Two major internal calcium stores, the cytosolic endoplasmic reticulum (ER) (i.e. IP₃/ryanodine-sensitive Ca²⁺ stores) and mitochondria, could be involved in the initial Ca²⁺ release. When cells were treated with a combination of cyclopiazonic acid (CPA, 15 µM) and ryanodine (10 µM) under Ca²⁺-free conditions (Fig. 1C), a small [Ca²⁺]_i transient, probably due to depletion of the ER stores, was seen in the absence of metabolic inhibition. Complete depletion of the ER store by these two blockers in granule cells was confirmed by the lack of further Ca²⁺ release (data not shown) on addition of 0·5 mM carbachol (an agonist of the IP₃-sensitive calcium store in granule cells) in the presence of 10 µM ryanodine (depletion of the ryanodine-sensitive store) and 15 µM CPA (Irving et al. 1992). In the presence of ryanodine and CPA, the initial small [Ca²⁺]_i rise induced by metabolic inhibition was still seen (Fig. 1C and Table 1), therefore suggesting that the IP₃/ER store does not play a major role in this process. In contrast, under the same Ca²⁺-free conditions, this initial small peak was no longer seen in the presence of 100 µM carbonyl cyanide m-chlorophenyl-hydrazone (CCCP) (Fig. 1D and Table 1), a proton ionophore that collapses the negative mitochondrial membrane potential, and results in mitochondrial calcium release and inhibition of calcium uptake in granule cells (Budd & Nicholls, 1996). This indicates that it is the mitochondria that are responsible for the initial Ca²⁺ release seen during early metabolic inhibition.

Lack of involvement of Ca²⁺ and NMDA/non-NMDA channels, the plasmalemmal Na⁺_i-Ca exchanger and Ca²⁺-ATPase in the large increase in [Ca²⁺]_i during metabolic inhibition

Anoxia induces marked depolarization of the membrane potential in hippocampal slices (Fung & Haddad, 1997), and thus the Ca²⁺ influx seen during hypoxia/ischaemia could possibly be due to the opening of voltage-dependent Ca²⁺ channels. Granule cells possess L- and N-type, but not T/P- type, Ca²⁺ channels (Pearson et al. 1995). However, in our experiments, 10 µM nifedipine (an L-type Ca²⁺ channel blocker) or 1 µM -conotoxin GVIA (an N-type Ca²⁺ channel blocker) had no effect on the Ca²⁺ influx after 10 min perfusion with metabolic inhibitors (Table 2).

Table 2. Effects of various treatments on MI-induced [Ca²⁺ ]_i increase

Treatment	[Ca2+]i (nM)
MI (10 min)	250 ± 30 (28)
Nifedipine	190 ± 30 (4)
-Conotoxin	200 ± 40 (4)
SK&F 96365	200 ± 45 (4)
Gadolinium	235 ± 70 (4)
Flufenamic acid	235 ± 37 (4)
MK-801	268 ± 14 (4)
DNQX	248 ± 18 (5)
MK-801 + DNQX	262 ± 11 (4)
Na+ free	191 ± 27 (4)
MI (10 min)	218 ± 46 (4)
K+ free	11 ± 3 *
MI (10 min)	196 ± 25 (4)
pHo 8·5	548 ± 21 *
MI (10 min)	188 ± 10 (4)
Eosin B	312 ± 21 *

The possible involvement of the two other calcium-permeable channels was also tested, namely the calcium release-activated calcium (CRAC) channel (present in granule cells; Simpson et al. 1995) and the non-selective cation channel. Neither the CRAC channel blocker SK&F 96365 (30 µM) nor the two non-selective cation channel blockers flufenamic acid or gadolinium (both 100 µM) had any effect on the metabolic inhibition-induced increase in [Ca²⁺]_i (Table 2).

As NMDA/non-NMDA channels are involved in hypoxia/ischaemia-induced Ca²⁺ increase in hippocampal and cortical neurons (Dubinsky & Rothman, 1991; Uematsu et al. 1991), this possibility was investigated in granule cells. When NMDA receptors were activated by 100 µM NMDA in Mg²⁺-free solution containing 10 µM glycine, the resultant [Ca²⁺]_i increase could be blocked by 10 µM MK-801, whereas the metabolic inhibition-induced [Ca²⁺]_i increase was not (Fig. 2A and Table 2). Similarly, the [Ca²⁺]_i increase induced by the non-NMDA agonist kainic acid (100 µM) was blocked by 6 µM DNQX, but the metabolic inhibition-induced [Ca²⁺]_i increase was not (Fig. 2B and Table 2). Moreover, pretreatment with both blockers had no effect on the subsequent metabolic inhibition-induced [Ca²⁺]_i increase (Fig. 2C and Table 2), strongly suggesting that, under the present experimental conditions, neither NMDA nor non-NMDA receptors are involved in the metabolic inhibition-induced [Ca²⁺]_i increase.

		View larger version [in this window] [in a new window]
Figure 2. NMDA and non-NMDA receptors are not involved in the metabolic inhibition-induced [Ca2+]i influx Calcium influx in the presence of the NMDA receptor antagonist MK-801 (10 µM) (A), the non-NMDA receptor antagonist DNQX (6 µM) (B) and after pretreatment with both inhibitors (C).

Ischaemia has been demonstrated to cause depolarization and activation of the TTX-sensitive Na⁺ channel in hippocampal slices (Fung & Haddad, 1997), and the subsequent increase in [Na⁺]_i can lead to an increase in [Ca²⁺]_i through activation of the Na⁺_i-Ca exchanger (Goldberg & Choi, 1993). This possibility was examined in the following experiments. In granule cells, the resting [Na⁺]_i, measured using SBFI AM, was 7·8 ± 0·6 mM (n = 70). Within 3·2 ± 0·2 min (n = 19) of the onset of metabolic inhibition, [Na⁺]_i started to increase, reaching a level of 30·0 ± 1·5 mM (n = 12) at 15 min (Fig. 3A); this effect was reversed by replacement of all external sodium ions with N-methyl-D-glucamine (NMDG) (Fig. 3A). Comparison of the results seen in Na⁺-containing (Fig. 1A) and Na⁺-free medium (Fig. 3A) shows that the increase in [Na⁺]_i, but not that in [Ca²⁺]_i (Fig. 3B), was inhibited under Na⁺-free conditions (Table 2), suggesting that the Na⁺_i-Ca exchanger may not be involved. The above results are also summarized in Table 5.

