Philip E. Chen and Alexander R. Johnston contributed equally to the work.
Influence of a threonine residue in the S2 ligand binding domain in determining agonist potency and deactivation rate of recombinant NR1a/NR2D NMDA receptors
Article first published online: 24 JUN 2004
The Journal of Physiology
Volume 558, Issue 1, pages 45–58, July 2004
How to Cite
Chen, P. E., Johnston, A. R., Mok, M. H. S., Schoepfer, R. and Wyllie, D. J. A. (2004), Influence of a threonine residue in the S2 ligand binding domain in determining agonist potency and deactivation rate of recombinant NR1a/NR2D NMDA receptors. The Journal of Physiology, 558: 45–58. doi: 10.1113/jphysiol.2004.063800
- Issue published online: 24 JUN 2004
- Article first published online: 24 JUN 2004
- (Received 3 March 2004; accepted after revision 22 April 2004; first published online 23 April 2004)
NR1/NR2D NMDA receptors display unusually slow deactivation kinetics which may be critical for their role as extrasynaptic receptors. A threonine to alanine point mutation has been inserted at amino acid position 692 of the NR2D subunit (T692A). Recombinant NR1a/NR2D(T692A) NMDA receptors have been expressed in Xenopus laevis oocytes and their pharmacological and single-channel properties examined using two-electrode voltage-clamp and patch-clamp recording techniques. Glutamate dose–response curves from NR1a/NR2D(T692A) receptor channels produced an approximately 1600-fold reduction in glutamate potency compared to wild-type NR1a/NR2D receptors. There was no change in Hill slopes or gross reduction in mean maximal currents recorded in oocytes expressing either wild-type or mutant receptors. The mutation did not affect the potency of the co-agonist glycine. The shifts in potency produced by NR2D(T692A) containing receptors when activated by other glutamate-site agonists such as aspartate or NMDA were 30- to 60-fold compared to wild-type. Single-channel conductance levels of NR1a/NR2D(T692A) mutant receptors were indistinguishable from wild-type NR2D-containing channels. Additionally NR1a/NR2D(T692A) receptors showed the transitional asymmetry that is characteristic of NR2D-containing NMDA receptors. Rapid applications of glutamate on outside-out patches containing NR1a/NR2D(T692A) receptors produced macroscopic current deactivations that were about 60-fold faster than wild-type NR1a/NR2D receptors. Our results suggest that this conserved threonine residue plays a crucial role in ligand binding to NMDA NR2 receptor subunits and supports the idea that the slow decay kinetics associated with NR1a/NR2D NMDA receptors can be explained by the slow dissociation of glutamate from this NMDA receptor subtype.
Ionotropic NMDA glutamate receptors are thought to exist as hetero-oligomers of NR1 and NR2 subunits (Kutsuwada et al. 1992; Monyer et al. 1992). There are multiple splice variants of the NR1 subunit (Sugihara et al. 1992) and four different NR2 subunits (NR2A–D) (Monyer et al. 1992; Ishii et al. 1993). The NR1 subunit is ubiquitously expressed in the CNS and each of the four NR2 subunits show distinct temporal and spatial expression patterns (Watanabe et al. 1992; Akazawa et al. 1994). The NR2 subunits generate functional diversity, determining single-channel properties and kinetic behaviour (Stern et al. 1992; Monyer et al. 1994; Wyllie et al. 1996, 1998; Vicini et al. 1998), voltage-dependent magnesium block (Kuner & Schoepfer, 1996) and pharmacological properties of the heteromeric NMDA receptors (Kutsuwada et al. 1992). In addition, NMDA receptors can also contain NR3A or NR3B receptor subunits which have also been shown to modulate channel function and which when coassembled with NR1 subunits form cation channels that are gated by glycine alone (Chatterton et al. 2002). For reviews on the structure–function properties of NMDA receptors see Cull-Candy et al. (2001) and Dingledine et al. (1999).
In a recent study, Brickley et al. (2003) concluded that incorporation of the NR2D subunit into heteromeric NMDA receptor complexes might restrict such receptors to extrasynaptic sites (see also Momiyama, 2000). Given that NR2D mRNA levels are at their highest early in development (Watanabe et al. 1992; Monyer et al. 1994), the activation of these receptors by glutamate may be important in mediating effects that are distinct from the ‘classical’ role played by ionotropic glutamate receptors in ‘fast’ synaptic transmission. Recombinant and native NR2D-containing NMDA receptors show slow deactivation times (time constant of decay of around 5000 ms) following brief applications of glutamate (Monyer et al. 1994; Vicini et al. 1998; Wyllie et al. 1998; Misra et al. 2000). However, it is unclear what mechanism determines the slow kinetics of NR2D-containing receptors.
The proposed topology of NMDA receptor subunits includes an extracellular amino terminus, three membrane spanning segments (M1, M3 and M4), one membrane-embedded inverted loop (M2) that forms the channel pore and an intracellular carboxy terminus (for a review see Dingledine et al. 1999). Two polypeptide segments, S1 (∼150 residues preceding M1) and S2 (the extracellular loop between M3 and M4), are found in glutamate receptor subunits and show sequence homology with bacterial periplasmic amino acid binding proteins (Stern-Bach et al. 1994). Functional binding studies and structural data have shown that the S1/S2 segments form a ligand-binding pocket and confer ligand-specificity in AMPA/kainate receptors and in S1S2 fusion proteins (Kuusinen et al. 1995; Armstrong et al. 1998; Keinanen et al. 1998; Lampinen et al. 1998). A sketch of the proposed structure of ionotropic glutamate receptors is shown in Fig. 1.
