Control of dorsal raphe serotonergic neurons by the medial prefrontal cortex: Involvement of serotonin-1A, GABA(A), and glutamate receptors

Abstract

Anatomical evidence indicates that medial prefrontal cortex (mPFC) neurons project to the dorsal raphe nucleus (DR). In this study, we functionally characterized this descending pathway in rat brain. Projection neurons in the mPFC were identified by antidromic stimulation from the DR. Electrical stimulation of the mPFC mainly inhibited the activity of DR 5-HT neurons (55 of 66). Peristimulus time histograms showed a silence of 150 +/- 9 msec poststimulus (latency, 36 +/- 1 msec). The administration of WAY-100635 and picrotoxinin partly reversed this inhibition, indicating the involvement of 5-HT(1A) and GABA(A) receptors. In rats depleted of 5-HT with p-chlorophenylalanine, the electrical stimulation of mPFC mainly activated 5-HT neurons (31 of 40). The excitations (latency, 17 +/- 1 msec) were antagonized by MK-801 and NBQX. Likewise, MK-801 prevented the rise in DR 5-HT release induced by electrical stimulation of mPFC. The application of 8-OH-DPAT in mPFC significantly inhibited the firing rate of DR 5-HT neurons and, in dual-probe microdialysis experiments, reduced the 5-HT output in mPFC and DR. Furthermore, the application of WAY-100635 in mPFC significantly antagonized the reduction of 5-HT release produced by systemic 8-OH-DPAT administration in both areas. These results indicate the existence of a complex regulation of DR 5-HT neurons by mPFC afferents. The stimulus-induced excitation of some 5-HT neurons by descending excitatory fibers releases 5-HT, which inhibits the same or other DR neurons by acting on 5-HT(1A) autoreceptors. Afferents from the mPFC also inhibit 5-HT neurons through the activation of GABAergic interneurons. Ascending serotonergic pathways may control the activity of this descending pathway by acting on postsynaptic 5-HT(1A) receptors.

Fig. 1.

Extracellular recording of a representative mPFC neuron projecting to the DR. Electrical stimulation of the DR (arrow) evoked antidromic responses in mPFC neurons. Theasterisk denotes an antidromic spike missing because of collision with an spontaneous action potential.

Fig. 2.

Typical examples of responses of DR neurons to the electrical stimulation of the mPFC. A shows a PSTH corresponding to an inhibition of a 5-HT neuron. B shows a PSTH in which a subthreshold activation (9%) occurred before inhibition. C and D show an orthodromic activation of a 5-HT neuron (note the different time scale and bin size in D). The arrow indicates the stimulus artifact. PSTH made up of 250 consecutive trials of single pulse stimuli (0.2 msec, 1 mA) delivered at 0.9 Hz. Bin width 4 msec (A–C) and 1 msec (D). Note the different ordinate scale in A and Bversus C and D.

Fig. 3.

Effects of electrical stimulation of the mPFC on the 5-HT release in the DR. A, Three different stimulation conditions (S1, 0.9 Hz, 1,7 mA, 0.2 msec; S2, 10 Hz, 0.5 mA, 1 msec; S3, 20 Hz, 0.5 mA, 1 msec; shown by horizontal bars) increased the 5-HT release in the DR (filled circles; n = 4). Control rats (open circles; n = 6) were implanted with electrodes, but no current was passed.B, The electrical stimulation of mPFC using the same conditions than in PSTH experiments (S4, 0.9 Hz, 2 mA, 0.2 msec;first bar) produced a transient elevation of the 5-HT release in DR (see also the effect of S1 in A), whereas the stimulation at S3 conditions (20 Hz, 0.5 mA, 1 msec) produced a sustained increase in 5-HT release for at least 60 min (filled circles; n = 7). Control rats (open circles; n = 6) were implanted with electrodes, but no current was passed. The local application of MK-801 (300 μm; filled triangles;n = 7) by reverse dialysis prevented the increase in 5-HT release induced by the second stimulation conditions. *p < 0.05 versus controls;⁺p < 0.05 versus stimulation (no MK-801)

Fig. 4.

