Brain Research, 542 (1991) 259-265 Elsevier
259
BRES 16363
Glutamate release and spreading depression in the fascia dentata in response to microdialysis with high K÷: role of glia John C. Szerb Department of Physiology and Biophysics, Dalhousie University, Halifax, N.S. (Canada) (Accepted 18 September 1990)
Key words: Glutamate release; Ghitamine release; Microdialysis; Spreading depression; High potassium concentration; Field excitatory postsynaptic potential; Fluoroacetate; Glia
To see electrophysiologlcal and neuroehemieal events during microdialysiswith high [K+], direct current (DC) and excitatory postsynaptic field potentials (fEPSPs) due to perforant path stimulation were recorded in the granule cell layer of the fascia dentata, while 3, 25, 50 or 100 mM KCI was perfused through a microdialysis probe placed 1.5 mm from the recording electrode. Glutamate and glutamine content of the dialysate was measured by high performance liquid chromatography. Raising [K+] from 3 to 25 mM reduced the effiux of glutamine, without affecting that of glutamate or the electrical activity. In about 50% of experiments, 50 mM K + induced large (20-30 mV) negative waves of spreading depression (SD), and a suppression of fEPSPs. In the other 50%, without SD, fEPSPs did not change. Glutamate eftlux increased 3-fold in both groups. SD waves were produced in all experiments with 100 mM K+ which evoked a more than 10-fold increase in glutamate release. Glutamine effiux decreased equally, by about 50%, with the 3 concentrations of K+. Mierodialysis with 20 mM fluoroacetate, a glial metabolic poison, decreased the spontaneous effiux of glutamine and glutamate and increased the incidence of SD waves. Results suggest that perfusion of 50 or 100 mM K+ through a microdialysis probe causes spreading depression which blocks surrounding electrical activity. The activity of glia partly protects against spreading depression caused by high [K+].
INTRODUCTION Microdialysis is widely used to measure the concentration of small, diffusible molecules, such as transmitters in brain extracellular spaces 4,5. In addition to observing the spontaneous release of transmitters, high potassium concentrations are often introduced through the microdialysis probe to demonstrate transmitter release evoked by depolarization 19,~,32,35,37. However, elevated potassium is also known to cause an increase, then a suppression of neuronal excitability 8,13. Furthermore, local application of high concentrations of potassium results in a large negative shift in extracellular direct current (DC) potentials, accompanied by complete electrical silence 12,14,~8,21. This phenomenon can travel several millimeters and is known as spreading depression (SD). The objective of these experiments was to see what changes in electrical activity accompany the perfusion of the microdialysis probe with different concentrations of potassium ions and to relate these changes with the release of glutamate and glutarnine. The fascia dentata of the hippocampal formation was chosen for this purpose, because it receives glutamatergic input through the
perforant path 6'3~ whose stimulation evokes large field potentials in this structure. In addition, the role of astroglia in the electrophysiological and neurochemical effects of high [K +] was investigated by the perfusion of an inhibitor of glial tricarboxylic acid cycle, fluoroacetate. Like acetate, fluoroacetate is taken up preferentially both in vitro and in vivo by astroglia 11'22 where fluoroacetate is condensed with oxaloacetate to form fluorocitrate which blocks the tricarboxylic acid cycle27. Since the metabolism of glia represents only a small fraction of the total, fluoroacetate does not significantly reduce overall 0 2 consumption in brain slices. However, it inhibits glutamine synthesis 2°, which takes place exclusively in the small glial compartment 24. In agreement with these biochemical findings, intrastriatal injection of 1 nmole fluorocitrate causes reversible swelling of astrocytes, without affecting neurons. However, the toxic effect of 2 nmole fluorocitrate extends to neurons 25. In the present experiments the glial inhibitory action of fluoroacetate, which, unlike fluorocitrate, does not chelate Ca 2+, was used to estimate the role of astroglia in the response of the fascia dentata to microdialysis with elevated [K+].
Correspondence: J.C. Szerb, Department of Physiology and Biophysics, Dalhousie University, Halifax, N.S. B3H 4H7, Canada. 0006-8993/91/$03.50 © 1991 Elsevier Science Publishers B.V. (Biomedical Division)
260 MATERIALS AND METHODS
Male Sprague-Dawley rats, weighing 300-400 g, anesthetized with 1.5 g/kg urethane i.p., were placed in a stereotaxic apparatus. A concentric bipolar stimulating electrode was lowered into the angular bundle (coordinates relative to bregma: 8.0 mm posterior, 4.0 mm lateral and 5.0 mm ventral). Then a glass electrode filled with saline (resistance 2 MI2) was introduced into the fascia dentata (3.9 mm posterior, 2.5 mm lateral relative to the bregma) through a hole in the holder of the microdialysis probe. During the insertion of the recording electrode the angular bundle was stimulated at 0.5 Hz with pulses of 0.08 ms, 0.8-1.5 mA and the depth of the maximal positivity of field excitatory postsynaptic potentials (fEPSPs), corresponding to the granule cell layer 6, was established. Then, while leaving the recording electrode in place, a 3.0-mm-long Carnegie Medicin microdialysis probe (o.d. 0.5 mm) was inserted so that it was located 1.5 mm rostromedially from the recording electrode with its tip 1.5 mm below the tip of the glass electrode. Potentials between the glass electrode and a glass tube filled with agar-saline placed in the neck muscles were recorded through chlorided silver wires. After DC amplification, the signal was divided between the two amplifiers of a double-beam Tektronix 5111A oscilloscope. One amplifier was set at DC amplification, the other at a 0.1 Hz time constant. The rear-end output of the DC amplifier was used to display DC potentials on a chart recorder, while for measuring evoked potentials, the output from the other amplifier was led to a LabMaster A/D board attached to a personal computer. By means of a semiautomated program 2 the evoked potentials were digitized at 20 kHz. Potentials from 12 stimuli delivered at 0.1 Hz were averaged every 5 min and the averages
were saved on floppy disks. The amplitude of the fEPSPs was compared to a + 10 mV, 1 ms calibration pulse (Fig. 2) applied with every stimulus from a Grass square-wave calibrator. The microdialysis probe was perfused at a rate of 3.4/ai/min with a filtered artificial CSF solution of the following composition (in mM): NaCI, 122; KCI, 3.0; CaCI2, 1.3; NaHCO3, 25; MgSO 4, 1.2; KH2PO4, 0.4; glucose 5. When the concentration of KCI was raised or when 20 mM sodium fluoroacetate was added to this solution, the NaC1 content was correspondingly reduced. Perfusion solutions were switched by means of a Carnegie Medicin liquid switch. The amino acid content of the perfusates collected over 15 min was measured by an automated high-performance liquid chromatography (HPLC) system. Every 50 min the samples were mixed for 15 s with 100/al o-phthalaldehyde reagent 33 and then loaded into a 50-/al loop. Reversed-phase chromatography was performed on 4.6 x 25 cm, 5/am microsphere C-18 columns with two buffers: for the first 20 min buffer A (0.05 M Na phosphate/acetate pH 7.5 buffer containing 5% methanol, 7% acetonitrile and 2% tetrahydrofuran) was run, followed by buffer B (same composition as buffer A but containing 55% methanol) for 25 min, then water for 1.5 min before returning to buffer A. a-Amino adipic acid was the internal standard. In every experiment the dialysis probe, after being placed into the fascia dentata, was perfused for 2 h before the collection of samples and recording of DC and evoked potentials was begun. The angular bundle was stimulated at 0.1 Hz throughout the experiment. Following the collection of three 15-min samples (9 averaged EPSPs), the KCI content of the perfusion fluid was raised during two 30-min periods according to one of the following two schedules: schedule A, 25 mM during the first period, 50 mM during the second; schedule B, 50 mM during first period, 100 mM during the
A
,°myI ÷| 15rain *
j 100 m M K
50mMK +
B
+
I
1 miJl
+I 10mV
Fig. 1. A: DC potential tracing obtained from an entire experiment in which 50 and 100 mM KCI were perfused through the microdialysis probe in the absence of fluoroacetate. B: faster tracing obtained from a different experiment showing spreading depression waves during the application of 100 mM K + in the absence of fluoroacetate. Negativity upward.