For NMDA receptors, mutation of amino acid residues within the S1 and S2 domains has shown that the NR1 subunit controls glycine potency (Kuryatov et al. 1994; Hirai et al. 1996). Furthermore, recent structural data of the S1S2 domains within the NR1 subunits has revealed a number of similarities to that of the GluR2 binding pocket (Furukawa & Gouaux, 2003). Mutation of several residues in either the S1 or S2 domains of either NR2A or NR2B subunits results in reduced glutamate potency (Laube et al. 1997; Anson et al. 1998). In particular, one mutation, a threonine to alanine mutation at position 671 on the NR2A subunit reduced glutamate potency by three orders of magnitude, with little effect on glycine potency (Anson et al. 1998). Analysis of single-channel currents produced by NR2A(T671A)-containing NMDA receptors indicated that the major effect of this mutation was to disrupt the ability of glutamate to remain bound to the NR2A receptor subunit. Furthermore, this threonine residue is conserved in all AMPA, kainate and NMDA receptor subunits that bind glutamate.
To address the importance of the length of time glutamate is bound to the receptor, a mutation analogous to the T671A mutation in the NR2A subunit was introduced into the NR2D subunit (T692A). This mutation should disrupt the binding of glutamate. The mutant NR2D subunit cRNA was coexpressed with wild-type NR1a cRNA in Xenopus laevis oocytes and the effect of the mutation on glutamate and glycine potency was examined under two-electrode voltage-clamp configuration. Single-channel properties of the receptor were analysed in outside-out membrane patches and concentration jump experiments were performed to examine the macroscopic current decay of NR1a/NR2D(T692A) receptors following a brief application of glutamate.
Expression plasmid constructs and mutagenesis
Heterogous expression of NMDA receptors
cRNA was synthesized as run-off transcripts from linearized plasmid DNA using the Promega (Madison, WI, USA) RiboMAx RNA synthesis kit. Reactions were supplemented as previously described (Anson et al. 1998). NR1 and NR2D cRNAs were mixed in a nominal ratio of 1: 9 before injection. Xenopus laevis were killed, in accordance with current Home Office guidelines, by exposure to an overdose of anaesthetic. Oocytes were removed, defolliculated and injected with either NR1a/NR2D wild-type or NR1a/NR2D(T692A) cRNA mixtures. Oocytes were incubated at 19°C for 2–3 days in modified Barth's solution (mm): 88 NaCl, 1 KCl, 2.4 NaHCO3, 0.33 Ca(NO3)2, 0.82 MgSO4, 0.44 CaCl2, 15 Tris-Cl, adjusted to pH 7.4 with NaOH, supplemented with 50 IU ml−1 penicillin (BRL, Bethesda, MD, USA), 50 μg ml−1 streptomycin (BRL), and 30 μmd(–)-2-amino-5-phosphonopentanoic acid (APV) (Tocris Cookson, UK), followed by storage at 4°C until they were used for electrophysiological measurements (2–7 days after injection).
Responses to glutamate, aspartate, NMDA or glycine were measured with a two-electrode voltage-clamp amplifier (TEC05, npi electronics, Tamm, Germany or Geneclamp, Axon Instruments) at −60 mV using 0.5–2 mΩ electrodes filled with 3 m KCl. Oocytes were perfused in a modified Ca2+-free normal frog Ringer solution (NFR) solution supplemented by 1.8 mm BaCl2 as described by Anson et al. (1998). For glutamate, aspartate and NMDA dose–response measurements, the NFR was further supplemented with 20 μm glycine and for glycine dose–response measurements the NFR was supplemented with either 30 μm glutamate for wild-type receptors or 10 mm glutamate for mutant receptors.
Application of solutions was either controlled manually or by the computer program, CellWorks Lite (npi electronics). Data were filtered at 100 Hz, digitized at 300 Hz and reduced to a final sampling rate of 10 Hz (oversampling with equal weight averaging of 30 samples) before storage.
Data analysis for dose–response curves
Data analysis was performed using the program CVFit (see http://www.ucl.ac.uk/Pharmacology/dc.html). Dose–response curves were fitted individually for each oocyte with the Hill equation:
where nH is the Hill coefficient, Imax the maximum current, [A] is the concentration of agonist, and EC50 is the concentration of agonist that produces a half-maximum response. Each data point was then normalized to the fitted maximum of the dose–response curve. The normalized values were then pooled and averaged for each construct and fitted again with the Hill equation, with the maximum and minimum for each curve being constrained to asymptote to 1 and 0, respectively.
Before making patch-clamp recordings the vitelline membrane of each oocyte was removed in a hypotonic solution containing (mm): 200 sodium methylsulphate, 20 KCl, 1 MgCl, 10 Hepes (pH 7.4 with KOH). Outside-out patches were held at −100 mV and recordings were made in an external solution containing (mm): 125 NaCl, 3 KCl, 1.25 NaH2PO4, 20 Hepes, 0.85 CaCl2 plus 20 μm glycine (pH 7.4 with NaOH). Patch pipettes were made from thick-walled borosilicate glass (Clark Electromedical Instruments, Pangbourne, UK), and contained (mm): 141 potassium gluconate, 2.5 NaCl, 10 Hepes, 11 EGTA (pH 7.4, with KOH). After fire-polishing the tips, patch pipettes had resistances of 10–20 mΩ. Single-channel activity of NR2D(T692A)-containing NMDA receptors was evoked by either glutamate (200 μm), aspartate (100 μm) or NMDA (100 μm) and recorded with an Axopatch-1D amplifier (Axon Instruments, Foster City, CA, USA) and recorded on digital audio tape (Biologic DTR 1205, Biologic Instruments, Claix, France).
