Zinc toxicity on cultured cortical neurons: Involvement of N-methyl-d -aspartate receptors

Pergamon

0306-4522(94)E0005-0

NeuroscienceVol. 60, No. 4, pp. 104%1057,1994 ElsevierScienceLtd Copyright 0 1994IBRO Printed in Great Britain. All rights reserved 0306-4522/94$7.00+ 0.00

ZINC TOXICITY ON CULTURED CORTICAL NEURONS: INVOLVEMENT OF N-METHYL-D-ASPARTATE RECEPTORS J.-Y. KOH and D. W. CHOI* Department of Neurology and Center for the Study of Nervous System Injury, Box 8111, Washington University School of Medicine, 660 S. Euclid Ave.. St Louis, MO 63110, U.S.A. Abstract-Neuronal injury induced by the excessive release of endogenous Zn2+ at central glutamatergic synapses may contribute to the pathogenesis of epileptic brain damage. We explored the possibility that N-methyl-D-aspartate receptors might be involved in Zn2+ neurotoxicity. Exposure of murine cortical cell cultures to 300-1000 pM concentrations of Zn2+ for 15 min resulted in widespread neuronal degeneration, accompanied by the release of lactate dehydrogenase to the bathing medium. Both non-competitive and competitive N-methyl-D-aspartate antagonists attenuated this degeneration. However, the participation of N-methyl-D-aspartate receptors in Zn2+ neurotoxicity was atypical. Removal of extracellular Ca2+ attenuated N-methyl-D-aspartate neurotoxicity but potentiated Zn2+ neurotoxicity, whereas increasing extracellular Ca2+ potentiated N-methyl-D-aspartate neurotoxicity but attenuated Zn2+ neurotoxicity. Furthermore, the nature of the antagonism of Zn2+ neurotoxicity induced by N-methyl-D-aspartate antagonists was qualitatively different from that seen with other N-methyl-D-aspartate receptor-mediated events. The block of Zn2+ neurotoxicity induced by the non-competitive N-methyl-D-aspartate antagonist MK-801 was better overcome by increasing Zn2+ concentration than the block induced by the competitive antagonists o-aminophosphonovaierate and CGS-19755. We hypothesize that N-methyl-o-aspartate receptor-gated channels contribute to Zn2+ toxicity by providing a route of Zn2+ influx into neurons. Consistent with this idea, intracellular Zn2+ visualized by rose during Zn2+ exthe fluorescent Zn2+ chelator, N-(6-methoxy-&quinolyl)-p-toluenesulfonamide, posure; this rise was increased by N-methyl-o-aspartatd and reduced by either N-methyl-D-aspartate antagonists or high Ca2+. Since Zn2+ affects activity of many neuronal proteins, intracellular Zn2+ overload may produce lethal disturbances in neuronal cell homeostasis.

The mammalian brain contains a substantial quantity of chelatable zinc in certain vesicles of excitatory synaptic boutons.‘4,‘6 Growing evidence suggests that Ca*+-dependent Zn*+ release into the extracellular space accompanies neuronal activity, and may build up to concentrations of several hundred micromoles.3~5~2’~27 The function of synaptic Zn2+ has not been established, although modulation of glutamatergic neurotransmission seems one possibility. Zn*+ blocks neuroexcitation mediated by N-methyl-o-aspartate (NMDA) receptors, while potentiating that mediated by ~~-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptors.28.33 Attenuation of GABA receptor-mediated currents has also been reported.33 Thus, the amount of Zn2+ co-released with gluta-

*To whom correspondence should be addressed. AMPA, cc-amino-3-hydroxy-5-methyl-4-isoxazolepropionate; APB, aminophdsphonobutyrate; APV, aminophosphonovalerate: 0X-19755. cis4phosphonomethyl-i-piperidine carboxylic acid; HEPES, N-2-hydroxy-ethyl-piperazine-N’-2-ethanesulfonic acid; LDH, lactate dehydrogenase; MK-801, ( + )-5methyl-lo,1 I-dihydro-5H-dibenzo[a,d]cyclo-hephen5, IO-iminemaleate: NMDA, N-methyl-o-aspartate; TSQ, N-(6-methoxy-8-quinolyl)-p-toluenesulfonamide.