261
A 3raM K
+
+
B
/L
50raM K +
3raM K +
+
Fig. 2. Effect of perfusion of microdialysis probes with 50 mM K+ on fEPSPs evoked by angular-bundle stimulation recorded 1.5 mm from the probe. No fluoroacetate present. Upper tracings before, middle tracings during, bottom tracings 25 rain after the perfusion of 50 mM K+. A: effect of 50 mM K+ during a spreading depression wave. B: effect of 50 mM K+ when no spreading depression appeared. Averages of 12 potentials shown with calibration signal of +10 mV, 1 ms. Positivity upward.
second. The two periods of perfusion with raised KCI were separated by 45 min perfusion with 3 mM KCI. Similarly, the second period of high KCI was followed by 45 min perfusion with 3 mM KCI. The adoption of two different schedules of high KC1 application allowed the testing of the effects of 3 concentrations of KCI without unduly prolonging the experiments. Otherwise, results obtained with a third period of elevated KCI could have been of doubtful value, due to the deterioration of the preparation. Results obtained with the 3 concentrations of elevated [K+] will be presented separately. In experiments with fluoroacetate, perfusion with 20 mM Na fluoroacetate was begun at the time of the insertion of the probe into the fascia dentata. Therefore, in these experiments there was a 2-h exposure of the fascia dentata to fluoroacetate before the collection of the samples started. With fluoroacetate present, only schedule A (25 and 50 mM KCi) was followed.
RESULTS
D C and e v o k e d potentials Insertion of the microdialysis probe produced a transient large negative D C shift and the disappearance of the evoked potentials. Benveniste et ai. 5 have previously described a transient spreading depression produced by the introduction of a microdialysis probe. After 3 - 5 min, the D C potential returned to its original level and the
evoked potentials started to recover and progressively became even larger than those recorded before the insertion of the probe. During perfusion of 3 m M K +, low frequency small, 1-2 mV, oscillation of the D C potential could often be observed, without any apparent relationship to respiration or the stimulation of the perforant path (Fig. 1A).
TABLE I
Number of spreading depression waves during 30-rain microdialysis with different KCI concentrations, including the lO-min period following the return to 3 mM KCI Values are the mean + S.E.M. The number of experiments is given in parentheses.
[K +] (raM)
Control
Fluoroacetate
25 50 100
0 (6) 1.6+0.6(11) 6.2 _+0.9 (6)**
1.0 _+0.3 (6) 5.7+0.3(6)* -
* Highly significantly different from 25 mM K+ with fluoroacetate and from 50 mM K+ without fluoroacetate. * * Highly significantly different from 50 mM K+ without fluoroacetate.
262
50mMK + ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
25 m M K
0.1). However, the a m o u n t of glutamine in the dialysate decreased significantly during perfusion with 25
263 spontaneous SD waves in 5 out of 6 experiments, but there was no obvious increase in their frequency during the application of 25 mM K ÷ (Fig. 3).
100 m M K +
2420UJ ~
16-
:;E 0
840"-" 100-
~'~ s0.-.-
20"
1510-
5-
0
(10)
(5) (7)
0
30
60
TIME
( min
90
)
Fig. 5. Effect of 100 m M K + on the release of glutamate and
glutamine, on fEPSP amplitude, and on the incidence of SD waves in the absence of fluoroacetate. Symbols as in Fig. 3, but scale for glutamate release reduced by two, as compared to Figs. 3 and 4. n --6.
mM K ÷ (Fig. 3). This concentration of K + did not affect the amplitude of fEPSPs, nor did it induce any SD (Fig. 3). Inclusion of 20 mM fluoroacetate into the perfusion fluid had several effects: it highly significantly reduced the spontaneous release of glutamate and glutamine but also increased glutamate release evoked by 25 mM KCI, to become highly significant (Fig. 3). In addition, fluoroacetate progressively depressed the amplitude of the fEPSPs, even before perfusion with 25 mM K ÷, so that depression of fEPSPs by 25 mM K ÷ was not evident (Fig. 3). Furthermore, fluoroacetate induced occasional
Effect o f 50 m M K + Out of the 11 experiments with 50 mM K ÷, SD waves appeared in 6 during perfusion with 50 mM K + (Fig. 4), while in the other 5 there was no SD. Results from these two groups are shown separately. There was no difference between the two groups in the resting and evoked release of glutamate, 50 mM K ÷ inducing significant increases in both groups (Fig. 4). The release of glutamine in the two groups was also similar. However, fEPSPs were decreased by about 75% when SD waves were present, while without SD, 50 mM K + did not depress fEPSPs (Figs. 2 and 4). Fluoroacetate again significantly decreased the spontaneous release of glutamate and glutamine, although it did not enhance the release of glutamate evoked by 50 mM K ÷, as compared to controls (Fig. 4). Fluoroacetate, however, profoundly affected electrical activity: in these experiments fEPSPs were smaller even before the application of 50 mM K +, since they were a continuation of those with 25 mM K +, in which fEPSPs gradually decreased (Fig. 3). Immediately upon the application of 50 mM K ÷, SD waves were regularly produced and these were subsequently repeated 4-9 times (Fig. 4). Concurrently, fEPSPs were almost completely suppressed, but returned, at least partly, when perfusion with 50 mM K ÷ was discontinued. Unlike with 25 mM K ÷, the incidence of SD was much greater during perfusion with 50 mM K + than before or after. Effect of 100 m M K + The release of glutamate was increased about 4 times more by 100 mM K ÷ than by 50 mM K ÷, although the depression in glutamine release was about equal (Figs. 4 and 5). Simultaneously, even in the absence of fluoroacetate, fEPSPs were severely depressed and large numbers of SD waves were induced. All these effects were reversible (Fig. 5). A comparison of the incidence of SD waves shows that their frequency was higher during perfusion of increasing concentrations of potassium (Table I). In addition, fluoroacetate made potassium more effective in producing SD: with fluoroacetate present, 50 mM K + induced about the same number of SD waves as did 100 mM K ÷ in the absence of fluoroacetate (Table I).
