Introduction

Top Abstract Introduction Experimental Procedures Results and Discussion References

Excitotoxicity has been linked causally with glutamate-elicited Ca²⁺ influx (Hartley et al., 1993; Eimerl and Schramm, 1994) and Ca-dependent mitochondrial depolarization (Budd and Nicholls, 1996; Schinder et al., 1996; White and Reynolds, 1996; Stout et al., 1998). Activation of ionotropic glutamate receptors depolarizes the plasma membrane (PM) and opens several Ca-permeable pathways: ionic channels of glutamate receptors (Mayer and Westbrook, 1987), voltage-sensitive Ca²⁺ channels (Reichling and MacDermott, 1993), and the Ca²⁺ influx resulting from the operation of the plasma membrane Na⁺/Ca²⁺ exchanger in the reverse mode (Kiedrowski et al., 1994; Hoyt et al., 1998; Kiedrowski, 1999). In vitro experiments aimed at defining the pathway of the excitotoxic Ca²⁺ entry have shown that Ca²⁺ influx via N-methyl-D-aspartate (NMDA) receptors mediates glutamate excitotoxicity (Choi et al., 1988; Manev et al., 1989; Tymianski et al., 1993). Those experiments were conducted on cultured neurons incubated in media containing Na⁺ and K⁺ concentrations characteristic of neurons under resting conditions. It is well established, however, that within a few minutes of brain ischemia, extracellular concentrations of K⁺ ([K⁺]_E) reach 60 mM or higher, and extracellular Na⁺ concentrations ([Na⁺]_E) decrease to about 60 mM (Somjen, 1979; Hansen et al., 1980; Hansen, 1985; Erecinska and Silver, 1994). This implies that the Na⁺ and the K⁺ concentration gradients across the PM (Na/K gradient) are profoundly destabilized when the glutamate excitotoxicity in vivo is executed. Yet, how destabilization of the Na/K gradient affects the mechanisms of glutamate excitotoxicity has never been formally studied. One can suspect that the role of the Na/K gradient may be important in excitotoxicity because depolarization of the PM strongly affects the NMDA-induced Ca²⁺ influx (1999). Therefore, the present study was designed to determine how the elevation of [K⁺]_E combined with the decrease of [Na⁺]_E affect excitotoxicity.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results and Discussion References

Neuronal Cultures. Primary cultures of cerebellar granule cells (CGCs) were prepared from 8-day-old Sprague-Dawley rats and were plated, using basal Eagle's medium supplemented with 25 mM KCl, 10% bovine fetal serum, 2 mM glutamine, and 50 µg/ml gentamycin, as described in Kiedrowski (1999). Cultures at 8 to 11 days in vitro were used for the experiments.

Media. Experimental solutions were based on Locke's buffer that contained 154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO₃, 1.3 mM CaCl₂, 1 mM MgCl₂, 5 mM glucose, and 10 mM HEPES, pH 7.4, adjusted with Tris. The desired concentrations of K⁺ or Li⁺ in these experimental solutions were achieved by an equimolar substitution of Na⁺ with K⁺ or Li⁺. CGCs were exposed to these solutions at 37°C using a static bath. To apply a new solution, the previous solution was aspirated and the cells were washed with the new solution several times. [K⁺]_E was never decreased below 5.6 mM to prevent inhibition of Na⁺/K⁺ ATPase due to lack of extracellular K⁺. Glutamate (1 mM unless indicated otherwise) or NMDA (300 µM) was applied in Mg-free Locke's containing 10 µM glycine. Depolarizing pulses of 60 mM K⁺ were delivered in the presence of Mg²⁺ (1 mM) and MK-801 (10 µM) to prevent Ca²⁺ influx caused by activation of NMDA receptors by endogenous glutamate. Mitochondrial depolarization was carried out by application of Locke's buffer containing 10 µM carbonyl cyanide m-chlorophenylhydrazone (CCCP) and 3 µg/ml oligomycin (Budd and Nicholls, 1996).