Abbreviations:

NSC 60/L,

mate could influence dynamically the activation of receptor subtypes, shifting activation away from NMDA receptors and towards AMPA receptors. Antagonism of NMDA receptors may reflect two effects: at low micromolar concentrations, ZnZ+ produces a voltage-independent reduction in NMDA receptor-gated channel opening frequency, whereas at higher concentrations it may also produce a voltage-dependent fast flicker block of the NMDA receptor-gated channel.‘* Consistent with these effects on glutamate receptor-mediated currents, Zn2+ also attenuates NMDA toxicity.23 In addition to modulating membrane currents and toxicity induced by glutamate agonists, intense exposure to Zn 2+ alone is also neurotoxic. Exposure to 600 PM Zn2+ for 15 min was sufficient to destroy large numbers of cultured cortical neurons.10,35 This direct Zn2+ toxicity could contribute to neuronal injury in pathological conditions associated with intense activation of excitatory pathways, for example prolonged seizures. Loss of Timm’s staining for chelatable Zn*+ has been found in hippocampal mossy fibers after seizure induced by perforant path stimulation.29 We have previously reported that depolarizing agents such as AMPA, kainate or high K+ all potentiated Zn‘2+-induced death of cultured cortical

1049

J.-Y. KOH and K. W. CH~I

1050

neurons.32 Because this potentiation was associated with enhanced neuronal intracellular Zn’+ concentrations as determined by N-(6-methoxy-&quinolyl)p -toluenesulfonamide (TSQ) fluorescence, and attenuated by the addition of Co” after Zn” washout, we hypothesize that Zn*’ toxicity could be mediated by Zn2+ entry through voltaged-gated Ca2+ channels. The purpose of the present study was to examine the related possibility that NMDA receptorgated channels might provide another route for excess Zn” entry into neurons, leading to neuronal death. Abstracts have appeared.“,‘J EXPERIMENTAL PROCEDURES

Cortical cell cultures

Mixed cortical cell cultures, containing both neuronal and glial elements, were prepared as described previously’ from fetal Swiss Webster mice (Simonsen Laboratory) at 14-17 days gestation. Dissociated cortical cells were plated in Primaria (Falcon) 15mm multiwell vessels (0.4 x 10” cells/well’) in Eagle’s minima1 essential medium (Earle’s salts) supplemented with 10% heat-inactivated horse serum, 10% fetal bovine serum, glutamine (2 mM) and glucose (l&20 mM). Cultures were kept at 37°C in a humidified CO,-containing incubator (pH 7.1-7.4). After five to 10 days in vitro, non-neuronal cell division was halted by one to three days of exposure to IOms M cytosine arabinoside, and the cells were shifted into a maintenance medium identical to the plating medium, but lacking fetal serum. Subsequent media replacement was carried out twice per week. Only mature (1424 days in vitro) cortical cultures were selected for study; all comparisons were made whenever possible on sister cultures derived from a single plating. Cortical glial cell cultures were prepared using the same protocol as above, but using cortices removed from early postnatal (postnatal days ll3) mice instead of fetal mice. since neurons removed from older animals do not survive the plating period.4 Exposure

Visualization

of neuronnl irtjq

Overall cell injury was estimated in all experiments by examination of cultures with phase-contrast microscopy at x 100 to x400. This examination was usually performed one day after Zn2 + exposure, at which point the process of cell death was largely complete. In some experiments, this examination was verified by subsequent bright-field examination of Trypan Blue staining (0.4% for 5 min), a dye staining debris and non-viable cells. In most experiments, overall neuronal cell injury was also quantitatively assessed by the measurement of lactate dehydrogenase (LDH), released by damaged or destroyed cells, in the extracellular

of Zn”

uptukc

Intracellular Zn’+ accumulation after brief Zn2’ exposure was visualized by a specific fluorescent Zn2+ chelator, under the fluorescent microscope.‘7.2’ Cultures were exposed for l-2 min to 50-IOOpM Zn2+ in exposure solution lacking Ca*+. Cultures were then washed with HEPES buffer, 7.5 ~tl of TSQ solution (O.S%, in ethanol) were added (final concentration about 0.01%) and mixed by gentle swirling. After 5 min, cultures were examined under fluorescence microscopy with a UV filter (excitation 355-375 nm, barrier 520 nm) and photographed. TSQ--Zn2+ emits at 495 nm; addition of high concentrations of Ca2+ or Mg’+ has been reported to produce no interference with this Zn*+ signal in tissue sections.” Drugs und chemical.~ Zn” was added as ZnCl, (Baker, reagent grade). TSQ was obtained from Molecular Probes. MK-801, dextrorphan and CCS-19755 were kind gifts of Merck, Sharp & Dohme, Hoffman-La Roche and C&a-Geigy, respectively. RESULTS