Single-channel currents were replayed from digital audio tape, filtered at 2 kHz and digitized at 20 kHz. The amplitudes of single channel currents and their durations were measured using the SCAN program and analysed using the EKDIST program (Colquhoun & Sigworth, 1995; available from http://www.ucl.ac.uk/Pharmacology/dc.html). A resolution of 40 μs was imposed on both openings and closings in each data record. Distributions of fitted amplitudes that had durations greater than 415 μs were fitted with a mixture of two Gaussian components using method of maximum likelihood. Distributions of open periods, shut times, burst lengths and total open time per burst were fitted with either mixtures of exponential components using the method of maximum likelihood. Bursts of single-channel openings were defined by a critical gap length (tcrit) that was calculated to give equal numbers of misclassified events, i.e. the number of gaps that were incorrectly assigned as being within a burst when they were in fact between bursts was equal to the number that were classified as being between bursts when they were in fact within bursts (see Magleby & Pallotta, 1983; Clapham & Neher, 1984). To examine the frequency of direct transitions between two conductance levels the durations of the open period before and after the direct transition needed to be longer than two filter rise times (332 μs) to be included in the analysis. A critical amplitude (A crit) was determined to separate the two conductance levels that produced equal numbers of misclassified events. This amplitude is denoted by the dashed line in the asymmetry plots. In such histograms the occurrence of a transition between two states (represented as the ith amplitude level followed by the (i+1)th amplitude, where i is an integer) is indicated by a single point.
Concentration jump experiments
Fast concentration jumps were achieved by the rapid switching, across the tip of a patch pipette, of two solutions flowing from either side of a piece of theta glass. The control solution was supplemented with 20 μm glycine, and the test solution contained 10 mm glutamate (+ 20 μm glycine). Outside-out patches, held at −100 mV, were exposed to glutamate for 100 ms, and jumps were made every 10 s. Measurement of liquid junction potentials at the end of experiments indicated that the exchange of control and test solutions occurred within 300 μs. The macroscopic agonist-evoked currents were filtered at 2 kHz and sampled at 10 kHz onto a computer hard disk via a CED 1401 Plus interface (CED Instruments, Cambridge, UK) using the computer program CJUMP. Individual jumps were averaged and the resulting mean current fitted (least squares) with a single exponential curve using the CJFIT program (available from http://www.ucl.ac.uk/Pharmacology/dc.html).
The NR2D(T692A) mutation
Figure 1A shows a partial amino acid sequence alignment of bacterial amino acid binding proteins and mammalian glutamate receptor subunits. Within the S2 region of NMDA NR2 receptor subunits and AMPA/kainate receptor this threonine residue is conserved. Figure 1B shows a cartoon of the topology of the NR2D subunit and the location of the S1 and S2 regions that form a ‘clamshell’ structure to bind glutamate and the location of the T692 residue in the NR2D subunit is highlighted.
NR1a/NR2D(T692A) receptor-channels show a reduction in glutamate potency
Figure 2 shows individual responses recorded under two-electrode voltage clamp to increasing concentrations of glutamate (in the presence of 20 μm glycine) from an oocyte expressing wild-type NR1a/NR2D NMDA receptor-channels (Fig. 2A) and from an oocyte expressing NR2D subunits carrying the T692A point mutation (Fig. 2B). It is apparent from these panels that considerably higher concentrations of glutamate are required to evoke responses in the oocyte expressing the mutant receptor-channel. However currents evoked by glutamate in oocytes expressing NR1a/NR2D(T692A) receptors (Imax= 290 ± 48 nA, n= 9) were slightly greater than those evoked by glutamate in oocytes expressing wild-type NR1a/NR2D receptors (Imax= 153 ± 34 nA, n= 5). The fact that we did not observe any reduction in the maximal current in oocytes injected with a similar amount of cRNA coding for mutated NR2D receptor compared to those injected with the cRNA for the wild-type subunit indicates that this mutation is unlikely to affect, to any great extent, the gating of the channel. A series of concentration–response curves were generated from oocytes expressing either NR1a/NR2D and NR1a/NR2D(T692A) receptor-channels. The pooled data are shown in Fig. 2C and illustrate clearly that the NR2D(T692A) mutation shifts the glutamate dose–response curve far to the right compared to the wild-type. The glutamate potency for the NR1a/NR2D(T692A) receptor-channels (glutamate EC50= 703 ± 43 μm, n= 9) is reduced by 1562-fold compared to wild-type (glutamate EC50= 0.45 ± 0.04 μm, n= 5). Despite the shift in glutamate potency there is little change in the Hill slope for the concentration–response curve of the mutant receptors when compared to wild-type receptor-channels (1.50 ± 0.13 and 1.31 ± 0.15, respectively). To characterize further the effects of the threonine to alanine mutation in the NR2D receptor subunit we examined the potency of two other NMDA receptors agonists on both wild-type NR2D- and NR2D(T692A)-containing receptor-channels. For both NMDA and aspartate there is a clear rightward shift in potency for responses recorded from receptor-channels carrying the point mutation. Overall the EC50 and Hill slope (nH) values for these curves are: for aspartate (Fig. 2D), EC50(WT)= 3.3 ± 0.3 μm, nH= 1.13 ± 0.1; EC50(T692A)= 191 ± 20 μm, nH= 1.29 ± 0.1; and for NMDA (Fig. 2E), EC50(WT)= 3.7 ± 0.2 μm, nH= 1.37 ± 0.1; EC50(T692A)= 117 ± 7 μm, nH= 1.58 ± 0.1 (n= 5–9 for each curve). Thus the mutation causes an approximately 1600-fold shift in glutamate potency but more modest shifts in aspartate and NMDA potencies of approximately 60-fold and 30-fold, respectively. In contrast to its effects on glutamate, NMDA and aspartate potencies, the mutation has little effect on glycine potency (Fig. 2F), with EC50(WT)= 0.08 ± 0.01 μm and EC50(T692A)= 0.11 ± 0.01 μm (n= 4). This is in agreement with our previous demonstration that a series of mutations in the S1 and S2 ligand binding domains of the NR2A NMDA receptor subunit have little effect on glycine potency (Anson et al. 1998).