Viability Test. After excitotoxic challenge the cells were returned to a conditioned basal Eagle's medium (CM), i.e., the medium (containing 10% bovine fetal serum and 25 mM K⁺), in which the cells were cultured. The CM was supplemented with 10 µM propidium iodide, and the viability of CGCs was quantified after approximately 24 h by detecting the propidium iodide fluorescence as described in Kiedrowski (1999) and illustrated in Fig. 1.

View larger version (128K): [in this window] [in a new window]   Fig. 1.   Glutamate is neurotoxic at 5.6 mM [K+]E but not at 60 mM [K+]E. CGCs were exposed to glutamate (1 mM glutamate plus 10 µM glycine and Mg-free medium) for 90 min at 5.6 mM [K+]E (A) or 60 mM [K+]E (B). Oblique illumination images of cells before the exposure (pre-GLU), 24 h after the exposure (post-GLU), the propidium iodide fluorescence image 24 h after the exposure (PI), and the overlay image that was obtained by subtracting digitally the image "PI" from the image "post-Glu" are shown. Scale bar, 30 µm. The combined data from three independent experiments are presented in Table 1.

Simultaneous Assay of Cytoplasmic [Ca²⁺] ([Ca²⁺]_C) and Plasma Membrane Potential (E_m). [Ca²⁺]_C and E_m were monitored in single CGCs loaded with 4 µM fura-2 acetoxymethyl ester and exposed to 100 nM bis(1,3-dibutylbarbituric acid)trimethine oxonol [DiBAC₄(3)], as recently described (Kiedrowski, 1999). Briefly, the fluorescences of fura-2 and DiBAC₄(3) were monitored at 37°C using the Attofluor digital imaging system (Atto Instruments, Rockville, MD), a Zeiss Axiovert 10 microscope, and a Zeiss Achrostigmat objective (40×, NA 1.30). The images of fluorescence emitted at over 520 nm after excitation at 334 nm (F₃₃₄) and 380 nm (F₃₈₀) for fura-2, and at 488 nm (F₄₈₈) for DiBAC₄(3), were saved every 10 to 20 s. The DiBAC₄(3) fluorescence measured in regions of interest in the peripheral parts of neuronal somata just underneath the PM was normalized and was used as a relative index of E_m. The fura-2 F₃₃₄/F₃₈₀ ratio measured in the central part of somata was used as an index of [Ca²⁺]_C and was calibrated in situ. It must be stressed, however, that changes in the fura-2 F₃₃₄/F₃₈₀ ratio do not always reflect [Ca²⁺]_C changes, i.e., when the fluorescent properties of fura-2 are affected by the NMDA-induced changes in cytoplasmic pH (Kiedrowski, 1999).

Assay of Cytoplasmic [Na⁺] ([Na⁺]_C). CGCs were loaded for 60 min at 37°C with 20 µM Na⁺-binding benzofuran isophthalate acetoxymethyl ester (SBFI AM) dissolved in CM. The stock concentration of SBFI AM was 5 mM in dimethyl sulfoxide. SBFI fluorescence was monitored at 37°C using the same excitation and emission settings as described above for fura-2. The F₃₃₄/F₃₈₀ ratio was calibrated for [Na⁺]_C in situ at the end of the experiments using five to six Na⁺ concentrations in a range from 0 to 163 mM. The buffers used to calibrate [Na⁺]_C contained 5 µM gramicidin D and were prepared by appropriate mixing of high-concentration solutions of Na⁺ described previously (Kiedrowski, 1999) with an analogous Na-free solution in which all Na⁺ was substituted with K⁺. [Na⁺]_C values were calculated using a nonlinear least-squares fit of the data to the Michaelis-Menten equation as described by Kasner and Ganz (1992).

⁴⁵Ca²⁺ Uptake. Uptake of ⁴⁵Ca²⁺ was measured as described previously (Kiedrowski, 1999).

Materials. Fura-2 AM, SBFI AM, and DiBAC₄(3) were obtained from Molecular Probes (Eugene, OR), MK-801 [(+)5-methyl-10,11-dihydro-5H-dibenzocyclohepten-5,10-imine] and CPP [R-3-(2-carboxypiperazin-4-yl)-propyl-1-phosphonic acid] were purchased from Research Biochemicals Inc. (Natick, MA) or Tocris (Ballwin, MO), and ⁴⁵Ca²⁺ was obtained from Amersham (Arlington Heights, IL). The culture media and all other chemicals were purchased from Sigma (St. Louis, MO).