		View larger version [in this window] [in a new window]
Figure 3. Role of the Na+i-Ca2+o exchanger in the metabolic inhibition-induced [Ca2+]i increase Intracellular sodium (A and C) and calcium (B and D) measurements. A, under metabolic inhibition (Glc-free + CN-), a significant [Na+]i increase was seen which was reversed upon changing to Na+-free medium (all external Na+ replaced with N-methyl-D-glucamine). B, [Ca2+]i was increased with metabolic inhibition in Na+-free medium. C, inhibition of Na+-K+-ATPase with either strophenthidin (100 µM) or K+-free medium increased [Na+]i. D, [Ca2+]i was higher under metabolic inhibition than in K+-free medium.

In order to raise Na⁺_i to a level similar to that seen under conditions of metabolic inhibition and determine whether this [Na⁺]_i could induce an increase in [Ca²⁺]_i, the Na⁺-K⁺-ATPase was inhibited using K⁺-free medium; the resulting increase in [Na⁺]_i is shown in Fig. 3C. However, the [Ca²⁺]_i increase induced by K⁺-free medium was much less than that seen during metabolic inhibition (Fig. 3D and Table 2), again suggesting that the Na⁺_i-Ca exchanger was not involved. Interestingly, a commonly used Na⁺-K⁺-ATPase blocker, strophenthidin (100 µM), did not induce an increase in [Na⁺]_i comparable to that produced by K⁺-free medium (Fig. 3C); the reason for this is unknown, but it may be due to K⁺-free medium hyperpolarizing the membrane potential, resulting in a greater sodium influx. Other possibilities cannot be ruled out completely.

Since intracellular ATP levels can be depleted during brain ischaemia (for review see Blaustein, 1988), we tested whether the metabolic inhibition-induced [Ca²⁺]_i increase was due to inhibition of plasma membrane Ca²⁺-ATPase. Before testing this possibility, we conducted a set of control experiments consisting of two successive treatments with 5 mM CN^--containing glucose-free medium, which resulted in similar calcium responses (236 ± 14 and 231 ± 19 nM, P > 0·05, Student's paired t test, n = 4).

Ca²⁺-ATPase is a Ca-H exchanger and, in granule cells, is blocked at a pH_o of 8·5 (reduced [H⁺]_o) (Khodorov et al. 1995). When Ca²⁺-ATPase was blocked at pH_o 8·5, the metabolic inhibition-induced [Ca²⁺]_i increase was significantly higher than at pH_o 7·4 (Fig. 4A and Table 2). A similar result was obtained using another potent Ca²⁺-ATPase blocker, Eosin B (20 µM, Gatto & Milanick, 1993) (Fig. 4B and Table 2). The above results therefore suggest that Ca²⁺-ATPase is still operating during early metabolic inhibition.

		View larger version [in this window] [in a new window]
Figure 4. Role of calcium extrusion mechanisms in the metabolic inhibition-induced [Ca2+]i increase When Ca2+-ATPase was blocked either by pHo 8·5 (A) or Eosin B (20 µM, B), the metabolic inhibition-induced [Ca2+]i increase is larger. The mitochondrial blocker CCCP (100 µM) released stored calcium in the mitochondria both during (C) and after (D) metabolic inhibition. In the buffer system for pHo 8·5 medium, 10 mM bicine replaced 10 mM Hepes.

In neurons, mitochondria have a high capacity to take up the excess Ca²⁺ during calcium overload (Gunter et al. 1994; Kiedrowski & Costa, 1995; Budd & Nicholls, 1996; Wang & Thayer, 1996). The following experiment was performed to test whether this process was blocked during early metabolic inhibition, thus resulting in the marked [Ca²⁺]_i increase. When 100 µM CCCP was added during the upstroke of the calcium increase (Fig. 4C, arrowhead), a further transient [Ca²⁺]_i increase (867 ± 156 nM, n = 4), presumably due to mitochondrial release (Kiedrowski & Costa, 1995; Budd & Nicholls, 1996), was seen, and [Ca²⁺]_i returned to the original level when CCCP was removed (Fig. 4C). Furthermore, a rapid rise in [Ca²⁺]_i (722 ± 100 nM, n = 4; Fig. 4D) was seen when 100 µM CCCP was added during the recovery period from metabolic inhibition. This suggests that mitochondrial calcium uptake was at least partially functioning under the experimental conditions.

Role of PLA₂ and membrane lipids in the large metabolic inhibition-induced [Ca²⁺]_i increase

PLA₂ activity in rat neocortical neurons is increased during brain ischaemia (Umemura et al. 1992). When this possibility was tested in granule cells, the marked Ca²⁺ influx was completely and reversibly blocked by either of two potent PLA₂ inhibitors, mepacrine (50 µM, Fig. 5A and Table 3) and antiflammin-1 (200 nM, Fig. 5B and Table 3) (Löffler et al. 1985; Lloret & Moreno, 1992). It was also blocked by divalent cations, Ni²⁺ (Fig. 5C and Table 3) and Co²⁺ (Fig. 5D and Table 3). Melittin (50 nM), a potent PLA₂ activator (Choi et al. 1992; Clapp et al. 1995), induced a marked increase in the [Ca²⁺]_i, which was similarly blocked by these divalent cations (Fig. 5E and F and Table 3).

		View larger version [in this window] [in a new window]
Figure 5. Role of PLA2 in the metabolic inhibition-induced [Ca2+]i increase Effect of two PLA2 inhibitors, mepacrine (50 µM, A) and antiflammin-1 (200 nM, B). Ni2+ (C) and Co2+ (D) (both at 1 mM) markedly inhibit the calcium influx induced by metabolic inhibition. The melittin (50 nM, a potent PLA2 activator)-induced calcium influx can be also inhibited by Ni2+ (E) and Co2+ (F) (both 1 mM).

