The NR2B-specific Interactions of Polyamines and Protons with theN-Methyl-d-aspartate Receptor*

Many compounds exhibit NR2B-specific modulation of the N-methyl-d-aspartate receptor, although their mechanism(s) of action are largely unknown. Using chimeric NR2A/NR2B subunits, we have located a region of NR2B (amino acids 138–238) which regulated glycine-independent polyamine stimulation. Mutation of glutamate 201 in this region affected stimulation by polyamines in the order E201D < E201A < E201N < E201R. The relief of proton inhibition of theN-methyl-d-aspartate-induced currents mediated by these mutant receptors correlated with the reduction in glycine-independent polyamine stimulation. Electrophysiological evidence with a triple mutant of NR2A further supports the hypothesis that polyamine stimulation may be linked to the relief of tonic inhibition by protons and demonstrates the crucial role of amino acids 200 and 201 in polyamine stimulation. Polyamines and protons, therefore, share common NR2B determinants.

The N-methyl-D-aspartate (NMDA) 1 receptor is a multimeric ligand-gated ion channel that plays a key role in glutamatergic transmission in the central nervous system (1)(2)(3)(4). The activated NMDA receptor increases the neuronal membrane permeability to Ca 2ϩ and has been implicated in epilepsy (5), Huntington's disease (6), and the delayed neuronal death following cerebral ischemia (7). NMDA receptor activation requires both glutamate and glycine and is modulated by many channel-blocking agents and noncompetitive inhibitors (8). Dizocilpine (MK-801) (9) and phencyclidine block the channel in the open conformation (10) and have been vital for the pharmacologic characterization of these receptors, although the psychomimetic effects conferred by these agents prohibit their clinical use (4). Agents that modulate NMDA receptors at other sites, including the noncompetitive antagonist ifenprodil (11), the endogenous polyamine spermidine (12), and the -site ligand haloperidol (13), may provide better models for novel therapeutic design because they do not produce psychomimetic effects.
Polyamines are endogenous compounds in the central nervous system, although their function in the brain is largely unknown (29). These compounds modulate the NMDA receptor by at least four distinct mechanisms, possibly occurring at distinct receptor sites (27,30,31). In subsaturating concentrations of the coagonist glycine, polyamines enhance the binding of glycine to NMDA receptors (glycine-dependent stimulation) (32,33). In saturating glycine concentrations, polyamines stimulate receptor opening (glycine-independent stimulation) at concentrations below 200 mM (28,34), whereas at higher concentrations, polyamines block NMDA receptors in a voltage-dependent manner (35,36) and decrease the affinity of the receptor for glutamate (31). Glycine-independent stimulation depends on subunit composition, with receptors containing NR1 splice variants lacking the 5Ј-insert (such as NR1A) exhibiting stimulation, whereas receptors containing NR1 subunits with the 5Ј-insert (NR1B) do not exhibit glycine-independent stimulation (21,30). In addition, glycine-independent stimulation is seen only in receptors containing NR2B and is not exhibited by NR2A-, NR2C-, or NR2D-containing receptors (27,37). Several specific residues of NR1 affect glycine-independent stimulation. A residue between the M3 and M4 putative transmembrane regions (Asp-669) (37) and the NH 2 -terminal residues Glu-342 and Glu-339 of NR1A (38) have all been implicated in the control of polyamine sensitivity. Mutation of these residues causes a loss or reduction in glycine-independent polyamine stimulation. Specific residues of the NR2B subunit which are involved in glycine-independent polyamine stimulation have not been reported, although we have previously localized the determinants of NR2B-specific polyamine stimulation to the NH 2 -terminal third of this subunit (21).
The current mediated by NMDA receptors is sensitive to protons in a subunit-specific manner (39). Like polyamines, proton sensitivity is altered by the presence or absence of the 5Ј-insert. NR1B (5Ј-insert present)-containing receptors show an EC 50 for proton inhibition of pH 6.3, whereas NR1A (lacking insert)-containing receptors have a greater proton sensitivity (EC 50 ϭ pH of 7.3) (39). Proton inhibition depends on NR2 subunit expression, although coexpression of either NR2A or NR2B subunits with NR1A yields receptors with a half-maximal pH inhibition of 7.3; NR2C-containing receptors are insensitive to protons (half-maximal pH ϭ 6.8) (39). The NR1A mutations (Asp-669, Glu-342, and Glu-339), implicated in glycine-independent polyamine stimulation (37)(38)(39), also affect proton sensitivity, suggesting that these modulatory effects may be linked.
