Glutamate receptor subtypes mediating synaptic activation of prefrontal cortex neurons: relevance for schizophrenia

Abstract

Schizophrenia may involve hypofunction of NMDA receptor (NMDAR)-mediated signaling, and alterations in parvalbumin-positive fast-spiking (FS) GABA neurons that may cause abnormal gamma oscillations. It was recently hypothesized that prefrontal cortex (PFC) FS neuron activity is highly dependent on NMDAR activation and that, consequently, FS neuron dysfunction in schizophrenia is secondary to NMDAR hypofunction. However, NMDARs are abundant in synapses onto PFC pyramidal neurons; thus, a key question is whether FS neuron or pyramidal cell activation is more dependent on NMDARs. We examined the AMPAR and NMDAR contribution to synaptic activation of FS neurons and pyramidal cells in the PFC of adult mice. In FS neurons, EPSCs had fast decay and weak NMDAR contribution, whereas in pyramidal cells, EPSCs were significantly prolonged by NMDAR-mediated currents. Moreover, the AMPAR/NMDAR EPSC ratio was higher in FS cells. NMDAR antagonists decreased EPSPs and EPSP-spike coupling more strongly in pyramidal cells than in FS neurons, showing that FS neuron activation is less NMDAR dependent than pyramidal cell excitation. The precise EPSP-spike coupling produced by fast-decaying EPSCs in FS cells may be important for network mechanisms of gamma oscillations based on feedback inhibition. To test this possibility, we used simulations in a computational network of reciprocally connected FS neurons and pyramidal cells and found that brief AMPAR-mediated FS neuron activation is crucial to synchronize, via feedback inhibition, pyramidal cells in the gamma frequency band. Our results raise interesting questions about the mechanisms that might link NMDAR hypofunction to alterations of FS neurons in schizophrenia.

Figure 1.

PFC pyramidal cells and FS interneurons were identified based on their firing pattern and morphological properties. A, Top, Firing properties of a typical FS interneuron identified in layer 5. Bottom, Firing properties of a typical layer 5 pyramidal neuron. Some pyramidal cells had weakly bursting properties (not shown). B, Left, Adaptation ratio (last interspike interval/first interspike interval, see Materials and Methods) plotted as a function of spike duration at half peak amplitude for 30 pyramidal cells (triangles), 60 FS neurons (black circles), and 45 non-FS cells (gray circles). Right, Plot of the AHP amplitude versus spike duration at half-width. Cells classified as FS (see Materials and Methods) had spike duration ≤0.6 ms, AHP amplitude ≥15 mV, and adaptation ratio ≤1.2. For examples of non-FS cell properties, see supplemental Figure 1 (available at www.jneurosci.org as supplemental material). C, Examples of the morphological properties of pyramidal cells (left) and FS neurons (right) filled with biocytin during recording. All the FS cells had properties consistent with the basket cell morphological class (no chandelier neurons were found in this data sample). Axons shown in gray, dendrites in black.

Figure 2.

Contribution of AMPARs and NMDARs to sEPSCs in PFC neurons. A, Left, Examples of sEPSCs recorded from a pyramidal cell (gray) or a FS interneuron (black). Note the faster decay time course of individual sEPSCs in the FS neuron. Right, Averages of the sEPSCs recorded from the pyramidal cell and FS neuron are shown superimposed. No Mg²⁺ was added to the extracellular solution. B, Average sEPSCs recorded from a pyramidal cell (left) and a FS neuron (right) in control conditions (thin black), or after 15 min of d,l-AP5 (100 μm) application (gray). The NMDAR-mediated sEPSC waveform obtained by subtraction is shown as well (thick black). C, Bar graphs summarizing the differences in the decay time constant of the control average sEPSCs (AMPAR- and NMDAR-mediated) and the sEPSCs recorded after AP5 application (AMPAR-mediated). AP5 significantly accelerated the sEPSC decay in pyramidal cells but not in FS neurons. *p < 0.05 (sEPSC decay time, pyramidal cells, n = 11, control: 7.36 ± 0.75 ms, AP5: 4.63 ± 0.22 ms; FS cells, n = 10, sEPSC: 2.04 ± 0.20 ms, AP5: 1.84 ± 0.12 ms; two-factor ANOVA, cell type: F_(1,38) = 89.9 p < 0.00001; NMDAR: F_(1,38) = 11.9 p < 0.002; interaction: F_(1,38) = 8.7 p < 0.01; post hoc Fisher LSD tests control versus AP5: pyramidal cells, p < 0.0001, FS neurons: p = 0.736). D, Bar graph showing the AMPA/NMDA charge ratio, estimated from the sEPSC waveforms for both pyramidal cells and FS neurons. *p < 0.01 (AMPA/NMDA charge ratio, pyramidal cells, n = 11: 2.30 ± 0.48; FS neurons, n = 10: 5.05 ± 0.67; Mann–Whitney U test, U = 2.72, p < 0.005).