Table 3. Effects of PLA₂ inhibitors or metal ions on MI-, melittin- or fatty acid-induced [Ca²⁺]_i increases

Treatment	[Ca2+]i (nM)
MI (10 min)	203 ± 32 (4)
+ Mepacrine	20 ± 2 *
MI (10 min)	168 ± 28 (4)
+ Antiflammin-1	14 ± 3 *
Melittin	229 ± 23 (5)
+ Ni2+	124 ± 21 *
Melittin	207 ± 15 (5)
+ Co2+	72 ± 11 *
MI (10 min)	340 ± 43 (7)
+ Ni2+	32 ± 9 *
MI (10 min)	163 ± 21 (4)
+ Co2+	14 ± 5 *
LPC	430 ± 36 (5)
+ Ni2+	184 ± 24 *
LPC	512 ± 73 (4)
+ Co2+	85 ± 18 *
AA	265 ± 30 (5)
+ Ni2+	127 ± 17 *
AA	194 ± 30 (4)
+ Co2+	93 ± 8 *

In the brain, PLA₂ can be activated when the [Ca²⁺]_i is increased to the range 0·01-1 µM (Farooqui et al. 1997). Since the initial small Ca²⁺ increase was due to mitochondrial Ca²⁺ release, the question then arose as to whether the calcium response was initiated by the small mitochondrial calcium release, which then activated PLA₂ and caused the large Ca²⁺ influx. However, this possibility seems unlikely, since, following 10-15 min pretreatment with 100 µM CCCP in the absence or presence (Fig. 6A, arrowhead) of 2 mM [Ca²⁺]_o, the [Ca²⁺]_i was changed little (Table 1), and the subsequent metabolic inhibition-induced Ca²⁺ influx was even greater (Fig. 6B, second arrowhead, 961 ± 155 nM, n = 4) than that seen in the absence of CCCP (250 ± 30 nM, n = 28, Fig. 1A and Table 1). This suggests that there is no direct coupling between the initial mitochondrial calcium release and the PLA₂ activation-mediated calcium influx.

		View larger version [in this window] [in a new window]
Figure 6. Role of the mitochondria in the calcium response A, a small [Ca2+]i increase, induced by CCCP addition (Ca2+-free medium), is followed by a similar change in [Ca2+]i (arrowhead) in the presence of 2 mM [Ca2+]o. B, after CCCP pretreatment (first arrowhead), 5 mM CN- in glucose-free medium induced a fast calcium influx (second arrowhead) in 2 mM [Ca2+]o-containing medium. The concentration of CCCP was 100 µM.

PLA₂ activation results in the release of free fatty acids, including arachidonic acid (AA), and lysophospholipids, including lysophosphatidylcholine (LPC), and both induce calcium overload during cardiac ischaemia (Jones et al. 1989; Donck et al. 1992). The effect of AA or LPC on the metabolic inhibition-induced [Ca²⁺]_i increase was tested. LPC (3 µM) caused a marked increase in the [Ca²⁺]_i in Ca-containing solution similar to that seen during metabolic inhibition, and this effect was blocked by either 1 mM Ni²⁺ (Fig. 7A and Table 3) or Co²⁺ (Fig. 7B and Table 3). A similar result was seen using AA (10 µM, Fig. 7C and D and Table 3). Inhibitors of cyclo-oxygenase (10 µM indomethacin), lipo-oxygenase (10 µM MK-886 + 10 µM baicalein) or cytochrome P450 (10 µM econazole) did not affect the metabolic inhibition-induced calcium increase (data not shown), indicating that AA metabolites are probably not involved.

		View larger version [in this window] [in a new window]
Figure 7. Role of the membrane degradation products LPC and AA, in the metabolic inhibition-induced [Ca2+]i increase Ni2+ (1 mM) inhibited the LPC (3 µM, A)- and AA (10 µM, C)-induced [Ca2+]i increase. Co2+ (1 mM) also inhibited LPC (3 µM, B)-and AA (10 µM, D)-induced [Ca2+]i increase.

Role of free radicals in metabolic inhibition-induced [Ca²⁺]_i increase

In cerebellar granule cells, addition of CN^- has been shown to activate NMDA channels and simultaneously generate reactive oxygen species (ROS), which can be removed by superoxide dismutase (SOD) and L-NAME (a nitric oxide synthase inhibitor) (Gunasekar et al. 1996). Moreover, stimulation of the NMDA receptor/channel results in NMDA-dependent O₂^-· production and neurotoxicity in granule cells (Lafon-Cazal et al. 1993). Since ROS overproduction can cause an increase in [Ca²⁺]_i in other cells (Dubinsky & Rothman, 1991; Uematsu et al. 1991), we tested whether overproduction of ROS/NO contributed to the metabolic inhibition-induced Ca²⁺ influx seen in granule cells.

When the cells were continuously perfused with 5 mM CN^--containing glucose-free medium (30 min), a marked increase was seen in the [Ca²⁺]_i, which slowly plateaued at 1188 ± 247 nM (n = 8) (Fig. 8A and Tables 1 and 4). In the presence of SOD + L-NAME (Fig. 8B), the metabolic inhibition-induced [Ca²⁺]_i increase was not significantly different, possibly in part due to variable calcium peaks in this set of experiments (Table 4). However, using either U-78517 F (20 µM, a potent lipid peroxidation inhibitor and intracellular O₂^-· scavenger, Fig. 8C) (Hall et al. 1991) or N-(2-mercaptopropionyl)-glycine (N-MPG) (an OH· scavenger, Fig. 8D) (Bolli et al. 1989), the calcium rise in the upstroke was halted and the peak [Ca²⁺]_i levels were significantly lower (Table 4). The above results therefore suggest that free radicals may play a role in the metabolic inhibition-induced calcium increase.

