Neuregulin and dopamine modulation of hippocampal gamma oscillations is dependent on dopamine D4 receptors

Abstract

The neuregulin/ErbB signaling network is genetically associated with schizophrenia and modulates hippocampal γ oscillations--a type of neuronal network activity important for higher brain processes and altered in psychiatric disorders. Because neuregulin-1 (NRG-1) dramatically increases extracellular dopamine levels in the hippocampus, we investigated the relationship between NRG/ErbB and dopamine signaling in hippocampal γ oscillations. Using agonists for different D1- and D2-type dopamine receptors, we found that the D4 receptor (D4R) agonist PD168077, but not D1/D5 and D2/D3 agonists, increases γ oscillation power, and its effect is blocked by the highly specific D4R antagonist L-745,870. Using double in situ hybridization and immunofluorescence histochemistry, we show that hippocampal D4R mRNA and protein are more highly expressed in GAD67-positive GABAergic interneurons, many of which express the NRG-1 receptor ErbB4. Importantly, D4 and ErbB4 receptors are coexpressed in parvalbumin-positive basket cells that are critical for γ oscillations. Last, we report that D4R activation is essential for the effects of NRG-1 on network activity because L-745,870 and the atypical antipsychotic clozapine dramatically reduce the NRG-1-induced increase in γ oscillation power. This unique link between D4R and ErbB4 signaling on γ oscillation power, and their coexpression in parvalbumin-expressing interneurons, suggests a cellular mechanism that may be compromised in different psychiatric disorders affecting cognitive control. These findings are important given the association of a DRD4 polymorphism with alterations in attention, working memory, and γ oscillations, and suggest potential benefits of D4R modulators for targeting cognitive deficits.

Fig. 1.

Analysis of KA-induced γ oscillations in rat hippocampal slices after application of DA and selective D1- and D2-type receptor agonists/antagonists. Representative sample traces (Left) and power spectra (Right) of KA-induced γ oscillations in rat slices; the color of lines in power spectra correspond to the type of treatment indicated over each sample trace (Right). (A) DA (10 μM) was added after the induction of γ oscillations with 100 nM KA (n = 5). No change was observed with DA. (B) Effect of the D1-type selective antagonist SCH23390 (300 nM) on KA-induced γ oscillations in presence of 10 μM DA (n = 7). (C) Lack of effect of the D1/D5 receptor agonist SKF81297 (1 μM; n = 14). (D) Lack of effect of the D2/D3 receptor-selective agonist 7-OH-DPAT (100 nM; n = 10). (E) Quantification of normalized KA-induced γ oscillation power in response to the above treatments. Inclusion of the D1/D5 receptor-selective antagonist SCH23390 to the bath solution significantly increased normalized γ oscillation power by 42% (P = 0.013).

Fig. 2.

Effects of PD168077 on KA-induced γ oscillations are blocked by a specific D4R, but not D2/D3 receptor, antagonists. Representative sample traces (Left) and power spectra (Right) of KA-induced γ oscillations in rat slices; line colors in power spectra correspond to the type of treatment indicated over each sample trace (Right). (A) The D4R-selective agonist PD168077 (100 nM) significantly increases the power of KA-induced γ oscillations (n = 13). (B) Prior application of the D4R antagonist L-745,870 (500 nM; n = 6) prevents the PD168077-induced increase in γ oscillation power. (C) Prior application of the D2/D3 receptor antagonist sulpiride (500 nM; n = 6) does not prevent the PD168077-induced increase in γ oscillation power. (D) Summary analysis of normalized data shown in A, B, and C; power of KA-induced γ oscillations set as 100%.

Fig. 3.

D4R mRNA and protein expression in the hippocampus is highest in GABAgergic interneurons. (A) Representative ISH low-magnification dark-field images of the hippocampus showing D4R mRNA-expressing neurons. Areas outlined by rectangles are magnified in A1 and A2 and show accumulation of grains over sparse cells in the CA1 and the subiculum. (B–E) DISH analysis using ³⁵S-labeled D4R and digoxigenin-labeled GAD67 probes in the CA3 (B and C) and the subiculum (D and E), revealing colocalization of both transcripts in a subset of neurons (arrows). (F–K) Double IFH in adult hippocampal sections for D4R and either the neuronal marker NeuN or GAD67. (F–H) Colocalization with NeuN shows that D4R immunofluorescence is confined to neurons located mostly in the strata pyramidale and oriens, the former consistent with basket cells. (I–K) In CA3, D4R-immunolabeling is observed in GAD67-positive interneurons (arrows); however, occasionally expression is observed in cells that are negative or labeled lightly for GAD67. (Scale bars, 100 μm.)

Fig. 4.

D4 receptors are expressed in parvalbumin- and ErbB4-positive interneurons. Sample images of double IFH for D4R, and PV or ErbB4. D4R immunolabeling is observed in PV+ neurons in both CA1 (A–C) and CA3 (D–F) areas of the hippocampus (arrows). The localization of these PV+ neurons within or near the pyramidal layer suggests they represent basket cells. (G–I) A large proportion of D4R-positive cells in CA3 express ErbB4 (arrows). Note that there are D4R-immunopositive cells in both CA1 and CA3 areas that do not express PV or ErbB4 (arrowheads); refer to Results for quantification in CA1–CA3 and dentate. (Scale bars, 100 μm.)

Fig. 5.

Convergence of ErbB4 and D4R signaling in parvalbumin-positive interneurons. Representative sample traces (Left) and power spectra (Right) of KA-induced γ oscillations in rat slices. NRG-1β (2 nM) was tested in combination with (A) KA alone (n = 8), (B) KA and L-745,870 (50 nM; n = 8), or (C) KA and clozapine (2 μM; n = 5). (D) Summary analysis of normalized data from A–C. Values for KA treatment are set as 100%. (E–H) Sample images of triple IFH for D4R, ErbB4 and PV in hippocampal CA3. Overlay image showing PV+ neurons that also coexpress D4 and ErbB4 receptors (arrows). (Scale bar, 100 μm.)