect on the interval between the SD episodes, the shape of the DC
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
181
Fig. 3. The effects of local injection of barium, glutamate, or TBOA on high K+-induced cortical SD. In each panel, the DC and K+ recordings obtained before and 1 h after injection of 0.1 mM barium chloride (A), 3 mM glutamate (B), or 2 mM TBOA, a glutamate uptake blocker (C), are presented. There was no change in the measured parameters of potassium regulation during SD after injection of any of these substances. Calibrations indicated on the figure are the same for all panels.
potential, or the [K+]o changes recorded during the SD events (n = 5; Fig. 3A). The changes recorded after administration of FC or FA could be due to an increase in extracellular glutamate concentration. To test this hypothesis, additional animals were injected with glutamate or a glutamate uptake blocker, TBOA. The peak [K+]o and the recovery rate of the [K+]o were not changed by local injections of 3 mM glutamate (Fig. 3B; n = 5, 99.5 F 9.5 of the baseline recovery rate, p>0.05) or 2 mM TBOA (Fig. 3C; n = 5, 102.3 F 10.0% of the baseline recovery rate, p>0.05). Addition of 2 mM TBOA had no significant effect on the ability of 0.05 mM FC to alter the rate of recovery of the [K+]o (n = 5). These results suggest that the effects of FC and FA on [K+]o are not due to alterations in glutamate levels or glutamate uptake. Measurements of [Ca2 +]o also demonstrated significant changes after injection of FC or FA. In the untreated cortex, or after injection of PBS, a decrease in [Ca2 +]o began before the
decrease in DC potential that signaled the spreading depression (Fig. 4). This decrease in [Ca2 +]o reached a peak and began to recover before the peak negativity of the DC potential. The [Ca2 +]o did not simply recover to the baseline levels, but actually peaked a second time above the baseline levels. The overshoot was almost a mirror image of the initial decrease in [Ca2 +]o, but somewhat smaller in peak amplitude. Membrane depolarization of astrocytes is proportional to the [K+]o and also contributes to calcium fluxes [13]. Therefore, the [Ca2 +]o experiments were done using the concentration of FC that did not significantly alter the peak [K+]o (i.e., 0.05 mM). After local injection of 0.05 mM FC, there was no significant change in the baseline levels of [Ca2 +]o or in the peak decrease in [Ca2 +]o that was measured in the early part of the SD wave. The peak was 0.69 F 0.13 mM before injection of FC and 0.8 F 0.11 mM 2 h after injection of 0.05 mM FC (n = 7). However, there was a dramatic change in the waveform after this early peak (Fig. 4). Before injection of
Fig. 4. The DC and Ca2 + recordings obtained before and after injection of FC (0.05 mM) during high K+-induced cortical SD. Changes in the [Ca2 +]o tracing were measurable within 30 min of the injection of FC. Changes were maximal between 1 and 2 h after injection. Calibrations are indicated on the figure.
182
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
FC, the peak increase in [Ca2 +]o during the later part of the SD wave was 0.37 F 0.25 mM. However, within minutes of injection of FC, this late positive peak began to get smaller. At 2 h after injection of 0.05 mM FC (the time of peak FC effect), there was a second negative peak in the [Ca2 +]o tracing of 1.15 F 0.31 mM (n = 7). So, what was initially a positive peak had become a negative peak. As might be expected, there was variability in the shape of the [Ca2 +]o between the two negative peaks now present. During most SD events, there was partial recovery before the second downward movement. These results were confirmed with local injection of 20 mM FA (n = 5). Histological examination of the brains from these animals indicated that the distribution and overall shape of the astrocytes were not changed by the local injection of FC or FA. One set of animals was examined 2 h after toxin injection (n = 3 for 0.02, 0.05, and 0.5 mM FC). Additional animals were examined 4 – 6 h after injection (n = 5 for 0.05 mM FC; n = 6 for 20 mM FA). The results from all of these animals were the same. Compared to the region around the vehicle injection (Fig. 5A), there was no increase in the thickness of the astrocytic processes, nor was there a loss of
Fig. 5. Effects of injection on the morphology of GFAP-positive cells in the cortex. Each panel presents a section of the cortex adjacent to the injection site from animals sacrificed 3 h after injection of phosphate-buffered saline (A) or FC (B) (0.05 mM). Each photograph was obtained at the same depth from the dura. The scale bar indicates 100 Am.
astrocytes in the region surrounding the injection of FC or FA (Fig. 5B).
4. Discussion The experiments presented here demonstrate that astrocytes participate in the regulation of [K+]o during spreading depression. Moderate doses of FC and FA, which will specifically inhibit glial function, slowed the rate of [K+]o recovery from elevated levels. Together, these data indicate that loss (or reduction) of glial cell function results in a modest, but significant, decrease in the rate of clearing of potassium from the extracellular space. Slowing of the recovery phase could be due to restricted diffusion due to cell swelling, but this is not consistent with the histology for either FC/FA, or a change in the recovery from the swelling that occurs during spreading depression [2]. The present results are comparable to the changes in [K+]o regulation recorded during and after seizure activity in the hippocampus [29], where local administration of FC/FA resulted in a slowing of the half-time of recovery of [K+]o. The results are also consistent with the conclusions of Largo et al. [16], where FC was dialyzed into the CA1 region of the hippocampus. The lack of effect of locally applied barium and the relatively small effects of the glial toxins suggest that the predominant uptake after spreading depression is through Na+ –K+ ATPase activity. This is supported by the fact that spreading depression is associated with a decrease in extracellular sodium levels, which will stimulate sodium pump activity [1]. At the doses that produced a slowing of the recovery of the [K+]o, there was also a consistent change in the overall shape of the tracings of the [K+]o that is harder to interpret. The increase in the width of the curve is most probably due to the slower recovery rate. There does not appear to be a slowing of the initial rise in [K+]o, suggesting that potassium movement out of neurons into the extracellular space has not changed. After administration of