		View larger version [in this window] [in a new window]
Figure 8. Effect of antioxidants on the metabolic inhibition-induced [Ca2+]i increase A, control, showing response to metabolic inhibitors (~30 min). B-D, response in the presence of antioxidants. The concentrations of L-NAME, superoxide dismutase (SOD), U-78517F and N-MPG were 300 µM, 100 units ml-1, 20 µM and 10 mM, respectively.

Role of the Na-Ca exchanger, Na⁺-H⁺ exchanger, Na⁺-K⁺-2Cl^- cotransporter, voltage-gated Na⁺ channels and PLA₂ in metabolic inhibition-induced [Na⁺]_i increase

It is possible that the large Na⁺ influx seen in granule cells during early metabolic inhibition could occur via the Na-Ca exchanger as a result of the large [Ca²⁺]_i increase; however, the magnitude of the metabolic inhibition-induced increase in [Na⁺]_i seen in calcium-free medium (Fig. 9A and Table 5) was essentially similar to that seen in calcium-containing medium (Fig. 3A and Table 5). A further possibility was that metabolic inhibition induced a decrease in the pH_i (Wu & Vaughan-Jones, 1994) and consequently stimulated the Na⁺-H⁺ exchanger to produce an increase in [Na⁺]_i. However, addition of 60 µM HOE 694 (a Na⁺-H⁺ exchanger blocker) had no effect on the metabolic inhibition-induced [Na⁺]_i increase (Fig. 9B and Table 5). Furthermore, the Na⁺-K⁺-2Cl^- cotransporter or the voltage-gated sodium channel could not account for the Na influx, since neither 10 µM bumetanide (a co-transporter blocker) (Table 5, with bumetanide autofluorescence corrected) nor 1 µM TTX (Fig. 9C and Table 5) had any effect. Interestingly, 50 µM mepacrine completely and reversibly inhibited the Na influx (Fig. 9D and Table 5, with mepacrine autofluorescence corrected). Moreover, Ni²⁺ blocked both the metabolic inhibition- and melittin-induced Na influxes (Fig. 9E and F and Table 5).

		View larger version [in this window] [in a new window]
Figure 9. Possible mechanisms involved in the [Na+]i increase during metabolic inhibition A, metabolic inhibition increased [Na+]i under Ca2+-free conditions. In the presence of Ca2+ (2 mM) the metabolic inhibition-induced [Na+]i increase was not inhibited either by addition of the Na+-H+ exchange blocker HOE 694 (60 µM, B), or the Na+ channel blocker TTX (1 µM, C). However, the [Na+]i increase was inhibited by the PLA2 blocker mepacrine (50 µM, D), and Ni2+ (1 mM, E). Addition of Ni2+ also inhibited the effect of the PLA2 stimulator melittin on [Na+]i (50 nM, F).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In cultured rat cerebellar granular cells, metabolic inhibition induces an increase in the [Ca²⁺]_i, which can be divided into two stages, an initial small increase due to calcium release from the mitochondria, and a subsequent large increase due to calcium influx, probably involving PLA₂ activation, as does the observed large influx of Na. These results, obtained from cultured cells, differ from the findings from previous in vivo or brain slice studies. These discrepancies are discussed below.

Role of the mitochondria in the [Ca²⁺]_i response seen during metabolic inhibition

Recent studies have shown that the mitochondria play an important role in Ca²⁺ buffering during glutamate-induced Ca²⁺ overload in granule cells (Kiedrowski & Costa, 1995; Budd & Nicholls, 1996). In the present study, when mitochondrial calcium uptake was blocked by CCCP (Kiedrowski & Costa, 1995; Budd & Nicholls, 1996) in the presence of 2 mM [Ca²⁺]_o (Fig. 6A and B, first arrowheads), the [Ca²⁺]_i was higher (130 nM, Table 1) than that seen in the absence of CCCP (i.e. basal state, 30 nM), suggesting that mitochondria may play an important role in regulating the basal [Ca²⁺]_i. Moreover, metabolic inhibition caused a greater increase in [Ca²⁺]_i when mitochondrial calcium uptake was blocked (Fig. 6B, second arrow), providing further evidence that mitochondria have a high capacity to take up the overloaded calcium, as reported by other investigators (Kiedrowski & Costa, 1995; Budd & Nicholls, 1996).

To the best of our knowledge, the present study is the first to show that calcium release from mitochondria, but not from the ryanodine/IP₃-sensitive stores (ER), is responsible for the initial increase in [Ca²⁺]_i during metabolic inhibition in neurons (Fig. 1C and D). This finding raised the interesting possibility of the initial small Ca²⁺ release resulting in PLA₂ activation, thus generating the subsequent large Ca²⁺ influx. However, this seems unlikely, since pretreatment with CCCP alone, which, as mentioned above, caused little change in [Ca²⁺]_i, did not result in the large rapid Ca²⁺ influx seen during metabolic inhibition (compare Figs 1A, 6A and 6B). We therefore suggest that the initial mitochondrial calcium release is not coupled to the subsequent large calcium influx seen during metabolic inhibition. This also suggests that neither mitochondrial depolarization nor inhibition of mitochondrial ATP synthesis by CN^- can account for the large calcium influx.

Another interesting point is that mitochondrial calcium uptake seems to be at least partially functioning after short-term metabolic inhibition, as shown by the fact that it was still possible for CCCP to induce a large calcium release from the mitochondria both during (Fig. 4C) and after (Fig. 4D) metabolic inhibition. The larger CCCP-induced calcium release seen in Ca²⁺-containing medium compared with Ca²⁺-free medium (Fig. 1D) is probably due to greater calcium influx, resulting in greater mitochondrial uptake, during metabolic inhibition. In the carotid body, CN^- depolarizes the mitochondrial potential, resulting in the release of stored calcium (Duchen & Biscoe, 1992). However, we found that the mitochondria are still able to take up the influxed calcium after CN^- treatment (Fig. 4C), which may depolarize the mitochondrial potential in granule cells. One possible explanation is that CN^- causes less depolarization than that produced by CCCP, which we cannot accurately measure using rhodamine-123. Other possibilities cannot be ruled out completely.