In the present study we have probed the NR2B-specific interaction of the polyamine spermidine and protons with the NMDA receptor to understand further the allosteric modulation of these agents at the molecular level. Chimeric ⑀ 1 /⑀ 2 subunits were used to localize the NR2B-specific determinants of glycine-independent polyamine stimulation to the NH 2 -terminal region of NR2B. Mutation of a glutamate residue (Glu-201(⑀ 2 )) in this region altered glycine-independent polyamine stimulation. In addition, replacing the three amino acids of ⑀ 1 (MQN) with the corresponding ⑀ 2 residues (LEE) formed a subunit that partially conferred glycine-independent polyamine stimulation. This mutant showed an increase in proton sensitivity compared with wild type NR1A/⑀ 1 receptors, further suggesting that glycine-independent polyamine stimulation may be linked to proton inhibition. Understanding the mechanism(s) of modulation of the NMDA receptor at the molecular level will provide vital information for the design of agents with higher therapeutic potential for the treatment of ischemia or other neurological diseases. Taq  Chimeric NR2 Subunit Construction-The constructions of the ⑀ 1 /⑀ 2 chimeras CH5 and CH6 were described previously (21). A chimera that contains the ⑀ 2 sequence between residues 138 and 464 was constructed by replacing the corresponding sequence of ⑀ 1 between the restriction sites BamHI (414) and AflII (1393). Because there was an additional BamHI site in the 5Ј-untranslated region of our ⑀ 1 clone, we needed to eliminate the 5Ј-untranslated region. By using the primers 5UTSALP2 (5Ј-CCACCTTCTCCGTCGACAGGGACCCTAAGTGGC-3Ј), which introduces an SalI site at the immediate 5Ј end of ⑀ 1 , and the previously reported primer E1AFLII3 (21), the first 1.3 kilobases of ⑀ 1 were synthesized by PCR (⑀ 1 as the template). Substituting the 1.3-kilobase SalI/AflII fragment into the parent clone yielded the plasmid pE1BAM, with a unique BamHI at base 414. The primers BamHI53 (5Ј-GGCA-GATAAGGATCCGTCCTCCATGTTCTTC-3Ј) and E2AFLII3 (21) were then used to amplify (by PCR) a 979-bp fragment of ⑀ 2 , which could be cloned into pE1BAM at the unique BamHI and AflII sites, thus yielding the product CH1.

Materials-Restriction enzymes and
Three additional chimeras were made by utilizing an NdeI site unique to the coding sequence of CH8. Unfortunately, the vector of CH8 (pRK7) contained an additional NdeI site, so the coding sequence of CH8 was subcloned into Bluescript between the SalI and EcoRI sites. A PCR fragment was obtained using the primers E1NDE153 (5Ј-GGTCT-CATTTAGTCTCATATGACGACTGGGACTAC-3Ј) and E1AFLII3 (see above) with ⑀ 1 as the template. The resulting 552-bp fragment was cloned into CH8 (Bluescript) at the NdeI/AflII sites. The SalI/AflII fragment from the resulting plasmid was subcloned back into ⑀ 1 , yielding the chimera CH2. CH9 was derived from CH2 by ligating the SalI/XhoI fragment of CH1 into the same sites in CH2. Likewise, CH10 was made by ligating the SalI/XhoI fragment of CH5 into the same sites of CH2. The sequences of all chimeras were verified by double stranded dideoxy sequencing.
Cell Culture and Transfections-HEK 293 cells were obtained from American Type Culture Collection (ATCC) and were propagated as described previously (21,28). Cells were transfected with a 1:1 ratio of NR1A and either chimeric or mutant NR2 subunits by the calcium phosphate precipitation method (40,41). Cells were protected from NMDA receptor-mediated cell death by the addition of 10 M MK-801 during all steps of the transfections.
125 I-MK-801 Binding Experiments-Cell membranes were prepared as described previously (21,26,28,41). Each modulator concentration was performed in duplicate with a corresponding blank containing 10 M cold MK-801. Membranes were incubated in saturating glycine (100 M) and glutamate (100 M), 300 pM 125 I-MK-801, and the desired concentration of either ifenprodil or spermidine for 3 h to allow the ligand to reach equilibrium. For the experiments assessing glycine effects, membranes were washed without glycine twice before incubations were commenced with 20 mM Hepes, pH 7.5, 100 M glutamate, 100 M spermidine, and the designated concentration of glycine. Membranes were harvested (Brandel Harvester) onto polyethyleneiminecoated glass fiber filters (Schleicher & Schuell) and subsequently counted with a Beckman (model 5500B) ␥-counter.
