Triheteromeric NMDA receptors at hippocampal synapses

Abstract

NMDA receptors are composed of two GluN1 (N1) and two GluN2 (N2) subunits. Constituent N2 subunits control the pharmacological and kinetic characteristics of the receptor. NMDA receptors in hippocampal or cortical neurons are often thought of as diheteromeric, meaning that they contain only one type of N2 subunit. However, triheteromeric receptors with more than one type of N2 subunit also have been reported, and the relative contribution of diheteromeric and triheteromeric NMDA receptors at synapses has been difficult to assess. Because wild-type hippocampal principal neurons express N1, N2A, and N2B, we used cultured hippocampal principal neurons from N2A and N2B knock-out mice as templates for diheteromeric synaptic receptors. However, summation of N1/N2B and N1/N2A EPSCs could not account for the deactivation kinetics of wild-type EPSCs. To make a quantitative estimate of NMDA receptor subtypes at wild-type synapses, we used the deactivation kinetics and the effects of the competitive antagonist NVP-AAM077. Our results indicate that three types of NMDA receptors contribute to wild-type EPSCs, with at least two-thirds being triheteromeric receptors. Functional isolation of synaptic triheteromeric receptors revealed deactivation kinetics and pharmacology that were distinct from either diheteromeric receptor subtype. Because of differences in open probability, synaptic triheteromeric receptors outnumbered N1/N2A receptors by 5.8 to 1 and N1/N2B receptors by 3.2 to 1. Our results suggest that triheteromeric NMDA receptors must either be preferentially assembled or preferentially localized at synapses.

Figure 1.

Model system for diheteromeric synaptic NMDA receptors. A, EPSCs from N2A KO (gray trace) and N2B KO neurons (black trace), superimposed and peak-scaled, highlighting the difference in their deactivation time course. B, Weighted time constant of decay (τ_w) from N2A and N2B KO neuron EPSCs (n = 35). C, D, N2B and N2A KO EPSCs are sensitive to zinc (100 nm, free) and ifenprodil (3 μm), as expected for A-type or B-type receptors, respectively. E, EPSCs from wild-type (left) and DKO neurons (right). The wild-type EPSC has an (R)-CPP-sensitive component (black trace) and an NBQX-sensitive component (red trace), indicative of NMDA and AMPA receptors, respectively. The EPSCs from the DKO neuron have no contribution from NMDA receptors, as indicated by the similarity of the control and (R)-CPP traces. F, G, Whole-cell application of NMDA (100 μm) always (18/18) resulted in robust currents from wild-type (WT) neurons. In contrast, NMDA application to DKO neurons produced no current in most neurons (38/44) and small currents in six neurons.

Figure 2.

Deactivation kinetics of NMDA receptor-mediated EPSCs. A, Fast and slow time constants (τ_fast and τ_slow) from fitting the A-type, B-type, and wild-type (WT) neuron EPSC deactivation are plotted individually (red traces) ± 1 SD (gray regions). The relative contributions ± 1 SD of each time constant are shown at left. The black traces represent the EPSC deactivation from a representative neuron from each cell type. For each group, n = 35 neurons. B, Traces on the left show the superimposed and peak-scaled two-exponential functions using the mean deactivation parameters from A-type EPSCs (black trace) and B-type EPSCs (gray trace). The deactivation from a representative wild-type EPSC is plotted in red. The plot on the right shows that at 1000 ms after the peak, the A-type EPSC has returned to baseline, whereas EPSCs from wild-type and B-type neurons contribute a small amount of current at this long latency. The dashed rectangle on the left shows the region indicated in the plot on the right. Dashed lines indicate the range (200 ms) over which the amplitude measurement was made. C, Cumulative distribution of the estimate of the maximum amount of B-type current in the wild-type EPSCs (n = 51). The mean (red line) and 1 SD (dashed lines) are shown. D, Black trace shows the plot of a two-exponential function using the mean parameters from wild-type EPSC deactivation superimposed on a simulated EPSC deactivation made from the sum of B-type and A-type exponential functions (gray) ± 1 SD (gray region). The black dashed line shows the mean (23.9% B-type plus 76.1% A-type). The red trace is the fit of the mean, constrained with the mean τ_fast and τ_slow from wild-type EPSCs. Note that ∼20% of the simulated EPSC deactivation could not be fit with these parameters.

Figure 3.

Using a competitive antagonist to separate NMDA receptor subtypes. A, NVP-AAM077 dose-inhibition curves from A-type (closed circles), B-type (open circles), and wild-type (WT; red circles) EPSCs. Data from all neuron types were well fit with a single site isotherm (see Materials and Methods). Fits to data from A-type and B-type EPSCs are shown with dashed lines. The single-site isotherm fit of the wild-type EPSC dose-inhibition data (red trace) was much better (χ² = 0.34) than a fit using a two-site isotherm (χ² = 6.71) with the IC₅₀ values from A-type and B-type EPSCs and the contribution of each unconstrained (black trace). For this fit, the fractional contribution from B-type receptors was 0.54 ± 0.07. Each NVP concentration represented the mean reduction from 4 to 14 neurons from each of the three cell types. B, τ_w from A-type, B-type, and wild-type neuron EPSCs as a function of NVP concentration. As expected, τ_w of wild-type EPSCs increased with increasing NVP concentration, consistent with selective elimination of A-type receptors. Note that τ_w of wild-type EPSCs never became as slow as τ_w from B-type EPSCs. C, At high NVP concentration, the wild-type EPSC deactivation (black trace) was still slower than the mean B-type deactivation (gray trace). D, In wild-type neurons, the deactivation of the current resulting from subtracting the EPSC in 10 nm NVP (black trace) from the control EPSC is slower than the mean A-type EPSC deactivation (gray trace).