Effect of the systemic administration of the 5-HT_1A receptor antagonist WAY-100635 on the poststimulus inhibition of a DR 5-HT neuron in response to stimulation of the mPFC.A, DR serotonergic neuron inhibited by mPFC stimulation (arrow) (suppression of 230 msec to 30% of prestimulus firing rate). B, WAY-100635 (5 μg/kg, i.v.) partially blocks the inhibition (to 50 msec, 77% prestimulus firing rate). Bin size: 10 msec, 250 sweeps.

Fig. 5.

Effect of mPFC stimulation (S3, 20 Hz, 1 msec, 0.5 mA, for 3 min; shown by a horizontal bar) on the spontaneous firing of 5-HT neurons. A andD show integrated firing rate histograms of two representative neurons in control (A) andpCPA-treated (D) rats.B and E show the corresponding spike trains of the same neurons before and 2 min after S3 stimulation. During the second minute of the stimulation, the firing rate of 5-HT neurons in pCPA-treated rats was already significantly increased versus basal (1.1 ± 0.2 vs 0.6 ± 0.2 spikes/sec;p < 0.02). However, the maximal effect of the stimulation was noted later. C and F are bar graphs with the mean ± SEM values of firing rate before and after (4th minute) S3 stimulation (n = 12 controls and 7 pCPA-treated rats). *p < 0.02 versus baseline.

Fig. 6.

Effect of the GABA_A receptor antagonist picrotoxinin on the poststimulus inhibition of a DR 5-HT neuron in response to mPFC stimulation. A, PSTH of a DR 5-HT neuron inhibited by mPFC stimulation (arrow) (suppression of 260 msec to 24% of prestimulus firing rate).B, Picrotoxinin (1 mg/kg, i.v.) partially blocked the mPFC-induced inhibition. C, An additional intravenous picrotoxinin dose of 1 mg/kg further reduced the mPFC-induced inhibition (to 140 msec, 57% prestimulus firing rate). Bin size: 10 msec, 250 sweeps.

Fig. 7.

Involvement of ionotropic glutamate receptors in the mPFC-induced activations of DR 5-HT neurons. A shows the baseline orthodromic activation (52%, latency 18–36 msec) of a 5-HT neuron after the stimulation of mPFC. B andC show, in the same neuron, the reversal of the excitation produced by increasing doses of the NMDA receptor antagonist (±)MK-801 (0.66 + 0.33 mg/kg, i.v.). Bin size: 4 msec, 240 sweeps.D shows the baseline orthodromic activation (21.2%, latency 26–48 msec) of another DR 5-HT neuron after the stimulation of mPFC. E and F show the reversal of the excitation produced by increasing doses of the AMPA–KA receptor antagonist NBQX (1 + 1 mg/kg, i.v.) in the neuron shown inD. Bin size: 4 msec, 225 sweeps.

Fig. 8.

Effect of the local application of WAY-100635 (100 μm) in the DR or mPFC on the reduction of 5-HT release elicited by 8-OH-DPAT (0.1 mg/kg, s.c.) in these areas of rats implanted with two dialysis probes. The graph in A shows the effect of two sequential injections of 8-OH-DPAT (arrows) on the 5-HT release in the mPFC (open circles). The perfusion of WAY-100635 (horizontal bar) in the DR (B, filled circles) or mPFC (C, filled squares) partly antagonized the effect of the second 8-OH-DPAT injection on 5-HT release in mPFC. The bar graph in Dshows the peak ratios between the maximal effects on 5-HT release in the mPFC elicited by the two 8-OH-DPAT injections in identical conditions (open bars; e.g., as in A) or during the infusion of WAY-100635 in the DR (striped bars; as in B) or mPFC (black bars, as in C). The bar graph in E shows the peak ratios of the maximal effects on 5-HT release elicited by the two 8-OH-DPAT injections (as in D) but in this case, referred to the effects on 5-HT release in the DR (for simplicity, the actual dialysate graphs in the DR are not shown). Data are means ± SEM of five or six rats per group.^*p < 0.05 versus controls;⁺p < 0.05 versus WAY-100635 in mPFC.