Such large shifts in potency are indicative of the mutation having an effect on agonist binding rather than gating (see Discussion and also Anson et al. 1998, 2000). To provide further evidence of this we investigated the effect of the T692A mutation on single-channel currents and receptor deactivation following rapid changes in agonist concentration.
Single-channel amplitudes and transition asymmetry
Figure 3 compares the single-channel events evoked by glutamate and recorded at a holding potential of −100 mV from an outside-out membrane patch excised from either an oocyte expressing wild-type NR1a/NR2D channels (Fig. 3A) or one expressing NR1a/NR2D(T692A) channels (Fig. 3B). It is clear that the point mutation does not affect the amplitude of these events when compared to wild-type. NR1a/NR2D(T692A) channels displayed two conductance levels that are typical of NR2D-containing NMDA receptors. Amplitude histograms from two recordings are shown in Fig. 3C and D and each is fitted with a mixture of two Gaussian components. Overall the mean amplitudes (and relative areas) for the larger (main) conductance level (recorded at −100 mV) were 4.05 ± 0.19 pA (61 ± 4%) and 3.97 ± 0.10 pA (69 ± 3%) for wild-type (n= 4) and mutant (n= 6) receptor-channels, respectively. The mean values of the lower (sub) conductance level were 2.24 ± 0.07 and 2.03 ± 0.07 pA for wild-type and mutant receptor-channels, respectively. The values for the amplitudes and areas for each component in the distributions of wild-type and mutant receptor-channels are not significantly different (P > 0.05 in all tests). The values we report here for the amplitude of the main and subconductance levels are greater than those we reported in our previous study of recombinant NR1a/NR2D NMDA receptor-channels, also expressed in oocytes (see Wyllie et al. 1996). However, in our earlier study the internal recording solution was Na+-based (rather than K+-based, as used in the present study) and this may account for the differences we have observed.
A characteristic feature of both recombinant and native NR2D-containing NMDA receptors is that there is an asymmetry of transition frequencies between the main and subconductance levels in that direct transitions beginning in the main level and going to the subconductance level are approximately twice as frequent as those starting in the subconductance level and going to the main conductance level (for example see Wyllie et al. 1996; Cull-Candy & Usowicz, 1987; Misra et al. 2000). These two types of transitions are illustrated in Fig. 3E. The NR1/NR2D(T692A) channels also show the transitional asymmetry. Figure 3E and F shows typical asymmetry plots for wild-type and mutant channels, respectively. Overall for wild-type channels 63 ± 4% of direct transitions between the two conductance levels started in the main level and went to the subconductance level, whereas for the NR2D(T692A)-containing channels 67 ± 2% of direct transitions were of this type. These values are not significantly different from each other (P > 0.4).
We also examined the properties of NR1a/NR2D(T692A) channels when activated by either aspartate (100 μm) or NMDA (100 μm) (n= 3 patches for each agonist) (Fig. 4). The amplitudes of NR1a/NR2D(T692A) single-channel currents evoked by aspartate or NMDA were not significantly different from those produced by glutamate and on average were 3.96 ± 0.27 pA (relative area 62 ± 2%) and 2.02 ± 0.17 pA (relative area 38 ± 2%) for aspartate-activated events and 3.98 ± 0.36 pA (60 ± 4%) and 1.99 ± 0.25 pA (40 ± 4%) for NMDA-activated single-channel currents. Typical amplitude histograms for aspartate- and NMDA-evoked single-channel currents are illustrated in Fig. 4A and B. As was the case for glutamate, aspartate and NMDA, each gave rise to events that displayed temporal asymmetry in direct transitions (for examples see Fig. 4C and D). The mean values for the percentage of transitions starting in the main conductance level and going to the subconductance level were 64 ± 0.3% for aspartate-evoked events and 60 ± 3% for NMDA-evoked events. These values are not statistically different from those obtained from recordings made with glutamate as the agonist.