Electrophysiology-Xenopus oocytes were prepared for injection as described previously (42) and were maintained in ND-96 (96 mM NaCl, 2 mM KCl, 1.8 mM BaCl 2 , and 10 mM Hepes, pH 7.5) supplemented with 0.55 g/liter pyruvate and 50 mg/ml gentamycin. NR1A, ⑀ 2 , and E201N cRNA were synthesized in vitro after digestion with either XbaI (NR1) or EcoRI (NR2 types). Recordings were performed 2-4 days after coinjection of 1 ng of NR1A and 5 ng of ⑀ 1 , ⑀ 2 or E201 mutant cRNA. Oocytes were perfused continuously with ND-96 solution (10 ml/min, 22°C) during two electrode voltage clamp experiments. Oocytes were perfused with ND-96 supplemented with the desired concentration of drug for 1 min before inducing current with 100 M NMDA and 100 M glycine (identical concentration of drug included). I-V curves were performed by stepping the holding potential in Ϫ10-mV increments from Ϫ120 to ϩ40 mV. Leak currents were subtracted in all cases. For experiments examining pH sensitivity, oocytes were equilibrated at the pH to be examined before application of agonists.

RESULTS
Because the molecular basis for the NR2B-specific effects of polyamines is currently unknown, chimeric NR2A/NR2B subunits could be valuable tools for determining the subunit-specific interactions of these modulators. We designed six chimeric ⑀ 1 (NR2A)/⑀ 2 (NR2B) subunits that, when expressed with NR1A, form intact receptors that allowed the localization of the effects of polyamines to a specific region of the NH 2 terminus of NR2B.
Glycine-independent Polyamine Stimulation Localizes to the NH 2 Terminus of NR2B-We have previously localized the determinants of glycine-independent polyamine stimulation to the NH 2 -terminal region between amino acids 198 and 364, using chimeras CH8, CH25, CH5, and CH6 (21). Fig. 1 shows the polyamine stimulation curves for chimeras CH5 and CH6 (top panel), demonstrating that the peak stimulation of 1A/ CH5 receptors is approximately 80% of wild type stimulation, whereas 1A/CH6 showed only 63% wild type levels of stimulation. These and our previous studies suggested that although components downstream from amino acid 464 (perhaps in the channel region) may be important for polyamine stimulation, the region between amino acids 138 and 356 must be important for polyamine effects. Results from two additional chimeras make the localization of the stimulatory region less clear (Fig.  1, bottom panel). 1A/CH9 receptors (containing amino acids 138 -238 of ⑀ 2 ) lacked polyamine stimulation and mimicked closely the profile of 1A/⑀ 1 receptors. Surprisingly, the polyamine stimulation of 125 I-MK-801 binding mediated by 1A/ CH10 receptors (Fig. 1, bottom), which possess even less NR2B sequence (amino acids 198 -238), was identical to 1A/⑀ 2 receptors. These results suggest that multiple regions of NR2B are required to form an intact polyamine stimulatory site and that tertiary and quaternary structural components must be involved in spermidine stimulation, although the region between amino acids 198 and 238 is likely to contain important determinants for NR2B-specific polyamine stimulation.
Glu-201 Is Implicated in Glycine-independent Polyamine Stimulation-Because the data from the chimera experiments implicated the region between amino acids 198 and 238 for polyamine stimulation, the protein sequences of NR2A were compared with the known NR2B sequences to search for residues in this region which were unique to NR2B. This search revealed a conserved negatively charged amino acid (Glu-201 in ⑀ 2 and NR2B) which was an asparagine (Asn-202) in both the NR2A and ⑀ 1 sequences ( Fig. 2A).