FC or FA, there was the additional appearance of a second peak with a slower rise time, which also contributed to the broadening of the [K+]o tracing. This change in the shape of the DC potential during spreading depression was also observed during dialysis of FC into the hippocampus [17] or bath application of FA to hippocampal slices [18]. The physiological changes underlying this change in shape cannot be deduced from the present experiments, but it has been suggested that astrocytes have a role in the moderation or propagation of spreading depression [17,18,20]. A slower propagation could explain the broader waves with smaller amplitude. The alterations in the peak [K+]o are more difficult to interpret, especially since the effects were dose-dependent. Changes in the peak [K+]o could be due to changes in the baseline [K+]o, but no baseline changes were recorded in the present experiments. The increase in peak [K+]o (with the low doses of FC/FA) could be due to reduced glial uptake, if
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
that uptake was an active process, since FC/FA will inhibit synthesis of glutamine and ATP [3,7,22,26]. The decrease in peak [K+]o is more difficult to interpret, but may be due to the slower propagation of spreading depression in the presence of reduced glial function [17,18,20]. It is also possible that the vasodilatation associated with spreading depression (for review, see Ref. [19]) produces a transient increase in the removal of potassium from the extracellular space. In addition, it has recently been shown that astrocytes play a role in the local regulation of cerebral blood flow [32]. It is not yet known whether a reduction in glial function would alter local blood flow during spreading depression. The changes in [Ca2 +]o during spreading depression suggest that there are at least two competing processes ongoing under normal conditions. The early movement of calcium into cells, either neurons, glia, or both, reduces the [Ca2 +]o. This appears to begin before a measurable change in DC potential, suggesting that the calcium influx may be the trigger for the SD event. That this early decrease in calcium is not affected by local injection of FC/FA suggests that it does not require ATP generated by glial cells. Since the [Ca2 +]o is much higher in the extracellular space than inside of both neurons and glial cells, this movement of calcium out of the extracellular space is most likely secondary to the opening of membrane calcium channels. The [Ca2 +]o recovery consists of an overshoot, suggesting an active extrusion of calcium from cells in an amount greater than that which had moved into the cells. This suggests that calcium is being released from internal stores and actively pumped out of cells—either neurons or glia. Glial cells are known to have significant internal stores of calcium, which can be released in response to decreases in extracellular calcium levels [24,31]. This, combined with the fact that FC/FA will block the synthesis of ATP, thus reducing any active pumping, makes it most likely that the astrocytes are responsible for the recovery of the [Ca2 +]o and overshoot during the time frame of the SD episode. That the [Ca2 +]o levels eventually recover back to baseline indicates that either passive movement of calcium is sufficient to restore proper levels, that neurons can extrude calcium sufficiently to restore levels, or that there is some residual pumping ability within the glial cells that contributes to the restoration of [Ca2 +]o levels—albeit more slowly than normal. A number of factors suggest that the doses of FC/FA used in these experiments are not toxic to neurons. First, concentrations of 0.005 – 0.1 mM FC, or less, have been reported to act as a selective glial toxin in neuronal and glial cell cultures [11]. Injection of 1 nmol of FC produces selective and reversible glial cell damage in the striatum, without any ultrastructural evidence of neuronal damage [22]. Therefore, it has been suggested that this dose of FC (1 nmol) produces a selective and temporary impairment of glial cell energy metabolism and could be used as a model for the study of the importance of glial cells for brain function in vivo [22]. The doses used in the present study,
183
which were 0.02– 0.08 nmol of FC, are clearly within the glial-selective range. Secondly, in SD models, a lengthening of the time between SD events has been shown to correlate with neuronal damage [16]. In this study, the doses/concentration of FC/FA that were used did not increase the interval between SD events, suggesting that they were not causing neuronal damage. Therefore, the effects reported here are most likely due to selective changes in glial function. In summary, this study demonstrates that inhibition of glial function consistently slowed the recovery of the [K+]o and altered the active recovery of calcium during spreading depression. These data suggest that astrocytes normally participate in the regulation of extracellular potassium and calcium concentrations in the cortex.
Acknowledgements This work was supported by a grant to J.L.S. from the NINDS (NS39941).
References [1] T. Amedee, A. Robert, J.A. Coles, Potassium homeostasis and glial energy metabolism, Glia 21 (1997) 46 – 55. [2] M. Anderova, S. Kubinova, T. Mazel, A. Chvatal, C. Eliasson, M. Pekny, E. Sykova, Effect of elevated K+, hypotonic stress, and cortical spreading depression on astrocyte swelling in GFAP-deficient mice, Glia 35 (2001) 189 – 203. [3] A.M. Benjamin, Z.H. Verjee, Control of aerobic glycolysis in the brain in vitro, Neurochem. Res. 5 (1980) 921 – 934. [4] J. Berg-Johnsen, R.E. Paulsen, F. Fonnum, I.A. Langmoen, Changes in evoked potentials and amino acid content during fluorocitrate action studied in rat hippocampal slices, Exp. Brain Res. 96 (1993) 241 – 246. [5] D.D. Clarke, W.J. Nicklas, S. Berl, Tricarboxylic acid-cycle metabolism in brain. Effect of fluoroacetate and fluorocitrate on the labeling of glutamate, aspartate, glutamine and g-aminobutyrate, Biochem. J. 120 (1970) 345 – 351. [6] R.J. Do Carmo, H. Martins-Ferreira, Spreading depression of Lea˜o probed with ion-selective microelectrodes in isolated chick retina, An. Acad. Bras. Cienc. 56 (1984) 401 – 421. [7] F. Fonnum, A. Johnsen, B. Hassel, Use of fluorocitrate and fluoroacetate in study of brain metabolism, Glia 21 (1997) 106 – 113. [8] A. Furuta, J.D. Rothstein, J.L. Martin, Glutamate transporter protein subtypes are expressed differentially during rat central nervous system development, J. Neurosci. 17 (1997) 8363 – 8375. [9] A.J. Hansen, T. Zeuthen, Extracellular ion concentrations during spreading depression and ischemia in the rat brain cortex, Acta Physiol. Scand. 113 (1981) 437 – 445. [10] B. Hassel, R.E. Paulson, A. Johnsen, F. Fonnum, Selective inhibition of glial cell metabolism in vivo by fluorocitrate, Brain Res. 576 (1992) 120 – 124. [11] B. Hassel, N. Westergaard, A. Schousboe, F. Fonnum, Metabolic differences between primary cultures of astrocytes and neurons from cerebellum and cerebral cortex: effects of fluorocitrate, Neurochem. Res. 20 (1995) 413 – 420. [12] R. Jauch, O. Windmuller, T.-N. Lehmann, U. Heinemann, S. Gabriel, Effects of barium, furosemide, ouabaine and 4,4V-diisothiocyanatostilbene-2,2V-disulfonic acid (DIDS) on iontophoretically-induced changes in extracellular potassium concentration in hippocampal slices