Kinetic properties of NR1a/NR2D(T692A) receptor-channels
Table 1 summarizes the mean values for open periods, shut times and burst lengths for glutamate-evoked NR1a/NR2D(T692A) channel activity.
|τ1 (μs)||τ2 (ms)||τ3 (ms)||τ4 (ms)||τ5 (ms)||Distribution mean (ms)|
|Open period||71 ± 16||0.667 ± 0.060||1.51 ± 0.12||—||—||0.92 ± 0.09|
|(13 ± 4%)||(47 ± 6%)||(40 ± 7%)|
|Shut time||39 ± 5||0.675 ± 0.174||15.0 ± 3.3||217 ± 93||915 ± 470||111 ± 62|
|(26 ± 4%)||(14 ± 2%)||(18 ± 8%)||(25 ± 6%)||(17 ± 5%)|
|Burst length||180 ± 55||2.24 ± 0.97||66.7 ± 30.2||—||—||48.6 ± 26.2|
|(7 ± 3%)||(34 ± 10%)||(59 ± 8%)|
|Open time per burst||194 ± 77||1.71 ± 0.37||6.45 ± 1.25||—||—||3.60 ± 1.03|
|(8 ± 3%)||(49 ± 13%)||(43 ± 13%)|
Open periods. Distributions of open periods of NR1a/NR2D(T692A) channels were best fitted with a mixture of three exponential components, as we have previously found for wild-type NR1a/NR2D receptors (Wyllie et al. 1998). A typical open period distribution is illustrated in Fig. 5A. The dashed line represents the mean distribution of open periods for wild-type NR1a/NR2D receptors, reported by Wyllie et al. (1998) and scaled appropriately to take account of the number of events included in the particular distribution shown. Comparison of the mean values for wild-type and mutant channels indicates that the T692A mutation gives rise to a shortening of mean open period for events contained within the third (longest) component of open period distributions. The mean value for the third component reported in our previous study was 2.58 ± 0.14 ms (Wyllie et al. 1998) compared with 1.51 ± 0.12 ms for NR1a/NR2D(T692A) channels. This accounts for the difference in the overall mean value for open periods between wild-type and mutant channels: 1.50 ± 0.13 ms for wild-type and 0.92 ± 0.09 ms for mutant. The mean values for the time constants of the first and second components contained in open period distributions of wild-type and mutant channels are not significantly different. In addition, the mean values for the relative areas of each of the components contained within open period distributions of wild-type and mutant channels are not significantly different.
Shut times. Shut time distributions of NR1a/NR2D(T692A) channels were best fitted with a mixture of five exponential components whereas six components are required to describe wild-type distributions (Wyllie et al. 1998). A typical shut time distribution for NR1a/NR2D(T692A) channels is shown in Fig. 5B. The fourth and fifth components within each shut time distribution were rather variable from patch to patch, with τ4 ranging between 53 and 557 ms and τ5 ranging between 161 and 2548 ms. Given this variability, we made the assumption that these two components represented shut times that occurred between separate activations (bursts) of channel activity. This is in contrast to our findings for wild-type NR1a/NR2D channels where only the last component of the shut time distribution was considered to represent shut times between separate activations. In these shut time distributions the time constant of the fifth component varied only 2-fold (140–290 ms; see Wyllie et al. 1998). The fact that two shut time components contained within the NR1a/NR2D(T692A) distributions are considered to be between bursts is analogous with the observation we made with NR1a/NR2A channels carrying the homologous T671A point mutation, where we concluded that the latter two components represented gaps between separate activations of these receptor-channels (see Anson et al. 2000). A tcrit was calculated as the gap most likely to separate the shut times contained within the third and fourth components of each distribution and thus allow us to identify individual activations (bursts) of channel openings. The mean tcrit calculated by the method of misclassifying equal numbers of events from each component of the distribution (Clapham & Neher, 1984) was 20.8 ± 4.4 ms. In the distribution shown in Fig. 5B the tcrit is 23 ms.
Burst lengths. Following identification of an appropriate tcrit, distributions of burst lengths were constructed. The distribution resulting from imposing a tcrit of 23 ms on the data illustrated in Fig. 5A and B is shown in Fig. 5C. Burst length distributions were best fitted with a mixture of three exponential components. Distributions of wild-type NR1a/NR2D individual channel activations show extremely long-lasting bursts of activity – the slowest two components in such distributions have time constants of around 1400 and 5000 ms (see Wyllie et al. 1998). Such long-lasting bursts are missing in NR1a/NR2D(T692A) distributions where the longest component has a mean of 66.7 ms. Thus the T692A mutation results in an approximately 80-fold reduction in the time constant of the slowest component contained within burst length distributions. Analysis of distributions of total open time per burst indicated that these were best described by a mixture of three exponential components and had an overall mean of 3.60 ± 1.03 ms which gave a probability that a channel was open during a burst of 0.074 ± 0.03. This open probability is not significantly different from that observed for wild-type NR1a/NR2D channels of 0.04 ± 0.02 (P > 0.05) (see Wyllie et al. 1998). The top three traces in Fig. 5D show three bursts of NR1a/NR2D(T692A) channel openings while the lower trace in this panel is an example of a wild-type NR1a/NR2D burst but on a 100-fold slower time base.
Deactivation of macroscopic currents mediated by NR1a/NR2D(T692A) channels
Assuming we have correctly identified individual activations of NR1a/NR2D(T692A) channels we would predict that the deactivation of NR1a/NR2D(T692A) receptor-mediated currents would be considerably faster than those seen for wild-type channels, which decay with a time constant of between 3 and 5 s (see Monyer et al. 1994; Vicini et al. 1998; Wyllie et al. 1998; Misra et al. 2000).