As reported

tures

previously, exposure of cortical culto 300-1000 PM concentrations of Zn2+ for

to zinc

Exposure to Zn’+ was via the bathing medium, utilizing defined solutions lacking serum, glutamate or lactate dehydrogenase. Care was taken to wash out the normal medium from cultures prior to addition of Zn2+ exposure solutions. Exposure to Zn*+ was carried out in room air, using a Tris-buffered salt solution with the following standard composition (in mM): NaCl 120, KCI 5.4, MgCl, 0.8, CaCl, 1.8. Tris-HCl (pH 7.6 at 25°C) 25, glucose 15. In some indicated experiments, Ca2+ concentration was changed (O-20 mM) or both Na+ and Ca2+ were removed and Nat was replaced with isomolar choline ( - Na/ - Ca). After 15 or 30 mitt, the exposure solution was washed out thoroughly and replaced with Eagle’s minimal essential medium plus glucose (25 mM), prior to returning the dishes to the 37°C incubator. Little or no cortical cell damage was produced by this protocol if Zn2+ was omitted. Assessment

fluid one day after exposure IO Zn*‘. A small amount of LDH was always present in the media of cultures carried through the exposure protocol but without addition of Zn’+. This background amount, determined on sister cultures within each experiment, was subtracted from values obtained in treated cultures. The absolute value of the LDH efflux produced by ZnL ’ exposure was quite consistent within sister cultures of a single plating, but differed somewhat between platings. largely as a function of neuronal density (which varied despite constant original plating densities, presumably reflecting small variations in cell preparation or serum characteristics). Therefore, each observed LDH value was scaled to a mean value produced by exposure to I mM Zn’ f in other sister cultures, in which condition near complete neuronal death but little glial damage usually occurs.‘”

DOSE

RESPONSE

OF Zn TOXICIlY

(15

MN)

200 ,,,j

0 -3.5

-3 LOG

MIXED

-2.5

[Zn]

concentration-toxicity relationship. Mean Fig. I. Zn” LDH values (& S.E.M., n = 4 cultures per point) present in the bathing medium one day after I5 min exposure of either mixed neuronal and glial cultures, or pure glial cultures, to the indicated concentrations of Zn*+. LDH values were scaled to the mean value in mixed cultures exposed to 1 mM Zn2+ (= IOO), a condition producing near complete neuronal death with little glial death. The background LDH value found in cultures exposed to sham wash alone was subtracted from all values to yield the signal specifically associated with Zn*+ exposure. Mixed cultures show biphasic LDH release, reflecting the occurrence of neuronal injury prior to glial injury.

1051

Zinc toxicity and NMDA receptors EFFZCT OF NON-COYPETITkE

NMM

ANTAGONISTS

ON

Table

Zn TOXICIM

is effective only when added during the zinc exposure

I. Aminophosphonovalerate

% LDH release, mean + S.E.M. (n)

Conditions

100 + 11 (4)

Zn exposure only (1 mM) + APV (1 mM) during the Zn exposure + APV (1 mM) after the Zn exposure

29 5 7 (4)* lOOk 17(4)

*Difference from Zn exposure only in sister cultures (P < 0.01, two-tailed r-test with Bonferroni correction for two comparisons). 01

-6

-7

-5

-4

-3

-2

-1

LOG [ANTAGONIST]

Fig. 2. Non-competitive NMDA antagonists attenuate Zn2+ neurotoxicity. Cultures were exposed for 15 min to 1 mM Zn*+ in the presence of indicated concentrations of MK801, phencyclidine (PCP) or dextrorphan (DX), and LDH (mean f S.E.M., n = 4 cultures per point) was determined the next day. LDH values were scaled to the mean value in sister cultures exposed to 1 mM Zn*+ alone (= 100). (Inset) Sister cultures were exposed for 15 min to 1 mM Zn*+ (control, CTRL). MK-801 (1OOpM) was applied during (“ACUTE”) or immediately after (“LATE”) the exposure to Zn*+. Only acute addition of MK-801 attenuated Zn2+ neurotoxicity. Asterisk denotes difference from control at P i 0.05 (two-tailed f-test with Bonferroni correction). 15 min resulted in a concentration-dependent degeneration of neurons, but not glia, by the following day.” Neuronal damage was quantitatively estimated by measuring LDH activity in the bathing medium, released by injured neurons’* (Fig. 1), which correlated well with the data previously obtained by cell counts.iO Exposure to higher Zn*+ concentration (150&3000 PM) additionally resulted in glial damage. Pure glial cultures were injured by the same levels of Zn2+ exposure as glial cells in mixed cultures (Fig. 1). When the non-competitive NMDA antagonists dextrorphan,8,‘3 phencyclidime’ and MK-80134 were added during Zn2+ exposure, resultant neurotoxicity was attenuated in a concentration-dependent fashion,

