- Email: [email protected]
An endogenous metal appears to regulate NMDA receptor mediated 45Ca influx and toxicity in cultured cerebellar granule cells
Neuroscience Letters, 137 (1992) 198-202 © 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
198
NSL 08496
An endogenous metal appears to regulate N M D A receptor mediated 45Ca influx and toxicity in cultured cerebellar granule cells Sara Eimerl a n d Michael S c h r a m m Department of Biological Chemistry and the Otto Loevi Centerfor Neurobiology, The Hebrew University of Jerusalem, Jerusalem (Israel) (Received 6 November 1991; Revised version received 27 December 1991: Accepted 31 December 1991)
Key words." Potentiator of N-methyl-D-aspartate toxicity; Cysteine toxicity; Calcium influx; N-Methyl-o-aspartate receptor Glutamate induced 45Ca influx and toxicity were enhanced 2 10 fold by EDTA. A chelator concentration of 10 ~M, which was equivalent to less than 1% of the Mg 2+ and Ca 2÷ concentration in the medium, was effective. The chelator revealed no activity on its own and caused potentiation only when present simultaneously with the agonist of the N M D A receptor. Cysteine, which is known to bind certain metals tightly through its sulfhydryl group, and another chelator, O-phenanthroline, produced the same effect as EDTA. The findings indicate that when the N-methyl-n-aspartate receptor is activated, an endogenous metal can become bound to a chelator or to a physiological metal binding agent, such as cysteine, leading to enhanced Ca 2+ influx into the neuron and toxicity.
In recent years a considerable amount of information on the N-methyl-D-aspartate (NMDA) receptor in CNS neurons [18, 24, 26] and on Ca 2+ influx through its channel [6, 10, 14, 15, 23] has accumulated. While this Ca 2+ message appears to be involved in plasticity and memory processing [16, 20], excessive influx of C a 2+ is apparently the major cause of glutamate toxicity [2, 5, 10, 15, 17, 19]. Thus, for a variety of reasons, there is interest in factors and processes which modulate Ca 2+ influx through the N M D A receptor. Recently we have shown that a low concentration of serum, and of serum albumin strongly potentiate glutamate and N M D A toxicity in cerebetlar granule cells [7, 8, 22]. During these studies it appeared that metal chelators and cysteine, at low concentration, are also strong potentiators of glutamate and N M D A toxicity and of 45Ca influx. These findings are presented below, including the intriguing observation that EDTA is effective only if present during N M D A receptor activation. Cell culture and glutamate toxicity studies were performed as previously described [8]. Rat cerebellar granule cells, at day 8 postnatal, were cultured 9-18 days in 24-well culture boxes (5 x 105 cells/well) in 0.5 ml basal Eagle medium - 25 mM KCI 10% fetal calf serum. To measure glutamate toxicity o r 45Ca uptake, the culture Correspondence." M. Schramm, Department of Biological Chemistry, Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem 91904, lsrael.
medium was removed and the cells were washed twice with 0.5 ml Locke 25 medium (in mM): NaC1 129, KC1 25, MgC12 1, Ca 2.3, NaHCO3 4, HEPES 10, glucose 5, pH 7.5. Fresh Locke 25 was added to measure glutamate toxicity o r 45Ca2+ influx, both performed at 25°C. Some of the experiments were performed in Locke 5 which contains 5 mM KCI, 150 mM NaCI, without Mg 2+, and the other components as above. Toxicity. N M D A receptor agonist and other additions were made to give a final volume of 0.5 ml Locke. Incubation time was 5 30 min. The medium was then replaced with fresh Locke 25 containing 0.2 ¢tM MK-801 to stop further action of the N M D A receptor channel. After 1 h, cell viability was measured by reduction of a tetrazolium salt [3]. Control systems, containing the same additions without an N M D A receptor agonist, or with the agonist, but in presence of 0.2 ¢tM MK-801, were also run. The control containing Locke without agonist served as a measure of 100% viability (0% cells killed). A survey by phase contrast microscopy [22] showed that this control contained <5% dead cells. The percentage of cells killed was calculated for each experimental system relative to the control without agonist. 45Ca influx. Additions were made to give a final volume of 0.25 ml, including 2 x 105 cpm 45CAC12, which was added just before addition of an agonist. After ~ 3 0 min at 25°C, the cells were rapidly washed 3 times at 25°C with 0.5 ml Locke containing 2 /IM (_+) MK-801. A higher concentration of MK-801 in the 45Ca assay was
199
d "~ 70 I-'0 I,'-
T
It#l-NO
c
L
3.0
r'zr,I - N 0
AGONIST
E i'M
U-
J_
,
°50
d
I
AGONIST
1
-
•
o
+
IZl
Z
I
-o
,~
J 30 co _J
Izl
~
to
to
IO-I
0.