Role of PLA₂ and membrane phospholipids in the large influxes of Ca²⁺ and Na⁺

[Ca²⁺]_i increase. During early ischaemia (within 15 min), both phospholipase C (PLC) and PLA₂ are activated in the rat neocortex (Umemura et al. 1992). In rat hepatocytes, PLA₂ can also be activated by metabolic inhibition (Sakaida et al. 1992). The present study shows that two potent PLA₂ blockers, mepacrine and antiflammin-1, were able to completely block the large Ca²⁺ influx (Fig. 5A and B). Moreover, melittin, a potent activator of endogenous PLA₂ (Choi et al. 1992), mimicked the effect of metabolic inhibition (Fig. 5E and F). These results therefore strongly suggest that PLA₂, which is probably stimulated during hypoxia/ischaemia, is involved in the large Ca²⁺ influx seen during metabolic inhibition. However, the possibility that PLC activation also contributes to the calcium influx cannot be entirely excluded.

Brain ischaemia-induced PLA₂ activation can result in the accumulation of free fatty acids produced by membrane phospholipid degradation. Of these, AA produces the greatest increase (within 5 min) and the levels of AA are found to remain high even following cessation of brain ischaemia in many in vivo animal studies (Abe et al. 1987; Umemura et al. 1992). The role of lysophospholipids in the metabolic inhibition-induced calcium increase in neurons, however, is less clear. We suggest that activation of PLA₂ results in the accumulation of LPC and AA, which then contribute to the large Ca²⁺ influx, since addition of either of these products produced a calcium response similar to that induced by metabolic inhibition. Downstream AA metabolites are probably not important, as cyclo-oxygenase, lipo-oxygenase or cytochrome P450 inhibitors did not affect the metabolic inhibition-induced calcium increase. In addition to LPC and AA, other fatty acids or lysophospholipids may also contribute to the Ca²⁺ influx.

Role of ROS. In cerebellar granule neurons, generation of O₂^-· is significantly increased following NMDA activation (Lafon-Cazal et al. 1993). Moreover, CN^- has been shown to activate NMDA channels and simultaneously generate reactive oxygen species (ROS) in granule cells (Gunasekar et al. 1996). However, none of these studies examined possible changes in [Ca²⁺]_i during the generation of ROS/NO. Our results suggest that intracellular production of O₂^-· and OH· may play a role in the metabolic inhibition-induced [Ca²⁺]_i increase, since either U-78517F (a potent O₂^-· scavenger and inhibitor of lipid peroxidation) or N-MPG (an intracellular OH· scavenger) significantly inhibited the large calcium influx (Table 4). However, addition of SOD (a scavenger of extracellular O₂^-·) and L-NAME (an NO · scavenger) had little effect (Table 4). Since AA and LPC induced the calcium influx similar to that seen during metabolic inhibition, this suggests that lipid peroxidation-induced ROS overproduction (possibly intracellular O₂^-· and OH·) (Easton & Fraser, 1998) is also involved in the large calcium influx seen during metabolic inhibition.

Table 4. Effects of metal ions on the AA- or LPC-induced [Ca²⁺]_i increase

Treatment	[Ca2+]i (nM)
MI (30 min)	1188 ± 247 (8)
L-NAME + SOD	1309 ± 428 (4)
U-78517F	241 ± 37 (4)
N-MPG	302 ± 32 (4)

[Na⁺]_i increase. In brain slices, anoxia induces a decrease in the extracellular Na⁺ levels, and removal of Na prevents anoxia-induced morphological changes in hypoglossal neurons (Jiang et al. 1992). The possibility that anoxia-induced depolarization results in a TTX-sensitive Na⁺ influx in hippocampal slices has been suggested (Fung & Haddad, 1997), since TTX significantly attenuates both the depolarization rate and the rate of input resistance decline. Another study in cultured neocortical neurons reports that the Na-Ca exchanger plays a more important role than the TTX-sensitive Na⁺ channel, as amiloride (a non-specific blocker of the Na-Ca exchanger), rather than TTX, prevents the anoxia-induced morphological changes (Chidekel et al. 1997).

In this study, we found the [Na⁺]_i increased to 30 mM after 15 min metabolic inhibition (Fig. 3A). The involvement of the Na-Ca exchanger, Na⁺-H⁺ exchanger, Na⁺-K⁺-2Cl^- cotransporter and TTX-sensitive Na⁺ channel (Fig. 9 and Table 5) was ruled out. However, PLA₂ activation appears to be involved, since mepacrine completely and reversibly blocked the Na⁺ response (Fig. 9D), and melittin mimicked the effect of metabolic inhibition on the [Na⁺]_i (Fig. 9F). These facts, together with the similar time course (3 min) of the onsets of the [Ca²⁺]_i and [Na⁺]_i increases, suggest that the Ca²⁺ and Na⁺ influxes probably occur via LPC/AA-activated non-selective cation or leak channels. Further investigations are required to clarify this issue.

Table 5. Influx of extracellular Na⁺ under various conditions

Treatment	[Na+]i or [Na+]i/t
Basal	7·8 ± 0·6 mM (70)
MI	30·0 ± 1·5 mM (12)
Ca2+-free	36·7 ± 2·0 mM (4)
HOE 694	29·7 ± 7·2 mM (4)
Bumetanide	27·5 ± 1·2 mM (4)
Mepacrine	-1·4 ± 0·5 mM (4)
MI + Ni2+	2·0 ± 0·7 mM (4)
Melittin + Ni2+	10·9 ± 3·1 mM (4)
MI	1·8 ± 0·6 mM min-1
MI + TTX	1·7 ± 0·6 mM min-1

Role of NMDA/non-NMDA channels, voltage-sensitive or insensitive Ca²⁺ channels, the Na⁺_i-Ca exchanger and Ca²⁺-ATPase in the large [Ca²⁺]_i increase in metabolic inhibition

Activation of NMDA/non-NMDA channels during hypoxia/anoxia can result in an increased [Ca²⁺]_i in hippocampal neurons (Dubinsky & Rothman, 1991) and the cerebral cortex (Uematsu et al. 1991). However, in cultured granule cells, neither of these channels is involved in the calcium increase seen following metabolic inhibition (Fig. 2 and Table 2). Voltage-dependent and -independent Ca²⁺ channels also seem not to be involved, since the specific blockers of these channels had no inhibitory effect (Table 2). In the present study, Ni²⁺, a known blocker of voltage-dependent Ca²⁺ channels, non-selective leak channels (see Nilius et al. 1993 for review) and the Na-Ca exchanger (Du et al. 1997), was able to inhibit the metabolic inhibition-, LPC-, AA- or melittin-induced Ca²⁺ increase (Figs 5 and 7 and Table 2). On the basis of the above data, we speculate that the blocking effect of the metal ions on the Ca²⁺/Na⁺ response in granule cells is probably via non-selective cation or leak channels; this point requires further investigation.

