N-Methyl-d-aspartate Induces Neurogranin/RC3 Oxidation in Rat Brain Slices*

Neurogranin/RC3 (Ng), a postsynaptic neuronal protein kinase C (PKC) substrate, binds calmodulin (CaM) at low level of Ca2+. In vitro, rat brain Ng can be oxidized by nitric oxide (NO) donors and by oxidants to form an intramolecular disulfide bond with resulting downward mobility shift on nonreducing SDS-polyacrylamide gel electrophoresis. The oxidized Ng, as compared with the reduced form, is a poorer substrate of PKC but like the PKC-phosphorylated Ng has a lower affinity for CaM than the reduced form. To investigate the physiological relevance of Ng oxidation, we tested the effects of neurotransmitter,N-methyl-d-aspartate (NMDA), NO donors, and other oxidants such as hydrogen peroxide and oxidized glutathione on the oxidation of this protein in rat brain slices. Western blot analysis showed that the NMDA-induced oxidation of Ng was rapid and transient, it reached maximum within 3–5 min and declined to base line in 30 min. The response was dose-dependent (EC50 ∼100 μm) and could be blocked by NMDA-receptor antagonist 2-amino-5-phosphonovaleric acid and by NO synthase inhibitorN G-nitro-l-arginine methyl ester and N G-monomethyl-l-arginine. Ng was oxidized by NO donors, sodium nitroprusside,S-nitroso-N-acetylpenicillamine, andS-nitrosoglutathione, and H2O2 at concentrations less than 0.5 mm. Oxidation of Ng in brain slices induced by sodium nitroprusside could be reversed by dithiothreitol, ascorbic acid, or reduced glutathione. Reversible oxidation and reduction of Ng were also observed in rat brain extracts, in which oxidation was enhanced by Ca2+ and the oxidized Ng could be reduced by NADPH or reduced glutathione. These results suggest that redox of Ng is involved in the NMDA-mediated signaling pathway and that there are enzymes catalyzing the oxidation and reduction of Ng in the brain. We speculate that the redox state of Ng, similar to the state of phosphorylation of this protein, may regulate the level of CaM, which in turn modulates the activities of CaM-dependent enzymes in the neurons.

Rat brain neurogranin (Ng), 1 also known as RC3 and BICKS, is a 78-amino acid neuron-specific, postnatal-onset, postsynap-tic protein kinase C (PKC) substrate (for review, see Ref. 1). Ng resembles a presynaptic PKC substrate neuromodulin (Nm, also known as GAP-43, F1, or B-50) in its high affinity binding of calmodulin (CaM) at low levels of Ca 2ϩ (for review, see Refs. [1][2][3]. Both Ng and Nm are phosphorylated by PKC at a single site located within a conserved 19-amino acid region, which also contains the predicted CaM-binding domain (4 -8). Phosphorylation of these two proteins by PKC weakens their binding affinities for CaM. In addition to the PKC-mediated regulation, both rat brain Ng and Nm are potential targets of nitric oxide (NO) and other oxidants (9,10). Treatment of cultured neurons with NO donors interferes with thioester-linked long chain fatty acylation of Nm and causes inhibition of neurite outgrowth (9). In vitro, oxidation of purified rat brain Ng with NO donors or oxidants results in intramolecular disulfide formation and causes an attenuation of its binding affinity for CaM (10). These previous findings point to an interesting possibility that both rat brain Ng and Nm are common targets of both PKC-and NO-mediated signaling pathways. For rat brain Ng, the PKC-mediated phosphorylation of this protein in brain slices and cultured hippocampal neurons in response to phorbol ester (4,11), electrical stimulation (11,12), and excitatory amino acids (13) has been demonstrated; however, the oxidation of this protein in intact cell has not been shown. Although the physiological function of Ng is not clear, cumulating evidences suggest that Ng may be involved in the regulation of CaM targets (14), transduction of the Ca 2ϩ signals (15), long term potentiation (LTP) (11, 16 -18), and neuronal differentiation and synaptogenesis during telencephalic development (19).
