Neuroscience Letters, 133 (1991) 11-14
11
© 1991 ElsevierScientific Publishers Ireland Ltd. All rights reserved0304-3940/91/$ 03.50 NSL 08179
Oxidized glutathione modulates N-methyl-D-aspartate- and depolarization-induced increases in intracellular Ca 2 ÷ in cultured rat forebrain neurons K a r i R. G i l b e r t 1, Elias A i z e n m a n 2 a n d I a n J. R e y n o l d s 1 Departments of 1Pharmacology and 2Physiology. University of Pittsburgh School of Medicine, Pittsburgh, PA 15261 ( U.S.A.)
(Received 3 May 1991; Revisedversion received28 June 1991; Accepted 7 August 1991) Key words: Intracellular Caz +; N-Methyl-D-aspartatereceptor; Glutathione; Redox modulation
We have investigated the interaction of reduced and oxidized glutathione (GSSG) with intracellular Ca2+ increases produced by N-methyl-Daspartate (NMDA), kainate and KC1 in primary cultures of forebrain neurons derived from fetal rats. Responses to NMDA, applied with glycine, were inhibited by GSSG (10 mM), but were unaffected by reduced glutathione and L-cysteine. Inhibition by GSSG was still apparent after cells were oxidized by 5,5'-dithio-bis-2-nitrobenzoicacid, and this effect showed spontaneous but only partial reversal. This suggests that modulation of the redox site on the NMDA receptor could not account for all of the effectsproduced by GSSG. However,the observation that complete recovery from GSSG treatment required exposure of cells to dithiothreitol suggests that oxidation of the redox site contributes to the action of GSSG. GSSG also inhibited responses produced by 50 mM KC1 but not those produced by 50/zM kainate. The effects of GSSG on KCI responses were fully and rapidly reversible. These results suggest that high concentrations of GSSG may modulate NMDA receptors, and that some of the actions of GSSG may be mediated by the redox site on the receptor complex.
The N-methyl-D-aspartate ( N M D A ) preferring subtype of glutamate receptor is subject to modulation by a complex array of endogenous neurochemicals, including glycine, M g 2 +, Zn 2 ÷, H ÷ and polyamines [3]. A recent addition to the list of modulatory influences is the site that is subject to reduction and oxidation by sulfhydryl redox reagents [2]. Reduction o f this site by dithiothreitol (DTT) results in increased activation of the N M D A receptor, while oxidation of this site decreases receptor activation [2, 8]. While the full spectrum of physiological consequences of modulation o f this site have yet to be realized, it is clear that redox reagents can influence several N M D A receptor mediated events, including delayed cell death [1, 5, 6], epileptiform activity in hippocampal slices [12], and long-term potentiation [111. The central role of N M D A receptors in numerous physiological and pathophysiological processes highlights the importance of understanding the contribution of endogenous agents to the responsiveness o f the N M D A receptor. The identity of endogenous sulfhydryl modifying agents that might interact with the N M D A Correspondence: I.J. Reynolds, Department of Pharmacology, E1354 BST, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, U.S.A. Fax: (1) (412) 648-1945.