Site-directed mutagenesis demonstrated that glycine-independent polyamine stimulation preferred a negatively charged residue at amino acid 201 (Fig. 2B). Conservation of negative charge at position 201 had a small effect on glycine-independent polyamine stimulation. 1A/E201D receptors exhibit polyamine stimulation levels that are 88% of 1A/⑀ 2 receptors, whereas 1A/E201A receptors only exhibit 63% the level of wild type stimulation (Fig. 2B). Glycine-independent polyamine stimulation was eliminated with substitution of a positively charged amino acid (E201R). Thus, the loss of a negatively charged residue at amino acid 201 reduced polyamine stimulation, whereas substitution of a positively charged residue at this position created receptors that were inhibited by polyamines.
To ensure that the effects noted were truly independent of the glycine concentration, the effects of spermidine were assessed in the presence of multiple glycine concentrations. The amount of polyamine stimulation of wild type ⑀ 2 -containing receptors was unchanged by varying the glycine concentration (Fig. 3, A and B). Similarly, the glycine dependence curves resulting from radioligand binding assays are identical for 1A/⑀ 1 , 1A/⑀ 2 , and 1A/E201R receptors. This further confirms that alterations in the effects of glycine cannot explain the differences seen in our study.
Further Characterization of the Effects of E201 Mutants on Polyamine Stimulation-Electrophysiological measurements of the effects of spermidine on the NMDA-induced currents in oocytes injected with NR1 and the E201 mutants were used to characterize further the voltage dependence of the observed changes in polyamine stimulation exhibited by these mutant receptors. It has been described previously that at depolarized potentials (more positive than Ϫ70 mV) polyamine stimulation is increased, whereas at more negative potentials voltage-dependent block predominates over potentiation by polyamines (27). The top panel of Fig. 4 shows the polyamine effects on NMDA-induced currents at a holding potential of Ϫ40 mV. At this holding potential there is a stimulation of both wild type 1A/⑀ 2 and 1A/E201D receptors, whereas no significant stimulation was found for 1A/E201A, 1A/E201N, 1A/E201R, or 1A/⑀ 1 receptors. At a holding potential of Ϫ110 mV (Fig. 4, middle panel) all wild type and mutant receptors exhibit voltage-dependent block, suggesting that mutation of Glu-201 has no effect on the mechanism of voltage-dependent block of these receptors. Consistent with previous studies (31), the stimulation of 1A/⑀ 2 receptors at a holding potential of Ϫ70 mV exhibited an intermediate level of stimulation (100 -125% over base line) between the results shown for Ϫ40 and Ϫ110 mV holding potentials, whereas 1A/⑀ 1 receptors exhibited no stimulation at any of the holding potentials tested (data not shown).
Polyamine stimulation increases at more acidic pH conditions (37). We therefore studied the polyamine stimulation of both wild type and mutant receptors at a pH of 6.8 (holding potential ϭ Ϫ70 mV) (Fig. 4, lower panel). Both 1A/⑀ 2 and 1A/E201D receptors showed a 2-3-fold increase in glycineindependent polyamine stimulation, whereas a slight increase was seen in E201N-and E201A-containing receptors. Both 1A/E201R and 1A/⑀ 1 receptors were unaffected by the shift in pH. This provides additional evidence that glycine-independ-  (39), the proton inhibition of the Glu-201 mutant receptors was investigated (Fig. 5, A and B). Previous measurements of the pH dependence of heterologous combinations of NR1A with either NR2A or NR2B show a trend toward subunit dependence for proton inhibition (39), whereas our experiments showed a slight difference in the IC 50 values for NR1A/⑀ 1 ([H ϩ ] ϭ 100 Ϯ 13 nM, pH ϭ 7.0) and NR1A/⑀ 2 ([H ϩ ], 50 Ϯ 4 nM, pH ϭ 7.3) receptors. Like the results for spermidine, receptors with a negatively charged residue at position 201 (E201D) exhibited no change in the pH dependence (IC 50 for E201D ϭ 50 Ϯ 6 nM, pH ϭ 7.3), whereas the mutation E201N (the residue found in ⑀ 1 ) demonstrated a pH dependence identical to that of 1A/⑀ 1 receptors. The pH dependence of NR1A/ E201A receptors was shifted to the left (IC 50 ϭ 160 Ϯ 20 nM, pH ϭ 6.8), whereas the greatest change was again seen with the E201R substitution (IC 50 ϭ 300 Ϯ 40 nM, pH ϭ 6.5). These results suggest that polyamine stimulation may share a mechanism with proton inhibition and support the hypothesis that glycine-independent polyamine stimulation occurs by relief of the tonic inhibition of NMDA receptor by protons.  was also mutated to determine the necessity of this residue for NR2B-specific modulation. Truncation of the glutamate side chain by the substitution of an alanine at this position, E200A, had no effect on glycine-independent polyamine stimulation in 125 I-MK-801 binding assays (Fig. 6A,  upper panel), whereas spermidine stimulated NMDA-induced currents 61% (Ϯ 9%) with this receptor, slightly less than that seen with 1A/⑀ 2 receptors (holding potential ϭ Ϫ40 mV, pH 7.5, data not shown). In the oocyte expression system, 1A/E200A receptors demonstrated a pH dependence identical to that of wild type 1A/⑀ 2 receptors (Fig. 6A, bottom panel). Therefore, the residue Glu-200 is not required for either glycine-independent polyamine stimulation or proton-dependent inhibition.