The endogenously formed oxidants, such as reactive nitrogen (nitric oxide, NO) and oxygen (superoxide, O 2 . , and hydrogen peroxide, H 2 O 2 ), and redox stress are thought to play important roles in both physiological and pathological responses (20 -25). Some of these responses are due to a direct effect of these oxidants and others arise from the reaction of NO with superoxide (O 2 . ) (26) or O 2 (27) to form peroxynitrite (ONOO Ϫ ) or N 2 O 3 , respectively. Protein thiols are critical sites of interaction with free radicals and their intermediates. Oxidation of thiol on a protein by H 2 O 2 forms sulfenic acid (RS-OH), which in turn interacts with a free thiol group to form disulfide products (28), and with NO to form S-nitrosothiol and subsequent disulfide formation (27,29). S-Nitrosation and subsequent disulfide formation of proteins are important physiological processes involved in the regulation of blood pressure, host defense, and neurotransmission (30).
In this study we are attempting to answer the question if rat brain Ng in intact cells can be oxidized by the commonly recognized oxidants such as NO donors, hydrogen peroxide, and oxidized glutathione, and more importantly, if any physiological agent that is known to stimulate the production of oxidant will enhance the oxidation of this protein. Since oxidation of rat brain Ng results in a preferential formation of intramolecular disulfide bond and leads to an increase in the electrophoretic mobility upon nonreducing SDS-PAGE (10), this unique feature serves as a convenient assay to monitor the oxidation of this protein by immunoblot analysis. Here, we demonstrate that rat brain Ng in the brain slices can be oxidized by several exogenously added oxidants. Furthermore, addition of neurotransmitter, N-methyl-D-aspartate (NMDA), to the brain slices caused a transient oxidation of this protein, and these effects of NMDA can be blocked by its antagonist, 2-amino-5-phosphonovaleric acid (AP5), as well as by NO synthase inhibitor, N Gnitro-L-arginine methyl ester (L-NAME) and N G -monomethyl-L-arginine (L-NMMA). Oxidation/reduction of Ng can also be shown in rat brain extracts, in which the oxidation of this protein is stimulated by Ca 2ϩ and the oxidized form can be reduced by NADPH.
Preparation of Rat Brain Slices and Treatment with NO Donors and Oxidants-Brain from adult Sprague-Dawley rat (200 -250 g) was removed immediately after decapitation and placed in ice-cold pH 7.4 artificial cerebrospinal fluid (ACSF) (125 mM NaCl, 2.5 mM KCl, 2 mM CaCl 2 , 26 mM NaHCO 3 , 1.25 mM NaH 2 PO 4 , 1 mM MgCl 2 , 25 mM glucose, bubbled with 95% O 2 , 5% CO 2 ) for 3-5 min. Cortical slices (400 m thickness) were prepared using a Vibratome 1000 sectioning device with the brain submerged in ice-cold ACSF bubbled with 95% O 2 , 5% CO 2 . The slices were maintained in ACSF in a storage bath for 1-3 h at 30°C to allow recovery from possible damage during slicing. Each slice was transferred to a 20 ml scintillation vial containing 1 ml of ACSF and various concentrations of NO donors, and other oxidants. Incubation was carried out at 30°C in a shaking water bath under 95% O 2 , 5% CO 2 atmosphere. To determine the reversal of Ng oxidation, brain slices were washed three times with 1 ml of ACSF in 15 min and then further incubated with 1 ml of ACSF containing DTT, reduced glutathione, or ascorbic acid. Each slice was homogenized in the cell lysis buffer (50 mM Tris-Cl, pH 7.5, 20 mM iodoacetamide, and 0.5% Nonidet P-40). Extracted protein (60 -80 g) was analyzed for Ng oxidation by immunoblot analysis.