	Results and Discussion

Top Abstract Introduction Experimental Procedures Results and Discussion References

High [K⁺]_E Inhibits Excitotoxicity. To test the impact of high [K⁺]_E on excitotoxicity, CGCs were exposed for 90 min at 37°C to glutamate (1 mM glutamate, 10 µM glycine, and Mg-free medium) or NMDA (300 µM NMDA, 10 µM glycine, and Mg-free medium) at 5.6 or 60 mM [K⁺]_E, and neuronal viability was assessed after 24 h (for details see Experimental Procedures). At 60 mM [K⁺]_E, neither glutamate nor NMDA was neurotoxic (Fig. 1 and Table 1); moreover the glutamate excitotoxicity induced at 5.6 mM [K⁺]_E was completely prevented if 10 µM MK-801 or 100 µM CPP, a noncompetitive or competitive inhibitor of NMDA receptors, respectively, was present in the medium during the exposure (data not shown). The data indicate that 60 mM [K⁺]_E prevents the excitotoxicity resulting from NMDA receptor activation.

High [K⁺]_E Inhibits NMDA-Induced Ca²⁺ Influx. When [K⁺]_E was increased from 5.6 to 60 mM, the cytoplasmic ⁴⁵Ca²⁺ accumulation elicited by glutamate or NMDA was inhibited in a dose-dependent manner (Fig. 2A). This inhibition of ⁴⁵Ca²⁺ accumulation by high [K⁺]_E could not be explained by improved Ca²⁺ extrusion, because the rate of ⁴⁵Ca²⁺ efflux from CGCs preloaded with ⁴⁵Ca²⁺ was the same regardless of [K⁺]_E (Fig. 2B). These data indicate that high [K⁺]_E inhibits the NMDA-induced Ca²⁺ influx.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   High [K+]E inhibits NMDA- or glutamate-elicited 45Ca2+ uptake (A) but does not affect 45Ca2+ extrusion (B). Presented are the means ± S.E.M. from three independent experiments. A, 45Ca2+ uptake was stimulated by 300 µM NMDA or 100 µM glutamate (GLU) applied in Mg-free Locke's buffer supplemented with 10 µM glycine. Basal 45Ca2+ uptake was measured in the presence of 10 µM MK-801 to inhibit the Ca2+ influx resulting from the activation of NMDA receptors by endogenous glutamate. B, CGCs were incubated for 5 min at 37°C with 20 µM glutamate applied with Mg-free Locke's buffer containing 45Ca2+ (1 µCi) and 10 µM glycine, then the cells were washed with a nonradioactive Locke's buffer containing either 5.6 or 60 mM K+, and the amounts of 45Ca2+ trapped in the cells at the indicated time intervals were determined.

High [K⁺]_E Prevents Mitochondrial Ca²⁺ Overload in Neurons Exposed to Glutamate. Glutamate excitotoxicity has been linked causally to Ca²⁺ influx (Hartley et al., 1993; Eimerl and Schramm, 1994), and the consequent Ca²⁺ influx-dependent mitochondrial depolarization (Budd and Nicholls, 1996; Khodorov et al., 1996; Schinder et al., 1996; White and Reynolds, 1996; Stout et al., 1998). To test whether high [K⁺]_E modifies mitochondrial buffering of Ca²⁺ in CGCs exposed to glutamate, how mitochondrial depolarization with 10 µM CCCP plus 3 µg/ml oligomycin (Budd and Nicholls, 1996) affects [Ca²⁺]_C was studied. CCCP plus oligomycin were expected to depolarize mitochondria, causing a release of the Ca²⁺ stored in the mitochondrial matrix, which can be detected as a rapid increase in [Ca²⁺]_C.