Mutation of Glu-200 Has No Effect on Either Polyamine Stimulation or Proton Inhibition-A conserved glutamate residue in NR2B subunits found adjacent to Glu-201 at position
The ulation. Surprisingly, E200Q,E201N when coexpressed with NR1A, was identical to wild type ⑀ 2 receptors with respect to glycine-independent polyamine stimulation and pH dependence (Fig. 6B). This provides evidence that a negatively charged residue is not required at Glu-201.
To confirm the effects of spermidine and pH on the M200L,N201E,Q202E mutant, I-V curves were performed (Fig.  7B). The NR1A/⑀ 1 receptor demonstrated voltage-dependent block by spermidine with no stimulation at any voltage, whereas NR1A/⑀ 2 was stimulated by spermidine with a superimposed voltage-dependent block being seen as flattening of the curve at more negative holding potentials and at higher spermidine concentrations. These results resemble previously reported curves for these receptors when the effects of spermine were tested (27). The NR1A/M200L,N201E,Q202E mutant was stimulated by spermidine at all voltages, although not to the same extent as 1A/⑀ 2 combinations. Voltage-dependent block was also noted at higher spermidine concentrations. No voltage-dependent effects of pH were noted on the I-V curve for the M200L,N201E,Q202 mutant (Fig. 7B) or in other subunit combinations (data not shown).

DISCUSSION
In the present study, chimeric ⑀ 1 /⑀ 2 receptors facilitated the further localization of the NR2B-specific determinants of glycine-independent stimulation on the NH 2 terminus of NR2B. We have shown previously that the NR2B-specific determinants of ifenprodil and polyamine interaction localize to the NH 2 -terminal third of NR2B (21) by using the chimeras 1A/ CH8 and 1A/CH25. Our detailed mapping suggests that multiple regions of NR2B may play a role in polyamine stimulation, with full stimulation requiring distinct tertiary structural elements from both the NH 2 -terminal and other regions of the subunit. In this and our previous study (21), we have localized the determinants of glycine-independent polyamine stimulation to the region around amino acid 198, with the chimera containing a minimal component of NR2B (only 40 amino acids of ⑀ 2 (198 -238)) exhibiting wild type levels of polyamine stimulation. The region between amino acids 138 and 238 is highly conserved between the NR2A and NR2B protein sequences (14, 16 -20). Searching this region for residues that were uniquely conserved in NR2B type receptors but not in NR2A type revealed a single acidic residue, Glu-201 (⑀ 2 ), which was the nonconserved residue asparagine in NR2A and ⑀ 1 . Because polyamines are highly basic compounds, their potential interaction with acidic residues could be postulated.