Treatment of Brain Slices with NMDA-Slices were kept at 30°C in a storage bath containing ACSF for ϳ3 h, and the solution was changed three times during storage. Ng in these slices was mostly in the reduced form (Ͼ90%). These cortical slices were incubated at 30°C in 1 ml of ACSF for 3 min with increasing concentrations of NMDA from 100 to 500 M for dose response or with 250 M NMDA at the indicated times for the time course experiments. To examine the effects of NMDA antagonist, AP5, the slices were preincubated with 0 -500 M concentration of this compound for 10 min before the addition of 250 M NMDA. To test the effects of NO synthase inhibitors, L-NAME and L-NMMA, on the NMDA-induced Ng oxidation, brain slices were incubated with 0 -200 M concentration of the inhibitors for 15 min, then 250 M NMDA was added for 3 min.
Preparation of Rat Brain Extracts-Adult rat brain, after removal of cerebellum and brain stem, was homogenized in 4 volumes (v/w) of ice-cold 20 mM potassium phosphate buffer, pH 7.5. The homogenate was centrifuged at 27,000 ϫ g for 20 min, and the supernatant was either used directly after keeping on ice for 2 h or added EDTA and ascorbic acid immediately after centrifugation to final concentrations of 1 and 10 mM, respectively. In the absence of EDTA and ascorbic acid, Ng in the brain extract is spontaneously oxidized. For testing the effects of Ca 2ϩ and NADPH on the oxidation and reduction of Ng, the extract containing KPO 4 /EDTA/ascorbic acid was used and for testing the effect of reduced glutathione on the oxidized Ng, the extract containing KPO 4 only was used.
Determination of Ng Oxidation-Conversion of the reduced Ng to the oxidized form was analyzed by applying the sample (containing 60 -80 g of protein/lane in SDS gel sample buffer without mercaptoethanol and tracking dye) to a 10 -20% gradient SDS (0.1%) gel, which had been prerun at 40 mA/slab gel for 15 min. After electrophoresis, the proteins were transferred onto nitrocellulose membrane. Immunoblot analysis was done using rabbit anti-rat brain Ng polyclonal antibody and horseradish peroxidase-conjugated goat anti-rabbit IgG and ECL reagent (5). Quantitative analysis of the conversion of the reduced Ng to the oxi-FIG. 1. NMDA-induced oxidation of Ng in rat brain slices. Cortical slices from adult rat brain (400 m thickness) were incubated with ACSF at 30°C under 5% CO 2 , 95% O 2 atmosphere with increasing concentrations of NMDA for 3 min (panel A) or with 250 M NMDA for the indicated times (panel B). Brain slices were kept frozen at Ϫ70°C after incubation and homogenized with lysis buffer (50 mM Tris-Cl, pH 7.5, 20 mM iodoacetamide, and 0.5% Nonidet P-40). Proteins (80 g) were separated by nonreducing 10 -20% gradient SDS gel, transferred to nitrocellulose membrane, and the oxidized and reduced Ng detected by immunoblot and the extent of oxidation quantified by densitometric scanning of the film. Each set of data represents the average Ϯ S.E. of six separate experiments. dized form was done by scanning the x-ray film and the data analyzed using the Fotodyne Gel-Pro Analyzer program. The extent of Ng oxidation was calculated by dividing the intensity of the oxidized form with the sum of the oxidized and reduced forms. To ensure that equal amount of protein from each sample was applied to the gel, the intensity of immunoreactive band corresponding to neuromodulin, which crossreacts with our Ng polyclonal antibody (10), was served as internal standard in the same blot. Data for multiple comparisons versus a control were analyzed with Bonferroni's t test and significance was accepted for p Ͻ 0.05.