At High [K⁺]_E NMDA Induces Less Pronounced Depolarization of PM than at Low [K⁺]_E Depolarization of the PM curtails the NMDA-induced Ca²⁺ influx by affecting the electrochemical driving force for Ca²⁺ influx (CaDF), defined as the difference between the PM potential (E_m) and the Ca²⁺ equilibrium potential (E_Ca) (Kiedrowski, 1999). Therefore, one may envision that the mechanism of inhibition of the NMDA-induced Ca²⁺ influx by high [K⁺]_E involves a decrease in the CaDF due to PM depolarization. This explanation would only be valid, however, if the PM depolarization induced by NMDA at 60 mM [K⁺]_E were indeed greater than at 5.6 mM [K⁺]_E. To test whether this is the case, the effects of [K⁺]_E on E_m and [Ca²⁺]_C were studied by monitoring the fluorescences of DiBAC₄(3) and fura-2, respectively, in single CGCs exposed to NMDA.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   CGCs exposed to NMDA are more depolarized at low than at high [K+]E. [Ca2+]C and Em were simultaneously monitored in CGCs exposed to the indicated [K+]E. Initially CGCs were incubated with CM without DiBAC4(3), then CM was replaced with Locke's buffer containing 100 nM DiBAC4(3) (arrows) and 5.6 mM K+ (K5.6). DiBAC4(3) (100 nM) was present in the extracellular medium at all times thereafter. Data are the means ± S.E.M. from 15 to 24 neurons in representative experiments that were repeated at least three times with similar results. The S.E.M. are small and therefore in most cases are obscured by the circles representing the data points. A, control, CGCs exposed to the indicated increase in [K+]E alone; 60 mM [K+]E (K60) was applied together with 10 µM MK-801 to prevent the activation of NMDA receptors by endogenous glutamate. B and C, CGCs treated with the indicated [K+]E and exposed to NMDA.

At High [K⁺]_E NMDA-Induced Na⁺ Influx Is Greatly Decreased. Because the NMDA-induced depolarization of the PM is caused by Na⁺ influx (Hösli et al., 1973), it is possible that the counterintuitive effect of [K⁺]_E on E_m in CGCs exposed to NMDA, i.e., the greater depolarization of the PM at lower [K⁺]_E (Fig. 3B), might be the result of a greater Na⁺ influx. To test this hypothesis, the effects of [K⁺]_E on [Na⁺]_C in CGCs were studied under control conditions and following exposure to NMDA. As shown in Fig. 4A, under control conditions (no NMDA added), changes in [K⁺]_E had only minor effects on [Na⁺]_C: an increase of [K⁺]_E from 5.6 to 35.6 mM caused a transient increase in [Na⁺]_C from a basal level of about 2 to 3 mM to 8 ± 0.8 mM, followed by a drop and stabilization at 6 ± 0.5 mM. Upon a subsequent increase of [K⁺]_E to 65.6 mM, [Na⁺]_C dropped to 4 ± 0.4 mM. When the same CGCs were then exposed to NMDA at different [K⁺]_E, there was an inverse relationship between [K⁺]_E and [Na⁺]_C: at 65.6 mM [K⁺]_E, [Na⁺]_C increased to 26 ± 2.1 mM; when [K⁺]_E was decreased to 35.6 mM, [Na⁺]_C increased to 48 ± 4.3 mM; [Na⁺]_C increased further, to as much as 89 ± 12.4 mM, when [K⁺]_E was decreased to the 5.6 mM level (Fig. 4A). The decrease in [Na⁺]_C that was induced by application of high [K⁺]_E in CGCs exposed to NMDA was promptly reverted by inhibiting Na⁺/K⁺ ATPase with 1 mM ouabain (Fig. 4B). This indicates that Na⁺/K⁺ ATPase activity is involved in lowering [Na⁺]_C when [K⁺]_E is increased. The fact that at 5.6 mM [K⁺]_E, NMDA exposure excessively elevates [Na⁺]_C in CGCs (Fig. 4A) implies that under these conditions the NMDA-induced Na⁺ influx exceeds the rate of Na⁺ extrusion by Na⁺/K⁺ ATPase; indeed Na⁺/K⁺ ATPase is saturated when [Na⁺]_C exceeds 30 mM (Collins et al., 1992). Very likely, this enormous, NMDA-induced elevation in [Na⁺]_C in a neuronal culture is an experimental artifact caused by the great excess of extracellular over cytoplasmic volume; because of that, the NMDA-induced Na⁺ influx does not lead to any significant drop in [Na⁺]_E. By contrast, the extracellular space is only about 20% of brain volume (Nicholson and Sykova, 1998), and therefore the Na⁺ influx during brain ischemia is associated with a significant drop in [Na⁺]_E, a limiting factor for Na⁺ influx.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effects of [K+]E on the NMDA-induced changes in [Na+]C in CGCs. A, decreasing [K+]E of CGCs exposed to NMDA leads to an increase in [Na+]C. [Na+]C was monitored in CGCs that were exposed to the indicated [K+]E changes first in the absence and then in the presence of NMDA, as indicated. Data are the means ± S.E. from 25 neurons monitored in a representative experiment. B, inhibition of Na+/K+ ATPase with 1 mM ouabain counteracts the [K+]E-dependent decrease in [Na+]C in CGCs exposed to NMDA. Data are the means ± S.E. from 40 neurons monitored in a representative experiment. The time bar applies to A and B.