DISCUSSION Neurochemical observations An increase in glutamate collected by microdialysis
264 during perfusion with elevated potassium observed in this study has been reported by a number of authors 19'26'32'37. It is clear from this study that the extent of this increase depended on the concentration of potassium: while 25 mM K ÷ produced no significant enhancement, 50 mM increased glutamate efflux 3-fold and 100 mM more than 10-fold. In contrast, the concurrent decrease in glutamine efflux, also reported by the above authors, was seen in the present experiments even with 25 mM K ÷, when there was no significant increase in glutamate collected. Furthermore, glutamine efflux decreased equally, by about 50%, with the 3 different concentrations of potassium, in spite of large differences in glutamate release. This suggests that in vivo, a decrease in glutamine efflux is not simply a reflection of its increased utilization as a precursor for transmitter amino acids, but a yet undefined homeostatic mechanism, such as increased transport from plasma 1'7, may maintain nearly constant extracellular glutamine levels in spite of different rates of utilization. Present observations suggest, however, that enhanced glial synthesis of glutamine 24 is less likely to have this homeostatic function, because fluoroacetate did not make the drop in glutamine efflux produced by high K ÷ perfusion more pronounced, although it decreased the steady efflux of glutamine, as expected from its toxic effect on glia. Fluoroacetate, similarly to fluorocitrate 25'26, not only decreased the efflux of glutamine and glutamate, but also increased the release of glutamate evoked by 25 mM K ÷. One factor producing this increased evoked release, also seen in hippocampal slices stimulated electrically3°, is probably a reduced uptake of released glutamate by the metabolically poisoned astroglia. In the present experiments a subsequent release of glutamate evoked by 50 mM K ÷, was not enhanced by fluoroacetate, probably because an inadequate supply of glutamine reduced glutamate synthesis and release 31 which cancelled out a decrease in glutamate uptake into glia.
Electrophysiological observations In every experiment, perfusion of the microdialysis probe with 100 mM KCI produced repeated large, 20-30 mV negative SD waves and a virtual disappearance of fEPSPs evoked by stimulation of the angular bundle. In approximately half of the experiments, 50 mM K + had similar effects. Since in experiments with 50 mM K ÷, fEPSPs disappeared only when SD was present, the two effects are likely to share a common mechanism, namely a depolarization of the neuronal membrane due to a large increase in [K÷]o which invariably accompanies SD 9' 14,16,29. This correlation also suggests that the depression of fEPSPs was not due to the direct depolarizing effect of large amounts of K ÷ diffusing from the microdialysis
probe, but to the SD produced by these ions. This then resulted in a self-propagating increase in [K+]o TM which, in the present experiments, travelled a distance of at least 1.5 mm, from the probe to the recording electrode. While the suppression of fEPSPs depended on the occurrence of SD during perfusion with 50 mM K +, the release of glutamate was not any larger when SD was induced. This is somewhat surprising because the evoked release of glutamate was highly dependent on the concentration of KC1 perfused and during SD [K+]o is known to rise to 30-60 mM 34. A likely explanation for this unexpected finding is that, although more glutamate was released during SD waves than in their absence, the amount of glutamate collected did not increase accordingly, because the diffusion of released glutamate to the microdialysis probe was impeded by cellular swelling which accompanies SD 15'23'28. This possibility is suggested by recent findings 4'17 showing that in vivo the recovery of solutes by microdialysis is limited by their rate of diffusion in extracellular spaces. With fluoroacetate present, the incidence of SD increased: sometimes they occurred even while peffusing 3 mM KCI and were generated consistently and with high frequency during perfusion with 50 mM KCI. One of the factors contributing to this effect of fluoroacetate may be the reduced spatial buffering of K + by glia 1°'23'34. Normally, [K+]o diffusing out of the probe is taken up by the surrounding glia and released at a distance, thereby diluting [K+]o around the probe. Metabolically poisoned glia can no longer perform this function and K + diffusing from the probe increases [K+]o more than in the absence of fluoroacetate. This higher [K+]o may have contributed to the potentiation by fluoroacetate of glutamate release evoked by 25 mM K + and to the increased incidence of SD while peffusing 50 mM K +. Whether extracellular accumulation of K + is also involved in the generation of the occasional SD seen in the presence of 3 mM K + and fluoroacetate cannot be ascertained from these experiments. In addition to an increased incidence of SD, fEPSPs tended to decrease progressively during microdialysis with fluoroacetate. Several factors may have contributed to this rundown: [Na+]i probably accumulates in glia exposed to fluoroacetate which leads to their osmotic swelling, hence to a displacement of the recording electrode from its original position. Furthermore, with prolonged exposure to fluoroacetate its non-specific neuronal toxicity could contribute to a decline of fEPSPs. In summary, evidence presented here shows that perfusion of microdialysis probes with elevated concentrations of K +, similar to those commonly used to evoke transmitter release, induces repetitive waves of spreading depression which travel at least 1.5 mm from the probe.