NMDA-induced Transient Oxidation of Ng in Brain
Slices-Ng expression in rat brain increases during development, and its level reaches maximum 2-3 weeks after birth (31,32). Adult rat brain slices were used to test the effect of NMDA on the oxidation of Ng. NMDA caused a dose-dependent oxidation of Ng in these slices (Fig. 1, panel A). Maximal response was seen at 250 M and at a higher concentration, 500 M, no further increase was seen. The EC 50 was ϳ100 M. A 3-6-fold increase in oxidation of this protein was observed following exposure to 250 M NMDA for 3 min. At this concentration, the NMDA-induced oxidation of Ng was transient, it reached a maximal level within 3-5 min and declined to base line after 30 min (Fig. 1, panel B). Pretreatment of brain slices with NMDA receptor antagonist, AP5, for 10 min blocked the NMDA-induced oxidation of Ng (Fig. 2). AP5 at concentrations over 250 M completely blocked the enhanced oxidation of Ng by the same concentration of NMDA. Pretreatment of brain slices with NO synthase inhibitor, L-NAME, resulted in a dose-de-pendent inhibition of the NMDA-induced oxidation of Ng (Fig.  3, panel A). The oxidation induced by 250 M of NMDA was completely inhibited by 100 M of L-NAME; the same concentration of the inactive enantiomer, D-NAME, was ineffective. Furthermore, the inhibition of the NMDA-induced oxidation of Ng by 100 M L-NAME could be blocked by over 500 M Larginine. Since NO synthase also generates O 2 . , we tested the effect of another inhibitor L-NMMA, which inhibits NO but not O 2 . production (33). L-NMMA inhibited the NMDA-induced Ng oxidation dose-dependently (Fig. 3, panel B), suggesting that NO, rather than O 2 . , is the primary active species for Ng oxidation. These results demonstrate that the oxidation of Ng in the neurons is triggered by the transient production of NO following NMDA binding to its receptor.

Oxidation of Ng in the Brain Slices by NO Donors and Other
Oxidants-To determine whether the exogenously supplied oxidants can also have similar effects as those induced by NMDA, we tested the actions of several oxidants. Incubation of brain slices with NO donors (SNP, SNAP, and SNOG), GSSG, and H 2 O 2 all resulted in the oxidation of Ng (Fig. 4). Among these compounds, SNP, which contains a NO group with strong NO ϩ character (34), appeared to be the most potent; approximately 20 -30% of Ng was converted to the oxidized form containing intramolecular disulfide bond in 30 min at 0.1 mM. The same concentration of the light-inactivated SNP was less effective than the freshly prepared SNP in oxidizing Ng. For example, at 0.1 mM, the freshly prepared SNP caused a 36% oxidation of Ng versus 5% for the light-inactivated SNP. The high potency of SNP in causing Ng oxidation was further demonstrated by an effective oxidation of this protein in the brain slices (up to 10 -20%) even at 20 M (data not shown). The S-nitrosothiols, SNAP and SNOG, which release NO in the presence of reduced sulfhydryl compounds to form mixed disulfide (35), were less potent than SNP. Significant oxidation of Ng was seen at a 0.5 mM concentration of these compounds. Previously, we showed that SNAP and SNOG, when tested in vitro at 5 mM, were ineffective in oxidizing Ng (10,36). Thus, the potencies of these NO donors in oxidizing Ng in the brain slices are different from those seen in vitro using purified Ng. It seems that the effects of these compounds in the brain slices may be mediated by other unknown intermediate(s).
At 0.1 mM, both H 2 O 2 and the GSSG were ineffective in oxidizing Ng in the brain slices (Fig. 4). Oxidation of Ng by these two oxidants were seen at 0.5 mM. The effects of GSSG and H 2 O 2 in the brain slices are apparently mediated by different mechanisms. H 2 O 2 is freely diffusible through cell membrane, and it can oxidize Ng directly or indirectly after conversion to other reactive oxygen species such as the hydroxyl radical OH ⅐ (37) and the superoxide anion, O 2 . , through a secondary effect (38). GSSG, on the other hand, is not permeable to the cell membrane and its oxidation potential is either transmitted through other intermediate(s) or by transport of the cystine moiety inside the cells following extracellular breakdown of GSSG (39).