EFFECT

OF COMPEIITIVE

NMbA

ON Zn TOXICITY

ANTAGONISTS

with approximate ICY values of 200, 30 and 5 PM, respectively (Fig. 2). At the highest concentrations, each antagonist reduced Zn*+ toxicity by 6@90%. In contrast to its protective action when added during Zn*+ exposure, even 100pM MK-801 added after termination of exposure had little protective effect (Fig. 2, inset). The competitive NMDA antagonists, D-Zamino-Sphosphonovalerate (D-APV) and CGS-19755, also produced concentration-dependent reduction in Zn*+ neurotoxicity, with lcso values of about 500 and 200 PM, respectively (Fig. 3). As with MK-801, the late addition of 1 mM D-APV after exposure termination did not reduce injury (Table 1). We considered the possibility that the protective effect of D-APV was due to its negatively charged phosphonate group lowering Zn*+ activity in solution. However, two compounds with similar phosphonate moieties but little activity at the NMDA receptor, L-APV and DL-aminophosphonobutyrate (APB),26,30 did not attenuate Zn*+ neurotoxicity (Table 2). The above experiments, with late addition of MK801 or D-APV, did not reveal an important injury contribution from NMDA receptor stimulation after Zn*+ exposure termination. To test the possibility that traditional NMDA neurotoxicity contributed to Zn*+-induced neuronal death, we examined the dependence of Zn2+ toxicity on extracellular Ca*+. When Ca2+ was taken out, neurotoxicity produced by 15 min exposure to 500 PM NMDA was markedly reduced,6 but the submaximal neurotoxicity produced by 15 min exposure to 500 PM Zn*+ was potentiated (Table 3). Conversely, increasing the concentration of extracellular Ca*+ to 20 mM potentiated submaximal NMDA neurotoxicity while attenuating Zn*+ neurotoxicity. Table 2. Selectivity of o-aminophosphono: valerate

0. -5

-4

-3

LOG [ANTAGONIST]

Fig. 3. Competitive NMDA antagonists attenuate Zn2+ neurotoxicity. Mean LDH (k S.E.M., n = 4) released the next day after 15 min exposure to 1 mM Zn2+ in the presence of indicated concentrations of D-APV or CGS19755 is shown. LDH values were scaled to the mean value in sister cultures exposed to 1 mM Zn2+ alone (= 100).

% LDH release mean f S.E.M. (n)

Conditions Zn only (1 + I mM + 1 mM + 1 mM

mM) D-APV L-APV DL-APB

100 k 46 k 82 f 86 f

11 (8) 25 (8); 18 (8) 28 (8)

*Difference from Zn only in sister cultures (P < 0.05, two-tailed f-test with Bonferroni correction for three comparisons).

1052 Table 3. Opposite

J.-Y. KOH and K. W. CHOI effects of extracellular and Zn2+ neurotoxicity % LDH

release,

Ca*+ on NMDA

mean k S.E.M.

1.8 mM Ca2+ 20 mM Ca*+ NMDA, 50fiM NMDA, 500pM Zn, 1 mM Zn, 5OOpM

51 100 100 21

+ 8 (4) i 19 (4) i 30 (4) *3(4)

0 Ca’+

106 & 39 (4)* N.D. N.D. 25 + 3 (4)* N.D. 52 + 9 (4)* N.D. 41 + 6 (4)*

potentiates zinc neurotoxicof Na+ and Ca2+ % LDH release, mean f S.E.M. (n)

Conditions

NMDA, 500 PM (control

solution) NMDA, 5OOpM (-Na/-Ca) Zn only, 300 FM (-Na/-Ca) Zn + 1 mM NMDA (-Na/-Ca) Zn + NMDA + 30 pM MK-801 (-Na/-Ca)

-

100

-

80

-

(n) : 9 $

60.