-I-
I+1
+
I -
"
r5
~ r~ 1~ F~ 25mM I
K +,
~
o-
+
~-o
+
~ +
X
0
6
+
z
+
6
EL 2 . 0
0
lmM Mg"~÷ i 2.3raM Ca ÷÷
r5
5raM K÷, no Mg÷+ I
1.0-~
...1
~ s ~ to ~.= g ~ , , .g + + ÷ + +
p-
I + +
z
o (..)
to
+
_.1 LL Z
2 5 m M K + , I mM Mg ++
i
5 m M K + n o M g ++
2.3mM Ca ++ Fig. 1. Toxicityinduced through the NMDA receptor is potentiated by chelators and by cysteine. Incubation was 10 min with 50BM glutamate or 300 BM NMDA, with or without additions, in Locke 25 or Locke 5 as shown. Experimental procedures are described in the text. phenan, phenanthroline; cyst, cysteine, aThe NMDA + EDTA columns in Locke 5 represent the average of 9 experiments using 4 differentculture batches. A fifth culture batch (two repeated experiments), not included in the figure, gave a higher toxicity with NMDA alone, but also a higher potentiation, 41 + 1% and 70 + 5%, respectively,b N o value for chelator without agonist is shown because the value was zero.
used to ensure very fast channel block, which is essential for 1-2 min kinetics. The cells were finally suspended in 0.5 ml 0.1 M N a O H and counted. All 45Ca assays were run on duplicate wells. Experiments were repeated with at least two different batches of cell cultures. The average of all experiments + S.E.M. are presented. Fig. 1 shows that glutamate and N M D A toxicity are potentiated by EDTA, O-phenanthroline and cysteine. These potentiating agents were completely harmless when tested in absence of agonists under the conditions of the experiment (Fig. 1, no agonist). Surprisingly, the potentiation by 10 # M E D T A was obtained in presence of a 230-fold molar excess of Ca 2+ and a 100-fold molar excess of Mg 2+. In fact, even 1 # M E D T A produced some potentiation (not shown). The effect of the chelators is also obtained in Locke 5 without Mg 2÷. In relative terms, the potentiation is seen to be stronger with N M D A as agonist than with glutamate. It is also evident in Fig. 1 that the potentiation was in all instances blocked by MK-801, the specific non-competitive antagonist of the N M D A receptor. The fact that all 3 potentiating agents are known to bind metals effectively and the fact that E D T A is a highly specific metal chelating agent suggests that potentiation is due to binding of an endogenous metal. The low concentration of the chelators and the large excess of Ca 2+ and Mg 2+ further indicates an association constant of the endogenous metal with the chelators which is orders of magnitude higher than that for Ca 2+.
Fig. 2. 45Ca influx mediated through the NMDA receptor channel, is potentiated by chelators and by cysteine. Agonist concentrations were as in Fig. 1. Incubation time was 2 min. The basal influx in absence of agonist was subtracted from all values, but is shown as hatched columns. Abbreviations are as shown in the legend of Fig. 1. *The value is significantly higher than that of NMDA without EDTA, P<0.025, Wilcoxon test. **The value is significantly higher than that of glutamate alone. P<0.005, Wilcoxon test. In order to see whether the effect of the potentiating agents is intimately connected with the function of the N M D A receptor, 45Ca influx induced by glutamate and by N M D A was measured in presence and absence of these agents. Fig. 2 shows that 45Ca influx, induced by either glutamate or N M D A , was potentiated by EDTA, phenanthroline and cysteine. The chelators in absence of agonist did not increase 45Ca influx above the low level, barely detectable, in the controls without the chelators. Similar effects of the chelators on glutamate induced 45Ca2+ influx were seen even when measured as early as 1 min after addition of N M D A (not shown). The pattern of potentiation of 45Ca influx is seen to be quite similar to that shown in Fig. 1 for toxicity. It should, however, be noted that potentiation of 45Ca influx in Locke 5, without Mg 2+, was less than in Locke 25 with Mg 2+, apparently because influx in absence of the potentiating agent was already high. In order to remove the endogenous metal to which the chelators bind, the cells were washed with Locke 25 containing 1 0 0 # M E D T A and the E D T A was then removed by washing with Locke 25. Surprisingly, this pretreatment with E D T A was entirely without effect. E D T A was still required to be added together with glutamate to produce potentiation of toxicity (Table I). Further extensive attempts to remove the metal by prolonged preincubation with EDTA, to render the cells maximally responsive upon subsequent addition of glutamate, all failed. It was therefore thought that the metal site might be masked and only becomes exposed when the N M D A
200 TABLE l ATTEMPTS TO ENHANCE TOXICITY BY PRE- AND POSTTREATMENTS WITH EDTA Unless otherwise noted experiments were performed in Locke 25, containing 1 mM Mg2÷and 2.3 mM CaCI2. Treatments 1
5 i ii iii
Toxicity induction by 300 HM NMDA for 15 min. Removal of the agonist, followed by incubation for 1 h with 100 HM EDTA in Locke 25 medium containing 0.2/,tM MK-801. Ten min preincubation with 100 HM EDTA at 25°C or 37°C, washing the cells twice, followed by toxicity induction by 50HM glutamate, 10 min. As above but 2.5 h preincubation with 100 HM EDTA. Addition of 100/IM EDTA to the culture overnight, prior to the experiment. Further treatment and toxicity induction as in 2. above. Sub-toxicity pretreatments in presence of 100 ,uM EDTA: 5 min 2 HM glutamate, second 5 min another 2 HM glutamate. 10 min 5 HM glutamate in Locke 25 with only 1 mM CaC12. 100 gtM NMDA, 10 min, or 50HM glutamate 2 min in Locke 25 with only 1 mM CaC12.Final toxicity induction after removal of the pretreatment medium: 50/./M glutamate 10 min.