receptor has not been clearly established. In a recent study, however, we demonstrated that oxygen radicals can oxidize the N M D A receptor in vitro [1]. In this study we have continued our investigation into the redox regulation of the N M D A receptor, and report that oxidized glutathione (GSSG), but not reduced glutathione (GSH) can alter N M D A receptor activity. Neurons were cultured from 17-day embryonic rat forebrain isolated as previously described [8]. Cells were plated onto poly-D-lysine coated glass coverslips at a concentration of 3 x 10s cells/ml in Dulbecco's modified Eagle's medium ( D M E M ) containing 10% v/v fetal bovine serum, 100 U/ml streptomycin and 5/zg/ml penicillin. This medium was replaced 24 h after plating to _DMEM with 10% horse serum and antibiotics, and the coverslips were inverted. Cells were kept in a humidified incubator at 37°C in 95% air/5% CO 2 for 12-21 days until intracellular free Ca a+ ([CaZ+]i) recordings were made. [Caz ÷]i recordings were made essentially as described earlier [8] using fura-2 microspectrofluorimetry. On the day the recordings were made, the coverslips were rinsed with H E P E S buffered salt solution (HBSS), which contained (in mM): NaC1 137, KC1 5, MgSO 4 0.9, CaCI 2 1.4, N a H C O 3 3, Na2HPO 4 0.6, K H z P O 4 0.4, glucose 5.6 and H E P E S 20, p H adjusted to 7.4 with N a O H . The cells
12
GSSG were obtained from Calbiochem (La Jolla, CA), while DTT was from Boehringer-Mannheim Biochemicals (Indianapolis, IN). We have previously shown that NMDA- and glycineinduced increases in [Caz +]i are very sensitive to changes in the redox state of the NMDA receptor [1, 8]. Fig. IA shows the effect of GSH, GSSG and L-cysteine on the response to NMDA (30 /~M) and glycine (1 /LM). In these experiments the putative redox reagents were applied for 2 min and then washed from the chamber prior to agonist addition. GSH had essentially no effect on agonist-induced [Ca2+]i changes, nor did L-cysteine, both tested at 10 mM. However, GSSG treatment substantially inhibited the response to NMDA (Fig. 1B). Ten mM GSSG substantially decreased NMDA responses in untreated and 2 mM DTT treated cells. Lower concentrations also inhibited NMDA responses. However, these effects were not significantly greater than that arising from spontaneous reversal of the DTT enhancement. As this effect was measured after the washout of the drug from the recording chamber these results suggest that at least a part of this response is due to a chemical modification of the receptor. GSSG (10
were then incubated in 5 ¢tM fura-2 AM (Molecular Probes, Eugene OR) in HBSS with 5 mg/ml bovine serum albumin for approximately 1 h at 37°C. After incubation the cells were rinsed for 10 min with HBSS and then mounted in a recording chamber. Solutions were perfused through the recording chamber at a rate of approximately 20 ml/min. Agonist-induced changes in [Caz +]i were recorded by exposing the neurons to agonists for approximately 40 s immediately followed by perfusion with agonist-free HBSS to wash out the effect. Pretreatment of neurons with redox agents was accomplished by exposing cells to reagents for 100-200 s followed by perfusion with HBSS for at least 30 s prior to agonist addition. Application of 10 mM GSH, GSSG, or L-cysteine elicited a small change (usually less than 100 nM) in [Ca2 +]i in 39 of 60 cells tested. This change in [Ca2+]i was allowed to return to baseline before agonist application. The effect of GSH and GSSG may be due to contamination by glutamate as it was prevented by the NMDA antagonist aminophosphonovalerate (AP5). NMDA, glycine, kainic acid, and L-cysteine were obtained from Sigma (St. Louis, MO). The redox agents 5,5'-dithiobis-2-nitrobenzoic acid (DTNB), GSH and
B
A
c 200
15¢
10~ 100
F t~
50 50
0
•
GSSG 1
GSH I ~ V [ 100
200
0
300
100
200
C Y S I 0mbl
O 300
400
I00
200
300
400
T i m e , sec
100
75 50
g
25
0 GSH
1
3
10
CYS
GSSG, mM
Fig. 1. A-C: oxidized glutathione (GSSG) but not reduced glutathione (GSH) or L-cysteine (CYS) inhibits N M D A - and glycine-stimulated increases in [Ca 2 +]i- N M D A (30/~M) and glycine (1 /tM) were added at the arrowheads. Responses in each trace were measured before and then after 2 min exposure to 10 m M of (A) GSH, (B) GSSG and (C) L-cysteine. Control responses for GSSG and GSH represent 'native', that is, unmodified by DTT or DTNB, cells, while control responses for L-cysteine are responses that were recorded after DTNB-treatment. The results shown represent typical responses measured in single cells from different coverslips that were repeated 4 8 times with similar results. D: mean effect of GSH, GSSG and L-cysteine on responses to 30 g M N M D A and 1 #M glycine. The results are expressed as a percentage of the response measured in the absence of modulator, and are the mean ( + S.E.M.) of 3 8 cells. Ten mM of each modulator was used except as indicated for GSSG. The range of N M D A induced [Ca2 *]~ increases in cells before treatment was 95-698 nM. Control responses for GSH represent 'native' responses, while control responses for GSSG and L-cysteine represent DTT- and DTNB-pretreated responses respectively. **Significantly different from control (P
11
© 1991 ElsevierScientific Publishers Ireland Ltd. All rights reserved0304-3940/91/$ 03.50 NSL 08179
Oxidized glutathione modulates N-methyl-D-aspartate- and depolarization-induced increases in intracellular Ca 2 ÷ in cultured rat forebrain neurons K a r i R. G i l b e r t 1, Elias A i z e n m a n 2 a n d I a n J. R e y n o l d s 1 Departments of 1Pharmacology and 2Physiology. University of Pittsburgh School of Medicine, Pittsburgh, PA 15261 ( U.S.A.)