184
[13] [14] [15]
[16]
[17]
[18]
[19] [20]
[21]
[22]
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184 from rats and from patients with epilepsy, Brain Res. 925 (2002) 18 – 27. H. Kettenmann, Physiology of glial cells, Adv. Neurol. 79 (1999) 565 – 571. R.P. Kraig, C. Nicholson, Extracellular ionic variations during spreading depression, Neuroscience 3 (1978) 1045 – 1059. K. Krnjevic, M.E. Morris, R.J. Reiffenstein, Stimulation-evoked changes in extracellular K+ and Ca2 + in pyramidal layers of the rat’s hippocampus, Can. J. Physiol. Pharmacol. 60 (1982) 1643 – 1657. C. Largo, P. Cuevas, G.G. Somjen, R. Martin Del Rio, O. Herreras, The effect of depressing glial function on rat brain in situ on homeostasis, synaptic transmission and neuronal survival, J. Neurosci. 16 (1996) 1219 – 1229. C. Largo, J.M. Ibrarz, O. Herreras, Effects of the gliotoxin fluorocitrate on spreading depression and glial membrane potential in rat brain in situ, J. Neurophysiol. 78 (1997) 295 – 307. C. Largo, G.C. Tombaugh, P.G. Aitken, O. Herreras, G.G. Somjen, Heptanol but not fluoroacetate prevents the propagation of spreading depression in rat hippocampal slices, J. Neurophysiol. 77 (1997) 9 – 16. M. Lauritzen, Pathophysiology of the migraine aura. The spreading depression theory, Brain 117 (1994) 199 – 210. M. Muller, G.G. Somjen, Intrinsic optical signals in rat hippocampal slices during hypoxia-induced spreading depression-like depolarization, J. Neurophysiol. 82 (1999) 1818 – 1831. C. Nicholson, Ion-selective microelectrodes and diffusion measurement as tools to explore the brain cell microenvironment, J. Neurosci. Methods 48 (1993) 199 – 213. R.E. Paulsen, A. Contestablie, L. Villani, F. Fonnum, An in vivo model for studying function of brain tissue temporarily devoid of glial cell metabolism: the use of fluorocitrate, J. Neurochem. 48 (1987) 1377 – 1385.
[23] G.G. Somjen, Mechanisms of spreading depression and hypoxic spreading depression-like depolarization, Physiol. Rev. 81 (2001) 1065 – 1096. [24] C. Stout, A. Charles, Modulation of intercellular calcium signaling in astrocytes by extracellular calcium and magnesium, Glia 43 (2003) 265 – 273. [25] J.L. Stringer, E.W. Lothman, Model of spontaneous hippocampal epilepsy in the anesthetized rat: electrographic, [K+]o, and [Ca2 +]o response patterns, Epilepsy Res. 4 (1989) 177 – 186. [26] R.A. Swanson, S.H. Graham, Fluorocitrate and fluoroacetate effects on astrocyte metabolism in vitro, Brain Res. 664 (1994) 94 – 100. [27] J.C. Szerb, Glutamate release and spreading depression in the fascia dentata in response to microdialysis with high K+: role of glia, Brain Res. 542 (1991) 259 – 265. [28] F. Vyskocil, N. Kriz, J. Bures, Potassium-selective microelectrodes used for measuring the extracellular brain potassium during spreading depression and anoxic depolarization in rats, Brain Res. 39 (1972) 255 – 259. [29] Z.-Q. Xiong, J.L. Stringer, Astrocytic regulation of the recovery of extracellular potassium after seizures in vivo, Eur. J. Neurosci. 11 (1999) 1677 – 1684. [30] Z.-Q. Xiong, J.L. Stringer, Sodium pump activity, not glial spatial buffering, clears potassium after epileptiform activity induced with 0-added calcium and 8 mM potassium in the dentate gyrus in vitro, J. Neurophysiol. 83 (2000) 1443 – 1451. [31] S. Zanotti, A. Charles, Extracellular calcium sensing by glial cells: low extracellular calcium induces intracellular calcium release and intercellular signaling, J. Neurochem. 69 (1997) 594 – 602. [32] M. Zonta, M.C. Angulo, S. Gobbo, B. Rosengarten, K.-A. Hossmann, T. Pozzan, G. Carmignoto, Neuron-to-astrocyte signaling is central to the dynamic control of brain microcirculation, Nat. Neurosci. 6 (2003) 42 – 50.
Research report
Astrocytes contribute to regulation of extracellular calcium and potassium in the rat cerebral cortex during spreading depression Xiao-Yuan Lian, Janet L. Stringer * Department of Pharmacology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA Accepted 3 April 2004 Available online 10 May 2004
Abstract This study used spreading depression (SD), which is characterized by redistribution of ions, to examine the role of astrocytes in the regulation of extracellular potassium ([K+]o) and calcium ([Ca2 +]o) levels. Recurrent spreading depression episodes were induced by application of 3 M potassium chloride to the cortex of adult anesthetized rats while monitoring the extracellular direct current (DC) potential shifts and changes in [K+]o or [Ca2 +]o 6 – 7 mm away. The reversible glial toxins, fluorocitrate (FC) and fluoroacetate (FA), were injected locally into the cortex at doses that are selective for reducing glial function. The peak changes and area under the curve for [K+]o and [Ca2 +]o, recovery rate for [K+]o, and interval between spreading depression episodes were measured before and at various times after administration of the toxins. Both fluorocitrate and fluroacetate slowed the recovery of the [K+]o and altered the recovery of the [Ca2 +]o. Local injection of glutamate uptake inhibitors or barium had no effect on the peak changes in [K+]o or the rate of recovery of the [K+]o. The slowing of the recovery rate is consistent with the hypothesis that glial cells play a role in the return of [K+]o to baseline after spreading depression in the cortex in vivo. The change in movement of calcium after administration of FC suggests that astrocytes normally extrude calcium during spreading depression, resulting in rapid recovery of the levels of [Ca2 +]o with an overshoot. These findings demonstrate that astrocytes contribute to the regulation of both potassium and calcium during and after a stress to the ionic homeostatic mechanisms. D 2004 Elsevier B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Epilepsy: basic mechanisms Keywords: Fluorocitrate; Fluoroacetate; Ion homeostasis