Results I
All above pre- and posttreatments had no effect on the toxicity induced by glutamate or NMDA, when compared with untreated systems. In all the above experiments, irrespective of the treatment, the neurons demonstrated effectivepotentiation of toxicity when replicate samples finally received EDTA together with a toxic concentration of the agonist.
receptor is activated. Accordingly, brief preincubations were performed in presence o f both E D T A and glutamate (Table I). However, after washing the cells to remove the agonist and the chelator the neurons still showed the same sensitivity to readdition o f glutamate and to potentiation by E D T A as cells which had not been pretreated. The conclusions f r o m these experiments are discussed below. It was t h o u g h t that the endogenous metal might be Zn 2+, since we have reported that added Z n 2+ attenuates glutamate induced toxicity in the cerebellar granule cells [8]. Others had shown earlier that Zn 2+ decreases the current t h r o u g h the N M D A receptor channel [4, 25], as well as glutamate toxicity [13]. Experiments were therefore conducted to see whether exogenously added Zn 2+ and the endogenous metal have similar affinities for sites on the neurons (Table II). The results indicate that the
endogenous metal is b o u n d with a much higher affinity than the added Zn 2÷. A l t h o u g h the added Zn 2+ strongly inhibited N M D A induced 45Ca influx, the inhibition was readily abolished by washing the cells twice. In contrast, washing the neurons 5 times appeared not to remove the endogenous metal since the N M D A induced 45Ca2+ influx remained the same as in neurons washed twice. Furthermore, both systems still demonstrated a b o u t the same extent o f potentiation by E D T A . It is thus apparent that the endogenous metal is b o u n d with a much higher affinity than the exogenous, added Zn 2+. The findings indicate that an endogenous metal, which can tightly associate with E D T A , O-phenanthroline and cysteine, attenuates both 45Ca influx t h r o u g h the N M D A receptor channel and glutamate toxicity. While the identity o f the metal remains u n k n o w n , it must have an affinity for the chelators which is several orders o f magnitude higher than that o f Ca 2÷ or M g 2+. This is because the latter are present in the m e d i u m at more than a 100fold m o l a r excess over the chelator. Also, a concentration o f E D T A higher than 10 ¢tM is required to achieve maximal potentiation (not shown). Therefore it is likely that the binding site on the cells has a very high affinity for the metal. On that basis the binding site for added Zn 2+ is not a likely candidate since it appeared to have a relatively low affinity for the metal. The effect o f added Zn 2+ was readily abolished by washing the neurons twice. In contrast, washing did not appear to remove the endogenous metal since it did not increase the N M D A induced 45Ca influx, nor did it diminish the potentiation when E D T A was finally added with N M D A . W h y did preincubation with E D T A fail to remove the endogenous metal? One possibility is that the metal can only be removed by the chelator when the N M D A receptor is in the fully activated state for a prolonged period. Alternatively, it is possible that the metal is indeed removed by the chelator when the receptor is even briefly activated, but that it is replaced within a short time from endogenous stores inside the neuron. The situation could be similar to that o f glycine, the requirement for which is often difficult to demonstrate because o f its continuous leakage from the cells [12]. There is also the possibility that the Locke m e d i u m contains enough o f the metal as a trace impurity to reload the sites after the E D T A is removed by washing the cells with this solution. The metal ion could be Fe 3+, C a 2+, Ni 2+, Pb 2+, C o 2+, Z n 2+, Cd 2+ or A13+. It can be calculated, for example, that a c o n t a m i n a t i o n o f the NaC1 reagent by 0.0003% o f A13+ will supply a b o u t 1/.tM AI 3+. At this level, it will be difficult to identify the metal. The potentiating effect o f cysteine deserves special notice. Cysteine was reported to cause toxicity on its own, t h r o u g h activation o f the N M D A receptor [21]. H o w -
201 TABLE II PREINCUBATION WITH Zn 2+ FOLLOWED BY WASHING THE NEURONS, OR JUST EXTENSIVE WASHING, DOES NOT ALTER N M D A INDUCED 45Ca INFLUX OR ITS POTENTIATION BY EDTA The percent given in the table in brackets show the inhibition of N M D A induced 4~Ca influx in presence of Zn 2÷. The entire experiment was performed in Locke 25. The results represent 3 4 repeats of the experiment with two different culture batches. For each experimental system a control was run which did not receive NMDA in the final incubation. These controls were 0.184).2 for the left column in the table, 0.16~.21 for the middle column and 0.154).3 for the right column. These controls were subtracted to give the numbers shown in the table. Pretreatment
45Ca influx (nmol/10 min/5 × 105 cells) 300gM NMDA
2 x washed 5 x washed 2 g M Zn 2+, 10 rain, then 2 × washed
0.79+0.14
300gM NMDA + 2/aM Zn 2+
0.19+0.03 (-75%) 0.71+_0.05 0.25+_0.04 (-64%) 0.69+_0.08 0.25+_0.03 (-64%)
300,uM NMDA + 10 g M EDTA 3.7+_0.25 3.1+_0.10 3.0+_0.03
