Dopamine D4 receptor activation increases hippocampal gamma oscillations by enhancing synchronization of fast-spiking interneurons

Abstract

Background: Gamma oscillations are electric activity patterns of the mammalian brain hypothesized to serve attention, sensory perception, working memory and memory encoding. They are disrupted or altered in schizophrenic patients with associated cognitive deficits, which persist in spite of treatment with antipsychotics. Because cognitive symptoms are a core feature of schizophrenia it is relevant to explore signaling pathways that potentially regulate gamma oscillations. Dopamine has been reported to decrease gamma oscillation power via D1-like receptors. Based on the expression pattern of D4 receptors (D4R) in hippocampus, and pharmacological effects of D4R ligands in animals, we hypothesize that they are in a position to regulate gamma oscillations as well.

Methodology/principal findings: To address this hypothesis we use rat hippocampal slices and kainate-induced gamma oscillations. Local field potential recordings as well as intracellular recordings of pyramidal cells, fast-spiking and non-fast-spiking interneurons were carried out. We show that D4R activation with the selective ligand PD168077 increases gamma oscillation power, which can be blocked by the D4R-specific antagonist L745,870 as well as by the antipsychotic drug Clozapine. Pyramidal cells did not exhibit changes in excitatory or inhibitory synaptic current amplitudes, but inhibitory currents became more coherent with the oscillations after application of PD168077. Fast-spiking, but not non-fast spiking, interneurons, increase their action potential phase-coupling and coherence with regard to ongoing gamma oscillations in response to D4R activation. Among several possible mechanisms we found that the NMDA receptor antagonist AP5 also blocks the D4R mediated increase in gamma oscillation power.

Conclusions/significance: We conclude that D4R activation affects fast-spiking interneuron synchronization and thereby increases gamma power by an NMDA receptor-dependent mechanism. This suggests that converging deficits on fast-spiking interneurons may lead to decreased network function and thus aberrant gamma oscillations and cognitive decline in schizophrenia.

Figure 1. D4 receptor activation increases gamma oscillation power.

LFP recordings of kainate-induced gamma oscillation and dopamine D4R agonist (PD168077, 100 nM) without and with prior application of D4R antagonist (L745,870, 500 nM or Clozapine (2 µM). A. Example traces of LFP recordings, showing an increase with D4R activation B. Power spectra of traces shown in A (kainate recording in black, subsequent recording with PD168077 in grey). C. Example traces of LFP recordings, showing the D4 effect is blocked with application of clozapine. D. Power spectra of traces shown in C. (kainate recording in black, clozapine as dotted line, and PD168077 in grey). E. Summary bar diagram of power (in % relative to initial kainate power, mean ± SEM). D4R activation significantly increases LFP gamma power, which is prevented by prior application of D4R antagonist or Clozapine. * P<0.05, ** P<0.01, *** P<0.001 (in unpaired t-tests).

Figure 2. Pyramidal cell EPSCs are unaffected by D4 receptor activation.

Intracellular recordings of EPSCs in pyramidal cells (clamped to −70 mV) with ongoing gamma oscillations. A. Example traces of EPSC recordings. B. Summary bar-diagram showing the amplitude of EPSCs (mean ± SEM). C. Power spectrum of EPSCs constructed from the traces shown in A (kainate recording in black, subsequent recording with PD168077 in grey). D. Summary bar-diagram of power normalized to kainate. E. Coherence spectrum of EPSCs versus LFP recordings from the traces shown in A (kainate recording in black, subsequent recording with PD168077 in grey). F. Summary bar-diagram of peak coherence values across experiments (mean ± SEM).

Figure 3. D4 receptor activation increases coherence of pyramidal cell IPSCs.

Intracellular recordings of IPSCs in pyramidal cells (clamped to 0 mV) with ongoing gamma oscillations. A. Example traces of IPSC recordings. B. Summary bar-diagram showing the amplitude of IPSCs (mean ± SEM). C. Power spectrum of IPSCs constructed from the traces shown in A. (kainate recording in black, subsequent recording with PD168077 in grey). D. Summary bar-diagram of LFP gamma power normalized to kainate. E. Coherence spectrum of IPSCs versus LFP recordings from recordings shown in A (kainate recording in black, subsequent recording with PD168077 in grey). F. Summary bar-diagram of peak coherence values across experiments (mean ± SEM).

Figure 4. Spike-phase coupling in pyramidal cells is unaffected by D4 receptor activation.

A. Example traces of concomitant recordings of a pyramidal cell and LFP oscillations before and after the addition of PD168077. B. Circular histograms based on recordings shown in A, indicating the number of action potentials discharged in each phase from −π to π. The radial axis indicates the number of action potentials. The higher and lower histograms represent recordings before and after PD168077 application, respectively. C. Summary bar-diagrams representing mean peak coherence and resultant vector length, respectively (mean ± SEM; kainate recording in black, subsequent recording with PD168077 in grey).

Figure 5. Spike-phase coupling in non-fast spiking interneurons is unaffected by D4 receptor activation.

A. Example traces of concomitant recordings of non-fast spiking interneurons and LFP oscillations before and after the addition of PD168077. B. Circular histograms based on recordings similar to those in A, indicating the number of action potentials discharged in each phase from −π to π. The radial axis indicates the number of action potentials. The higher and lower histograms represent recordings before and after PD168077 application, respectively. C. Summary bar-diagrams representing mean peak coherence and resultant vector length, respectively (mean ± SEM; kainate recording in black, subsequent recording with PD168077 in grey).

Figure 6. D4 receptor activation increases spike-phase coupling in fast-spiking interneurons.

A. Example traces of concomitant recordings of fast spiking interneurons and LFP oscillations before and after the addition of PD168077. B. Circular histograms based on recordings shown in A, indicating the number of action potentials discharged in each phase from −π to π. The radial axis indicates the number of action potentials. The higher and lower histograms represent recordings before and after PD168077 application, respectively. C. Summary bar-diagrams representing mean peak coherence and resultant vector length, respectively (mean ± SEM, * indicates p<0.05, ** p<0.01; kainate recording in black, subsequent recording with PD168077 in grey).

Figure 7. D4 receptor modulation of gamma oscillations is NMDAR-dependent.

EPSC recordings recorded from fast-spiking interneurons and the effect of NMDA receptor antagonists on LFP gamma power modulation by PD168077 A. Example traces of EPSCs (clamped to −70 mV, 50 µM picrotoxin, 50 µM AP5). B. Aggregated amplitudes across experiments plotted in an empirical cumulative distribution function (kainate recording in black, subsequent recording with PD168077 in grey). Note that the 2 graphs overlap to a large extent. C. Example traces of LFP recordings showing oscillations from (top to bottom) KA, the addition of AP5, the addition of PD 168077. D. Summary bar-diagram of LFP gamma power (in % relative to initial kainate power, means ± SEM). Addition of NMDA receptor antagonist AP5 does not have an effect on the gamma power but completely blocks the increase produced by D4R activation.

Figure 8. D4 receptor activation causes reduction in high-voltage outward current in fast-spiking interneurons.

Voltage steps recorded in fast-spiking interneurons (with 1 µM TTX present, 10 mV steps from −90 mV to 0 mV). A. Traces of current responses in control conditions (no kainate). B. Traces of current responses in the presence of PD168077. C. Traces of differences (digital subtraction of trace in B-A). D. Summary current vs. voltage plot across experiments. Black represents control, grey PD168077. *** P<0.001 (two-way ANOVA).