265 These large depolarizing waves could conceivably have an u n i n t e n d e d effect on more remote structures.
REFERENCES 1 Abdul-Ghani, A.S., Marton, M. and Dobkin, J., Studies on the transport of glutamine in vivo between the brain and blood in the resting state and during afferent electrical stimulation, J. Neurochem., 31 (1978) 541-546. 2 Aitken, P.G., Kainic acid and penicillin: differential effects on excitatory and inhibitory interactions in the CA1 region of the hippocampal slice, Brain Research, 325 (1985) 261-269. 3 Balestrino, M. and Somjen, G.G., Concentration of carbon dioxide, interstitial pH and synaptic transmission in hippocampal formation of the rat, J. Physiol., 396 (1988) 247-266. 4 Benveniste, H., Brain microdialysis, J. Neurochem., 52 (1989) 1667-1679. 5 Benveniste, H., Hansen, A.J. and Ottosen, N.S., Determination of brain interstitial concentrations by microdialysis, J. Neurochem., 52 (1989) 1741-1750. 6 Bliss, T.V.P., Douglas, R.M., Errington, M.L. and Lynch, M.A., Correlation between long-term potentiation and release of endogenous amino acids from dentate gyrus of anesthetized rats, J. Physiol., 377 (1986) 391-408. 7 Eriksson, L.S., Law, D.H., Hagenfeldt, L. and Wahren, J., Nitrogen metabolism of the human brain, J. Neurochem., 41 (1983) 1324-1328. 8 Fan, P., O'Regan, P.A. and Szerb, J.C., Effect of low glucose concentration on synaptic transmission in the rat hippocampal slice, Brain Res. Bull., 21 (1988) 741-747. 9 Futamachi, K.J., Mutani, R. and Prince, D.A., Potassium activity in rabbit cortex, Brain Research, 75 (1974) 5-25. 10 Gardner-Medwin, A.R., Analysis of potassium dynamics in mammalian brain tissue, J. Physiol., 335 (1983) 393-426. 11 Gonda, O. and Quastel, J.H., Transport and metabolism of acetate in rat brain cortex in vitro, Biochem. J., 100 (1966) 83-94. 12 Grafstein, B., Mechanism of spreading cortical depression, J. Neurophysiol., 19 (1956) 154-171. 13 Hablitz, J.J. and Lundervold, A., Hippocampai excitability and changes in extraceUular potassium, Exp. Neurol., 71 (1981) 410-420. 14 Haglund, M.M. and Schwartzkroin, P.A., Role of Na-K pump potassium regulation and IPSPs in seizures and spreading depression in immature rabbit hippocampal slices, J. Neurophysiol., 63 (1990) 225-239. 15 Hansen, A.J. and Olsen, C.E., Brain extracellular space during spreading depression and ischemia, Acta Physiol. $cand., 108 (1980) 355-365. 16 Hansen, A.J. and Zeuthen, T., Extracellular ion concentrations during spreading depression and ischemia in rat brain cortex, Actu Physiol. Scand., 113 (1981) 437-445. 17 Hsiao, J.K., Ball, B.A., Morrison, P.E, Mefford, I.N. and Bungay, EM., Effects of different semipermeable membranes on in vitro and in vivo performance of microdialysis probes, J. Neurochem., 54 (1990) 1449-1452. 18 Kawasaki, K., Cz~h, G. and Somjen, G.G., Prolonged exposure to high potassium concentration results in irreversible loss of synaptic transmission in hippocampal tissue slices, Brain Research, 457 (1988) 322-329. 19 Korf, J.. and Venema, K., Amino acids in rat striatal dialysates: methodological aspects and changes after electroconvulsive shock, J. Neurochem., 45 (1985) 1341-1348. 20 Lahiri, S. and Quastel, J.H., Fluoroacetate and the metabolism
Acknowledgements. This work was supported by the Medical Research Council of Canada. The author wishes to thank Patrick O'Regan and Craig Seaboyer for their help in performing these experiments.
of ammonia in brain, Biochem. J., 89 (1963) 157-163. 21 Lefio, A.A.P., Spreading depression. In D.P. Purpura, J.K. Penry, D.B. Tower, D.M. Woodbury and R.D. Walter (Eds.), Experimental Models of Epilepsy -- A Manual for the Laboratory Worker, Raven, New York, 1972, pp. 173-196. 22 Muir, D., Berl, S. and Clarke, D.D., Acetate and fluoroacetate as possible metabolic markers for glial metabolism in vivo, Brain Research, 380 (1986) 336-340. 23 Nicholson, C., Regulation of the ion microenvironment and neuronal excitability. In H.H. Jasper and N.M. van Gelder (Eds.), Basic Mechanisms of Neuronal Hyperexcitability, Liss, New York, 1983, pp. 185-216. 24 Norenberg, M.D. and Martinez-Hernandez, A., Fine structural localization of glutamine synthetase in astrocytes in rat brain, Brain Research, 161 (1979) 303-310. 25 Paulsen, R.E., Contestabile, A., Villani, L. and Fonnum, F., An in vivo model for studying the function of brain tissue devoid of glial cell metabolism: the use of fluorocitrate, J. Neurochem., 48 (1987) 1377-1385. 26 Paulsen, R.E and Fonnum, E, Role of glial cells for basal and Ca2+-dependent K+-evoked release of transmitter amino acids investigated with microdialysis, J. Neurochem., 52 (1989) 18231829. 27 Peters, R.A., Mechanism of the toxicity of the active constituent of Dichapetalum cymosum and related compounds, Adv. Enzymol., 18 (1957) 113-159. 28 Phillips, J.M. and Nicholson, C., Anion permeability in spreading depression investigated with ion-sensitive microelectrodes, Brain Research, 173 (1979) 567-671. 29 Somjen, G.G., ExtraceUular potassium in the mammalian nervous system, Annu. Rev. Physiol., 41 (1979) 159-177. 30 Szerb, J.C. and lssekutz, B., Increase in the stimulation-induced overflow of glutamate by fluoroacetate, a selective inhibitor of the glial tricarboxylic acid cycle, Brain Research, 410 (1987) 116-120. 31 Szerb, J.C. and O'Regan, P.A., Effect of glutamine on glutamate release from hippocampal slices induced by high K÷ or by electrical stimulation: interaction with different Caz+ concentrations, J. Neurochem., 44 (1985) 1724-1731. 32 Tossman, U., Segovia, J. and Ungerstedt, U., Extracellular levels of amino acids in striatum and globus pallidus of 6-hydroxydopamine-lesioned rats measured with microdialysis, Acta Physiol. Scand., 127 (1986) 547-551. 33 TurneU, D.C. and Cooper, J.D.H., Rapid assay for amino acids in serum or urine by pre-column derivatization and reversedphase chromatography, Clin. Chem., 28 (1982) 527-531. 34 Walz, W. and Hertz, L., Functional interactions between neurons and astrocytes. II. Potassium homeostasis at the cellular level, Prog. Neurobiol., 20 (1983) 133-183. 35 Westerink, B.H.C. and de Vries, J.B., On the origin of extracellular GABA collected by brain microdialysis and assayed by a simplified on-line method, Naunyn-Schmiedeberg's Arch. Pharmacol., 339 (1989) 603-607. 36 White, W.F., Nadler, J.V., Hamberger, A., Cotman, C.W. and Cummins, J.T., Glutamate as transmitter of hippocampal perforant path, Nature, 270 (1977) 356-357. 37 Young, A.M.J. and Bradford, H.F., Excitatory amino acid transmitters in the corticostriate pathways: studies using intracerebral microdialysis in vivo, J. Neurochem., 47 (1986) 13991404.