ever, milimolar concentrations of cysteine were required when the experiments were conducted in low bicarbonate medium, as used also in the present work. Under the conditions used here, cysteine had no toxic effect on its own, but it effectively potentiated glutamate toxicity and 45Ca2+ influx. Almost maximal potentiation could be already achieved at 20 gM cysteine (not shown). At such low concentrations it seems reasonable to ascribe its potentiating effect to its well known potency as a metal binding agent [9] and not to its reducing activity. To obtain enhancement of NMDA receptor activity by reduction, the more potent DTT is apparently required at a concentration of 1 mM [1]. The present data are insufficient to conclude that the metal binding site is part of the NMDA receptor itself. However, in agreement with such a supposition is the finding that the chelators not only potentiated glutamate toxicity, but also enhanced the early influx of 45Ca through the NMDA receptor channel. It is therefore possible that the metal regulates physiological functions of the NMDA receptor of cerebellar granule cells. There is no evidence yet for such a feature in other neurons. However, a similar partial blocking of the receptor channel by an ion is suspected for hippocampal neurons [11]. The authors are most grateful to Ms. Aviva Petcho for the preparation of the cerebellar granule cells and assis-
tance in the experiments. This work was supported by a grant from the Deutsche Forschungs Gemeinschaft. 1 Aizenman, E., Lipton, S.A. and Loring, R.H., Selective modulation of NMDA responses by reduction and oxidation, Neuron, 2 (1989) 1257-1263. 2 Balazs, R., Metabolic imbalance and nerve cell damage in the brain, Prog. Brain Res., 73 (1988) 447461. 3 Carmichael, J., De Graft, W.G., Gazdar, A.F., Minna, J.D. and Mitchell, J.B., Evaluation of a tetrazolium based semiautomated colorimetric assay: assessment of chemosensitivity testing, Cancer Res., 47 (1987) 936-942. 4 Christine, C.W. and Choi, D.W., Effect of zinc on NMDA receptormediated channel currents in cortical neurons, Neuroscience, 10 (1990) 108-116. 5 Choi, D.W. and Rothman, S.M., The role of glutamate neurotoxicity in hypoxic-ischemia neuronal death, Annu. Rev. Neurosci., 13 (1990) 171-182. 6 Connor, J.A., Wadman, W.J., Hockberger, P.E. and Wong, R.K.S., Sustained dendritic gradients of Ca 2÷ induced by excitatory amino acids in CA1 hippocampal neurons, Science, 240 (1988) 649~553. 7 Eimerl, S. and Schramm, M., Acute glutamate toxicity and its potentiation by albumin are determined by the Ca z+ concentration, Neurosci. Lett., 130 (1991) 125-127. 8 Eimerl, S. and Schramm, M., Acute glutamate toxicity in cultured cerebellar granule cells: agonist potency, effects of pH, Zn 2+ and the potentiation by serum albumin, Brain Res., 560 (1991) 282-290. 9 Friedman, M., The chemistry and biochemistry of the sulfhydryl group in amino acids, peptides and proteins, Pergamon, Oxford, 1973, p. 26. 10 Garthwaite, G. and Garthwaite, J., Amino acid neurotoxicity: intracellular sites of calcium accumulation associated with the onset of irreversible damage to rat cerebellar neurons in vitro, Neurosci. Lett., 71 (1986) 53-58. 11 Jahr, C.E. and Stevens, C.F., A quantitative description of NMDA receptor-channel kinetic behaviour, J. Neurosci., 10 (1990) 18301837. 12 Johnson, J.W. and Ascher, P., Glycine potentiates the NMDA response in cultured mouse brain neurons, Nature, 325 (1987) 529531. 13 Koh, J. and Choi, D.W., Zinc alters excitatory amino acid neurotoxicity in cortical neurons, J. Neurosci., 8 (1988) 2164-2171. 14 MacDermott, A.B., Mayer, M.L., Westbrook, G.L., Smith, S.J. and Barker, J.L., NMDA-receptor activation increases cytoplasmic calcium concentration in cultured spinal cord neurons, Nature, 321 (1986) 519-522. 15 Madden, P.K., Clark, W.M., Marcoux, F.W., Probert, A.W., Weber, M.L., Rivier, J. and Zivin, J.A., Treatment with conotoxin, an 'N-type' calcium channel blocker, in neuronal hypoxic-ischemic injury, Brain Res., 537 (1990) 256-262. 16 Madison, D.V., Malenka, R.C. and Nicoll, R.A., Mechanisms underlying long-term potentiation of synaptic transmission, Annu. Rev. Neurosci., 14 (1991) 379-397. 17 Manev, H., Favaron, M., Guidotti, A. and Costa, E., Delayed increase of Ca 2+ influx elicited by glutamate: role in neuronal death, Mol. Pharmacol., 36 (1989) 106-112. 18 Mayer, M.L. and Miller, R.J., Excitatory amino acid receptors, second messengers and regulation of intracellular Ca 2÷ in mammalian neurons, Trends Pharmacol. Sci., 11 (1990) 254-260. 19 Meldrum, B. and Garthwaite, J., Excitatory amino acid neurotoxicity and neurodegenerative disease, Trends Pharmacol. Sci., 11 (1990) 379-387.
202 20 Morris, R.G.M., Anderson, E., Lynch, G.S. and Baudry, M., Selective impairment of learning and blockade of long-term potentiation by an N-methyl-D-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. 21 Olney, J.W., Zornmski, C., Price, M.T. and Labruyere, J., c-Cysteine, a bicarbonate-sensitive endogenous excitotoxin, Science, 248 (1990) 596-599. 22 Schramm, M., Eimerl, S. and Costa, E., Serum and depolarizing agents cause acute neurotoxicity in cultured cerebellar granule cells: role of the glutamate receptor responsive to N-methyl-D-aspartate, Proc. Natl. Acad. Sci. U.S.A., 87 (1990) 1193-1197. 23 Wahl, P., Schousboe, A., Honore, T. and Drejer, J., Glutamateinduced increase in intracellular Ca 2+ in cerebral cortex neurons is
transient in immature cells but permanent in mature cells, J. Neurochem., 53 (1989) 1316-1319. 24 Watkins, J.C., Krogsgaard-Larsen, E and Honore, T., Structureactivity relationships in the development of excitatory amino acid receptor agonists and competitive antagonists, Trends Pharmacol. Sci., 11 (1990) 25-33. 25 Westbrook, G.L. and Mayer, M.L., Micromolar concentrations of Zn 2+ antagonize NMDA and GABA responses of hippocampal neurons, Nature, 328 (1987) 640-643. 26 Wroblewski, J.T. and Danysz, W., Modulation of glutamate receptors: molecular mechanisms and functional implications, Annu. Rev. Pharmacol., 29 (1989) 441 -474.