Na⁺_i exchange for Ca, with the resultant Ca²⁺ overload, was one possible explanation for the metabolic inhibition-induced [Ca²⁺]_i increase, since a marked [Na⁺]_i increase was indeed observed. However, this possibility can be discounted, since the use of Na⁺-free medium (to block the Na⁺_i-Ca exchanger, Fig. 3B) or K⁺-free medium (to induce high [Na⁺]_i, Fig. 3D) did not affect Ca²⁺ influx.

Another possibility was inhibition of the plasmalemmal Ca²⁺-ATPase. This is also unlikely, since, in the presence of pH_o 8·5 or Eosin B (Fig. 4A and B and Table 2), the Ca²⁺ influx was augmented, suggesting that Ca²⁺-ATPase is still active during early metabolic inhibition. The possibility that the increase in [Ca²⁺]_i, seen in Fig. 4A, was due to extracellular alkalinization (pH_o 8·5) and an increased conductance of the calcium channels is also unlikely, since specific calcium channel blockers had no effect on the calcium influx (Table 2). It is intriguing to note that granule cell Ca²⁺-ATPase was not inhibited during early metabolic inhibition (ATP_i, 0·35 mM, Ekholm et al. 1992); this is probably because the Km of ATP for Ca²⁺-ATPase is low (0·03 mM, Caroni & Carafoli, 1980). However, another possibility is that Ca²⁺-ATPase in granule cells may be not so important for calcium extrusion, since the rate of calcium recovery during wash-off of the metabolic inhibitors was similar either in the presence or absence of Ca²⁺-ATPase blockers (Fig. 4A and B). Therefore, other calcium extrusion mechanisms (e.g. the mitochondria) may be more important than Ca²⁺-ATPase during recovery from Ca²⁺ load.

Comparison between the mechanisms suggested in the present work and those suggested by in vivo or brain slice studies

In vivo studies and those involving brain slices have demonstrated that the extracellular ion composition is greatly altered during hypoxia/ischaemia of the brain. For example, a rapid depolarization of the membrane potential, an increase in the concentration of extracellular K⁺ (50-80 mM) and decreases in pH_o and P_O2 have been observed within 5-10 min of the onset of brain hypoxia/ischaemia (reviewed in Hansen, 1985). Many mechanisms have therefore been suggested (see Introduction) to explain the changes in [Ca²⁺]_i and [Na⁺]_i seen in the brain during such insult. In addition to these mechanisms, activation of NMDA/non-NMDA channels, resulting from accumulation of glutamate in the extracellular space, has recently been demonstrated (for review see Choi, 1988; Uematsu et al. 1991).

The present study, using cultured cells, shows that PLA₂ activation plays an important role in the increases in [Ca²⁺]_i and [Na⁺]_i seen during metabolic inhibition. The conditions are not entirely consistent with those in in vivo and slice studies, as the extracellular ion composition is constant and totally oxygen-free conditions are not achieved. This may also explain the lack of involvement of the NMDA/non-NMDA, voltage-dependent Ca²⁺ channels or involvement of oxygen free radicals during the metabolic inhibition. Moreover, the cultured neurons used may have very different properties from cells in the intact brain, e.g. neonatal (cultured) cells tolerate anoxia longer than adult cells, and receptor or channel densities may change during the culturing process. Because of these differences, the possibility that activation of PLA₂ may also be involved in hypoxia/ischaemia-induced [Ca²⁺]_i/[Na⁺]_i changes should be investigated in vivo or in slices.

In summary, the present study shows that, during metabolic inhibition, there is an initial small calcium release from mitochondria, followed by a large influx of extracellular Ca²⁺ and Na⁺, the latter probably being mediated by products, including fatty acids and free radicals, formed on PLA₂ activation. We have shown that the calcium extruding mechanisms in granule cells appear to be at least partially functioning during early metabolic inhibition. Complete recovery to the resting [Ca²⁺]_i seen following wash-off of the metabolic inhibitors (Figs 1A and 8A) therefore indicates that PLA₂ can be reversibly activated by short-term metabolic inhibition; this point requires further investigation. The mechanism suggested in this study as an explanation for the metabolic inhibition-induced Ca and Na influxes differs from those suggested in in vivo and slices studies on neurons; this may be due to the different working conditions used. However, as granule cells are the most abundant neurons in the brain, the findings in the present study may be important in the understanding of the mechanisms involved in calcium and sodium overload under metabolic inhibition. Inhibition of PLA₂ activity might have implications in preventing serious brain injury and oedema during hypoxia/ischaemia.