4

40. 20

*Difference from respective LDH release in I .8 mM Ca*+ in sister cultures (P < 0.05, two-tailed t-test). N.D., not done. Exposure to 20mM Ca*+ alone was not toxic. Table 4. N-methyl-D-aspartate ity in the absence

120

100+12(4) 8 + 20 (4)* 42 k 9 (4) 106 + 9 (4)* 7 * (14)*t

*Difference from NMDA (control solution) or Zn only, respectively (P < 0.05, two-tailed t-test). tDifference from Zn + NMDA (P < 0.05, two-tailed t-test).

These observations suggested that the activation of NMDA receptors contributed to Zn’+ neurotoxicity, but that this contribution reflected something other than excess Ca*+ entry though the NMDA receptorgated channel. Furthermore, addition of NMDA in the absence of extracellular Na’ (replaced by equimolar choline) and Ca*+ -not toxic by itself6-still enhanced Zn*+ toxicity, and this potentiated toxicity was blocked by MK-801 (Table 4). Additional evidence for an atypical contribution of NMDA receptors to Zn*+ neurotoxicity was the nature of the attenuation produced by NMDA antagonists. The reduction in Zn*+ neurotoxicity produced by 3 PM MK-801 could be overcome by increasing Zn2+ concentration, consistent with competitive antagonism. On the other hand, the reductions produced by 200pM CGS-19755 or 500pM D-APV could not be overcome by increasing Zn*+ concentration, consistent with non-competitive antagonism (Fig. 4). The protective effects of NMDA antagonists against Zn*+ toxicity were specific to neurons. Extending the duration of exposure to 1 mM Zn*+ from 15 to 30 min resulted in widespread glial degeneration, in addition to neuronal degeneration. Addition of 100 PM MK-801 or 1 mM D-APV in the exposure

o--o .-. b----b .- -.

Zn ONLY +UK-801 +D-AFV fCOS19755

P

0 -3.5

-3 LOG [Zn]

Fig. 4. Effect of NMDA antagonists on the ZnZi concentration-toxicity relationship. Cultures were exposed for 15 min to the indicated concentrations of Zn’+, either alone (open circles) or in the presence of 3 PM MK-801 (Wed circles), 500 PM D-APV (open triangles) or 200pM CGS19755 (filled triangles, broken line). All LDH values were scaled to the mean value found in sister cultures exposed to I mM ZnZ+ alone (= 100). Gliotoxicity was not observed in the conditions tested here.

solution substantially attenuated neuronal injury by both morphological criteria with Trypan Blue staining and LDH efflux assay, but had little protective effect on glial injury, either in mixed or pure glial cultures (Table 5). A possible basis for involvement of NMDA receptors in Zn2+ neurotoxicity would be Zn*+ permeation into neurons through NMDA receptor-gated channels. Addition of the fluorescent selective Zn*+ chelator, TSQ, ” immediately after exposure to 100 PM Zn’+ in Ca*+-free solution for 2mir1, suggested a marked increase in intracellular free Zn*+ in neuronal cell bodies and processes, but not in glial cells (Fig. 5). No increase in TSQ fluorescence was seen if Zn*+ was omitted. Addition of 30pM MK-801 or 1 mM the neuronal Zn2+ D-APV with Zn*+ attenuated flubrescence (Fig. 5; three of three experiments). Addition of 500 PM NMDA increased the minimal fluorescent signal associated with exposure to 50 PM Zn*+ for 1.5 min (Fig. 6; three of three experiments). Consistent with its ability to attenuate Zn2+ neurotoxicity, increasing Ca*+ in the exposure solution to 10 mM also attenuated neuronal Zn*’ fluorescence (Fig. 7). DISCUSSION

The central finding of neurotoxicity produced concentrations of Zn2+ ated by several selective

the present study is by brief exposure can be substantially antagonists of the

Table 5. Lactate dehydrogenase release (u/ml), mean + S.E.M. (n), after 30 min antagonists exposure to 1 mM Zn’+ with and without N-methyl-D-aspartate Zn only Mixed Glial

1

1202 & 55 (6) 313 + 18(4)

+I mM APV 410 f 15 (6)* 264 k 16 (4)t

Zn only 2 791 f 29 (4) 304 * 13 (4)

+30/1M

MK-801

388 & 14 (4)* 319+22(4)t

*Difference from Zn only in sister cultures (P < 0.01, two-tailed r-test). tNo difference from Zn only in sister cultures (P > 0.2, two-tailed r-test).