(Received 3 May 1991; Revisedversion received28 June 1991; Accepted 7 August 1991) Key words: Intracellular Caz +; N-Methyl-D-aspartatereceptor; Glutathione; Redox modulation
We have investigated the interaction of reduced and oxidized glutathione (GSSG) with intracellular Ca2+ increases produced by N-methyl-Daspartate (NMDA), kainate and KC1 in primary cultures of forebrain neurons derived from fetal rats. Responses to NMDA, applied with glycine, were inhibited by GSSG (10 mM), but were unaffected by reduced glutathione and L-cysteine. Inhibition by GSSG was still apparent after cells were oxidized by 5,5'-dithio-bis-2-nitrobenzoicacid, and this effect showed spontaneous but only partial reversal. This suggests that modulation of the redox site on the NMDA receptor could not account for all of the effectsproduced by GSSG. However,the observation that complete recovery from GSSG treatment required exposure of cells to dithiothreitol suggests that oxidation of the redox site contributes to the action of GSSG. GSSG also inhibited responses produced by 50 mM KC1 but not those produced by 50/zM kainate. The effects of GSSG on KCI responses were fully and rapidly reversible. These results suggest that high concentrations of GSSG may modulate NMDA receptors, and that some of the actions of GSSG may be mediated by the redox site on the receptor complex.
The N-methyl-D-aspartate ( N M D A ) preferring subtype of glutamate receptor is subject to modulation by a complex array of endogenous neurochemicals, including glycine, M g 2 +, Zn 2 ÷, H ÷ and polyamines [3]. A recent addition to the list of modulatory influences is the site that is subject to reduction and oxidation by sulfhydryl redox reagents [2]. Reduction o f this site by dithiothreitol (DTT) results in increased activation of the N M D A receptor, while oxidation of this site decreases receptor activation [2, 8]. While the full spectrum of physiological consequences of modulation o f this site have yet to be realized, it is clear that redox reagents can influence several N M D A receptor mediated events, including delayed cell death [1, 5, 6], epileptiform activity in hippocampal slices [12], and long-term potentiation [111. The central role of N M D A receptors in numerous physiological and pathophysiological processes highlights the importance of understanding the contribution of endogenous agents to the responsiveness o f the N M D A receptor. The identity of endogenous sulfhydryl modifying agents that might interact with the N M D A Correspondence: I.J. Reynolds, Department of Pharmacology, E1354 BST, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, U.S.A. Fax: (1) (412) 648-1945.
