GluA2-Lacking AMPA Receptors and Nitric Oxide Signaling Gate Spike-Timing-Dependent Potentiation of Glutamate Synapses in the Dorsal Raphe Nucleus

Abstract

The dorsal raphe nucleus (DRn) receives glutamatergic inputs from numerous brain areas that control the function of DRn serotonin (5-HT) neurons. By integrating these synaptic inputs, 5-HT neurons modulate a plethora of behaviors and physiological functions. However, it remains unknown whether the excitatory inputs onto DRn 5-HT neurons can undergo activity-dependent change of strength, as well as the mechanisms that control their plasticity. Here, we describe a novel form of spike-timing-dependent long-term potentiation (tLTP) of glutamate synapses onto rat DRn 5-HT neurons. This form of synaptic plasticity is initiated by an increase in postsynaptic intracellular calcium but is maintained by a persistent increase in the probability of glutamate release. The tLTP of glutamate synapses onto DRn 5-HT is independent of NMDA receptors but requires the activation of calcium-permeable AMPA receptors and voltage-dependent calcium channels. The presynaptic expression of the tLTP is mediated by the retrograde messenger nitric oxide (NO) and activation of cGMP/PKG pathways. Collectively, these results indicate that glutamate synapses in the DRn undergo activity-dependent synaptic plasticity gated by NO signaling and unravel a previously unsuspected role of NO in controlling synaptic function and plasticity in the DRn.

Keywords: AMPA; LTP; NMDA; dorsal raphe; nitric oxide; serotonin.

Figure 1.

Pairing presynaptic stimulation with bAPs induces tLTP of AMPAR-EPSCs. A, Stimulation protocol (left) and voltage response (right) used to induce tLTP. B, Summary graph of the time course and the magnitude of the potentiation of AMPAR-EPSCs induced by pairing presynaptic stimulations with bAPs (•, n = 26), presynaptic stimulation alone (O, n = 10), and bAPs (□, n = 10). Right graph illustrates averaged AMPAR-EPSC traces taken at the time point indicated in the left graph. Note that the induction of the LTP requires pairing of pre- and postsynaptic stimulations.

Figure 2.

Decrease in PPR and CV indicate a presynaptic expression of tLTP. A, Summary graph of tLTP assessed using pairs of stimuli. B, Histogram summary of the average PPR (EPSC₂/EPSC₁) obtained before and during the tLTP. Inset depicts superimposed EPSC traces evoked by paired-pulse stimulation taken at the time point indicated by number. Scale bars: 50 pA, 20 ms. C, Histogram summary of the average CV obtained before and during the tLTP. Note that the tLTP is associated with a significant decrease in PPR (*, p < 0.05, n = 11) and CV (*, p < 0.05, n = 15).

Figure 3.

The tLTP requires a rise in postsynaptic intracellular Ca²⁺ but not the activation of NMDA receptors. A, Buffering postsynaptic intracellular Ca²⁺, but not blockade of NMDA receptors, abolishes the tLTP. Lower panel is a summary graph of tLTP obtained in control condition (O, n = 11), in the presence of D-AP 5 (•, 50 µM, n = 10), and with intracellular solution containing BAPTA (□, 20 mM, n = 10). Upper panel illustrates superimposed averaged AMPAR-EPSC traces taken at time points indicated in lower graph. Scale bars: 20 pA, 5 ms. B, Blockade of voltage-dependent Ca²⁺ channels abolishes tLTP. Lower panel is a summary graph of the time course and magnitude of tLTP obtained in control (O, n = 10) and in the presence of nifidepine (Δ, 20 µM, n = 11). Upper panel illustrates superimposed AMPAR-EPSC traces taken before and during the tLTP in control (left traces) and in the presence of nifidepine (right traces). Scale bars: 25 pA, 10 ms.

Figure 4.