	REFERENCES

Top Abstract Introduction Methods Results Discussion References

Abe, K., Kogure, K., Yamamoto, H., Imazawa, M. & Miyamoto, K. (1987). Mechanism of arachidonic acid liberation during ischemia in gerbil cerebral cortex. Journal of Neurochemistry 48, 503-509	[Abstract]
Allen, D. & Orchard, C. H. (1986). Myocardial contractile function during ischemia and hypoxia. Circulation Research 60, 153-158.	[Abstract]
Blaustein, M. P. (1988). Calcium transport and buffering in neurons. Trends in Neurosciences 11, 438-443	[Medline]
Bolli, R., Jeroudi, M. O., Patel, B. S., Aruoma, O. I., Halliwell, B., Lai, E. K. & McCay, P. B. (1989). Marked reduction of free radical generation and contractile dysfunction by antioxidant therapy begun at the time of reperfusion. Circulation Research 65, 607-622	[Abstract]
Budd, S. L. & Nicholls, D. G. (1996). A reevaluation of the role of mitochondria in neuronal Ca2+ homeostasis. Journal of Neurochemistry 66, 403-411	[Abstract]
Caroni, P. & Carafoli, E. (1980). An ATP-dependent Ca2+ pumping system in dog heart sarcolemma. Nature 283, 765-767	[Medline]
Cervos-Navarro, J. & Diemer, N. H. (1991). Selective vulnerability in brain hypoxia. Critical Reviews in Neurobiology 6, 149-182	[Medline]
Chidekel, A. S., Friedman, J. E. & Haddad, G. G. (1997). Anoxia-induced neuronal injury: role of Na+ entry and Na+-dependent transport. Experimental Neurology 146, 403-413	[Medline]
Choi, D. W. (1988). Calcium-mediated neurotoxicity: relationship to specific channel types and role in ischemic damage. Trends in Neurosciences 11, 465-469	[Medline]
Choi, O. H., Padgett, W. L. & Daly, J. W. (1992). Effects of the amphiphilic peptides melittin and mastoparan on calcium influx, phosphoinositide breakdown and arachidonic acid release in rat pheochromocytoma PC12 cells. Journal of Pharmacology and Experimental Therapeutics 260, 369-375	[Abstract]
Clapp, L. E., Klette, K. L., DeCoster, M. A., Bernton, E., Petras, J. M., Dave, J. R., Laskosky, M. S., Smallridge, R. C. & Tortella, F. C. (1995). Phospholipase A2-induced neurotoxicity in vitro and in vivo in rats. Brain Research 693, 101-111	[Medline]
Courtney, M. J., Lambert, J. J. & Nicholas, D. G. (1990). The interactions between plasma membrane depolarization and glutamate receptor activation in the regulation of cytoplasmic free calcium in cultured cerebellar granule cells. Journal of Neuroscience 10, 3873-3879	[Abstract]
Donck, L. V., Verellen, G., Geerts, H. & Borgers, M. (1992). Lysophosphatidylcholine-induced Ca2+-overload in isolated cardiomyocytes and effect of cytoprotective drugs. Journal of Molecular and Cellular Cardiology 24, 977-988.	[Medline]
Du, X.-J., Bobik, A., Little, P. J., Esler, M. D. & Dart, A. M. (1997). Role of Ca in metabolic inhibition-induced norepinephrine release in rat brain synaptosome. Circulation Research 80, 179-188	[Abstract/Full Text]
Dubinsky, J. M. & Rothman, S. M. (1991). Intracellular calcium concentrations during 'chemical hypoxia' and excitotoxic neuronal injury. Journal of Neuroscience 11, 2545-2551	[Abstract]
Duchen, M. R. & Biscoe, T. J. (1992). Relative mitochondrial membrane potential and [Ca2+]i in type 1 cells isolated from the rabbit carotid body. The Journal of Physiology 450, 33-61	[Abstract]
Easton, A. S. & Fraser, P. A. (1998). Arachidonic acid increases cerebral microvascular permeability by free radicals in single pial microvessels of the anaesthetized rat. The Journal of Physiology 507, 541-547	[Abstract/Full Text]
Ekholm, A., Asplund, B. & Siesjo, B. K. (1992). Perturbation of cellular energy state in complete ischemia: relationship to dissipate ion fluxes. Experimental Brain Research 90, 47-53	[Medline]
Farooqui, A. A., Yang, H. C., Rosenberger, T. A. & Horrocks, L. A. (1997). Phospholipase A2 and its role in brain tissue. Journal of Neurochemistry 69, 889-901	[Abstract]
Friedman, J. E. & Haddad, G. G. (1993). Major differences in Ca2+ response to anoxia between neonatal and adult rat CA1 neurons: role of Cao and Nao. Journal of Neuroscience 13, 63-72	[Abstract]
Friedman, J. E. & Haddad, G. G. (1994). Anoxia induces an increase in intracellular sodium in rat central neurons in vitro. Brain Research 663, 329-334	[Medline]
Fung, M.-L. & Haddad, G. G. (1997). Anoxia-induced depolarization in CA1 hippocampal neurons: role of Na-dependent mechanism. Brain Research 762, 97-102	[Medline]
Gallo, V., Giotti, M. T., Coleti, A., Aloisi, F. & Levi, G. (1982). Selective release of glutamate from cerebellar granule cells differentiating in culture. Proceedings of the National Academy of Sciences of the USA 79, 7919-7923	[Medline]
Gatto, C. & Milanick, M. A. (1993). Inhibition of the red blood cell calcium pump by eosin and other fluorescein analogues. American Journal of Physiology 264, C1577-1586	[Medline]
Goldberg, M. P & Choi, D. W. (1993). Combined oxygen and glucose deprivation in cortical cell culture: calcium-dependent and calcium-independent mechanisms of neuronal injury. Journal of Neuroscience 13, 3510-3524	[Abstract]
Grynkiewicz, G., Poenie, M. & Tsien, R. Y. (1985). A new generation of Ca2+indicators with greatly improved fluorescence properties. Journal of Biochemistry 260, 3440-3450.
Gunasekar, P. G., Sun, P. W., Kanthasamy, A. G., Borowitz, J. L. & Isom, G. E. (1996). Cyanide-induced neurotoxicity involves nitrite oxide and reactive oxygen species generation after N-methyl-D-aspartate receptor activation. Journal of Pharmacology and Experimental Therapeutics 277, 150-155	[Abstract]