259
BRES 16363
Glutamate release and spreading depression in the fascia dentata in response to microdialysis with high K÷: role of glia John C. Szerb Department of Physiology and Biophysics, Dalhousie University, Halifax, N.S. (Canada) (Accepted 18 September 1990)
Key words: Glutamate release; Ghitamine release; Microdialysis; Spreading depression; High potassium concentration; Field excitatory postsynaptic potential; Fluoroacetate; Glia
To see electrophysiologlcal and neuroehemieal events during microdialysiswith high [K+], direct current (DC) and excitatory postsynaptic field potentials (fEPSPs) due to perforant path stimulation were recorded in the granule cell layer of the fascia dentata, while 3, 25, 50 or 100 mM KCI was perfused through a microdialysis probe placed 1.5 mm from the recording electrode. Glutamate and glutamine content of the dialysate was measured by high performance liquid chromatography. Raising [K+] from 3 to 25 mM reduced the effiux of glutamine, without affecting that of glutamate or the electrical activity. In about 50% of experiments, 50 mM K + induced large (20-30 mV) negative waves of spreading depression (SD), and a suppression of fEPSPs. In the other 50%, without SD, fEPSPs did not change. Glutamate eftlux increased 3-fold in both groups. SD waves were produced in all experiments with 100 mM K+ which evoked a more than 10-fold increase in glutamate release. Glutamine effiux decreased equally, by about 50%, with the 3 concentrations of K+. Mierodialysis with 20 mM fluoroacetate, a glial metabolic poison, decreased the spontaneous effiux of glutamine and glutamate and increased the incidence of SD waves. Results suggest that perfusion of 50 or 100 mM K+ through a microdialysis probe causes spreading depression which blocks surrounding electrical activity. The activity of glia partly protects against spreading depression caused by high [K+].
INTRODUCTION Microdialysis is widely used to measure the concentration of small, diffusible molecules, such as transmitters in brain extracellular spaces 4,5. In addition to observing the spontaneous release of transmitters, high potassium concentrations are often introduced through the microdialysis probe to demonstrate transmitter release evoked by depolarization 19,~,32,35,37. However, elevated potassium is also known to cause an increase, then a suppression of neuronal excitability 8,13. Furthermore, local application of high concentrations of potassium results in a large negative shift in extracellular direct current (DC) potentials, accompanied by complete electrical silence 12,14,~8,21. This phenomenon can travel several millimeters and is known as spreading depression (SD). The objective of these experiments was to see what changes in electrical activity accompany the perfusion of the microdialysis probe with different concentrations of potassium ions and to relate these changes with the release of glutamate and glutarnine. The fascia dentata of the hippocampal formation was chosen for this purpose, because it receives glutamatergic input through the
perforant path 6'3~ whose stimulation evokes large field potentials in this structure. In addition, the role of astroglia in the electrophysiological and neurochemical effects of high [K +] was investigated by the perfusion of an inhibitor of glial tricarboxylic acid cycle, fluoroacetate. Like acetate, fluoroacetate is taken up preferentially both in vitro and in vivo by astroglia 11'22 where fluoroacetate is condensed with oxaloacetate to form fluorocitrate which blocks the tricarboxylic acid cycle27. Since the metabolism of glia represents only a small fraction of the total, fluoroacetate does not significantly reduce overall 0 2 consumption in brain slices. However, it inhibits glutamine synthesis 2°, which takes place exclusively in the small glial compartment 24. In agreement with these biochemical findings, intrastriatal injection of 1 nmole fluorocitrate causes reversible swelling of astrocytes, without affecting neurons. However, the toxic effect of 2 nmole fluorocitrate extends to neurons 25. In the present experiments the glial inhibitory action of fluoroacetate, which, unlike fluorocitrate, does not chelate Ca 2+, was used to estimate the role of astroglia in the response of the fascia dentata to microdialysis with elevated [K+].
Correspondence: J.C. Szerb, Department of Physiology and Biophysics, Dalhousie University, Halifax, N.S. B3H 4H7, Canada. 0006-8993/91/$03.50 © 1991 Elsevier Science Publishers B.V. (Biomedical Division)
260 MATERIALS AND METHODS
Male Sprague-Dawley rats, weighing 300-400 g, anesthetized with 1.5 g/kg urethane i.p., were placed in a stereotaxic apparatus. A concentric bipolar stimulating electrode was lowered into the angular bundle (coordinates relative to bregma: 8.0 mm posterior, 4.0 mm lateral and 5.0 mm ventral). Then a glass electrode filled with saline (resistance 2 MI2) was introduced into the fascia dentata (3.9 mm posterior, 2.5 mm lateral relative to the bregma) through a hole in the holder of the microdialysis probe. During the insertion of the recording electrode the angular bundle was stimulated at 0.5 Hz with pulses of 0.08 ms, 0.8-1.5 mA and the depth of the maximal positivity of field excitatory postsynaptic potentials (fEPSPs), corresponding to the granule cell layer 6, was established. Then, while leaving the recording electrode in place, a 3.0-mm-long Carnegie Medicin microdialysis probe (o.d. 0.5 mm) was inserted so that it was located 1.5 mm rostromedially from the recording electrode with its tip 1.5 mm below the tip of the glass electrode. Potentials between the glass electrode and a glass tube filled with agar-saline placed in the neck muscles were recorded through chlorided silver wires. After DC amplification, the signal was divided between the two amplifiers of a double-beam Tektronix 5111A oscilloscope. One amplifier was set at DC amplification, the other at a 0.1 Hz time constant. The rear-end output of the DC amplifier was used to display DC potentials on a chart recorder, while for measuring evoked potentials, the output from the other amplifier was led to a LabMaster A/D board attached to a personal computer. By means of a semiautomated program 2 the evoked potentials were digitized at 20 kHz. Potentials from 12 stimuli delivered at 0.1 Hz were averaged every 5 min and the averages
were saved on floppy disks. The amplitude of the fEPSPs was compared to a + 10 mV, 1 ms calibration pulse (Fig. 2) applied with every stimulus from a Grass square-wave calibrator. The microdialysis probe was perfused at a rate of 3.4/ai/min with a filtered artificial CSF solution of the following composition (in mM): NaCI, 122; KCI, 3.0; CaCI2, 1.3; NaHCO3, 25; MgSO 4, 1.2; KH2PO4, 0.4; glucose 5. When the concentration of KCI was raised or when 20 mM sodium fluoroacetate was added to this solution, the NaC1 content was correspondingly reduced. Perfusion solutions were switched by means of a Carnegie Medicin liquid switch. The amino acid content of the perfusates collected over 15 min was measured by an automated high-performance liquid chromatography (HPLC) system. Every 50 min the samples were mixed for 15 s with 100/al o-phthalaldehyde reagent 33 and then loaded into a 50-/al loop. Reversed-phase chromatography was performed on 4.6 x 25 cm, 5/am microsphere C-18 columns with two buffers: for the first 20 min buffer A (0.05 M Na phosphate/acetate pH 7.5 buffer containing 5% methanol, 7% acetonitrile and 2% tetrahydrofuran) was run, followed by buffer B (same composition as buffer A but containing 55% methanol) for 25 min, then water for 1.5 min before returning to buffer A. a-Amino adipic acid was the internal standard. In every experiment the dialysis probe, after being placed into the fascia dentata, was perfused for 2 h before the collection of samples and recording of DC and evoked potentials was begun. The angular bundle was stimulated at 0.1 Hz throughout the experiment. Following the collection of three 15-min samples (9 averaged EPSPs), the KCI content of the perfusion fluid was raised during two 30-min periods according to one of the following two schedules: schedule A, 25 mM during the first period, 50 mM during the second; schedule B, 50 mM during first period, 100 mM during the
A
,°myI ÷| 15rain *
j 100 m M K
50mMK +
B
+
I
1 miJl
+I 10mV
Fig. 1. A: DC potential tracing obtained from an entire experiment in which 50 and 100 mM KCI were perfused through the microdialysis probe in the absence of fluoroacetate. B: faster tracing obtained from a different experiment showing spreading depression waves during the application of 100 mM K + in the absence of fluoroacetate. Negativity upward.