1. Introduction Astrocytes have been postulated to play a role in maintaining, or restoring, extracellular ion concentrations, particularly potassium, during and after intense neuronal activity. During stimulus trains or seizure activity in the hippocampus, the extracellular potassium concentration ([K+]o) will reach a plateau level that is sustained until the end of the neuronal activity [29]. Previously, it has been shown that astrocytes participate in the return of the [K+]o to normal levels after seizures in the dentate gyrus, but not in the determination of the plateau level reached during seizures [29]. The plateau level appears to be determined more by active uptake into neurons by the Na+ – K+ ATPase, than * Corresponding author. Tel.: +1-713-798-7294; fax: +1-713-7983145. E-mail address: [email protected] (J.L. Stringer). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2004.04.011
by uptake into glial cells [30]. In the absence of neuronal activity, uptake of potassium (after iontophoretic application) appears to be mediated largely through barium-sensitive uptake mechanisms in glia [12]. Thus, it appears that astrocytes can participate in the regulation of [K+]o, but that the mechanisms responsible are variable depending on the conditions. Much less is known about the role of astrocytes in the maintenance of the extracellular calcium concentration ([Ca2 +]o). [Ca2 +]o decreases during intense neuronal activity, including seizures [15,25]. This is thought to reflect the entry of calcium into neurons through voltage-gated calcium channels. Although most populations of astrocytes express voltage-gated calcium channels, it is not known whether calcium also moves into glial cells during neuronal activity. Intracellular calcium in astrocytes increases in response to membrane depolarization, which will also occur during neuronal activity [13]. However, much of the increase in
178
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
internal calcium is thought to come from internal stores. Whether astrocytes contribute to recovery of [Ca2 +]o during neuronal activity, which would require active extrusion of calcium, is not known. Spreading depression (SD) is characterized by rapid depolarization of both neurons and glia with redistribution of ionic gradients between the intracellular and extracellular spaces. Spreading depression evolves as a regenerative process and propagates as a wave in brain tissues [23]. The major extracellular ionic shifts that occur during SD include an increase in potassium [28], and decreases in sodium, chloride, and calcium [6,9,14]. During SD, changes in the ionic concentrations are greater than those recorded during seizure activity, making SD a good model for studying the regulation of the extracellular ionic environment in vivo under conditions of extreme stress, but without the neuronal firing characteristic of seizure activity. In the present experiments, we examined the regulation of [K+]o and [Ca2 +]o during cortical SD after local injections of the reversible glial toxins fluorocitrate (FC) or fluoroacetate (FA). FC and FA are selectively taken up by astrocytes and inhibit aconitase, thus inhibiting the generation of ATP through the Krebs cycle [10,26].
2. Materials and methods All animal experiments were carried out in accordance with the NIH guide for the care and use of laboratory animals (NIH publication No. 8023, revised 1996) and with the approval of the local Animal Use Committee. All efforts were made to minimize the number of animals used and any suffering that might occur. Adult male Sprague – Dawley rats (170 –240 g) were anesthesized with urethane (1.2 – 1.5 g/kg, i.p.) and placed into a stereotaxic frame. Body temperature was maintained at 37 F 0.1 jC with a heating blanket. The skull was exposed and two burr holes were placed on the right side of the skull. The burr hole used for 3 M KCl application was made 2 mm anterior and 2 mm lateral to bregma, and the other burr hole for recordings of
direct current (DC) potential shifts and [K+]o (or [Ca2 +]o) was made 2 –3 mm lateral and 4– 5 mm posterior to bregma. The animals were grounded through a subcutaneous Ag/ AgCl wire in the scapular region. Recordings of extracellular DC potential shift and [K+]o (or [Ca2 +]o) were made with double-barreled ion-selective electrodes manufactured as described previously [21,25]. One barrel of the double-barreled electrodes was silanized with 15% tri-N-butylchlorosilane (Alfrebro, Monroe, OH, USA) in chloroform. For potassium-sensitive electrodes, the tip was filled with Corning 477 317 potassium (Corning Medical, Medfield, MA, USA) or WPI IE190 (World Precision Instruments, Sarasota, FL, USA) and then backfilled with 1 M potassium acetate. No difference was found in the results obtained with the two different exchangers. For calcium electrodes, the tip was filled with WPI IE200 and backfilled with 0.1 M CaCl2. The reference barrels were filled with 2 M NaCl. The reference and ion signals were amplified (Axoprobe 1A; Axon Instruments, USA) and displayed on a chart recorder. Electrodes were calibrated with a series of concentrations of standard solutions before and after each experiment and, on rare occasions when the two calibrations differed by more than 10%, the results from that experiment were not utilized in the analysis. The electrodes had a 45- to 59-mV change in potential for every log increase in ion concentration, although most electrodes were not perfectly linear on the log scale. The recording electrodes were placed in the cortex (down 1– 1.5 mm from dura). Cortical spreading depression was induced by application of 3 M KCl to the surface of the cortex in the anterior burr hole after removal of the dura. During each episode of SD, the [K+]o increased to a peak level and then recovered to baseline (Fig. 1). The peak change in [K+]o was defined as the maximum change in [K+]o (in millimolar) during each SD event. The recovery rate of the [K+]o was defined as the change in [K+]o (in millimolar) from the peak level to 80% of the return to baseline divided by the time for this change to occur, giving a change in [K+]o per second (in millimolar per second). The peak changes in calcium were measured as the maximal decrease or increase in [Ca2 +]o during the spread-
Fig. 1. Changes in DC and extracellular potassium (K+) and extracellular calcium (Ca2 +) during high potassium-induced spreading depression in normal rat cortex. Typical recordings of DC, K+ (A), and Ca2 + (B) are presented and the measurements of each of the parameters are indicated (see Materials and methods for details).