that the to high attenuNMDA

Zinc toxicity and NMDA receptors subcla ss of glutamate receptors. However, four observatl ions suggest that the basis for a dependence of Zn ‘+ neurotuxicity on NMDA receptor activation nnay not be the usual occurrence of NMDA or-mediated injury following acute CNS inreceptm sults. Firs :t, NMDA antagonists were only effective if added during Zn*+ exposure. Following comparable wideq Iread neuronal injury induced by brief exposure

1053

to glutamate*O or combined oxygen and glucose deprivation,‘9 NMDA antagonists retain partial protective efficacy, even when added after the insult, consistent with ongoing “late” neurotoxic stimulation of NMDA receptors by endogenous agonists released from damaged neurons. The absence of protective effect seen with NMDA antagonists after brief Zn*+ exposure does not exclude late NMDA receptor-mediated injury, but it indicates that such

Fig. 5. NMDA antagonists attenuate Zn*+ uptake visualized by TSQ. Sister cultures were exposed for 2min to 1OOpM Zn2+ in Ca2+-free solution alone (B), in the presence of 1mM D-APV (C) or in the presence of 30 PM MK-801 (D). Phase-contrast photomicrographs of identified fields were taken before Zn2+ exposure (left cofumn). Immediately after Zn2+ exposure and TSQ application, the same fields were relocated and photographed (right column). A shows that exposure of cells to the exposure solution without Zn*+ does not induce TSQ fluorescence. Scale bar = 2OOhm.

1054

J.-Y. KOH and K. W. CHOI

injury is not required for Zr?+-induced neuronal death. Second, changes in extracellular Ca’+ concentration had the opposite effect on NMDA neurotoxicity and Zr?’ neurotoxicity; although Cal-+ enhanced the former, it attenuated the latter. This opposite dependence argues against the idea that Ca’+ influx through NMDA receptor-gated channels contributes importantly to Zn*+ neurotoxicity. Rather, Ca* + appeared to interfere with Zn2 + toxicity. Third, under conditions where NMDA neurotoxicity is eliminated (in the absence of extracellular Na+ and Ca2+), the addition of exogenous NMDA still increased Zn2+ neurotoxicity. This observation suggests that the ability of NMDA receptor activation to facilitate Zn*+ neurotoxicity can be uncoupled from NMDA receptor-mediated neuronal damage. Finally, the nature of the antagonism of Zn’+ neurotoxicity induced by NMDA antagonists was

qualitatively different from that seen with other NMDA receptor-mediated events. The antagonism of Zr?+ neurotoxicity by the non-com~tit~ve NMDA antagonist MK-801 could be overcome by increasing Zn*+ concentration, whereas antagonism by the competitive antagonists II-APV and CGS 1975.5 could not be overcome by increasing Zn*+. To explain these observations, we propose that NMDA receptor activation potentiates Zn*+ neurotoxicity by facilitating the permeation of Zn2’ into neurons through NMDA receptor-gated channels. The protective effect of Ca2+ or MK-801 might be amounted for by interference with such Zn2 ’ permeation; the competitive nature of MK-801 block would be consistent with an ability of Zn2+ to override this interference. In contrast, Zn2+ could not open NMDA receptor-gated channels closed by the glutamate binding site antagonists, D-APV or CGS19755.

Addition of 500 FM NMDA to tne LII- exposuresorution increased TSQ fluorescence. (C) NMDA alone ‘500 PM) did not induce TSQ fluorescence. S# * ho* - mn a*-

Zinc toxicity

(lower row) photomicrographs extracellular

1055

receptors

(upper row) and corresponding TSQ fluorescence exposed to 1OOpM Zn2+ for 2min, either with Ca2+ = 10 mM (B). Ca2+ (10 mM) blocked Ca2+ = 0 mM (no added Ca *+., A) or extracellular Zn*+ fluorescence almost completely. Scale bar = 200 pm.