261
A 3raM K
+
+
B
/L
50raM K +
3raM K +
+
Fig. 2. Effect of perfusion of microdialysis probes with 50 mM K+ on fEPSPs evoked by angular-bundle stimulation recorded 1.5 mm from the probe. No fluoroacetate present. Upper tracings before, middle tracings during, bottom tracings 25 rain after the perfusion of 50 mM K+. A: effect of 50 mM K+ during a spreading depression wave. B: effect of 50 mM K+ when no spreading depression appeared. Averages of 12 potentials shown with calibration signal of +10 mV, 1 ms. Positivity upward.
second. The two periods of perfusion with raised KCI were separated by 45 min perfusion with 3 mM KCI. Similarly, the second period of high KCI was followed by 45 min perfusion with 3 mM KCI. The adoption of two different schedules of high KC1 application allowed the testing of the effects of 3 concentrations of KCI without unduly prolonging the experiments. Otherwise, results obtained with a third period of elevated KCI could have been of doubtful value, due to the deterioration of the preparation. Results obtained with the 3 concentrations of elevated [K+] will be presented separately. In experiments with fluoroacetate, perfusion with 20 mM Na fluoroacetate was begun at the time of the insertion of the probe into the fascia dentata. Therefore, in these experiments there was a 2-h exposure of the fascia dentata to fluoroacetate before the collection of the samples started. With fluoroacetate present, only schedule A (25 and 50 mM KCi) was followed.
RESULTS
D C and e v o k e d potentials Insertion of the microdialysis probe produced a transient large negative D C shift and the disappearance of the evoked potentials. Benveniste et ai. 5 have previously described a transient spreading depression produced by the introduction of a microdialysis probe. After 3 - 5 min, the D C potential returned to its original level and the
evoked potentials started to recover and progressively became even larger than those recorded before the insertion of the probe. During perfusion of 3 m M K +, low frequency small, 1-2 mV, oscillation of the D C potential could often be observed, without any apparent relationship to respiration or the stimulation of the perforant path (Fig. 1A).
TABLE I
Number of spreading depression waves during 30-rain microdialysis with different KCI concentrations, including the lO-min period following the return to 3 mM KCI Values are the mean + S.E.M. The number of experiments is given in parentheses.
[K +] (raM)
Control
Fluoroacetate
25 50 100
0 (6) 1.6+0.6(11) 6.2 _+0.9 (6)**
1.0 _+0.3 (6) 5.7+0.3(6)* -
* Highly significantly different from 25 mM K+ with fluoroacetate and from 50 mM K+ without fluoroacetate. * * Highly significantly different from 50 mM K+ without fluoroacetate.
262
50mMK + ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
25 m M K
0.1). However, the a m o u n t of glutamine in the dialysate decreased significantly during perfusion with 25
263 spontaneous SD waves in 5 out of 6 experiments, but there was no obvious increase in their frequency during the application of 25 mM K ÷ (Fig. 3).
100 m M K +
2420UJ ~
16-
:;E 0
840"-" 100-
~'~ s0.-.-
20"
1510-
5-
0
(10)
(5) (7)
0
30
60
TIME
( min
90
)
Fig. 5. Effect of 100 m M K + on the release of glutamate and
glutamine, on fEPSP amplitude, and on the incidence of SD waves in the absence of fluoroacetate. Symbols as in Fig. 3, but scale for glutamate release reduced by two, as compared to Figs. 3 and 4. n --6.
mM K ÷ (Fig. 3). This concentration of K + did not affect the amplitude of fEPSPs, nor did it induce any SD (Fig. 3). Inclusion of 20 mM fluoroacetate into the perfusion fluid had several effects: it highly significantly reduced the spontaneous release of glutamate and glutamine but also increased glutamate release evoked by 25 mM KCI, to become highly significant (Fig. 3). In addition, fluoroacetate progressively depressed the amplitude of the fEPSPs, even before perfusion with 25 mM K ÷, so that depression of fEPSPs by 25 mM K ÷ was not evident (Fig. 3). Furthermore, fluoroacetate induced occasional
Effect o f 50 m M K + Out of the 11 experiments with 50 mM K ÷, SD waves appeared in 6 during perfusion with 50 mM K + (Fig. 4), while in the other 5 there was no SD. Results from these two groups are shown separately. There was no difference between the two groups in the resting and evoked release of glutamate, 50 mM K ÷ inducing significant increases in both groups (Fig. 4). The release of glutamine in the two groups was also similar. However, fEPSPs were decreased by about 75% when SD waves were present, while without SD, 50 mM K + did not depress fEPSPs (Figs. 2 and 4). Fluoroacetate again significantly decreased the spontaneous release of glutamate and glutamine, although it did not enhance the release of glutamate evoked by 50 mM K ÷, as compared to controls (Fig. 4). Fluoroacetate, however, profoundly affected electrical activity: in these experiments fEPSPs were smaller even before the application of 50 mM K +, since they were a continuation of those with 25 mM K +, in which fEPSPs gradually decreased (Fig. 3). Immediately upon the application of 50 mM K ÷, SD waves were regularly produced and these were subsequently repeated 4-9 times (Fig. 4). Concurrently, fEPSPs were almost completely suppressed, but returned, at least partly, when perfusion with 50 mM K ÷ was discontinued. Unlike with 25 mM K ÷, the incidence of SD was much greater during perfusion with 50 mM K + than before or after. Effect of 100 m M K + The release of glutamate was increased about 4 times more by 100 mM K ÷ than by 50 mM K ÷, although the depression in glutamine release was about equal (Figs. 4 and 5). Simultaneously, even in the absence of fluoroacetate, fEPSPs were severely depressed and large numbers of SD waves were induced. All these effects were reversible (Fig. 5). A comparison of the incidence of SD waves shows that their frequency was higher during perfusion of increasing concentrations of potassium (Table I). In addition, fluoroacetate made potassium more effective in producing SD: with fluoroacetate present, 50 mM K + induced about the same number of SD waves as did 100 mM K ÷ in the absence of fluoroacetate (Table I).