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
ing depression event. The area under the curve was measured from digital tracings using ImageJ (NIH) and the baseline measurement was set at 100%. The time interval (in minutes) between two contiguous SD events was measured as the time between the peaks of the DC traces. When neuronal damage is present, it has been reported that SD waves become smaller and propagate slower [17], which would result in an increase in the interval between SD events. Therefore, this measurement was used as an indication of possible neuronal damage with the local injections. If there was a lengthening of the interval between SD events, it was assumed that there was neuronal damage present and the data from that animal were not used in the analysis of possible glial functions. Two selective, and reversible, toxins were used to reduce astrocytic function—FC and FA. FC is selectively taken up by glial cells and is a ‘suicide’ substrate for the enzyme aconitase [5], temporarily depressing glial function [4,7, 10,22]. FA is taken up into cells and converted into FC. The FC solution was prepared as described by Paulsen et al [22]. Briefly, 8 mg of the barium salt of DL-fluorocitric acid (Sigma, St. Louis, MO, USA) was dissolved in 0.1 M HCl, precipitated by addition of 0.1 M Na2SO4, then buffered with 0.1 M Na2HPO4 and centrifuged at 1000 g for 5 min. The supernatant containing FC was diluted with phosphate-buffered saline to the final concentration (pH 7.3 –7.4). Animals were randomly assigned to receive one of four concentrations of FC (0.01, 0.02, 0.05, or 0.5 mM). Baseline recordings were obtained and then 2 Al of one concentration of FC was injected into the right cortex (down 1 –1.5 mm) via a needle placed into the groove between the two halves of the doublebarreled electrode. FA (pH 7.4) was made by diluting a 1-M aqueous stock solution of the sodium salt of monofluoroacetic acid (Sigma) immediately before use to a final concentration of 10 or 20 mM with phosphate-buffered saline [18]. Animals were randomly assigned to receive one of the two concentrations of FA. Baseline recordings were obtained and then 2 Al of FA was injected into the right cortex. The interval, peak changes in [K+]o and [Ca2 +]o, and recovery rate of [K+]o were monitored for at least 2 h after injection of either FC or FA. Doses of FC and FA were based on reports in the literature [7,18,22,27] and our preliminary experiments. The action of FC appears to be maximal in the first 2– 4 h after injection [22]. Therefore, in the present study, the effects of FC and FA were measured within 3 h after toxin application. Controls consisted of animals injected with matched volumes of the vehicle. After completion of recording, all animals were perfused through the heart with 4% neutral-buffered paraformaldehyde. Brains were postfixed overnight. Sections 35 Am thick were cut horizontally using a Vibratome (Technical Products) and then processed for glial fibrillary acidic protein (GFAP) immunohistochemistry (1:500; Dako, Carpinteria, CA, USA) with diaminobenzidine for visualization. This verified that the local injections of FC or FA had no effect on the morphology of the astrocytes.
179
Astrocytes are known to take up glutamate that has been released from neurons [8,18,28]. This glutamate is converted into glutamine, for conversion back into glutamate in neurons. Therefore, a loss of astrocytic function may result in a disturbance in glutamate homeostasis. In fact, dialysis of the hippocampus with FC has been reported to cause, first, a decrease and then an increase in the extracellular levels of glutamate during spreading depression [16]. To control for this possible effect of the glial toxins, additional animals were locally injected with 4 Al of 3 mM glutamate (Sigma), or 4 Al of 2 mM DL-threo-h-benzyloxyaspartate (TBOA; Tocris Cookson, Ellisville, MO, USA), a nonspecific glutamate uptake inhibitor, or 4 Al of the combination of 2 mM TBOA and 0.05 mM FC. Solutions of TBOA and glutamate were made in phosphate-buffered saline, pH 7.3 – 7.4. DC and [K+]o recordings were obtained before and then up to 4 h after local injections of these solutions. Because of the variability in the SD episodes, the parameters were measured and averaged for four to six SD episodes. SD events more than 10 min before or after the time point were not included in the analysis. The peak [K+]o and [Ca2 +]o and recovery rate for [K+]o before and after FC/ FA or vehicle were determined in each animal and, within an experimental group, compared with a paired t test. The means for these numbers are reported in the Results section. The change in each parameter at each time point after injection of FC or FA was then calculated for each animal. The results after injection of vehicle or the various doses of FC or FA were compared using one-way analysis of variance (ANOVA) followed by the Dunnett’s test for multiple comparisons to control. Results for FC and FA were analyzed separately. There was no difference between the statistical significance determined by the paired t test and the ANOVA. For example, if an FC injection produced a statistically significant change in the recovery rate of potassium compared to baseline with the paired t test, then the results from this drug group were also statistically different from the injection of vehicle as determined with the ANOVA.
3. Results To determine the reproducibility and stability of the SD events and the corresponding shifts in extracellular potassium concentration ([K+]o), spreading depression events were monitored continuously for 8 h in one animal. The recovery rate of the [K+]o and peak [K+]o were stable for 4 h. As controls for the injection and to further test the stability and reproducibility of the SD, animals (n = 5) were followed before and after injection of vehicle. After 90 min of recording of the DC and [K+]o (or [Ca2 +]o) changes during SD, 4 Al of phosphate-buffered saline was injected into the cortex. The baseline measurements were quite stable with a variability of less than 15% between consecutive SD events. The measurements of [K+]o and [Ca2 +]o changes were also stable for up to 3 h after injection of PBS (n = 6, data not
180
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
shown). There was no statistically significant change in the peak [K+]o, recovery rate of potassium, area under the curve for both [K+]o and [Ca2 +]o, or peak changes in [Ca2 +]o after injection of vehicle ( p>0.05). Beyond 3 h, there was a gradual decline in the peak [K+]o and an increase in the time between SD episodes. The physiological basis for these changes is not known. Concentrations of FC lower than 0.01 mM had no effect on any of the measured parameters for [K+]o (data not shown). Injection of 0.01 (n = 6) or 0.02 mM FC (n = 6) significantly increased the peak [K+]o measured during SD (47 F 2.7 mM, 1 h after FC injection compared to 35 F 2.0 mM before FC injection, p < 0.05), while having no effect on the other measured parameters, including the interval between SD events. Injection of 0.05 mM FC (Fig. 2; n = 8) had no significant effect on the peak change in [K+]o (35.3 F 1.5, 1 h after injection of FC compared to 36.7 F 1.7 mM baseline, p>0.05), but the recovery of [K+]o was significantly slowed (61.0 F 7.0% of the baseline recovery rate, p < 0.01) and the area under the curve was significantly increased ( p < 0.005; 1 h after FC, the area was 214 F 22% compared to the baseline of 100%) . With this dose of FC, there was a trend towards a decrease in the interval between SD events, but this did not reach statistical significance (baseline: 4.6 F 0.5; 1 h after FC: 3.9 F 0.5; 2 h: 3.3 F 0.7). Injection of 0.5 mM FC (n = 6) significantly decreased the peak change in [K+]o (22.1 F 1.1 mM, 1 h after injection of FC compared to 40.8 F 1.0 mM baseline, p < 0.01) and significantly slowed the recovery rate of [K+]o (21.5 F 2.9% of the baseline recovery rate, p < 0.01). At doses < 0.5 mM, there was no change in the baseline level of [K+]o between episodes of SD. At doses above 0.5 mM FC, the interval was significantly increased 2 h after FC (baseline: 4.8 F 0.9 min; 1 h after FC: 3.4 F 0.3; 2 h: 3.0 F 0.0.3) and baseline levels of [K+]o rose gradually. Therefore, doses above 0.5 mM were not used for the current analysis. There were some changes in the shape of the [K+]o tracing after injection of FC that are not represented by the measured parameters. Before injection of FC, the shape of the [K+]o curve was symmetric with a smooth increase and then decreased (Fig. 1; Fig. 2, ‘‘before’’). Early after administration of FC, there was a broadening at the peak of the [K+]o recording (Fig. 2). At higher doses of FC, or after more time had passed, there appeared to be two phases to the increase in [K+]o. An early fast phase, which appeared unchanged from baseline, was followed by a slower phase. The results with FC were confirmed with local injections of FA, which is not a barium salt. Injection of 2 Al of 10 mM FA (n = 6) significantly increased the peak change in [K+]o at 0.5, 1, and 2 h after injection (49.4 F 2.3 mM, 1 h after injection compared to 28.5 F 1.22 mM baseline, p < 0.01). Injection of 2 Al of 20 mM FA (n = 6) had no significant effect on the peak [K+]o reached during SD (25.6 F 1.5 mM, 1 h after injection compared to 28.2 F 2.3 baseline, p>0.05). The recovery rate was significantly decreased at 1 h (60.3 F 7.1% of the baseline recovery
Fig. 2. The DC and K+ recordings obtained before and after injection of FC or FA during high K+-induced cortical SD. Results from three different animals are presented. In (A), 2 h after administration of FC (0.05 mM), there was an increase in the width of the [K+]o trace and a decrease in the interval between SD events. In (B), the change in the shape of the potassium transient is illustrated at 1 and 2 h after injection of FC. In (C), the shape of the [K+]o curve before and at three time points after injection of 20 mM FA is shown. Calibrations are indicated on the figure and are the same in all panels.