Figure 3.

Contribution of AMPARs and NMDARs to eEPSCs in PFC neurons. A, Left, Plot of the response amplitude versus stimulus number for eEPSCs recorded from a pyramidal neuron, in control conditions and after application before of the AMPAR antagonist CNQX (20 μm). Note that a significant eEPSC component was observed after CNQX application and was blocked by adding d,l-AP5 (100 μm). No Mg²⁺ was added to the extracellular solution. The inset shows recordings from the neuron depicted in the plot (calibration bars: 50 pA, 20 ms). Right, From recordings in a different pyramidal cell, shown superimposed are the average eEPSCs recorded in control conditions and after CNQX addition (thick and thin black, respectively) and the AMPAR-eEPSC waveform obtained by subtraction (gray). B, Left, In an experiment similar to that shown in A, a very small eEPSC was left in a FS neuron after CNQX application. Right, Average EPSCs and AMPA EPSC obtained by subtraction are shown superimposed for an example FS neuron. C, Bar graph summarizing differences in the decay time constant of the control eEPSCs (AMPAR- and NMDAR-mediated) and the AMPAR-mediated eEPSCs, obtained by waveform subtraction. **p < 0.001 (eEPSC decay time constant, pyramidal cells, n = 10, control eEPSC: 8.86 ± 0.98 ms, AMPA eEPSC: 4.78 ± 0.38 ms; FS cells, n = 13, control eEPSC: 2.85 ± 0.46 ms, AMPA eEPSC: 2.15 ± 0.24 ms; two-factor ANOVA, cell type: F_(1,42) = 59.4 p < 0.00001; NMDAR: F_(1,42) = 17.5 p < 0.0002; interaction: F_(1,42) = 8.4 p < 0.01; post hoc Fisher LSD tests control eEPSC vs AMPA eEPSC: pyramidal cells, p < 0.0001, FS neurons, p = 0.704). D, Bar graph summarizing data of the AMPA/NMDA charge ratio, estimated from the eEPSC waveforms for both pyramidal cells and FS neurons. Relative to the AMPA charge, the NMDA charge was significantly smaller in FS neurons, producing a larger AMPA/NMDA charge ratio. *p < 0.05 (AMPA/NMDA charge ratio, pyramidal cells, n = 23, 0.87 ± 0.28; FS neurons, n = 18, 5.21 ± 1.79; Mann–Whitney U test: Z = 2.022, p < 0.05).

Figure 4.