198
NSL 08496
An endogenous metal appears to regulate N M D A receptor mediated 45Ca influx and toxicity in cultured cerebellar granule cells Sara Eimerl a n d Michael S c h r a m m Department of Biological Chemistry and the Otto Loevi Centerfor Neurobiology, The Hebrew University of Jerusalem, Jerusalem (Israel) (Received 6 November 1991; Revised version received 27 December 1991: Accepted 31 December 1991)
Key words." Potentiator of N-methyl-D-aspartate toxicity; Cysteine toxicity; Calcium influx; N-Methyl-o-aspartate receptor Glutamate induced 45Ca influx and toxicity were enhanced 2 10 fold by EDTA. A chelator concentration of 10 ~M, which was equivalent to less than 1% of the Mg 2+ and Ca 2÷ concentration in the medium, was effective. The chelator revealed no activity on its own and caused potentiation only when present simultaneously with the agonist of the N M D A receptor. Cysteine, which is known to bind certain metals tightly through its sulfhydryl group, and another chelator, O-phenanthroline, produced the same effect as EDTA. The findings indicate that when the N-methyl-n-aspartate receptor is activated, an endogenous metal can become bound to a chelator or to a physiological metal binding agent, such as cysteine, leading to enhanced Ca 2+ influx into the neuron and toxicity.
In recent years a considerable amount of information on the N-methyl-D-aspartate (NMDA) receptor in CNS neurons [18, 24, 26] and on Ca 2+ influx through its channel [6, 10, 14, 15, 23] has accumulated. While this Ca 2+ message appears to be involved in plasticity and memory processing [16, 20], excessive influx of C a 2+ is apparently the major cause of glutamate toxicity [2, 5, 10, 15, 17, 19]. Thus, for a variety of reasons, there is interest in factors and processes which modulate Ca 2+ influx through the N M D A receptor. Recently we have shown that a low concentration of serum, and of serum albumin strongly potentiate glutamate and N M D A toxicity in cerebetlar granule cells [7, 8, 22]. During these studies it appeared that metal chelators and cysteine, at low concentration, are also strong potentiators of glutamate and N M D A toxicity and of 45Ca influx. These findings are presented below, including the intriguing observation that EDTA is effective only if present during N M D A receptor activation. Cell culture and glutamate toxicity studies were performed as previously described [8]. Rat cerebellar granule cells, at day 8 postnatal, were cultured 9-18 days in 24-well culture boxes (5 x 105 cells/well) in 0.5 ml basal Eagle medium - 25 mM KCI 10% fetal calf serum. To measure glutamate toxicity o r 45Ca uptake, the culture Correspondence." M. Schramm, Department of Biological Chemistry, Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem 91904, lsrael.
medium was removed and the cells were washed twice with 0.5 ml Locke 25 medium (in mM): NaC1 129, KC1 25, MgC12 1, Ca 2.3, NaHCO3 4, HEPES 10, glucose 5, pH 7.5. Fresh Locke 25 was added to measure glutamate toxicity o r 45Ca2+ influx, both performed at 25°C. Some of the experiments were performed in Locke 5 which contains 5 mM KCI, 150 mM NaCI, without Mg 2+, and the other components as above. Toxicity. N M D A receptor agonist and other additions were made to give a final volume of 0.5 ml Locke. Incubation time was 5 30 min. The medium was then replaced with fresh Locke 25 containing 0.2 ¢tM MK-801 to stop further action of the N M D A receptor channel. After 1 h, cell viability was measured by reduction of a tetrazolium salt [3]. Control systems, containing the same additions without an N M D A receptor agonist, or with the agonist, but in presence of 0.2 ¢tM MK-801, were also run. The control containing Locke without agonist served as a measure of 100% viability (0% cells killed). A survey by phase contrast microscopy [22] showed that this control contained <5% dead cells. The percentage of cells killed was calculated for each experimental system relative to the control without agonist. 45Ca influx. Additions were made to give a final volume of 0.25 ml, including 2 x 105 cpm 45CAC12, which was added just before addition of an agonist. After ~ 3 0 min at 25°C, the cells were rapidly washed 3 times at 25°C with 0.5 ml Locke containing 2 /IM (_+) MK-801. A higher concentration of MK-801 in the 45Ca assay was
199
d "~ 70 I-'0 I,'-
T
It#l-NO
c
L
3.0
r'zr,I - N 0
AGONIST
E i'M
U-
J_
,
°50
d
I
AGONIST
1
-
•
o
+
IZl
Z
I
-o
,~
J 30 co _J
Izl
~
to
to
IO-I
0.
-I-
I+1
+
I -
"
r5
~ r~ 1~ F~ 25mM I
K +,
~
o-
+
~-o
+
~ +
X
0
6
+
z
+
6
EL 2 . 0
0
lmM Mg"~÷ i 2.3raM Ca ÷÷
r5
5raM K÷, no Mg÷+ I
1.0-~
...1
~ s ~ to ~.= g ~ , , .g + + ÷ + +
p-
I + +
z
o (..)
to
+
_.1 LL Z
2 5 m M K + , I mM Mg ++
i
5 m M K + n o M g ++
2.3mM Ca ++ Fig. 1. Toxicityinduced through the NMDA receptor is potentiated by chelators and by cysteine. Incubation was 10 min with 50BM glutamate or 300 BM NMDA, with or without additions, in Locke 25 or Locke 5 as shown. Experimental procedures are described in the text. phenan, phenanthroline; cyst, cysteine, aThe NMDA + EDTA columns in Locke 5 represent the average of 9 experiments using 4 differentculture batches. A fifth culture batch (two repeated experiments), not included in the figure, gave a higher toxicity with NMDA alone, but also a higher potentiation, 41 + 1% and 70 + 5%, respectively,b N o value for chelator without agonist is shown because the value was zero.