rate, p < 0.01). Injections of FA were associated with the same changes in shape (Fig. 2C) that were described after injection of FC. There was broadening of the [K+]o tracing with the appearance of a longer latency peak. To further test that the residual barium was not the source of the changes after administration of FC, 100 AM barium chloride solution was injected into the cortex using identical methods. Local injection of BaCl2 had no effect on the interval between the SD episodes, the shape of the DC
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
181
Fig. 3. The effects of local injection of barium, glutamate, or TBOA on high K+-induced cortical SD. In each panel, the DC and K+ recordings obtained before and 1 h after injection of 0.1 mM barium chloride (A), 3 mM glutamate (B), or 2 mM TBOA, a glutamate uptake blocker (C), are presented. There was no change in the measured parameters of potassium regulation during SD after injection of any of these substances. Calibrations indicated on the figure are the same for all panels.
potential, or the [K+]o changes recorded during the SD events (n = 5; Fig. 3A). The changes recorded after administration of FC or FA could be due to an increase in extracellular glutamate concentration. To test this hypothesis, additional animals were injected with glutamate or a glutamate uptake blocker, TBOA. The peak [K+]o and the recovery rate of the [K+]o were not changed by local injections of 3 mM glutamate (Fig. 3B; n = 5, 99.5 F 9.5 of the baseline recovery rate, p>0.05) or 2 mM TBOA (Fig. 3C; n = 5, 102.3 F 10.0% of the baseline recovery rate, p>0.05). Addition of 2 mM TBOA had no significant effect on the ability of 0.05 mM FC to alter the rate of recovery of the [K+]o (n = 5). These results suggest that the effects of FC and FA on [K+]o are not due to alterations in glutamate levels or glutamate uptake. Measurements of [Ca2 +]o also demonstrated significant changes after injection of FC or FA. In the untreated cortex, or after injection of PBS, a decrease in [Ca2 +]o began before the
decrease in DC potential that signaled the spreading depression (Fig. 4). This decrease in [Ca2 +]o reached a peak and began to recover before the peak negativity of the DC potential. The [Ca2 +]o did not simply recover to the baseline levels, but actually peaked a second time above the baseline levels. The overshoot was almost a mirror image of the initial decrease in [Ca2 +]o, but somewhat smaller in peak amplitude. Membrane depolarization of astrocytes is proportional to the [K+]o and also contributes to calcium fluxes [13]. Therefore, the [Ca2 +]o experiments were done using the concentration of FC that did not significantly alter the peak [K+]o (i.e., 0.05 mM). After local injection of 0.05 mM FC, there was no significant change in the baseline levels of [Ca2 +]o or in the peak decrease in [Ca2 +]o that was measured in the early part of the SD wave. The peak was 0.69 F 0.13 mM before injection of FC and 0.8 F 0.11 mM 2 h after injection of 0.05 mM FC (n = 7). However, there was a dramatic change in the waveform after this early peak (Fig. 4). Before injection of
Fig. 4. The DC and Ca2 + recordings obtained before and after injection of FC (0.05 mM) during high K+-induced cortical SD. Changes in the [Ca2 +]o tracing were measurable within 30 min of the injection of FC. Changes were maximal between 1 and 2 h after injection. Calibrations are indicated on the figure.
182
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
FC, the peak increase in [Ca2 +]o during the later part of the SD wave was 0.37 F 0.25 mM. However, within minutes of injection of FC, this late positive peak began to get smaller. At 2 h after injection of 0.05 mM FC (the time of peak FC effect), there was a second negative peak in the [Ca2 +]o tracing of 1.15 F 0.31 mM (n = 7). So, what was initially a positive peak had become a negative peak. As might be expected, there was variability in the shape of the [Ca2 +]o between the two negative peaks now present. During most SD events, there was partial recovery before the second downward movement. These results were confirmed with local injection of 20 mM FA (n = 5). Histological examination of the brains from these animals indicated that the distribution and overall shape of the astrocytes were not changed by the local injection of FC or FA. One set of animals was examined 2 h after toxin injection (n = 3 for 0.02, 0.05, and 0.5 mM FC). Additional animals were examined 4 – 6 h after injection (n = 5 for 0.05 mM FC; n = 6 for 20 mM FA). The results from all of these animals were the same. Compared to the region around the vehicle injection (Fig. 5A), there was no increase in the thickness of the astrocytic processes, nor was there a loss of
Fig. 5. Effects of injection on the morphology of GFAP-positive cells in the cortex. Each panel presents a section of the cortex adjacent to the injection site from animals sacrificed 3 h after injection of phosphate-buffered saline (A) or FC (B) (0.05 mM). Each photograph was obtained at the same depth from the dura. The scale bar indicates 100 Am.
astrocytes in the region surrounding the injection of FC or FA (Fig. 5B).