NMDAR contribution to EPSPs is significantly stronger in PFC pyramidal cells than in FS neurons. A, Left, eEPSPs in a pyramidal cell at negative potentials (approximately −80 mV) and depolarized near spike threshold, in control conditions and after application of d,l-AP5 (100 μm). Here and in B and C, gray traces show individual eEPSPs and black traces the average of at least 10 eEPSPs. The numbers below the traces indicate the mean membrane potential measured 5 ms before stimulation. Right, The effects of depolarization and AP5 were estimated by computing the ratio between the eEPSP area at depolarized and hyperpolarized potentials (EPSP D/H area ratio). D/H area ratio = 1 means an absence of effect of depolarization (see Materials and Methods). The bar graph summarizes the experiments testing the effects of AP5 on EPSP D/H area ratio in pyramidal cells **p < 0.01 (eEPSP D/H area ratio for pyramidal cells, control: 2.98 ± 0.36, n = 17, AP5: 1.61 ± 0.24, n = 15; Mann–Whitney U test Z = 2.63, p = 0.0085). B, Left, eEPSPs evoked in a FS neuron at potentials near −80 mV and at a depolarized potential near spike threshold. Right, Bar graph summarizing the effects of membrane depolarization and d,l-AP5 (100 μm) application on eEPSPs recorded from FS cells (eEPSP D/H area ratio for FS cells, control: 2.10 ± 0.33, n = 11, AP5: 1.45 ± 0.13, n = 10; Mann–Whitney U test Z = 1.619, p = 0.105). FS interneurons were depolarized to similar subthreshold potentials than pyramidal cells (pyramidal cells, control: −57.82 ± 1.79 mV, AP5: −56.15 ± 1.66 mV; FS neurons, control: −55.19 ± 2.72 mV, AP5: −54.11 ± 4.11 mV; two-factor ANOVA followed by Fisher LSD comparisons, p > 0.25). C, Examples of sEPSPs recorded from a pyramidal cell (left panel) or a FS neuron (right panel) at different membrane potentials (average membrane potential values indicated below the traces). Gray traces show individual sEPSPs and black traces the average of at least 50 sEPSPs. D, Left, The average sEPSPs shown in C are displayed superimposed. Black traces and gray traces show the averages of sEPSPs recorded at hyperpolarized and depolarized membrane potentials, respectively. Right, Bar graph summarizing the effects of membrane depolarization and d,l-AP5 (100 μm) application on the sEPSP D/H area ratio. **p < 0.005 (sEPSP D/H area ratio for pyramidal cells, n = 8, control: 3.28 ± 0.33, AP5: 1.87 ± 0.22, Mann–Whitney U test, Z = 2.83, p = 0.0045; sEPSP D/H area ratio for FS cells, n = 8, control: 1.25 ± 0.51, AP5: n = 8: 1.30 ± 0.12; Mann–Whitney U test, Z = 0.210, p = 0.833). Pyramidal cells and FS neurons were depolarized to similar subthreshold membrane potentials before and during AP5 application (pyramidal cells, control: −53.5 ± 1.4 mV, AP5: −55.0 ± 1.5 mV, p = 0.655; FS cells, control: −55.9 ± 2.5 mV, AP5: −54.4 ± 3.2 mV, p = 0.641).

Figure 5.

The NMDAR contribution to subthreshold EPSP summation is significantly stronger in PFC pyramidal cells than in FS neurons. A, Left, Example of average eEPSPs evoked by repetitive stimulation (5 stimuli, 20 Hz) in a pyramidal cell recorded at approximately −80 mV (black trace) and at a depolarized membrane potential near threshold (gray trace). Right, Bar graph summarizing the differences in eEPSP summation at different membrane potentials in the absence or presence of the NMDAR antagonist d,l-AP5 (100 μm). *p < 0.02 (eEPSP₅/eEPSP₁ ratio, hyperpolarized: 1.82 ± 0.21, depolarized: 3.41 ± 0.62, n = 14, Z = 2.113 p < 0.05, Mann–Whitney U test; eEPSP₅/eEPSP₁ ratio hyperpolarized+AP5: 2.03 ± 0.46, depolarized+AP5: 2.65 ± 0.37, n = 12, Z = 1.386 p = 0.165, Mann–Whitney U test). The depolarized membrane potential in control and AP5 conditions did not differ significantly (control: −59.5 ± 2.7 mV; AP5: −57.4 ± 3.0 mV, p = 0.619). B, Left, Example of average eEPSPs evoked by repetitive stimulation of excitatory inputs (5 stimuli, 50 Hz) onto a FS cell recorded at approximately −80 mV (black traces) and at a depolarized membrane potential near threshold (gray traces). Right, Bar graph summarizing the differences in eEPSP summation at different membrane potentials in the absence or presence of the NMDAR antagonist d,l-AP5 (100 μm) (eEPSP₅/eEPSP₁ ratio, hyperpolarized: 1.33 ± 0.23 n = 11, depolarized: 1.59 ± 0.39 n = 10, Z: 0.281, p = 0.778, Mann–Whitney U test; eEPSP₅/eEPSP₁ ratio hyperpolarized+AP5: 1.24 ± 0.20 n = 9, depolarized+AP5: 1.46 ± 0.45, n = 7, Z = 0.0532, p = 0.958, Mann–Whitney U test). The depolarized membrane potential in control and AP5 conditions did not differ significantly (control: −55.9 ± 1.9 mV; AP5: −55.0 ± 2.9 mV, p = 0.791).

Figure 6.