DISCUSSION Neurochemical observations An increase in glutamate collected by microdialysis
264 during perfusion with elevated potassium observed in this study has been reported by a number of authors 19'26'32'37. It is clear from this study that the extent of this increase depended on the concentration of potassium: while 25 mM K ÷ produced no significant enhancement, 50 mM increased glutamate efflux 3-fold and 100 mM more than 10-fold. In contrast, the concurrent decrease in glutamine efflux, also reported by the above authors, was seen in the present experiments even with 25 mM K ÷, when there was no significant increase in glutamate collected. Furthermore, glutamine efflux decreased equally, by about 50%, with the 3 different concentrations of potassium, in spite of large differences in glutamate release. This suggests that in vivo, a decrease in glutamine efflux is not simply a reflection of its increased utilization as a precursor for transmitter amino acids, but a yet undefined homeostatic mechanism, such as increased transport from plasma 1'7, may maintain nearly constant extracellular glutamine levels in spite of different rates of utilization. Present observations suggest, however, that enhanced glial synthesis of glutamine 24 is less likely to have this homeostatic function, because fluoroacetate did not make the drop in glutamine efflux produced by high K ÷ perfusion more pronounced, although it decreased the steady efflux of glutamine, as expected from its toxic effect on glia. Fluoroacetate, similarly to fluorocitrate 25'26, not only decreased the efflux of glutamine and glutamate, but also increased the release of glutamate evoked by 25 mM K ÷. One factor producing this increased evoked release, also seen in hippocampal slices stimulated electrically3°, is probably a reduced uptake of released glutamate by the metabolically poisoned astroglia. In the present experiments a subsequent release of glutamate evoked by 50 mM K ÷, was not enhanced by fluoroacetate, probably because an inadequate supply of glutamine reduced glutamate synthesis and release 31 which cancelled out a decrease in glutamate uptake into glia.
Electrophysiological observations In every experiment, perfusion of the microdialysis probe with 100 mM KCI produced repeated large, 20-30 mV negative SD waves and a virtual disappearance of fEPSPs evoked by stimulation of the angular bundle. In approximately half of the experiments, 50 mM K + had similar effects. Since in experiments with 50 mM K ÷, fEPSPs disappeared only when SD was present, the two effects are likely to share a common mechanism, namely a depolarization of the neuronal membrane due to a large increase in [K÷]o which invariably accompanies SD 9' 14,16,29. This correlation also suggests that the depression of fEPSPs was not due to the direct depolarizing effect of large amounts of K ÷ diffusing from the microdialysis
probe, but to the SD produced by these ions. This then resulted in a self-propagating increase in [K+]o TM which, in the present experiments, travelled a distance of at least 1.5 mm, from the probe to the recording electrode. While the suppression of fEPSPs depended on the occurrence of SD during perfusion with 50 mM K +, the release of glutamate was not any larger when SD was induced. This is somewhat surprising because the evoked release of glutamate was highly dependent on the concentration of KC1 perfused and during SD [K+]o is known to rise to 30-60 mM 34. A likely explanation for this unexpected finding is that, although more glutamate was released during SD waves than in their absence, the amount of glutamate collected did not increase accordingly, because the diffusion of released glutamate to the microdialysis probe was impeded by cellular swelling which accompanies SD 15'23'28. This possibility is suggested by recent findings 4'17 showing that in vivo the recovery of solutes by microdialysis is limited by their rate of diffusion in extracellular spaces. With fluoroacetate present, the incidence of SD increased: sometimes they occurred even while peffusing 3 mM KCI and were generated consistently and with high frequency during perfusion with 50 mM KCI. One of the factors contributing to this effect of fluoroacetate may be the reduced spatial buffering of K + by glia 1°'23'34. Normally, [K+]o diffusing out of the probe is taken up by the surrounding glia and released at a distance, thereby diluting [K+]o around the probe. Metabolically poisoned glia can no longer perform this function and K + diffusing from the probe increases [K+]o more than in the absence of fluoroacetate. This higher [K+]o may have contributed to the potentiation by fluoroacetate of glutamate release evoked by 25 mM K + and to the increased incidence of SD while peffusing 50 mM K +. Whether extracellular accumulation of K + is also involved in the generation of the occasional SD seen in the presence of 3 mM K + and fluoroacetate cannot be ascertained from these experiments. In addition to an increased incidence of SD, fEPSPs tended to decrease progressively during microdialysis with fluoroacetate. Several factors may have contributed to this rundown: [Na+]i probably accumulates in glia exposed to fluoroacetate which leads to their osmotic swelling, hence to a displacement of the recording electrode from its original position. Furthermore, with prolonged exposure to fluoroacetate its non-specific neuronal toxicity could contribute to a decline of fEPSPs. In summary, evidence presented here shows that perfusion of microdialysis probes with elevated concentrations of K +, similar to those commonly used to evoke transmitter release, induces repetitive waves of spreading depression which travel at least 1.5 mm from the probe.