4. Discussion The experiments presented here demonstrate that astrocytes participate in the regulation of [K+]o during spreading depression. Moderate doses of FC and FA, which will specifically inhibit glial function, slowed the rate of [K+]o recovery from elevated levels. Together, these data indicate that loss (or reduction) of glial cell function results in a modest, but significant, decrease in the rate of clearing of potassium from the extracellular space. Slowing of the recovery phase could be due to restricted diffusion due to cell swelling, but this is not consistent with the histology for either FC/FA, or a change in the recovery from the swelling that occurs during spreading depression [2]. The present results are comparable to the changes in [K+]o regulation recorded during and after seizure activity in the hippocampus [29], where local administration of FC/FA resulted in a slowing of the half-time of recovery of [K+]o. The results are also consistent with the conclusions of Largo et al. [16], where FC was dialyzed into the CA1 region of the hippocampus. The lack of effect of locally applied barium and the relatively small effects of the glial toxins suggest that the predominant uptake after spreading depression is through Na+ –K+ ATPase activity. This is supported by the fact that spreading depression is associated with a decrease in extracellular sodium levels, which will stimulate sodium pump activity [1]. At the doses that produced a slowing of the recovery of the [K+]o, there was also a consistent change in the overall shape of the tracings of the [K+]o that is harder to interpret. The increase in the width of the curve is most probably due to the slower recovery rate. There does not appear to be a slowing of the initial rise in [K+]o, suggesting that potassium movement out of neurons into the extracellular space has not changed. After administration of FC or FA, there was the additional appearance of a second peak with a slower rise time, which also contributed to the broadening of the [K+]o tracing. This change in the shape of the DC potential during spreading depression was also observed during dialysis of FC into the hippocampus [17] or bath application of FA to hippocampal slices [18]. The physiological changes underlying this change in shape cannot be deduced from the present experiments, but it has been suggested that astrocytes have a role in the moderation or propagation of spreading depression [17,18,20]. A slower propagation could explain the broader waves with smaller amplitude. The alterations in the peak [K+]o are more difficult to interpret, especially since the effects were dose-dependent. Changes in the peak [K+]o could be due to changes in the baseline [K+]o, but no baseline changes were recorded in the present experiments. The increase in peak [K+]o (with the low doses of FC/FA) could be due to reduced glial uptake, if
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184
that uptake was an active process, since FC/FA will inhibit synthesis of glutamine and ATP [3,7,22,26]. The decrease in peak [K+]o is more difficult to interpret, but may be due to the slower propagation of spreading depression in the presence of reduced glial function [17,18,20]. It is also possible that the vasodilatation associated with spreading depression (for review, see Ref. [19]) produces a transient increase in the removal of potassium from the extracellular space. In addition, it has recently been shown that astrocytes play a role in the local regulation of cerebral blood flow [32]. It is not yet known whether a reduction in glial function would alter local blood flow during spreading depression. The changes in [Ca2 +]o during spreading depression suggest that there are at least two competing processes ongoing under normal conditions. The early movement of calcium into cells, either neurons, glia, or both, reduces the [Ca2 +]o. This appears to begin before a measurable change in DC potential, suggesting that the calcium influx may be the trigger for the SD event. That this early decrease in calcium is not affected by local injection of FC/FA suggests that it does not require ATP generated by glial cells. Since the [Ca2 +]o is much higher in the extracellular space than inside of both neurons and glial cells, this movement of calcium out of the extracellular space is most likely secondary to the opening of membrane calcium channels. The [Ca2 +]o recovery consists of an overshoot, suggesting an active extrusion of calcium from cells in an amount greater than that which had moved into the cells. This suggests that calcium is being released from internal stores and actively pumped out of cells—either neurons or glia. Glial cells are known to have significant internal stores of calcium, which can be released in response to decreases in extracellular calcium levels [24,31]. This, combined with the fact that FC/FA will block the synthesis of ATP, thus reducing any active pumping, makes it most likely that the astrocytes are responsible for the recovery of the [Ca2 +]o and overshoot during the time frame of the SD episode. That the [Ca2 +]o levels eventually recover back to baseline indicates that either passive movement of calcium is sufficient to restore proper levels, that neurons can extrude calcium sufficiently to restore levels, or that there is some residual pumping ability within the glial cells that contributes to the restoration of [Ca2 +]o levels—albeit more slowly than normal. A number of factors suggest that the doses of FC/FA used in these experiments are not toxic to neurons. First, concentrations of 0.005 – 0.1 mM FC, or less, have been reported to act as a selective glial toxin in neuronal and glial cell cultures [11]. Injection of 1 nmol of FC produces selective and reversible glial cell damage in the striatum, without any ultrastructural evidence of neuronal damage [22]. Therefore, it has been suggested that this dose of FC (1 nmol) produces a selective and temporary impairment of glial cell energy metabolism and could be used as a model for the study of the importance of glial cells for brain function in vivo [22]. The doses used in the present study,
183
which were 0.02– 0.08 nmol of FC, are clearly within the glial-selective range. Secondly, in SD models, a lengthening of the time between SD events has been shown to correlate with neuronal damage [16]. In this study, the doses/concentration of FC/FA that were used did not increase the interval between SD events, suggesting that they were not causing neuronal damage. Therefore, the effects reported here are most likely due to selective changes in glial function. In summary, this study demonstrates that inhibition of glial function consistently slowed the recovery of the [K+]o and altered the active recovery of calcium during spreading depression. These data suggest that astrocytes normally participate in the regulation of extracellular potassium and calcium concentrations in the cortex.
Acknowledgements This work was supported by a grant to J.L.S. from the NINDS (NS39941).