The NMDAR contribution to eEPSP–spike coupling is stronger in pyramidal cells. A, Left, Consecutive traces showing eEPSP trains evoked by 20 Hz stimulation in a pyramidal cell at a depolarized membrane potential near spike threshold in control conditions. Middle, Recordings from the same neuron in the presence of d,l-AP5 (100 μm). Here and in B, subthreshold and suprathreshold eEPSP trains are labeled in gray and black, respectively. The spikes are truncated for easier visualization of the eEPSPs. Right, Bar graph illustrating the significant reduction of spike probability by d,l-AP5 (100 μm) in pyramidal cells. *p < 0.05 (probability of spiking, n = 6, control: 0.67 ± 0.12, AP5: 0.36 ± 0.16, Student's t test, t = 3.14, p < 0.05). B, Left, Consecutive traces showing eEPSP trains evoked by 50 Hz stimulation in a FS neuron at a depolarized membrane potential near spike threshold in control conditions. Middle, Recordings in the presence of d,l-AP5 (100 μm). The spikes are truncated for easier visualization of the eEPSPs. Right, Bar graph illustrating the absence of significant effect of d,l-AP5 (100 μm) on spike probability in FS neurons (probability of spiking, n = 6, control: 0.60 ± 0.14, AP5: 0.63 ± 0.17, Student's t test, t = 0.233, p = 0.824).

Figure 7.

NMDAR contribution to disynaptic inhibition onto PFC pyramidal cells. A, Top, Three traces showing that long-distance lateral stimulation of synaptic inputs produced, in a pyramidal cell, a sequence of an EPSP followed by disynaptic IPSPs (arrows), which strongly truncated the decay phase of the EPSP. Note that the disynaptic IPSPs had variable latency from trial to trial. Middle, Three traces showing that in some trials of the same experiment shown above, input stimulation evoked an EPSP in the absence of disynaptic IPSPs. Bottom, Application of CNQX (20 μm) abolished all responses (EPSPs and IPSPs) evoked by focal stimulation. B, Repetitive input stimulation in some cases produced pyramidal neuron firing as a consequence of EPSP summation (gray). In other trials (black), the occurrence of disynaptic IPSPs (arrows) was sufficient to inhibit spiking. C, Left, Graph showing the number of spikes per EPSP train elicited in a pyramidal neuron plotted versus EPSP train number (trains delivered every 10 s, EPSP train frequency: 20 Hz). Note that application of the GABA_A receptor antagonist gabazine (2 μm) increased firing. Right, Graph summarizing the effect of gabazine application on the spiking probability. *p < 0.05 (spike probability: control, 0.19 ± 0.05, gabazine, 0.51 ± 0.12, p < 0.05, t = 3.618, n = 5, Student's t test). D, Left, Graph showing the number of spikes per EPSP train plotted versus train number. Note that application of the NMDAR channel blocker MK801 (10 μm) decreased firing. Right, Graph summarizing the effect of MK801 application on the spiking probability. *p < 0.05 (spike probability: control, 0.41 ± 0.12, MK801, 0.28 ± 0.10, p < 0.05, t = 3.048, n = 7, Student's t test).

Figure 8.

Fast AMPAR-mediated FS neuron activation is crucial for production of gamma oscillations via feedback mechanisms. A, Membrane properties of E cells (which represent pyramidal neurons) and I cells (which represent FS neurons) in the model network. B, Time course of synaptic conductances used to model excitatory and inhibitory synapses between model neurons. Note that the decay of the AMPA conductance is faster at excitatory synapses onto I cells (EI) than at excitatory synapses onto E cells (EE). In addition, the decay time of NMDA conductance is significantly longer than the decay of EI or EE. C, Raster plot showing the spike timing of E cells (left columns) and I cells (right columns) during a network oscillation episode produced in conditions of relatively low NMDAR conductance in the excitatory synapses onto I cells (g_ni = 0.002 mS/cm², see Materials and Methods). D, A raster plot similar to that shown in C, but obtained during an episode of network activity in conditions of higher NMDAR conductance in excitatory synapses onto I cells (g_ni = 0.008 mS/cm²). E, Plot of power spectral density (PSD) showing the effects of different strengths of g_ni on network activity. Note that larger values of g_ni produced a decrease in oscillation in power and synchrony. F, Plot of the time course of the total current entering a typical I cell (black trace), the synaptic output of that FS cell (red trace), and the activity of the pyramidal cells (the scaled total AMPA output of the pyramidal cell population, green trace). These variables, shown in arbitrary units in the y-axis (negative values for the black grace indicate hyperpolarizing/outward current), were computed for network activity produced with a g_ni = 0.002 mS/cm². Note the rhythmic input onto the I cell and the regular output from this I cell onto other cells in the network. G, A plot equivalent to that shown in F, for network activity produced with g_ni = 0.008 mS/cm². Note that in this case, the input currents onto the FS cell remain elevated between spikes (note the much higher levels of the black curves between peaks) and can result in spikes or bursts of spikes from the FS cell that are not locked to the excitatory activity (arrows).