Rapid applications of glutamate (10 mm) were applied for 100 ms to outside-out patches, held at −100 mV, and excised from oocytes expressing NR1a/NR2D(T692A) receptor-channels. We used a 100 ms application to ensure that we recorded a plateau to the current, as if the T692A mutation affects the dissociation rate of glutamate, briefer applications may not result in an equilibrium of agonist occupancy of the receptor being achieved. Figure 6A shows three individual sweeps recorded from a single experiment in which individual channel openings can be distinguished. Figure 6B shows the average of 75 jumps where the decay of the current following the removal of glutamate is fitted with a single exponential component with a time constant of 79 ms. The dashed line in Fig. 6B shows the comparable decay of wild-type NR1a/NR2D receptors. On average the deactivation of the NR1a/NR2D(T692A) channels could be described by a single exponential with a time constant of 72.6 ± 19.3 ms (n= 7). This is approximately 60-fold faster than the deactivation of wild-type NR1a/NR2D receptor-channels (Wyllie et al. 1998) and this difference is illustrated in Fig. 6C where the deactivation of the mutant and wild-type NR2D-containing NMDA receptors have been superimposed.
A conserved residue within the binding pocket that is involved in agonist binding
A number of structure–function studies have identified the residues in the S1 and S2 domains of the glutamate receptor subunits that control glutamate binding (Kuusinen et al. 1995; Paas et al. 1996; Laube et al. 1997; Anson et al. 1998, 2000; Wo et al. 1999). These findings have correlated well with evidence obtained from the analysis of the crystal structure of the GluR2 S1S2 protein fragment (Armstrong et al. 1998; Armstrong & Gouaux, 2000). Previously, we have reported the effects of mutating a threonine residue in the S2 domain of the NR2A subunit, which results in the disruption of glutamate binding to this receptor subunit and alters the kinetic behaviour of single-channel currents (Anson et al. 1998, 2000).
In the present study we have mutated the homologous residue in the NR2D subunit. This threonine residue is conserved throughout the ionotropic glutamate receptor subunit family in subunits that bind glutamate (see Fig. 1A), implying that this residue plays a pivotal role in glutamate binding, not only among NMDA NR2 receptor subunits but also within the AMPA and kainate receptor families. Indeed, X-ray crystallographic studies have confirmed the role of the threonine residue at position 655 (the homologous residue in the GluR2 subunit) in agonist binding of the γ-carboxyl group of glutamate to the GluR2 S1S2 protein fragment (Armstrong & Gouaux, 2000). It is intriguing to note that in both NR1 and NR3 subunits this threonine residue is absent and indeed in the NR3 subunit it is substituted with an alanine (Ciabarra et al. 1995; Sucher et al. 1995; Chatterton et al. 2002; see alignment in Fig. 1A) – thus, in part, this may explain the insensitivity of the NR3 subunits to activation by glutamate.
The reduction in agonist potency is consistent with a change in binding
The NR2D(T692A) mutation reduced the glutamate potency of NR1a/NR2D(T692A) receptors by three orders of magnitude, with little change in Hill slope compared to NR1a/NR2D wild-type receptors. The shift of around 1600-fold in glutamate potency is more than that seen with aspartate (approximately 60-fold shift) and NMDA (approximately 30-fold shift). As the EC50 for any agonist acting on a receptor is dependent on all the rate constants in the reaction scheme that describes its activation, it might be anticipated that we should have observed a change in the EC50 for the co-agonist, glycine, since the T692A mutation has, most likely, increased the dissociation rate constant for glutamate. However, the concentration–response curves for glycine were generated under conditions where NR2 receptor subunits were fully liganded by the saturating concentrations of glutamate used in these particular experiments (likewise, glutamate, aspartate and NMDA concentration–response curves were generated in the presence of saturating concentrations of glycine). Thus, the fact that we do not see a change in EC50 for glycine in these experiments is evidence against the idea that the T692A mutation affects the rate constants describing the binding of glycine to the NR1 receptor subunit.
Clearly a major difference between the agonists glutamate, aspartate and NMDA is in the length of the carbon chain backbone of the side chain of these ligands. The fact that the potencies of aspartate and NMDA are reduced to a lesser extent may suggest that the hydrogen bonding of these agonists to this threonine residue can be compensated for in a manner that is not achievable when glutamate occupies the binding site. Until a crystal structure for an NMDA NR2 binding site is elucidated it remains to be seen as to how various ligands orientate in the binding site and indeed whether different NMDA receptor agonists induce different degrees of domain closure, as has been reported to be the case when full and partial agonists occupy the ligand binding site of the GluR2 S1S2 protein fragment (Armstrong & Gouaux, 2000; Jin & Gouaux, 2003; Jin et al. 2003).
A reduction in glutamate potency could be explained not only by a change in binding but by also a change in gating (Colquhoun, 1998). However, such a large reduction in glutamate potency with little change in maximal response and Hill slope cannot be explained by a change in gating alone. If a change in gating had caused a 1600-fold shift in glutamate EC50, we would expect the maximal response to be reduced by a factor of about 2.5 million (see Colquhoun, 1998). This has not been observed here given that the maximal responses recorded from oocytes injected with similar amounts of cRNA were similar and is consistent with the data obtained from receptors containing the homologous mutation (T671A) in the NR2A subunit (Anson et al. 1998). Indeed wild-type NR2D-containing NMDA receptors have an extremely low probability of being open during an activation and are in fact closed for around 96% of the activation (see Wyllie et al. 1998). This observation implies that the ‘efficacy’ of glutamate at NR2D-containing receptors is extremely low. Consistent with the idea that gating is little affected by this mutation is our observation that the single-channel conductance of NR2D(T692A)-containing receptors is not different from wild-type. In as much as transitions between the main conductance and subconductance levels can be defined as a ‘gating’, the NR2D(T692A) mutation did not alter the temporal asymmetry of transition frequency between these two conductance levels when the channels were activated by any of the agonists used in this study. We did, however, observe a significant decrease in the mean value of the time constant of the third component of open period distributions of NR1a/NR2D(T692A) events when compared to wild-type. This reduction accounts for the difference in the overall mean value for open periods of the mutant (0.92 ms) and wild-type channels (1.50 ms; see Wyllie et al. 1998). Although this small decrease in open period duration cannot possibly account for the large change in glutamate potency observed, it may imply that coupling of the binding site to the activation gate of the channel has been disrupted in such a way as to increase the rate of channel closure from this open state. Given our data, we suggest that the most likely explanation for the differences in the potency shifts seen with glutamate, aspartate and NMDA is that they are attributable to the way these agonists interact with the binding site rather than any effect on gating properties of the channel.