used to ensure very fast channel block, which is essential for 1-2 min kinetics. The cells were finally suspended in 0.5 ml 0.1 M N a O H and counted. All 45Ca assays were run on duplicate wells. Experiments were repeated with at least two different batches of cell cultures. The average of all experiments + S.E.M. are presented. Fig. 1 shows that glutamate and N M D A toxicity are potentiated by EDTA, O-phenanthroline and cysteine. These potentiating agents were completely harmless when tested in absence of agonists under the conditions of the experiment (Fig. 1, no agonist). Surprisingly, the potentiation by 10 # M E D T A was obtained in presence of a 230-fold molar excess of Ca 2+ and a 100-fold molar excess of Mg 2+. In fact, even 1 # M E D T A produced some potentiation (not shown). The effect of the chelators is also obtained in Locke 5 without Mg 2÷. In relative terms, the potentiation is seen to be stronger with N M D A as agonist than with glutamate. It is also evident in Fig. 1 that the potentiation was in all instances blocked by MK-801, the specific non-competitive antagonist of the N M D A receptor. The fact that all 3 potentiating agents are known to bind metals effectively and the fact that E D T A is a highly specific metal chelating agent suggests that potentiation is due to binding of an endogenous metal. The low concentration of the chelators and the large excess of Ca 2+ and Mg 2+ further indicates an association constant of the endogenous metal with the chelators which is orders of magnitude higher than that for Ca 2+.
Fig. 2. 45Ca influx mediated through the NMDA receptor channel, is potentiated by chelators and by cysteine. Agonist concentrations were as in Fig. 1. Incubation time was 2 min. The basal influx in absence of agonist was subtracted from all values, but is shown as hatched columns. Abbreviations are as shown in the legend of Fig. 1. *The value is significantly higher than that of NMDA without EDTA, P<0.025, Wilcoxon test. **The value is significantly higher than that of glutamate alone. P<0.005, Wilcoxon test. In order to see whether the effect of the potentiating agents is intimately connected with the function of the N M D A receptor, 45Ca influx induced by glutamate and by N M D A was measured in presence and absence of these agents. Fig. 2 shows that 45Ca influx, induced by either glutamate or N M D A , was potentiated by EDTA, phenanthroline and cysteine. The chelators in absence of agonist did not increase 45Ca influx above the low level, barely detectable, in the controls without the chelators. Similar effects of the chelators on glutamate induced 45Ca2+ influx were seen even when measured as early as 1 min after addition of N M D A (not shown). The pattern of potentiation of 45Ca influx is seen to be quite similar to that shown in Fig. 1 for toxicity. It should, however, be noted that potentiation of 45Ca influx in Locke 5, without Mg 2+, was less than in Locke 25 with Mg 2+, apparently because influx in absence of the potentiating agent was already high. In order to remove the endogenous metal to which the chelators bind, the cells were washed with Locke 25 containing 1 0 0 # M E D T A and the E D T A was then removed by washing with Locke 25. Surprisingly, this pretreatment with E D T A was entirely without effect. E D T A was still required to be added together with glutamate to produce potentiation of toxicity (Table I). Further extensive attempts to remove the metal by prolonged preincubation with EDTA, to render the cells maximally responsive upon subsequent addition of glutamate, all failed. It was therefore thought that the metal site might be masked and only becomes exposed when the N M D A
200 TABLE l ATTEMPTS TO ENHANCE TOXICITY BY PRE- AND POSTTREATMENTS WITH EDTA Unless otherwise noted experiments were performed in Locke 25, containing 1 mM Mg2÷and 2.3 mM CaCI2. Treatments 1
5 i ii iii
Toxicity induction by 300 HM NMDA for 15 min. Removal of the agonist, followed by incubation for 1 h with 100 HM EDTA in Locke 25 medium containing 0.2/,tM MK-801. Ten min preincubation with 100 HM EDTA at 25°C or 37°C, washing the cells twice, followed by toxicity induction by 50HM glutamate, 10 min. As above but 2.5 h preincubation with 100 HM EDTA. Addition of 100/IM EDTA to the culture overnight, prior to the experiment. Further treatment and toxicity induction as in 2. above. Sub-toxicity pretreatments in presence of 100 ,uM EDTA: 5 min 2 HM glutamate, second 5 min another 2 HM glutamate. 10 min 5 HM glutamate in Locke 25 with only 1 mM CaC12. 100 gtM NMDA, 10 min, or 50HM glutamate 2 min in Locke 25 with only 1 mM CaC12.Final toxicity induction after removal of the pretreatment medium: 50/./M glutamate 10 min.