Reversal of the SNP-mediated Oxidation of Ng in the Brain Slices-The SNP-mediated oxidation of Ng in the brain slices could be reversed by reductants. Incubation of the brain slices with 0.1 mM SNP for 30 min resulted in approximately 30 -40% oxidation of Ng. Following washing of the slices with ACSF for 15 min and replacing the medium with fresh ACSF containing 1 mM DTT, ascorbic acid, or GSH for additional 30 min, the oxidized Ng was reduced (Fig. 5, lanes 4-6). Among these three reductants, DTT appears to be the most potent. This reductant is also very effective in reducing purified oxidized Ng in vitro. GSH may either function as antioxidant in the medium to promote Ng reduction inside the cell or by repletion of intra-FIG. 2. Effect of preincubation with AP5 on the NMDA-mediated oxidation of Ng in the brain slices. Rat brain slices were incubated with ACSF under 95% O 2 , 5% CO 2 atmosphere and increasing concentrations of AP5 for 10 min before the addition of 250 M NMDA. After additional 3 min of incubation, the slices were kept frozen at Ϫ70°C and homogenized with lysis buffer. Proteins (80 g) were separated by nonreducing 10 -20% gradient SDS gel, transferred to nitrocellulose membrane, and the oxidized and reduced Ng detected by immunoblot. The extent of oxidation was quantified by densitometric scanning of the film. Each set of data represents the average Ϯ S.E. of six separate experiments. cellular GSH through uptake of the breakdown products and intracellular resynthesis of GSH. Ascorbate cannot reduce the oxidized Ng in vitro, thus, it may function as an antioxidant, for instance to reduce thiyl radicals (40), produced by oxidation of Ng with NO that leads to the formation of disulfide. Replacement of the medium with ACSF without reductant did not cause a significant reversal (Fig. 5, lane 3); the extent of oxidation of Ng was comparable with that treated with 0.1 mM SNP during the entire 75 min of incubation (Fig. 5, lane 2). The Ng in the control slice incubated with ACSF alone showed only basal level of oxidation (Fig. 5, lane 1).
Oxidation/Reduction of Rat Brain Ng in the Crude Extract-Ng present in the rat brain extract prepared by homogenization in a buffer containing 20 mM potassium phosphate, pH 7.5, was spontaneously oxidized when kept on ice for over 2 h (Fig. 6, panel A, 0 min sample). Incubation of the extract with 3 mM GSH resulted in a time-dependent reduction of Ng. A near complete reduction of the oxidized Ng was seen after 30 min (Fig. 6, panel A). Ng in the brain extract prepared by homogenization in 20 mM phosphate buffer, pH 7.5, and followed by the addition of 1 mM EDTA and 10 mM ascorbic acid showed a partial oxidation. The partially oxidized Ng was used FIG. 3. Effect of NO synthase inhibitor L-NAME and L-NMMA on the NMDA-mediated oxidation of Ng in brain slices. Rat brain slices were incubated with ACSF under 95% O 2 , 5% CO 2 atmosphere and increasing concentrations of L-NAME (panel A) or L-NMMA (panel B) for 15 min before the addition of 250 M NMDA. After an additional 3 min of incubation, the slices were kept frozen at Ϫ70°C and homogenized with lysis buffer. Proteins (80 g) were separated by nonreducing 10 -20% gradient SDS gel, transferred to nitrocellulose membrane, and the oxidized and reduced Ng detected by immunoblot. The extent of oxidation was quantified by densitometric scanning of the film. Each set of data represents the average Ϯ S.E. of eight separate experiments.

FIG. 4. Oxidation of rat brain Ng in the brain slices by NO donors and other oxidants.
Cortical slices from adult rat brain were incubated in ACSF at 30°C under 5% CO 2 , 95% O 2 atmosphere and incubated without addition (lane 1) or with 0.1 and 0.5 mM NO donors: SNP (lanes 2 and 3), SNAP (lanes 4 and 5), and SNOG (lanes 10 and 11) and other oxidants: H 2 O 2 (lanes 6 and 7) and GSSG (lanes 8 and 9) for 30 min. Brain slices were kept frozen at Ϫ70°C and homogenized with lysis buffer. Proteins (80 g) were separated by nonreducing 10 -20% gradient SDS gel, transferred to nitrocellulose membrane, and the oxidized and reduced Ng detected by immunoblot. The extent of oxidation was quantified by densitometric scanning of the film. Each set of data represents the average Ϯ S.E. of eight separate experiments.