Single-channel and deactivation properties
Perhaps the most striking kinetic feature of NR1a/NR2D channel activations is their very long-lasting duration coupled with the fact that the channel remains closed for most of the activation (Wyllie et al. 1998). The slowest component in the burst length distribution is around 5000 ms, which is to be expected given the fact that deactivation of channel activity following brief exposure to glutamate of both native and recombinant NR2D-containing NMDA receptors takes several seconds (for example see Monyer et al. 1994; Vicini et al. 1998; Wyllie et al. 1998; Misra et al. 2000). In order to identify NR1a/NR2D(T692A) bursts of opening we chose to use a critical gap length that separated the third and fourth components contained within the shut time distributions. The first three components in such distributions showed the least variability, suggesting that they were not dependent on agonist concentration. This, together with the fact that deactivation rates of NR1a/NR2D(T692A) currents are considerably more rapid that those seen with wild-type, makes it unlikely that a different value of tcrit would correctly identify individual activations. The slowest component contained within distributions of NR1a/NR2D(T692A) burst lengths has a duration approximately 100 times less than the slowest component in corresponding distributions of wild-type burst lengths. From the values shown in Table 1 it can be seen that the overall probability of the channel being open during a burst is still low (< 0.1).
Deactivation times of macroscopic responses from NR1a/NR2D(T692A) receptors, after brief applications of glutamate, were around 60 times less than wild-type NR1a/NR2D receptors. Similarly an increase in the rate of decay was observed for NR1a/NR2A(T671A) receptors. Given that the deactivation of the macroscopic current is determined by the underlying kinetic behaviour of the single-channel events this increased deactivation rate is entirely expected. Indeed if one uses the model for NMDA receptor activation proposed by Lester & Jahr (1992) and increases the rate of dissociation by a factor of 1600 then the predicted change in the deactivation rate of a simulated macroscopic current is around 10-fold. This is not surprising, as the duration of individual activations will depend not only on the dissociation rate of the agonist but also on the duration and the number of openings (as well as the durations of shut times) contained within an activation. Therefore, just as changes in the EC50 are difficult to ascribe to alterations in either binding or gating parameters, the effect of altering the dissociation rate constant is difficult to quantify without knowledge of a kinetic scheme that adequately describes receptor activation. Recently, variations of the kinetic scheme originally proposed by Lester & Jahr (1992) describing the activation of recombinant NMDA receptors have been published (see Banke & Traynelis, 2003; Popescu & Auerbach, 2003). Neither of these newer schemes can account for the complex activation structure of NR2D-containing NMDA receptors and it remains to be seen whether an ‘all encompassing’ scheme, in which only the rate constants describing the entry to and exit from different states vary, can describe the heterogeneity of NMDA channel activations seen with different NR2 combinations. If this can be achieved then it may be easier to identify how alterations in the dissociation rate constant influence the observed EC50 for agonist action and more importantly the nature of individual channel activations. Nonetheless the channel properties we have described for NR1a/NR2D(T692A) receptors are entirely consistent with the notion that this mutation primarily affects the ability of glutamate to remain bound to the receptor.
NR2D-containing NMDA receptors appear to play an unusual and perhaps special functional role for ionotropic glutamate receptors due to their apparent exclusion from synapses (see for example Momiyama, 2000; Brickley et al. 2003) and their very slow deactivation kinetics. While our results do not address specifically the role played by such receptors they do indicate that the long-lasting activations most likely are a consequence of glutamate remaining bound to this receptor for the duration of the activation rather than initiating a series of conformational changes in the protein from which the channel can open and close in the absence of bound ligand.