Results I
All above pre- and posttreatments had no effect on the toxicity induced by glutamate or NMDA, when compared with untreated systems. In all the above experiments, irrespective of the treatment, the neurons demonstrated effectivepotentiation of toxicity when replicate samples finally received EDTA together with a toxic concentration of the agonist.
receptor is activated. Accordingly, brief preincubations were performed in presence o f both E D T A and glutamate (Table I). However, after washing the cells to remove the agonist and the chelator the neurons still showed the same sensitivity to readdition o f glutamate and to potentiation by E D T A as cells which had not been pretreated. The conclusions f r o m these experiments are discussed below. It was t h o u g h t that the endogenous metal might be Zn 2+, since we have reported that added Z n 2+ attenuates glutamate induced toxicity in the cerebellar granule cells [8]. Others had shown earlier that Zn 2+ decreases the current t h r o u g h the N M D A receptor channel [4, 25], as well as glutamate toxicity [13]. Experiments were therefore conducted to see whether exogenously added Zn 2+ and the endogenous metal have similar affinities for sites on the neurons (Table II). The results indicate that the
endogenous metal is b o u n d with a much higher affinity than the added Zn 2÷. A l t h o u g h the added Zn 2+ strongly inhibited N M D A induced 45Ca influx, the inhibition was readily abolished by washing the cells twice. In contrast, washing the neurons 5 times appeared not to remove the endogenous metal since the N M D A induced 45Ca2+ influx remained the same as in neurons washed twice. Furthermore, both systems still demonstrated a b o u t the same extent o f potentiation by E D T A . It is thus apparent that the endogenous metal is b o u n d with a much higher affinity than the exogenous, added Zn 2+. The findings indicate that an endogenous metal, which can tightly associate with E D T A , O-phenanthroline and cysteine, attenuates both 45Ca influx t h r o u g h the N M D A receptor channel and glutamate toxicity. While the identity o f the metal remains u n k n o w n , it must have an affinity for the chelators which is several orders o f magnitude higher than that o f Ca 2÷ or M g 2+. This is because the latter are present in the m e d i u m at more than a 100fold m o l a r excess over the chelator. Also, a concentration o f E D T A higher than 10 ¢tM is required to achieve maximal potentiation (not shown). Therefore it is likely that the binding site on the cells has a very high affinity for the metal. On that basis the binding site for added Zn 2+ is not a likely candidate since it appeared to have a relatively low affinity for the metal. The effect o f added Zn 2+ was readily abolished by washing the neurons twice. In contrast, washing did not appear to remove the endogenous metal since it did not increase the N M D A induced 45Ca influx, nor did it diminish the potentiation when E D T A was finally added with N M D A . W h y did preincubation with E D T A fail to remove the endogenous metal? One possibility is that the metal can only be removed by the chelator when the N M D A receptor is in the fully activated state for a prolonged period. Alternatively, it is possible that the metal is indeed removed by the chelator when the receptor is even briefly activated, but that it is replaced within a short time from endogenous stores inside the neuron. The situation could be similar to that o f glycine, the requirement for which is often difficult to demonstrate because o f its continuous leakage from the cells [12]. There is also the possibility that the Locke m e d i u m contains enough o f the metal as a trace impurity to reload the sites after the E D T A is removed by washing the cells with this solution. The metal ion could be Fe 3+, C a 2+, Ni 2+, Pb 2+, C o 2+, Z n 2+, Cd 2+ or A13+. It can be calculated, for example, that a c o n t a m i n a t i o n o f the NaC1 reagent by 0.0003% o f A13+ will supply a b o u t 1/.tM AI 3+. At this level, it will be difficult to identify the metal. The potentiating effect o f cysteine deserves special notice. Cysteine was reported to cause toxicity on its own, t h r o u g h activation o f the N M D A receptor [21]. H o w -
201 TABLE II PREINCUBATION WITH Zn 2+ FOLLOWED BY WASHING THE NEURONS, OR JUST EXTENSIVE WASHING, DOES NOT ALTER N M D A INDUCED 45Ca INFLUX OR ITS POTENTIATION BY EDTA The percent given in the table in brackets show the inhibition of N M D A induced 4~Ca influx in presence of Zn 2÷. The entire experiment was performed in Locke 25. The results represent 3 4 repeats of the experiment with two different culture batches. For each experimental system a control was run which did not receive NMDA in the final incubation. These controls were 0.184).2 for the left column in the table, 0.16~.21 for the middle column and 0.154).3 for the right column. These controls were subtracted to give the numbers shown in the table. Pretreatment
45Ca influx (nmol/10 min/5 × 105 cells) 300gM NMDA
2 x washed 5 x washed 2 g M Zn 2+, 10 rain, then 2 × washed
0.79+0.14
300gM NMDA + 2/aM Zn 2+
0.19+0.03 (-75%) 0.71+_0.05 0.25+_0.04 (-64%) 0.69+_0.08 0.25+_0.03 (-64%)
300,uM NMDA + 10 g M EDTA 3.7+_0.25 3.1+_0.10 3.0+_0.03