Figure 4.

Opening kinetics of synaptic NMDA receptor subtypes. A, Time course of the EPSC from a wild-type neuron (top, black trace) is accelerated in MK-801 (top, red trace) because channel openings are shortened. When the EPSC is integrated, the time at 60% charge transfer (Q₆₀) of the EPSC in MK-801 (25 μm) can be used as a relative measurement of how quickly NMDA receptor channels open after binding to synaptically-released glutamate. This is because MK-801 only binds to open channels, binds quickly, and does not dissociate during the course of the experiment. In MK-801, the ratio of charge at the time of the control EPSC peak (a) to the total charge (b) gives the probability of a channel having opened by the time of the peak. The red arrow indicates the Q₆₀ for this EPSC in MK-801. The Q₆₀ of the control EPSC is shown with the black arrowhead. Q₆₀ (B) and the P_o* (C) from A-type, B-type, and wild-type (WT) EPSCs. Wild-type EPSCs differed from A-type and B-type EPSCs in Q₆₀ (p < 10⁻⁵ and p < 0.001, respectively) and P_o* (p < 10⁻⁵ and p < 0.005, respectively). D, Q₆₀ of wild-type EPSCs (in red) as a function of NVP concentration. As shown, the Q₆₀ became slower as A-type receptors were blocked by NVP but never became as slow as B-type receptors. The Q₆₀ mean (dashed line) ± 1 SD (gray region) is shown for B-type (top) and A-type (bottom) EPSCs. Q₆₀ values for A-type and B-type receptors are from B. E, P_o* of wild-type EPSCs as a function of NVP concentration (in red). P_o* mean (dashed line) ± 1 SD (gray region) is shown for A-type (top) and B-type (bottom) EPSCs. P_o* values are from C.

Figure 5.

Quantitative estimate of the contribution of NMDA receptor subtypes to wild-type EPSCs. A, EPSCs from wild-type neurons were tested in five concentrations of NVP each and the data from each neuron were fit with a single-site isotherm. As shown, each fit (red lines) was steeper than predictions for mixtures of A-type and B-type receptors, ranging from 3:1 to 1:3 (gray shaded region). The solid black line describes where the contribution of A-type and B-type receptors is equal. The mean IC₅₀ from these fits was 30.0 ± 1.4 nm (n = 10). B, Dose-inhibition data from each neuron were combined with the NVP-induced changes in the deactivation kinetics and the known reduction of A- and B-type EPSCs in NVP to estimate the contribution of A-type, B-type, and AB-type receptors to the EPSCs from each cell (see Materials and Methods).

Figure 6.

Functional isolation of synaptic triheteromeric (AB-type) receptors. A, Outline of our strategy for isolation of AB-type NMDA receptor-mediated EPSCs in wild-type (WT) neurons. EPSCs were evoked at low frequency (0.1 Hz) while cells were perfused with NVP and MK-801 to prevent A-type channels from opening and to progressively block the remaining channels (B-type and AB-type). The NVP/MK-801 solution was removed, leaving EPSCs that result from A- and AB-type receptors (b). To enrich the remaining EPSC with AB-type receptors, the neurons were perfused with 50 nm NVP to block the majority of A-type receptors. B, EPSCs shown from different times in the isolation protocol (a–c) shown in A, indicating their relative amplitudes (left) and peak-scaled (right) to show how the deactivation changes during the course of the experiment. The inset (right) shows the EPSC at b compared with the deactivation of A-type EPSCs alone. The τ_w (C) and the estimate of the maximal B-type receptor contribution (D) do not differ between control and enriched AB-type receptors.

Figure 7.

Kinetic and pharmacological characteristics of AB-type receptors. A, Isolated AB-type EPSCs are reduced by zinc (left; 100 nm, free) or ifenprodil (right; 3 μm), but only zinc prolongs the EPSC deactivation, as shown by the peak-scaled EPSCs in the insets. Asterisk in the left panel indicates the point where the EPSC in zinc crosses the control EPSC, indicative of a prolongation, rather than simply revealing a slow EPSC component. WT indicates wild-type. B, For synaptic ligand-gated channels under nonequilibrium agonist conditions, the reduction in total charge as a function of reduction of the peak in response to competitive antagonists should fall on the unity line, as shown for NVP on A-type EPSCs (black circles and lines). The red dashed line is a linear fit to the data from NVP on A-type EPSCs, with the slope indicated in the figure (r² = 0.99). Zinc or ifenprodil each produce a slight potentiation of charge of A-type (open red circle) and B-type (open black circles) EPSCs from KO neurons, as indicated in the upward departure of these points from unity. In AB-type receptors, only zinc (100 nm, free) prolongs the EPSC (closed red circle), whereas ifenprodil (3 μm) does not (closed black circle). The NVP concentrations used were as follows (in nm): 1, 3, 10, 30, and 100.