Can the Inhibitory Effect of High [K⁺]_E on NMDA-Induced Ca²⁺ Influx Be Related to Plasma Membrane Na⁺/Ca²⁺ Exchange Operation? Because an increase in [K⁺]_E decreases [Na⁺]_C in neurons exposed to NMDA (Fig. 4, A and B), it must be considered that high [K⁺]_E may inhibit the Ca²⁺ influx resulting from the reverse operation of the plasmallemal Na⁺/Ca²⁺ exchanger (R/NaCaX). Because half-maximal activation of the R/NaCaX occurs at 38 mM [Na⁺]_C (Hilgemann, 1989), one may expect that at 60 mM [K⁺]_E, when [Na⁺]_C is 27 mM, the R/NaCaX will transport less Ca²⁺ than at 60 mM [K⁺]_E, when [Na⁺]_C is 80 mM (Fig. 4B).

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Effects of high [K+]E on the NMDA-induced Ca2+ influx may be linked to the inhibition of the reverse operation of the NaCaX. A, high [K+]E fails to inhibit the NMDA-elicited Ca2+ influx in CGCs pre-exposed to NMDA at low [K+]E. CGCs were exposed to NMDA and to the indicated [K+]E for 10 min; after the first 5 min, the medium was supplemented with 1 µCi 45Ca2+ and [K+]E was varied (left). The amount of 45Ca2+ trapped in the cells during the second 5 min of exposure to NMDA is expressed as a percentage of control (exp. 1). 45Ca2+ uptake data (right) are the means ± S.E. from three independent experiments. The background 45Ca2+ accumulation, measured at 5.6 mM K+ (no NMDA), was 19 ± 1.1 nmol/mg protein/5 min, and was subtracted from all the data. The control NMDA-induced 45Ca2+ accumulation measured before the background subtraction was 48 ± 5.0 nmol/mg protein/5 min (exp. 1). *P < .05, t test. Note that 60 mM K+ failed to inhibit 45Ca2+ accumulation in CGCs pre-exposed to NMDA at 5.6 mM K+ (exp. 3), but not at 60 mM K+ (exp. 2). B, Li+ mimics the inhibitory effect of K+ on the NMDA-induced 45Ca2+ uptake (left), but, in contrast to K+, fails to affect the basal 45Ca2+ accumulation (right). CGCs were exposed for 15 min to NMDA, while [K+] or [Li+] in the extracellular medium were increased as indicated on the x-axis; [Na+] was appropriately decreased to maintain osmoticity. The initial [K+] in the extracellular medium (before the addition of extra K+ or Li+) was 5.6 mM, and therefore the final [K+] was greater by 5.6 mM than indicated on the x-axis. The Li+ buffers contained the [Li+] indicated on the x-axis plus 5.6 mM K+. The effects of K+ or Li+ on basal 45Ca2+ accumulation were assessed in the presence of 1 mM Mg2+ plus 10 µM MK-801 (no NMDA or glycine added). Data are the means ± S.E. from five independent experiments.