receptor has not been clearly established. In a recent study, however, we demonstrated that oxygen radicals can oxidize the N M D A receptor in vitro [1]. In this study we have continued our investigation into the redox regulation of the N M D A receptor, and report that oxidized glutathione (GSSG), but not reduced glutathione (GSH) can alter N M D A receptor activity. Neurons were cultured from 17-day embryonic rat forebrain isolated as previously described [8]. Cells were plated onto poly-D-lysine coated glass coverslips at a concentration of 3 x 10s cells/ml in Dulbecco's modified Eagle's medium ( D M E M ) containing 10% v/v fetal bovine serum, 100 U/ml streptomycin and 5/zg/ml penicillin. This medium was replaced 24 h after plating to _DMEM with 10% horse serum and antibiotics, and the coverslips were inverted. Cells were kept in a humidified incubator at 37°C in 95% air/5% CO 2 for 12-21 days until intracellular free Ca a+ ([CaZ+]i) recordings were made. [Caz ÷]i recordings were made essentially as described earlier [8] using fura-2 microspectrofluorimetry. On the day the recordings were made, the coverslips were rinsed with H E P E S buffered salt solution (HBSS), which contained (in mM): NaC1 137, KC1 5, MgSO 4 0.9, CaCI 2 1.4, N a H C O 3 3, Na2HPO 4 0.6, K H z P O 4 0.4, glucose 5.6 and H E P E S 20, p H adjusted to 7.4 with N a O H . The cells
12
GSSG were obtained from Calbiochem (La Jolla, CA), while DTT was from Boehringer-Mannheim Biochemicals (Indianapolis, IN). We have previously shown that NMDA- and glycineinduced increases in [Caz +]i are very sensitive to changes in the redox state of the NMDA receptor [1, 8]. Fig. IA shows the effect of GSH, GSSG and L-cysteine on the response to NMDA (30 /~M) and glycine (1 /LM). In these experiments the putative redox reagents were applied for 2 min and then washed from the chamber prior to agonist addition. GSH had essentially no effect on agonist-induced [Ca2+]i changes, nor did L-cysteine, both tested at 10 mM. However, GSSG treatment substantially inhibited the response to NMDA (Fig. 1B). Ten mM GSSG substantially decreased NMDA responses in untreated and 2 mM DTT treated cells. Lower concentrations also inhibited NMDA responses. However, these effects were not significantly greater than that arising from spontaneous reversal of the DTT enhancement. As this effect was measured after the washout of the drug from the recording chamber these results suggest that at least a part of this response is due to a chemical modification of the receptor. GSSG (10
were then incubated in 5 ¢tM fura-2 AM (Molecular Probes, Eugene OR) in HBSS with 5 mg/ml bovine serum albumin for approximately 1 h at 37°C. After incubation the cells were rinsed for 10 min with HBSS and then mounted in a recording chamber. Solutions were perfused through the recording chamber at a rate of approximately 20 ml/min. Agonist-induced changes in [Caz +]i were recorded by exposing the neurons to agonists for approximately 40 s immediately followed by perfusion with agonist-free HBSS to wash out the effect. Pretreatment of neurons with redox agents was accomplished by exposing cells to reagents for 100-200 s followed by perfusion with HBSS for at least 30 s prior to agonist addition. Application of 10 mM GSH, GSSG, or L-cysteine elicited a small change (usually less than 100 nM) in [Ca2 +]i in 39 of 60 cells tested. This change in [Ca2+]i was allowed to return to baseline before agonist application. The effect of GSH and GSSG may be due to contamination by glutamate as it was prevented by the NMDA antagonist aminophosphonovalerate (AP5). NMDA, glycine, kainic acid, and L-cysteine were obtained from Sigma (St. Louis, MO). The redox agents 5,5'-dithiobis-2-nitrobenzoic acid (DTNB), GSH and
B
A
c 200
15¢
10~ 100
F t~
50 50
0
•
GSSG 1
GSH I ~ V [ 100
200
0
300
100
200
C Y S I 0mbl
O 300
400
I00
200
300
400
T i m e , sec
100
75 50
g
25
0 GSH
1
3
10
CYS
GSSG, mM
Fig. 1. A-C: oxidized glutathione (GSSG) but not reduced glutathione (GSH) or L-cysteine (CYS) inhibits N M D A - and glycine-stimulated increases in [Ca 2 +]i- N M D A (30/~M) and glycine (1 /tM) were added at the arrowheads. Responses in each trace were measured before and then after 2 min exposure to 10 m M of (A) GSH, (B) GSSG and (C) L-cysteine. Control responses for GSSG and GSH represent 'native', that is, unmodified by DTT or DTNB, cells, while control responses for L-cysteine are responses that were recorded after DTNB-treatment. The results shown represent typical responses measured in single cells from different coverslips that were repeated 4 8 times with similar results. D: mean effect of GSH, GSSG and L-cysteine on responses to 30 g M N M D A and 1 #M glycine. The results are expressed as a percentage of the response measured in the absence of modulator, and are the mean ( + S.E.M.) of 3 8 cells. Ten mM of each modulator was used except as indicated for GSSG. The range of N M D A induced [Ca2 *]~ increases in cells before treatment was 95-698 nM. Control responses for GSH represent 'native' responses, while control responses for GSSG and L-cysteine represent DTT- and DTNB-pretreated responses respectively. **Significantly different from control (P