- 1994). Differential expression of five N-methyl-D-aspartate receptor subunit mRNAs in the cerebellum of developing and adult rats. J Comp Neurol 347, 150–160. , , , & (
- 1998). Identification of amino acid residues of the NR2A subunit that control glutamate potency in recombinant NR1/NR2A NMDA receptors. J Neurosci 18, 581–589. , , , & (
- 2000). Single-channel analysis of a NMDA receptor possessing a mutation in the region of the glutamate binding site. J Physiol 527, 225–237. , , & (
- 2000). Mechanisms for activation and antagonism of an AMPA-sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core. Neuron 28, 165–181. & (
- 1998). Structure of a glutamate-receptor ligand-binding core in complex with kainate. Nature 395, 913–917. , , & (
- 2003). Activation of NR1/NR2B NMDA receptors. Nat Neurosci 6, 144–152. & (
- 2003). NR2B and NR2D subunits coassemble in cerebellat golgi cells to form a distinct NMDA receptor subtype restricted to extrasynaptic site. J Neurosci 23, 4958–4966. , , , & (
- 2002). Excitatory glycine receptors containing the NR3 family of NMDA receptor subunits. Nature 415, 793–798. , , , , , et al. (
- 1995). Cloning and characterization of chi-1: a developmentally regulated member of a novel class of the ionotropic glutamate receptor family. J Neurosc 15, 6498–6508. , , , , & (
- 1984). Substance P reduces acetylcholine-induced currents in isolated bovine chromaffin cells. J Physiol 347, 255–277. & (
- 1998). Binding, gating, affinity and efficacy: The interpretation of structure-activity relationships for agonists and of the effects of mutating receptors. Br J Pharm 125, 924–947. (
- 1995). Fitting and statistical analysis of single-channel records. In Single Channel Recording, ed. SakmannB & NeherE, 2nd edn, pp. 483–587. Plenum Press, New York . & (
- 2001). NMDA receptor subunits: diversity, development and disease. Curr Opin Neurobiol 11, 327–335. , & (
- 1987). Multiple-conductance channels activated by excitatory amino acids in cerebellar neurons. Nature 325, 525–528. & (
- 1999). The glutamate receptor ion channels. Pharmacol Rev 51, 7–61. , , & (
- 2003). Mechanisms of activation, inhibition and specificity: crystal structures of the NMDA receptor NR1 ligand-binding core. EMBO J 22, 2873–2885. & (
- 1996). The glycine binding site of the N-methyl-D-aspartate receptor subunit NR1: identification of novel determinants of co-agonist potentiation in the extracellular M3–M4 loop region. Proc Natl Acad Sci USA 93, 6031–6036. , , , & (
- 1993). Molecular characterisation of the family of the N-methyl-D-aspartate receptor subunits. J Biol Chem 268, 2836–2843. , , , , , et al. (
- 2003). Structural basis for partial agonist action at ionotropic glutamate receptors. Nat Neurosci 6, 803–810. , , , & (
- 2003). Probing the function, conformational plasticity, and dimer-dimer contacts of the GluR2 ligand-binding core: studies of 5-substituted willardiines and GluR2, S1S2 in the crystal. Biochemistry 42, 5201–5213. & (
- 1998). Characterization of the kainate-binding domain of the glutamate receptor GluR-6 subunit. Biochem J 330, 1461–1467. , & (
- 1996). Multiple structural elements determine subunit specificity of Mg2+ block in NMDA receptor channels. J Neurosci 16, 3549–3558. & (
- 1994). Mutational analysis of the glycine-binding site of the NMDA receptor: structural similarity with bacterial amino acid-binding proteins. Neuron 12, 1291–1300. , , & (
- 1992). Molecular diversity of the NMDA receptor channel. Nature 358, 36–41. , , , , , et al. (
- 1995). Molecular dissection of the agonist binding site of an AMPA receptor. EMBO J 14, 6327–6332. , & (
- 1998). AMPA receptors and bacterial periplasmic amino acid-binding proteins share the ionic mechanism of ligand recognition. EMBO J 17, 4704–4711. , , & (
- 1997). Molecular determinants of agonist discrimination by NMDA receptor subunits: analysis of the glutamate binding site on the NR2B subunit. Neuron 18, 493–503. , , , & (
- 1992). NMDA channel behaviour depends on agonist affinity. J Neurosci 12, 635–643. & (
- 1983). Burst kinetics of single calcium-activated potassium channels in cultured rat muscle. J Physiol 344, 605–623. & (
- 2000). Slow deactivation kinetics of NMDA receptors containing NR1 and NR2D subunits in rat cerebellar Purkinje cells. J Physiol 525, 299–305. , , & (
- 2000). Distinct synaptic and extrasynaptic NMDA receptors identified in dorsal horn neurones of the adult rat spinal cord. J Physiol 523, 621–628. (
- 1994). Developmental and regional expression in the rat brain and functional properties of four NMDA receptors. Neuron 12, 529–540. , , , & (
- 1992). Heteromeric NMDA receptors: molecular and functional distinction of subtypes. Science 256, 1217–1221. , , , , , et al. (
- 1996). Identification of the amino acid subsets accounting for the ligand binding specificity of a glutamate receptor. Neuron 17, 979–990. , , , & (
- 2003). Modal gating of NMDA receptors and the shape of their synaptic response. Nat Neurosci 6, 476–483. & (
- 1994). Agonist selectivity of glutamate receptors is specified by two domains structurally related to bacterial amino acid-binding proteins. Neuron 13, 1345–1357. , , , , & (
- 1992). Single-channel conductances of NMDA receptors expressed from cloned cDNAs: comparison with native receptors. Proc Roy Soc Lond B 250, 271–277. , , & (
- 1995). Developmental and regional expression pattern of a novel NMDA receptor-like subunit (NMDAR-L) in the rodent brain. J Neurosci 15, 6509–6520. , , , , , et al. (
- 1992). Structures and properties of seven isoforms of the NMDA receptor generated by alternative splicing. Biochem Biophys Res Comm 185, 826–832. , , , & (
- 1998). Functional and pharmacological differences between recombinant N-methyl-D-aspartate receptors. J Neurophysiol 79, 555–566. , , , , , et al. (
- 1992). Developmental changes in distribution of NMDA receptor channel subunit mRNAs. Neuroreport 3, 1138–1140. , , & (
- 1999). Cysteine mutagenesis and homology modeling of the ligand-binding site of a kainate-binding protein. J Biol Chem 274, 37210–37218. , , , & (
- 1998). Single-channel activations and concentration jumps: comparison of recombinant NR1a/NR2A and NR1a/NR2D NMDA receptors. J Physiol 510, 1–18. , & (
- 1996). Single-channel currents from recombinant NMDA NR1a/NR2D receptors expressed in Xenopus oocytes. Proc Roy Soc Lond B 263, 1079–1086. , , , & (