Mutation of Glu-201 revealed the importance of this residue as an NR2B-specific determinant of polyamine interaction with the NMDA receptor. Substitution of Glu-201 (⑀ 2 ) with the other negatively charged residue aspartate yielded receptors with glycine-independent polyamine stimulation identical to that of wild type receptors. Truncation of the side chain of Glu-201 by substitution of alanine (E201A) exhibited the mildest reduction in polyamine stimulation (63% reduction), which suggests that there is a steric constraint at this site for efficacious binding of modulators. Substitution with asparagine, the residue found in ⑀ 1 and NR2A, yielded receptors virtually identical to NR2A type receptors with respect to polyamine stimulation, whereas substitution to the positively charged arginine abolished glycine-independent polyamine stimulation. The arginine at this position could produce these effects by either a steric hindrance or by exerting an undesirable electrostatic repulsion with a nearby residue in the receptor complex, or it could act by FIG. 4. Electrophysiological analysis of the spermidine effects on the Glu-201 mutant receptors. The effects of increasing spermidine concentration on NMDA-induced currents at different holding potentials were studied. Oocytes were injected with NR1A and either ⑀ 1 (E), ⑀ 2 (q), E201A (ƒ), E201D (छ), E201N (ࡗ), or E201R (‚). Peak current was measured following a 1-min pretreatment of the desired concentration of spermidine in the presence of 100 M NMDA, 100 M glycine, and the desired concentration of spermidine. The top panel shows the polyamine curves for the Glu-201 mutants at a holding potential that promotes glycine-independent polyamine stimulation (Ϫ40 mV); the middle panel shows the spermidine curves at a holding potential that favors voltage-dependent block (Ϫ110 mV). Glycine-independent stimulation has been shown to be increased also at lower pH values (31). The bottom panel shows the spermidine stimulation of NMDA-induced currents at pH 6.8 (holding potential Ϫ70) for the mutant receptors. The spermidine curves are the result of four to six experiments.
repelling the positively charged spermidine molecule from the receptor complex.
The residue Glu-201 may play many possible roles in the NR2B-specific modulation by polyamines. Because polyamine interactions are linked to the binding of the coagonist glycine, mutations at residue Glu-201 could alter the glycine affinity of the receptor. This is unlikely because all of the Glu-201 mu-tants when coexpressed with NR1A exhibit comparable levels of 125 I-MK-801 binding and peak NMDA-induced current, which require the open channel conformation (43,44) and thus an intact glycine site. In addition, the residues that affect glycine affinity (Ser-669, Tyr-666, Phe-390, Tyr-392 etc.) are found exclusively on the NR1 subunit, in a distal region of the polypeptide sequence from the homologous region near Glu-201 (which causes a perfusion artifact as the proton block is removed with glutamate and glycine still bound). This was repeated throughout the pH titration for each receptor concentration. Oocytes were equilibrated at the desired pH before application of NMDA and glycine. Panel B, the pH dependence of NMDA-induced current curves for 1A/⑀ 1 (E), 1A/⑀ 2 (q), 1A/E201A (ƒ), 1A/E201D (छ), 1A/E201N (ࡗ), and 1A/E201R (‚) are shown. The proton dependence of 1A/E201D was identical to 1A/⑀ 2 , whereas 1A/E201A and 1A/E201N receptors shared similar proton dependence with 1A/⑀ 1 . The curve for 1A/E201R showed the most dramatic reduction on proton dependence, being shifted an entire pH unit to the left from 1A/⑀ 2 receptors. EC 50 values were calculated using the following equation: response ϭ (maximum Ϫ minimum)/(1 ϩ ([H ϩ ]/IC 50 ) n ) ϩ minimum. The curves shown were drawn by hand based on cumulation of the data and resulted from four to six experiments (for the mean Ϯ S.E., see "Results"). (45,46). Furthermore, the glycine dependence of the 1A/E201R receptor is identical to the 1A/⑀ 2 receptor. It is therefore unlikely that mutations at Glu-201 altered the glycine affinity of our expressed receptors.
Mutations of Glu-201 likely affect the actions of polyamines by a unique mechanism. The direct interaction of the positively charged spermidine with Glu-201 is unlikely because the double mutant E200Q,E201N, lacking a negatively charged amino acid, exhibited wild type levels of glycine-independent polyamine stimulation. It is possible that instead of a "stimulatory" sequence being present on the NR2B subunit, there may be a structural component of the NR2A subunit which interferes with polyamine stimulation. Perhaps the substitution of Glu-201 to arginine creates a structural change that inhibits efficacious polyamine stimulation, much the same way as an inhibitory region of NR2A might disrupt the allosteric effects of polyamines. This might explain why the three amino acid substitution in ⑀ 1 (M200L,N201E,Q202E) demonstrated glycineindependent polyamine stimulation. The introduction of negatively charged amino acids at positions Gln-201 and Asn-202 may be significant to disrupt the "inhibitory" region of NR2A and permit stimulation. In this alternative, polyamines could bind directly to the NR1 subunit, and this binding is regulated by the NR2 subunits, either directly or through allosteric modulation at the channel pore. Thus, the study of the mechanisms of polyamine stimulation may eventually provide information on the molecular interactions between NR1 and NR2 subunits.