Fig. 7. Ca*+ dependence

of Zn2+ uptake.

and NMDA

Phase-contrast

taken in cultures

Support for this hypothesis is provided by the finding that Zn*+ exposure was associated with an increase in neuronal fluorescence from the Zn*+ chelator, TSQ, and that TSQ fluorescence was affected appropriately by NMDA, Ca*+ or NMDA antagonists. Further studies will be required to determine why elevated intracellular Zn*+ might kill neurons. Since Zn*+ is a highly reactive ion with diverse effects on various protein functions,31 excess intracellular Zn*+ concentrations could have myriad deleterious effects on neuronal metabolism. Single channel recording suggests that Zn*+ may produce a flicker block by moving in and out of the NMDA receptor-gated channel.‘* As discussed by Ascher and Nowak* and Mayer and Westbrook, several divalent cations may permeate through the NMDA channel at rates determined by the ease with which surrounding water molecules can be shed. Even Mg*+ can probably permeate,* and Zn*+ can shed its water molecules faster than Mg*+, although slower than Ca2+.15

Thus, the present study complements our previously published study suggesting that Zn*+ can also induce neuronal death by permeating through voltage-gated Ca*+ channels.32 Under conditions where NMDA receptors were highly activated, but excitotoxicity was blocked by removal of extracellular Ca2+ and Na+, brief (15 min) exposure to 300 PM Zn*+ produced cortical neuronal death as extensive as that produced by exposure to 1 mM Zn*+ under control conditions. Further study will be required to determine the relative importance of voltage-gated Ca*+ channels, NMDA receptor-gated channels or other routes in mediating toxic Zn*+ entry under specific conditions. CONCLUSIONS

There are two major implications of the present study. First, it suggests a novel mechanism by which NMDA antagonists might reduce neuronal injury in pathological conditions involving excessive firing or

1056

J.-Y. KOH and K. W. CH~I

prolonged depolarization of glutamatergic nerve terminals. NMDA antagonists might not only be useful in reducing glutamate neurotoxicity, but might also have a beneficial effect in reducing Zn2+ neurotoxicity. Second, as discussed relative to the possibility of

ZnZ+ permeation though voltage-gated Ca2+ channels,32 the present observations provide another mechanism by which presynaptic terminal Zn2+ might enter postsynaptic neurons under pathologiCal’* or even physiological conditions.