Fig. 9.

Distribution of the ex vivo[³H]WAY-100635 binding along the rostrocaudal axis in the DR as measured autoradiographically in coronal midbrain sections. The histogram in A shows the mean ± SEM values of the [³H]WAY-100635 binding in 17 equally spaced coronal sections cut at different rostrocaudal levels (n = 4). B and Cshow, respectively, 10 autoradiograms of midbrain sections containing the DR and the amount of tritium label in each section. In this particular rat, tritium levels above background were found between approximately −7.6 to −8.7 mm from bregma. Note that the shape of the tritium label (dark spots in B) follows the track of the dialysis probe (30° inclination) and has a shape different from that of the DR, suggesting that not all 5-HT_1A receptors within the DR were labeled by the perfusion of [³H]WAY-100635.

Fig. 10.

Effects of the local infusion of 100 and 300 μm 8-OH-DPAT in prefrontal cortex (PFC) on the 5-HT release in DR (A) and PFC (B). Control animals (open circles) received aCSF for the whole experiment (syringe changes were also performed in this group). The application of 8-OH-DPAT in medial prefrontal cortex (filled circles) significantly reduced the 5-HT release in this area and the DR. In contrast, the perfusion of 8-OH-DPAT in the lateral prefrontal cortex (filled triangles), an area devoid of neurons projecting to the DR reduced locally the 5-HT release but not in the DR. Data are from seven or eight rats per group.

Fig. 11.

Effect of 8-OH-DPAT in mPFC on the firing rate of DR 5-HT neurons. A, The local application of 8-OH-DPAT (0.2 μl, 100 μm, 1 min horizontal bar) in mPFC reduced the firing rate of a DR neuron. A return to baseline firing was observed after the local application of WAY-100635 (0.2 μl, 100 μm, 1 min horizontal bar). B, Mean ± SEM values of the firing rate of 13 DR neurons before and after the application of 8-OH-DPAT in mPFC as above. In seven neurons, the application of a second 8-OH-DPAT dose in mPFC reduced the firing rate further. *p < 0.02 versus baseline.

Fig. 12.

Schematic representation of the putative relationships between projection neurons in mPFC and DR 5-HT neurons. Descending excitatory afferents from the mPFC control the activity of 5-HT neurons directly, via NMDA and AMPA–KA receptors, and indirectly, via activation of local inhibitory (5-HT_1A and GABA_A) receptors. The stimulus-induced excitation of neurons receiving a direct input from the mPFC (either 5-HT or GABAergic) releases 5-HT or GABA, which inhibit other 5-HT neurons via 5-HT_1A or GABA_A receptors. The involvement of 5-HT_1A receptors in the mPFC-induced inhibitions of 5-HT neurons is supported by the decrease in the proportion of inhibitions in rats depleted of 5-HT (Table 1) and by the reversal of the inhibitions induced by 5-HT_1A receptor blockade with WAY-100635. Additionally, GABAergic inputs may also occur via a serotonergic control of GABA interneurons (Liu et al., 2000). The firing activity of 5-HT neurons and the release of 5-HT in midbrain and forebrain are dependent on the activation of 5-HT_1Aautoreceptors in the DR. However, the activation by 8-OH-DPAT of postsynaptic 5-HT_1A receptors on cortical pyramidal neurons also reduces 5-HT cell firing in the DR and 5-HT release in the DR and mPFC, likely through the inhibition of the activity of excitatory inputs from the mPFC to the DR. Thus, pyramidal 5-HT_1Areceptors in mPFC are a second population of 5-HT_1Areceptors controlling serotonergic activity.