Activation of GluA2-lacking AMPARs is required for tLTP induction. A, GluA2-lacking AMPARs significantly contribute to the baseline amplitude of AMPAR-EPSCs. Lower panel depicts the effect of the selective GluA2-lacking AMPAR antagonist Napsm (30 µM, n = 8) on the baseline amplitude of AMPAR-EPSCs. Upper graph is superimposed AMPAR-EPSC traces taken during the course of the experiment as indicated by numbers in lower panel. B, Blockade of GluA2-lacking AMPARs abolishes tLTP. Lower graph is a summary of the time course and magnitude of tLTP obtained in control condition (O, n = 8) and in the presence of Napsm (30 µM, •, n = 8). Upper graph depicts superimposed AMPAR-EPSC traces collected before and during tLTP in control condition (left traces) and in the presence of NASPAM (right traces). Scale bars: 50 pA, 10 ms.

Figure 5.

Nitric oxide signaling mediates tLTP. A, The nitric oxide donor SNAP increases AMPAR-EPSC amplitude by enhancing glutamate release. Left panel illustrates summary graph of the effect of SNAP (200 µM, n = 8) on the amplitude of AMPAR-EPSCs. Middle panel illustrates superimposed average pairs of EPSC traces collected before and during SNAP application. B, Summary histogram of changes in PPR of AMPAR-EPSCs obtained in control and in the presence of SNAP. Note that SNAP (200 µM) significantly (*, p < 0.05, n = 8) reduced the PPR. C, Nitric oxide synthase inhibitor L-NAME abolishes tLTP. Left graph illustrates summary graph of tLTP obtained in control condition (O, n = 14) and in slices pretreated with L-NAME (100 µM, •, n = 14). Right panel represents sample average AMPAR-EPSC traces taken during the experiment as depicted by numbers in the left graph. D, The NO scavenger PTIO prevents tLTP. Left panel illustrates the time course and magnitude of tLTP obtained in control (O, n = 8) and in the presence of NO scavenger PTIO (100 µM, •, n = 8). Left graph is superimposed AMPAR-EPSC traces taken during the course of experiment as indicated by number in the left graph. Scale bars: 25 pA, 10 ms.

Figure 6.

Activation of cGMP-PKG pathway mediates tLTP. A, Inhibition of sGC blocks tLTP. Lower graph is a summary of tLTP obtained in control condition (O, n = 10) and in slices treated with the sGC inhibitor ODQ (100 µM, •, n = 10). Upper graph illustrates superimposed AMPAR-EPSC traces taken before and during tLTP as indicated by numbers in lower graph. B, Activation of sGC mimics tLTP. Lower panel depict a summary of the potentiation of AMPAR-EPSC induced by the selective sGC activator A350219 (100 µM, n = 7). Upper graph is superimposed AMPAR-EPSC traces taken before and during A350219 administration. C, The membrane permeable cGMP analog 8-pCPT-cGMP mimics tLTP. Lower graph illustrates the averaged potentiation of AMPAR-EPSCs induced by 8-pCPT-cGMP (100 µM, n = 7). Upper graph is superimposed traces of AMPAR-EPSCs taken before and during administration of 8-pCPT-cGMP. D, The membrane permeable cGMP analog 8-pCPT-cGMP occludes tLTP. Lower panel is a summary graph of the tLTP obtained in control condition (O, n = 8) and in the presence of 8-pCPT-cGMP (100 µM, •, n = 8). Upper graph depicts AMPAR-EPSCs traces taken at the time points indicated by numbers in lower graph. E, Inhibition of PKG does not alter the baseline amplitude of AMPAR-EPSCs. Lower panel is a summary graph of the effect of KY5823 (1 µM, n = 6) on the amplitude of AMPAR-EPSCs. Upper graph illustrates sample AMPAR-EPSC traces taken during the course of the experiments as indicated by numbers in the lower panel. F, Inhibition of PKG abolishes tLTP. Lower panel illustrates a summary of tLTP obtained in control (O, n = 8) and in slices treated with the PKG inhibitor KT5823 (1 µM, •, n = 5). Scale bars: 50 pA, 20 ms.