to demonstrate the interconversion between the oxidized and reduced forms by either the addition of Ca 2ϩ or NADPH. In the presence of KPO 4 /EDTA/ascorbic acid, the ratio of the oxidized and reduced Ng remained nearly the same upon incubation at 30°C for up to 30 min (Fig. 6, panel B). Addition of 1.1 mM CaCl 2 to the extract containing 20 mM KPO 4 , 1 mM EDTA, 10 mM ascorbic acid resulted in a rapid oxidation of Ng (Fig. 6, panel C). Previously, we have shown that Ng in the brain extract is completely dissociated from CaM in the presence of Ca 2ϩ (5). Thus, the free Ng becomes susceptible to oxidation by an unknown mechanism, even in the presence of 10 mM ascorbic acid. Incubation of the partially oxidized Ng in 20 mM phosphate, 1 mM EDTA, 10 mM ascorbic acid with 2 mM NADPH resulted in a reduction of this protein (Fig. 6, panel D), suggesting the presence of a Ng reductase. NADPH alone is ineffective in reducing the purified oxidized Ng (data not shown).

DISCUSSION
This study provides the first demonstration that rat brain Ng in intact cells is subjected to modification by oxidation and reduction. The redox of Ng in the neurons is likely a physiological event, since NMDA can induce a transient oxidation of this protein. Rat brain Ng has a high propensity to form a intramolecular disulfide bond, resulting in a downward mobility shift on nonreducing SDS-PAGE; oxidation of this protein in rat brain may conveniently serve as an indicator for the generation of oxidants in neurons. Based on our previous mutagenesis study (36) we have predicted that Ng oxidation results in intramolecular disulfide bond formation between Cys-3, Cys-4, or Cys-9 and Cys-51. Electrospray ionization mass spectrometry analysis showed that there were two pairs of disulfide bond in the oxidized Ng (data not presented). The resulting oxidized Ng, with its N-and C-terminal ends brought together by the disulfide bond, becomes more compact than the reduced Ng. The calculated molecular weight of rat brain Ng is 7500, but it migrates as a M r 14,000 -17,000 protein on SDS-PAGE depending on the conditions of electrophoresis. The oxidized Ng has an apparent molecular weight of approximately 10,000 on nonreducing SDS-PAGE. Oxidation of the purified Ng in vitro (10), in rat brain slices as induced by neurotransmitter NMDA and by the various oxidants, and that seen in the crude rat brain extracts, all result in the formation of the apparent M r ϭ 10,000 intramolecular disulfide form. It seems that formation of the intramolecular disulfide bonds of rat brain Ng is a kinetically favorable reaction.
Free radicals and redox stress are thought to participate in cellular signaling (41). Ng in the brain slices is oxidized by the addition of NMDA, NO donors, and several oxidants. The intramolecularly disulfide-bridged form of Ng has a lower affinity for CaM and becomes a poorer substrate of PKC than the reduced form (10). Both the phosphorylation and oxidation of Ng seem to have a similar consequence in reducing the affinity of this protein for CaM and thus render CaM available for many CaM-dependent enzymes, including CaM-dependent protein kinases, cyclic nucleotide phosphodiesterase, adenylyl cyclases, NO synthase, inositol 1,4,5-trisphosphate kinase, calcium pumps, CaM-dependent protein phosphatase, nicotinamide adenine dinucleotide kinase, and cytoskeletal proteins. Thus, the various oxidants either generated in situ, as in the case of NO from NMDA treatment, or present extracellularly may transduce their signals by oxidation of Ng and modulate the level of CaM. The phosphorylated Ng can be dephosphorylated by protein phosphatases 1 and 2A and by calcineurin (42) whereas the enzyme(s) responsible for the reduction of the oxidized Ng has yet to be identified. Nonetheless, there seems to be an enzyme system in the brain that can efficiently reduced the oxidized Ng as demonstrated by the transient effect of NMDA and that seen in the crude rat brain extracts.