Gunter, T. E., Gunter, K. K., Sheu, S. S. & Gavin, C. E. (1994). Mitochondrial calcium transport: physiological and pathological relevance. American Journal of Physiology 267, C313-339	[Medline]
Hall, E. D., Braughler, J. M., Yonkers, P. A., Smith, S. L., Linseman, K. L., Means, E. D., Scherch, H. M., Von Voigtlander, P. F., Lahti, R. A. & Jacobsen, E. J. (1991). U-78517F: a potent inhibitor of lipid peroxidation with activity in experimental brain injury and ischemia. Journal of Pharmacology and Experimental Therapeutics 258, 688-694	[Abstract]
Halliwell, B. (1992). Reactive oxygen species and the central nervous system. Journal of Neurochemistry 59, 1609-1623	[Abstract]
Hansen, A. J. (1985). Effect of anoxia on ion distribution in the brain. Physiological Reviews 65, 101-148	[Medline]
Harootunian, A. T., Kao, J. P. Y., Eckert, B. K. & Tsien, R. Y. (1989). Fluorescence ratio imaging of cytosolic free Na+ in individual fibroblasts and lymphocytes. Journal of Biological Chemistry 256, 19458-19467.
Irving, A. J., Collingridge, G. L. & Schofield, J. G. (1992). Interactions between Ca2+ mobilizing mechanisms in cultured rat cerebellar granule cells. The Journal of Physiology 456, 667-680	[Abstract]
Jiang, C., Agulian, S. & Haddad, G. G. (1992). Cl- and Na+ homeostasis during anoxia in rat hypoglossal neurons: intracellular and extracellular in vitro studies. The Journal of Physiology 448, 697-708	[Abstract]
Jones, R. L., Miller, J. C., Hagler, H. K., Chien, K. R., Willerson, J. T. & Buja, M. (1989). Association between arachidonic acid release and prevention of calcium loading during ATP depletion in cultured rat cardiac myocytes. American Journal of Pathology 135, 541-556	[Abstract]
Khodorov, B., Pinelis, V., Vergun, O., Storozhevykh, T., Fajuk, D., Vinskaya, N., Arsenjeva, E., Khaspekov, L., Lyzin, A., Isaev, N., Andreeva, N. & Victorov, I. (1995). Dramatic effects of external alkalinity on neuronal calcium recovery following a short-duration glutamate challenge: the role of the plasma membrane Ca2+/H+ pump. FEBS Letters 371, 249-252	[Medline]
Kiedrowski, L. & Costa, E. (1995). Glutamate-induced destabilization of intracellular calcium concentration homeostasis in cultured cerebellar granule cells: role of mitochondria in calcium buffering. Molecular Pharmacology 47, 140-147	[Abstract]
Lafon-Cazal, M., Pietri, S., Culcasi, M. & Bockaert, J. (1993). NMDA-dependent superoxide production and neurotoxicity. Nature 535, 537.
Lloret, S. & Moreno, J. J. (1992). In vitro and in vivo effects of the anti-inflammatory peptides, antiflammins. Biochemical Pharmacology 44, 1437-1441	[Medline]
Löffler, B. M., Bohn, E., Hesse, B. & Kunze, H. (1985). Effects of antimalarial drugs on phospholipase A and lysophospholipase activities in plasma membrane, mitochondrial, microsomal and cytosolic subcellular fractions of rat liver. Biochimica et Biophysica Acta 835, 448-455	[Medline]
Nilius, B., Droogmans, G., Gericke, M. & Schwarz, G. (1993). Non-selective ion pathways in human endothelial cells. EXS 66, 269-280.	[Medline]
Pearson, H. S., Sutton, K. G., Scott, R. H. & Dolphin, A. C. (1995). Characterization of Ca2+ channel currents in cultured rat cerebellar granule cells. The Journal of Physiology 482, 493-509	[Abstract]
Rose, C. R. & Ransom, B. R. (1997). Regulation of intracellular sodium in cultured rat hippocampal neurons. The Journal of Physiology 499, 573-587	[Abstract]
Sakaida, I., Thomas, A. P. & Farber, J. L. (1992). Phospholipid metabolism and intracellular Ca2+ homeostasis in cultured rat hepatocytes intoxicated with cyanide. American Journal of Physiology 263, C684-690	[Medline]
Simpson, P. B., Challiss, R. A. & Nahorski, S. R. (1995). Divalent cation entry in cultured rat cerebellar granule cells measured using Mn2+ quench of fura 2 fluorescence. European Journal of Neuroscience 7, 831-840	[Medline]
Uematsu, D., Araki, N., Greenberg, J. H., Sladky, J. & Reivich, M. (1991). Combined therapy with MK-801 and nimodipine for protection of ischemic brain damage. Neurology 41, 88-94	[Abstract]
Umemura, A., Mabe, H., Nagai, H. & Sugino, F. (1992). Action of phospholipase A2 and C on free fatty acid release during complete ischemia in rat neocortex. Effect of phospholipase C inhibitor and N-methyl-D-aspartate antagonist. Journal of Neurosurgery 76, 648-651	[Medline]
Wang, G. J. & Thayer, S. A. (1996). Sequestration of glutamate-induced Ca2+ loads by mitochondria in cultured hippocampal neurons. Journal of Neurophysiology 76, 1611-1621	[Medline]
Wu, M. L., Kao, E. F., Liu, I. H., Wang, B. S. & Lin-Shiau, S. Y. (1997). The capacitative Ca2+ influx in glial cells is inhibited by glycolytic inhibitors. Glia 21, 315-326	[Medline]
Wu, M. L. & Vaughan-Jones, R. D. (1994). Effect of metabolic inhibitors and second messengers upon Na-H exchanger in the sheep cardiac Purkinje fibres. The Journal of Physiology 478, 301-313	[Abstract]

Acknowledgements

The authors would like to thank the National Science Council for financial support (NSC 87-2314-B002-190).

Corresponding author

M.-L. Wu: Institute of Physiology, College of Medicine, National Taiwan University, No. 1, Sec. 1, Jen-Ai Road, Taipei, Taiwan, Republic of China.

Email: mlw{at}ha.mc.ntu.edu.tw

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Chen, W.-H. Articles by , M.-L. W. Search for Related Content PubMed PubMed Citation Articles by Chen, W.-H. Articles by , M.-L. W.