References [1] T. Amedee, A. Robert, J.A. Coles, Potassium homeostasis and glial energy metabolism, Glia 21 (1997) 46 – 55. [2] M. Anderova, S. Kubinova, T. Mazel, A. Chvatal, C. Eliasson, M. Pekny, E. Sykova, Effect of elevated K+, hypotonic stress, and cortical spreading depression on astrocyte swelling in GFAP-deficient mice, Glia 35 (2001) 189 – 203. [3] A.M. Benjamin, Z.H. Verjee, Control of aerobic glycolysis in the brain in vitro, Neurochem. Res. 5 (1980) 921 – 934. [4] J. Berg-Johnsen, R.E. Paulsen, F. Fonnum, I.A. Langmoen, Changes in evoked potentials and amino acid content during fluorocitrate action studied in rat hippocampal slices, Exp. Brain Res. 96 (1993) 241 – 246. [5] D.D. Clarke, W.J. Nicklas, S. Berl, Tricarboxylic acid-cycle metabolism in brain. Effect of fluoroacetate and fluorocitrate on the labeling of glutamate, aspartate, glutamine and g-aminobutyrate, Biochem. J. 120 (1970) 345 – 351. [6] R.J. Do Carmo, H. Martins-Ferreira, Spreading depression of Lea˜o probed with ion-selective microelectrodes in isolated chick retina, An. Acad. Bras. Cienc. 56 (1984) 401 – 421. [7] F. Fonnum, A. Johnsen, B. Hassel, Use of fluorocitrate and fluoroacetate in study of brain metabolism, Glia 21 (1997) 106 – 113. [8] A. Furuta, J.D. Rothstein, J.L. Martin, Glutamate transporter protein subtypes are expressed differentially during rat central nervous system development, J. Neurosci. 17 (1997) 8363 – 8375. [9] A.J. Hansen, T. Zeuthen, Extracellular ion concentrations during spreading depression and ischemia in the rat brain cortex, Acta Physiol. Scand. 113 (1981) 437 – 445. [10] B. Hassel, R.E. Paulson, A. Johnsen, F. Fonnum, Selective inhibition of glial cell metabolism in vivo by fluorocitrate, Brain Res. 576 (1992) 120 – 124. [11] B. Hassel, N. Westergaard, A. Schousboe, F. Fonnum, Metabolic differences between primary cultures of astrocytes and neurons from cerebellum and cerebral cortex: effects of fluorocitrate, Neurochem. Res. 20 (1995) 413 – 420. [12] R. Jauch, O. Windmuller, T.-N. Lehmann, U. Heinemann, S. Gabriel, Effects of barium, furosemide, ouabaine and 4,4V-diisothiocyanatostilbene-2,2V-disulfonic acid (DIDS) on iontophoretically-induced changes in extracellular potassium concentration in hippocampal slices
184
[13] [14] [15]
[16]
[17]
[18]
[19] [20]
[21]
[22]
X.-Y. Lian, J.L. Stringer / Brain Research 1012 (2004) 177–184 from rats and from patients with epilepsy, Brain Res. 925 (2002) 18 – 27. H. Kettenmann, Physiology of glial cells, Adv. Neurol. 79 (1999) 565 – 571. R.P. Kraig, C. Nicholson, Extracellular ionic variations during spreading depression, Neuroscience 3 (1978) 1045 – 1059. K. Krnjevic, M.E. Morris, R.J. Reiffenstein, Stimulation-evoked changes in extracellular K+ and Ca2 + in pyramidal layers of the rat’s hippocampus, Can. J. Physiol. Pharmacol. 60 (1982) 1643 – 1657. C. Largo, P. Cuevas, G.G. Somjen, R. Martin Del Rio, O. Herreras, The effect of depressing glial function on rat brain in situ on homeostasis, synaptic transmission and neuronal survival, J. Neurosci. 16 (1996) 1219 – 1229. C. Largo, J.M. Ibrarz, O. Herreras, Effects of the gliotoxin fluorocitrate on spreading depression and glial membrane potential in rat brain in situ, J. Neurophysiol. 78 (1997) 295 – 307. C. Largo, G.C. Tombaugh, P.G. Aitken, O. Herreras, G.G. Somjen, Heptanol but not fluoroacetate prevents the propagation of spreading depression in rat hippocampal slices, J. Neurophysiol. 77 (1997) 9 – 16. M. Lauritzen, Pathophysiology of the migraine aura. The spreading depression theory, Brain 117 (1994) 199 – 210. M. Muller, G.G. Somjen, Intrinsic optical signals in rat hippocampal slices during hypoxia-induced spreading depression-like depolarization, J. Neurophysiol. 82 (1999) 1818 – 1831. C. Nicholson, Ion-selective microelectrodes and diffusion measurement as tools to explore the brain cell microenvironment, J. Neurosci. Methods 48 (1993) 199 – 213. R.E. Paulsen, A. Contestablie, L. Villani, F. Fonnum, An in vivo model for studying function of brain tissue temporarily devoid of glial cell metabolism: the use of fluorocitrate, J. Neurochem. 48 (1987) 1377 – 1385.
[23] G.G. Somjen, Mechanisms of spreading depression and hypoxic spreading depression-like depolarization, Physiol. Rev. 81 (2001) 1065 – 1096. [24] C. Stout, A. Charles, Modulation of intercellular calcium signaling in astrocytes by extracellular calcium and magnesium, Glia 43 (2003) 265 – 273. [25] J.L. Stringer, E.W. Lothman, Model of spontaneous hippocampal epilepsy in the anesthetized rat: electrographic, [K+]o, and [Ca2 +]o response patterns, Epilepsy Res. 4 (1989) 177 – 186. [26] R.A. Swanson, S.H. Graham, Fluorocitrate and fluoroacetate effects on astrocyte metabolism in vitro, Brain Res. 664 (1994) 94 – 100. [27] J.C. Szerb, Glutamate release and spreading depression in the fascia dentata in response to microdialysis with high K+: role of glia, Brain Res. 542 (1991) 259 – 265. [28] F. Vyskocil, N. Kriz, J. Bures, Potassium-selective microelectrodes used for measuring the extracellular brain potassium during spreading depression and anoxic depolarization in rats, Brain Res. 39 (1972) 255 – 259. [29] Z.-Q. Xiong, J.L. Stringer, Astrocytic regulation of the recovery of extracellular potassium after seizures in vivo, Eur. J. Neurosci. 11 (1999) 1677 – 1684. [30] Z.-Q. Xiong, J.L. Stringer, Sodium pump activity, not glial spatial buffering, clears potassium after epileptiform activity induced with 0-added calcium and 8 mM potassium in the dentate gyrus in vitro, J. Neurophysiol. 83 (2000) 1443 – 1451. [31] S. Zanotti, A. Charles, Extracellular calcium sensing by glial cells: low extracellular calcium induces intracellular calcium release and intercellular signaling, J. Neurochem. 69 (1997) 594 – 602. [32] M. Zonta, M.C. Angulo, S. Gobbo, B. Rosengarten, K.-A. Hossmann, T. Pozzan, G. Carmignoto, Neuron-to-astrocyte signaling is central to the dynamic control of brain microcirculation, Nat. Neurosci. 6 (2003) 42 – 50.