In the rat brain slices, the oxidant-induced oxidation of Ng can also be reduced by exogenously added reductants, such as DTT, ascorbate, or GSH, suggesting that these reductants can influence the intracellular redox state in favor of the reduction of the oxidized Ng. Both DTT and GSH can reduce the oxidized Ng directly in the in vitro experiments, however, ascorbate cannot. Similarly, in the crude rat brain extracts, DTT (data not presented) and GSH are effective in reducing the oxidized Ng but not the ascorbate. It seems that these reductants may either reduce the intracellularly oxidized Ng directly or serve as antioxidants and thus favor the reduction of Ng by other enzyme(s). In this respect, it is interesting to note that addition of NADPH to the crude rat brain extracts containing endogenously oxidized Ng causes a reduction of Ng, although NADPH itself is not effective in reducing the purified Ng. It appears that there is a Ng reductase in the neurons that utilizes NADPH as electron donor to reduce the oxidized Ng. The Ng thiol-disulfide redox resembles those of thioredoxins and glutaredoxins, both are small molecular weight proteins (8000 -12,000) containing redox-active thiols (43). Ng does not exhibit any sequence homology to these proteins, including the short sequence regions containing redox active Cys residues. A major function of thioredoxin and glutaredoxin is to serve as a cofactor for ribonucleotide reductase, which catalyzes the conversion of ribonucleotides to deoxyribonucleotides (44). In the case of thioredoxin, the reduced cofactor is regenerated by a specific NADPH-dependent thioredoxin reductase (45). The oxidized glutaredoxin is reduced directly by cellular GSH, and the resulting GSSG is, in turn, reduced by NADPH via the glutathione reductase (46).
The physiological relevance of Ng oxidation is best illustrated by the NMDA-induced transient oxidation of Ng in rat brain slices, in which oxidation of Ng is inhibited by the receptor antagonist, AP5, as well as by the NO synthase inhibitors, L-NAME and L-NMMA. It has been shown that activation of NMDA receptors and the subsequent increase in [Ca 2ϩ ] i stimulate NO synthase (47)(48)(49), which generates both O 2 . and NO (33). Since both L-NAME, which inhibits O 2 . and NO production, and L-NMMA, which inhibits NO but not O 2 . production, are effective in inhibiting the NMDA-induced oxidation of Ng, we infer that NO is the primary active species for the oxidation of Ng. Thus, Ng oxidation serves as a downstream target of the NO signaling pathway. Our results on the reversible oxidation/ reduction of Ng further support the contention that the redox states of a protein may involve in the regulatory processes of NMDA. The transient nature of Ng oxidation following treatment with NMDA is indicative for the short burst of NO production after the administration of this neurotransmitter. Ng is a neuron-specific protein; its phosphorylation by PKC has been linked to the induction of LTP in rat hippocampal slices (11,16,17), and intracellular injection of antibodies against Ng prevents the maintenance phase of LTP (18). The tetanus-induced LTP requires activation of NMDA receptors (50) and an elevation of [Ca 2ϩ ] i (51,52), which result in the activation of PKCs and presumably stimulate phosphorylation of Ng located at the postsynaptic dendritic spine and shaft (53). However, phosphorylation of Ng was reported to be only at 60 min, but not at other time points tested (15,30, and 120 min), after tetanus-induced LTP (11). In contrast, phosphorylation of presynaptic PKC substrate, B-50/GAP-43/neuromodulin, reached maximum within 15-30 min (11). As the activation of NMDA receptors also stimulates NO synthase (47)(48)(49) and causes oxidation of Ng, it is likely that Ng is oxidized during the early phase of tetanic stimulation that renders Ng a poor substrate of PKC (10). However, the oxidized Ng functions similarly as the phosphorylated Ng in elevating the free CaM levels at the postsynaptic sites, where the Ca 2ϩ /CaM-dependent protein kinase II is also located. The activation of Ca 2ϩ / CaM-dependent protein kinase II has been linked to the postsynaptic mechanism for the enhancement of synaptic transmission during LTP; thus, the oxidation of Ng may potentiate the stimulatory effect during the early phase of LTP.