265 These large depolarizing waves could conceivably have an u n i n t e n d e d effect on more remote structures.
REFERENCES 1 Abdul-Ghani, A.S., Marton, M. and Dobkin, J., Studies on the transport of glutamine in vivo between the brain and blood in the resting state and during afferent electrical stimulation, J. Neurochem., 31 (1978) 541-546. 2 Aitken, P.G., Kainic acid and penicillin: differential effects on excitatory and inhibitory interactions in the CA1 region of the hippocampal slice, Brain Research, 325 (1985) 261-269. 3 Balestrino, M. and Somjen, G.G., Concentration of carbon dioxide, interstitial pH and synaptic transmission in hippocampal formation of the rat, J. Physiol., 396 (1988) 247-266. 4 Benveniste, H., Brain microdialysis, J. Neurochem., 52 (1989) 1667-1679. 5 Benveniste, H., Hansen, A.J. and Ottosen, N.S., Determination of brain interstitial concentrations by microdialysis, J. Neurochem., 52 (1989) 1741-1750. 6 Bliss, T.V.P., Douglas, R.M., Errington, M.L. and Lynch, M.A., Correlation between long-term potentiation and release of endogenous amino acids from dentate gyrus of anesthetized rats, J. Physiol., 377 (1986) 391-408. 7 Eriksson, L.S., Law, D.H., Hagenfeldt, L. and Wahren, J., Nitrogen metabolism of the human brain, J. Neurochem., 41 (1983) 1324-1328. 8 Fan, P., O'Regan, P.A. and Szerb, J.C., Effect of low glucose concentration on synaptic transmission in the rat hippocampal slice, Brain Res. Bull., 21 (1988) 741-747. 9 Futamachi, K.J., Mutani, R. and Prince, D.A., Potassium activity in rabbit cortex, Brain Research, 75 (1974) 5-25. 10 Gardner-Medwin, A.R., Analysis of potassium dynamics in mammalian brain tissue, J. Physiol., 335 (1983) 393-426. 11 Gonda, O. and Quastel, J.H., Transport and metabolism of acetate in rat brain cortex in vitro, Biochem. J., 100 (1966) 83-94. 12 Grafstein, B., Mechanism of spreading cortical depression, J. Neurophysiol., 19 (1956) 154-171. 13 Hablitz, J.J. and Lundervold, A., Hippocampai excitability and changes in extraceUular potassium, Exp. Neurol., 71 (1981) 410-420. 14 Haglund, M.M. and Schwartzkroin, P.A., Role of Na-K pump potassium regulation and IPSPs in seizures and spreading depression in immature rabbit hippocampal slices, J. Neurophysiol., 63 (1990) 225-239. 15 Hansen, A.J. and Olsen, C.E., Brain extracellular space during spreading depression and ischemia, Acta Physiol. $cand., 108 (1980) 355-365. 16 Hansen, A.J. and Zeuthen, T., Extracellular ion concentrations during spreading depression and ischemia in rat brain cortex, Actu Physiol. Scand., 113 (1981) 437-445. 17 Hsiao, J.K., Ball, B.A., Morrison, P.E, Mefford, I.N. and Bungay, EM., Effects of different semipermeable membranes on in vitro and in vivo performance of microdialysis probes, J. Neurochem., 54 (1990) 1449-1452. 18 Kawasaki, K., Cz~h, G. and Somjen, G.G., Prolonged exposure to high potassium concentration results in irreversible loss of synaptic transmission in hippocampal tissue slices, Brain Research, 457 (1988) 322-329. 19 Korf, J.. and Venema, K., Amino acids in rat striatal dialysates: methodological aspects and changes after electroconvulsive shock, J. Neurochem., 45 (1985) 1341-1348. 20 Lahiri, S. and Quastel, J.H., Fluoroacetate and the metabolism
Acknowledgements. This work was supported by the Medical Research Council of Canada. The author wishes to thank Patrick O'Regan and Craig Seaboyer for their help in performing these experiments.
of ammonia in brain, Biochem. J., 89 (1963) 157-163. 21 Lefio, A.A.P., Spreading depression. In D.P. Purpura, J.K. Penry, D.B. Tower, D.M. Woodbury and R.D. Walter (Eds.), Experimental Models of Epilepsy -- A Manual for the Laboratory Worker, Raven, New York, 1972, pp. 173-196. 22 Muir, D., Berl, S. and Clarke, D.D., Acetate and fluoroacetate as possible metabolic markers for glial metabolism in vivo, Brain Research, 380 (1986) 336-340. 23 Nicholson, C., Regulation of the ion microenvironment and neuronal excitability. In H.H. Jasper and N.M. van Gelder (Eds.), Basic Mechanisms of Neuronal Hyperexcitability, Liss, New York, 1983, pp. 185-216. 24 Norenberg, M.D. and Martinez-Hernandez, A., Fine structural localization of glutamine synthetase in astrocytes in rat brain, Brain Research, 161 (1979) 303-310. 25 Paulsen, R.E., Contestabile, A., Villani, L. and Fonnum, F., An in vivo model for studying the function of brain tissue devoid of glial cell metabolism: the use of fluorocitrate, J. Neurochem., 48 (1987) 1377-1385. 26 Paulsen, R.E and Fonnum, E, Role of glial cells for basal and Ca2+-dependent K+-evoked release of transmitter amino acids investigated with microdialysis, J. Neurochem., 52 (1989) 18231829. 27 Peters, R.A., Mechanism of the toxicity of the active constituent of Dichapetalum cymosum and related compounds, Adv. Enzymol., 18 (1957) 113-159. 28 Phillips, J.M. and Nicholson, C., Anion permeability in spreading depression investigated with ion-sensitive microelectrodes, Brain Research, 173 (1979) 567-671. 29 Somjen, G.G., ExtraceUular potassium in the mammalian nervous system, Annu. Rev. Physiol., 41 (1979) 159-177. 30 Szerb, J.C. and lssekutz, B., Increase in the stimulation-induced overflow of glutamate by fluoroacetate, a selective inhibitor of the glial tricarboxylic acid cycle, Brain Research, 410 (1987) 116-120. 31 Szerb, J.C. and O'Regan, P.A., Effect of glutamine on glutamate release from hippocampal slices induced by high K÷ or by electrical stimulation: interaction with different Caz+ concentrations, J. Neurochem., 44 (1985) 1724-1731. 32 Tossman, U., Segovia, J. and Ungerstedt, U., Extracellular levels of amino acids in striatum and globus pallidus of 6-hydroxydopamine-lesioned rats measured with microdialysis, Acta Physiol. Scand., 127 (1986) 547-551. 33 TurneU, D.C. and Cooper, J.D.H., Rapid assay for amino acids in serum or urine by pre-column derivatization and reversedphase chromatography, Clin. Chem., 28 (1982) 527-531. 34 Walz, W. and Hertz, L., Functional interactions between neurons and astrocytes. II. Potassium homeostasis at the cellular level, Prog. Neurobiol., 20 (1983) 133-183. 35 Westerink, B.H.C. and de Vries, J.B., On the origin of extracellular GABA collected by brain microdialysis and assayed by a simplified on-line method, Naunyn-Schmiedeberg's Arch. Pharmacol., 339 (1989) 603-607. 36 White, W.F., Nadler, J.V., Hamberger, A., Cotman, C.W. and Cummins, J.T., Glutamate as transmitter of hippocampal perforant path, Nature, 270 (1977) 356-357. 37 Young, A.M.J. and Bradford, H.F., Excitatory amino acid transmitters in the corticostriate pathways: studies using intracerebral microdialysis in vivo, J. Neurochem., 47 (1986) 13991404.