ever, milimolar concentrations of cysteine were required when the experiments were conducted in low bicarbonate medium, as used also in the present work. Under the conditions used here, cysteine had no toxic effect on its own, but it effectively potentiated glutamate toxicity and 45Ca2+ influx. Almost maximal potentiation could be already achieved at 20 gM cysteine (not shown). At such low concentrations it seems reasonable to ascribe its potentiating effect to its well known potency as a metal binding agent [9] and not to its reducing activity. To obtain enhancement of NMDA receptor activity by reduction, the more potent DTT is apparently required at a concentration of 1 mM [1]. The present data are insufficient to conclude that the metal binding site is part of the NMDA receptor itself. However, in agreement with such a supposition is the finding that the chelators not only potentiated glutamate toxicity, but also enhanced the early influx of 45Ca through the NMDA receptor channel. It is therefore possible that the metal regulates physiological functions of the NMDA receptor of cerebellar granule cells. There is no evidence yet for such a feature in other neurons. However, a similar partial blocking of the receptor channel by an ion is suspected for hippocampal neurons [11]. The authors are most grateful to Ms. Aviva Petcho for the preparation of the cerebellar granule cells and assis-
tance in the experiments. This work was supported by a grant from the Deutsche Forschungs Gemeinschaft. 1 Aizenman, E., Lipton, S.A. and Loring, R.H., Selective modulation of NMDA responses by reduction and oxidation, Neuron, 2 (1989) 1257-1263. 2 Balazs, R., Metabolic imbalance and nerve cell damage in the brain, Prog. Brain Res., 73 (1988) 447461. 3 Carmichael, J., De Graft, W.G., Gazdar, A.F., Minna, J.D. and Mitchell, J.B., Evaluation of a tetrazolium based semiautomated colorimetric assay: assessment of chemosensitivity testing, Cancer Res., 47 (1987) 936-942. 4 Christine, C.W. and Choi, D.W., Effect of zinc on NMDA receptormediated channel currents in cortical neurons, Neuroscience, 10 (1990) 108-116. 5 Choi, D.W. and Rothman, S.M., The role of glutamate neurotoxicity in hypoxic-ischemia neuronal death, Annu. Rev. Neurosci., 13 (1990) 171-182. 6 Connor, J.A., Wadman, W.J., Hockberger, P.E. and Wong, R.K.S., Sustained dendritic gradients of Ca 2÷ induced by excitatory amino acids in CA1 hippocampal neurons, Science, 240 (1988) 649~553. 7 Eimerl, S. and Schramm, M., Acute glutamate toxicity and its potentiation by albumin are determined by the Ca z+ concentration, Neurosci. Lett., 130 (1991) 125-127. 8 Eimerl, S. and Schramm, M., Acute glutamate toxicity in cultured cerebellar granule cells: agonist potency, effects of pH, Zn 2+ and the potentiation by serum albumin, Brain Res., 560 (1991) 282-290. 9 Friedman, M., The chemistry and biochemistry of the sulfhydryl group in amino acids, peptides and proteins, Pergamon, Oxford, 1973, p. 26. 10 Garthwaite, G. and Garthwaite, J., Amino acid neurotoxicity: intracellular sites of calcium accumulation associated with the onset of irreversible damage to rat cerebellar neurons in vitro, Neurosci. Lett., 71 (1986) 53-58. 11 Jahr, C.E. and Stevens, C.F., A quantitative description of NMDA receptor-channel kinetic behaviour, J. Neurosci., 10 (1990) 18301837. 12 Johnson, J.W. and Ascher, P., Glycine potentiates the NMDA response in cultured mouse brain neurons, Nature, 325 (1987) 529531. 13 Koh, J. and Choi, D.W., Zinc alters excitatory amino acid neurotoxicity in cortical neurons, J. Neurosci., 8 (1988) 2164-2171. 14 MacDermott, A.B., Mayer, M.L., Westbrook, G.L., Smith, S.J. and Barker, J.L., NMDA-receptor activation increases cytoplasmic calcium concentration in cultured spinal cord neurons, Nature, 321 (1986) 519-522. 15 Madden, P.K., Clark, W.M., Marcoux, F.W., Probert, A.W., Weber, M.L., Rivier, J. and Zivin, J.A., Treatment with conotoxin, an 'N-type' calcium channel blocker, in neuronal hypoxic-ischemic injury, Brain Res., 537 (1990) 256-262. 16 Madison, D.V., Malenka, R.C. and Nicoll, R.A., Mechanisms underlying long-term potentiation of synaptic transmission, Annu. Rev. Neurosci., 14 (1991) 379-397. 17 Manev, H., Favaron, M., Guidotti, A. and Costa, E., Delayed increase of Ca 2+ influx elicited by glutamate: role in neuronal death, Mol. Pharmacol., 36 (1989) 106-112. 18 Mayer, M.L. and Miller, R.J., Excitatory amino acid receptors, second messengers and regulation of intracellular Ca 2÷ in mammalian neurons, Trends Pharmacol. Sci., 11 (1990) 254-260. 19 Meldrum, B. and Garthwaite, J., Excitatory amino acid neurotoxicity and neurodegenerative disease, Trends Pharmacol. Sci., 11 (1990) 379-387.
202 20 Morris, R.G.M., Anderson, E., Lynch, G.S. and Baudry, M., Selective impairment of learning and blockade of long-term potentiation by an N-methyl-D-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. 21 Olney, J.W., Zornmski, C., Price, M.T. and Labruyere, J., c-Cysteine, a bicarbonate-sensitive endogenous excitotoxin, Science, 248 (1990) 596-599. 22 Schramm, M., Eimerl, S. and Costa, E., Serum and depolarizing agents cause acute neurotoxicity in cultured cerebellar granule cells: role of the glutamate receptor responsive to N-methyl-D-aspartate, Proc. Natl. Acad. Sci. U.S.A., 87 (1990) 1193-1197. 23 Wahl, P., Schousboe, A., Honore, T. and Drejer, J., Glutamateinduced increase in intracellular Ca 2+ in cerebral cortex neurons is
transient in immature cells but permanent in mature cells, J. Neurochem., 53 (1989) 1316-1319. 24 Watkins, J.C., Krogsgaard-Larsen, E and Honore, T., Structureactivity relationships in the development of excitatory amino acid receptor agonists and competitive antagonists, Trends Pharmacol. Sci., 11 (1990) 25-33. 25 Westbrook, G.L. and Mayer, M.L., Micromolar concentrations of Zn 2+ antagonize NMDA and GABA responses of hippocampal neurons, Nature, 328 (1987) 640-643. 26 Wroblewski, J.T. and Danysz, W., Modulation of glutamate receptors: molecular mechanisms and functional implications, Annu. Rev. Pharmacol., 29 (1989) 441 -474.