REFERENCES 1. Anis N. A., Berry S. C., Burton N. R. and Lodge D. (1983) The dissociative anaesthetics, ketamine and phencyclidine, selectively reduce excitation of central mammalian neurons by N-methyl-D-aspartate. Er. .I. Pharmac. 79, 565-575. 2. Ascher P. and Nowak L. (1987) The role of divalent cations in the N-methyl+aspartate responses of mouse central neurons in culture. J. Physiol. 339, 207-226. 3. Assaf Y. and Chung S. H. (1984) Release of endogenous Zn 2+ from brain during activity. Nature 308, 734-736. 4. Booher J. and Sensenbrenner M. (1972) Growth and cultivation of dissociated neurons from embryonic chick. rat and human in flask cultures. Neurobiology 2, 97-105. 5. Charton G., Rovira C., Ben-Ari Y. and Leviel V. (1985) Spontaneous and evoked release of endogenous Zn” in the hippocampal mossy fiber zone of the rat in situ. Expl. Bruin Res. 58, 202-205. 6. Cl&i D. %. (1987) Ionic dependence of glutamate neurotoxicity in cortical cell culture. J. Neurosci. 7, 369-379. 7. Choi D. W., Maulucci-Gedde M. A. and Kriegstein A. R. (1987) Glutamate neurotoxicitv in cortical cell culture. J. Neurosci. ‘7, 357-368. 8. Choi D. W., Peters S. and Viseskul V. (1987) Dextrorphan and levorphanol selectively block N-methyl-D-aspartate receptor-mediated neurotoxicity on cortical neurons. J. Pharmac. exp. Ther. 242, 713-7.19. 9. Choi D. W., Koh J. and Peters S. (1988) Pharmacology of glutamate neurotoxicity in cortical cell culture: attenuation by NMDA antagonsits. J. Neurosci. 8, 185-196. IO. Choi D. W., Yokoyama M. and Koh J. (1988) Zinc neurotoxicity in cortical cell culture. Neuroscience 24, 67.-79. Il. Choi D. W. and Koh J. (1988) Zinc central neurotoxicity may require open NMDA channels. Sot. Neurosci. Abstr. 14, 417. 12. Christine C. W. and Choi D. W. (1990) Effect of zinc on NMDA receptor-mediated channel currents in cortical neurons. J. Neurosci. 10, 108-116. 13. Church J., Lodge D. and Berry S. C. (1985) Differential effects of dextrorphan and levorphanol on the excitation of rat spinal neurons by amino acids. Eur. J. Pharmac. 111, 185-190. 14. Dansher G., Howell G., Perez-Clause11 J. and Hertel N. (1985) The dithizone, Timm’s sulphide silver and the selenium methods demonstrate a chelatable pool of zinc in the CNS. Hisrochemisfry 83, 419422. 15. Diebler H.. Eieen M.. Ilnenfritz G.. Maass G. and Winkler R. (1969) Kinetics and mechanisms of reactions of main group metal i&s with biological carriers. Pure appl. Chem. 20,‘93-il5. 16. Frederickson C. J., Klitenick M. A., Manton W. I. and Kirkpatrick J. B. (1983) Cytoarchitectonic distribution of zinc in the hippocampus of man and the rat. Brain Res. 246, 338-341. 17. Frederickson C. J., Kasarkis E. J., Ringo D. and Frederickson R. E. (1987) A quinoline fluorescence method for visualization and assaying the histochemically reactive zinc (bouton zinc) in the brain. J. Neurosci. Meth. 20, 91 103. 18. Frederickson C. J., Hernandez M. D. and McGinty J. F. (1989) Translocation of zinc may contribute to seizure-induced death of neurons. Bruin Res. 480, 317 -321. 19. Goldberg M. P. and Choi D. W. (1993) Combined oxygen and glucose deprivation in cortical cell culture: calcium-dependent and calcium-independent mechanisms of neuronal injury. J. Neurosci. 13, 3510 -3524. 20. Hartley D. M. and Choi D. W. (1989) Delayed rescue of N-methyl-D-aspartate receptor-mediated neuronal injury in cortical culture. J. Pharmac. e.xp. Ther. 250, 752.~758. 21. Howell G. A., Welch M. G. and Frederickson C. J. (1984) Stimulation-induced uptake and release of zinc in hippocampal slices. Nature 308, 736-738. 22. Koh J. and Choi D. W. (1987) Ouantitative determination of glutamate mediated cortical neuronal injury in cell culture by lactate dehydrogenase et&x assay. J. Neurosci. Meth. 20, 83-90. 23. Koh J. and Choi D. W. (1988) Zinc alters excitatory amino acid neurotoxicity on cortical neurons. J. Neurosci. 8, 2164~2171. 24. Koh J. and Choi D. W. (1989) Zinc may enter cortical neurons through NMDA channels. So<. Neurosci. Ahslr. 15, 534. 25. Mayer M. L. and Westbrook G. L. (1987) Permeation and block of N-methyl-D-aspartate acid receptor channels by divalent cations in mouse cultured central neurons. J. Physiol. 394, 501-527. 26. Monaghan D. T. and Cotman C. W. (1986) Identification and properties of N-methyl-o-aspartate receptors in rat brain synaptic plasma membranes. Proc. natn. Acad. Sci. U.S.A. 83, 7572-7536. 27. Perez-Clause11 J. and Dansher G. (1985) Intravesicular localization of zinc in rat telencephalic boutons. A histochemical study. Bruin Res. 337, 150-153. 28. Peters S.. Koh J. and Choi D. W. (1987) Zinc selectively blocks the action of N-methyl-D-aspartate on cortical neurons. Science 236, 589-593. 29. Sloviter R. S. (1985) A selective loss of hippocampal mossy fiber Timm stain accompanies granule cell seizure activity induced by peiforaht path stimulation. i&in Res. 330, 1% 153. 30. Stone T. W., Perkins M. N., Collins J. F. and Curry K. (1981) Activity of the enantiomers of 2-amino-phosphonovaleric acid (2-APV) as stereospecific antagonists of excitatory amino acids. Neuroscience 6, 2249.-2252. 31. Vallee B. L. (1959) Biochemistry, physiology, and pathology of zinc. Physiol. Rec. 39, 443490. 32. Weiss J. H., Hartley D. M., Koh J. and Choi D. W. (1993) AMPA receptor activation potentiates zinc neurotoxicity. Neuron IO. 4349.

Zinc toxicity and NMDA receptors

1057

33. Westbrook G. L. and Mayer M. L. (1987) Micromolar concentrations of Zn2+ antagonize NMDA and GABA responses of hippocampal neurons. Nature 328, 640643. 34. Wong E. H. F., Kemp J. A., Priestley T., Knight A. R., Woodruff G. N. and Iversen L. L. (1986) The anticonvulsant MK-801 is a potent N-methyl-o-aspartate antagonist. Proc. natn. Acad. Sci. U.S.A. 83, 7104-7108. 35. Yokoyama M., Koh J. and Choi D. W. (1986) Brief exposure to zinc is toxic to cortical neurons. Neurosci. Len. 71, 351-355. (Accepted

8 December

1993)