Glycine-independent polyamine stimulation and pH-dependent effects of the Glu-201 mutants closely correlated in our study. Possibly, spermidine is unable to relieve the proton inhibition for receptors such as 1A/E201R because the proton inhibition of this receptor has already been reduced by the mutation. The results of the E200Q,E201N mutant also correlated glycine-independent polyamine stimulation with proton inhibition, both exhibiting wild type levels. The region near Glu-201 is likely to be very near to, or form an integral part of, the proton sensor of the NMDA receptor. The proton sensor has been proposed to be on the NR1 subunit because homomeric receptors exhibit proton sensitivity and because experiments with the splice variants of NR1 which contain the 5Ј-insert (such as NR1B) have shown that this insert relieves the tonic inhibition by the receptor and glycine-independent polyamine FIG. 6. Mutants E200A and E200Q,E201N show the role of the residue at position 200. Panel A, the mutation of Glu-200 to alanine (E200A) had no effect on glycine-independent polyamine stimulation, being indistinguishable from ⑀ 2 -containing receptors in the MK-801 binding assay (upper panel). There was also no change in pH dependence of 1A/E200A receptors (lower panel), suggesting that Glu-200 is not an important residue for either polyamine stimulation or proton dependence. In both panels the mutant receptor 1A/E200A (OE) is shown along with the curves for 1A/⑀ 1 (E) and 1A/⑀ 2 (q) receptors. Data shown are the result of four or five repetitions. Panel B, the double mutant E200Q,E201N, when expressed with 1A, yielded receptors that exhibited no change in either polyamine or proton modulatory effects. Both glycine-independent polyamine stimulation, as measured by the 125 I-MK-801 displacement assay (top panel), and the proton dependence (bottom panel) of this mutant were identical to 1A/⑀ 2 receptors. The mutant receptor 1A/E200Q,E201N (f) is shown along with the curves for 1A/⑀ 1 (E) and 1A/⑀ 2 (q) receptors. Data shown were obtained and analyzed as described in the legends to Figs. 1 and 4. stimulation (39). Another mutation in the proposed extracellular segment between the putative membrane-spanning regions M2 and M3 in NR1 also affects both pH dependence and poly-amine stimulation in a way analogous to that of the 5Ј-insert (37). If one compares the homologous regions of NR1 and NR2 NH 2 termini, the region surrounding Glu-201 is very homolo- gous to the comparable site of insertion at the 5Ј-end of NR1B which relieves the proton inhibition (1,19). The positively charged residues in this 5Ј-insert have been implicated in the cause of the splice variant insensitivity to proton inhibition and polyamine stimulation (39,47). Perhaps the arginine residue in the mutant subunit E201R acts like the positively charged residues in the 5Ј-insert to disrupt the proton sensor of the ion channel. Additional mutations in both NR1 and NR2 subunits will be necessary to elucidate further the role of the region around Glu-201 in both proton sensitivity and polyamine interaction.
Further interesting results of this study were found with the three-amino acid substitution in ⑀ 1 (M200L,N201E,Q202E). This mutation exhibited both glycine-independent polyamine stimulation and an enhanced sensitivity to protons. This adds additional support to the hypothesis that glycine-independent polyamine stimulation occurs through the tonic relief of inhibition of protons. Although the location of the pH sensor of the NMDA receptor is not known, the region around amino acid 200 of the NR2 subunits may be spatially located near this site or allosterically coupled to this region of the NMDA receptor.
Polyamines are present in the mammalian nervous system at very high concentrations, although the physiologic function of the endogenous polyamine spermidine is largely unknown (29). Studies have shown that in normal brain, polyamines are not found in the synaptic cleft and are thus inaccessible to the extracellular face of NMDA receptors (48). Upon acidosis, which occurs during a hypoxic ischemic insult, polyamines synthesis is up-regulated (49), and there is a release of polyamines into the synaptic cleft where they can have direct interactions with the NMDA receptor (50). Polyamine stimulation of NMDA receptors is significantly enhanced at lower pH and at more depolarized membrane potentials (27,37), two of the characteristic conditions observed for the NMDA receptor during ischemia. Perhaps endogenous polyamines act as natural feedback modulators of the NR2B-containing subset of NMDA receptors (29). Another possible explanation is that another endogenous substance such as magnesium acts at the polyamine site in vivo (51). Understanding the interaction of polyamines with NMDA receptors at a molecular level may therefore lead to a better understanding of the events that occur during cerebral ischemia.