Serotonergic Suppression of Mouse Prefrontal Circuits Implicated in Task Attention

5-HT inhibits L6 pyramidal neurons of medial prefrontal cortex. Responses to 5-HT were probed in voltage-clamp and current-clamp by bath application of 5-HT. Representative voltage-clamp traces of the 5-HT response in L6 pyramidal neurons show that responses to 5-HT are stable and persist in the presence of synaptic blockers (A), are significantly suppressed by WAY100635 (B), and are similarly suppressed by a combination of WAY100635 and synaptic blockers (C). D, 5-HT elicits strong outward currents on L6 pyramidal neurons of mPFC (n = 28) that can be pharmacologically modulated (F_(2,53) = 11.8, p < 0.0001, one-way ANOVA). Post hoc analyses show that these currents persist in the presence of synaptic blockers (q = 0.2, p > 0.05, n = 7, Dunnett’s multiple comparison test) but are significantly suppressed by WAY100635 (q = 4.6, p < 0.0001, n = 19). Responses to 5-HT were probed in current-clamp in L6 pyramidal neurons of medial prefrontal cortex with current injection to elicit steady firing (∼2–3 Hz) at baseline. Representative current-clamp traces show that responses to 5-HT are inhibitory, repeatable, and unaffected by synaptic blockers (E) and not fully blocked by WAY100635 (F) or by WAY100635 and synaptic blockers (G). H, L6 pyramidal neurons of mPFC are strongly hyperpolarized by 5-HT (n = 26). Post hoc analyses of pharmacological effects on this hyperpolarization (F_(2,50) = 17, p < 0.0001, one-way ANOVA) show that this inhibition is unaffected by synaptic blockers (q = 0.8, p > 0.5, n = 6, Dunnett’s multiple comparison test) but greatly reduced by WAY100635 (q = 5.4, p < 0.0001, n = 21). I, Action potential firing was significantly affected by 5-HT (F_(4,103) = 37, p < 0.0001, one-way ANOVA). Post hoc analyses reveal that baseline firing was strongly suppressed by 5-HT (q = 9, p < 0.0001, n = 17, Dunnett’s multiple comparison test) and remained suppressed by 5-HT in synaptic blockers (q = 6.1, p < 0.0001, n = 6). The suppression was not blocked by WAY100635 (q = 8.8, p < 0.0001, n = 16) and returned to baseline levels after washout of 5-HT (q = 2.3, p > 0.05, n = 36).

Combined activation of serotonergic 5-HT_1A and 5-HT_2A receptors mediate inhibition of L6 neuronal excitability at suprathreshold potentials. Incremental current steps were injected into patched L6 pyramidal neurons, and their output in firing frequency was measured. Shown are representative recordings of the response to a 150-pA current step in single L6 pyramidal neurons (left) and the response to the same 150-pA current step in the presence of 5-HT (middle). The input–output relationship for each group is plotted (right). A, The input–output relationship of L6 pyramidal neurons is significantly right-shifted by 5-HT (F_(1,168) = 31, p < 0.0001, repeated-measures two-way ANOVA). Post hoc analysis showed significantly fewer elicited action potentials at every input step (p < 0.05, Sidak’s multiple comparisons test). B, L6 excitability is significantly suppressed by 5-HT in the presence of WAY100635 (p < 0.0001, F_(1,72) = 72, repeated-measures two-way ANOVA), an effect especially prominent at higher input steps (125- to 200-pA steps, p < 0.05, Sidak’s multiple comparisons test). C, L6 suppression by 5-HT is fully blocked by simultaneous blockade of both 5-HT_1A and 5-HT_2A receptors by specific antagonists WAY100635 and MDL100907 (F_(1,32) = 0.8, p = 0.4, repeated-measures two-way ANOVA). D, TCB-2, a selective 5-HT_2A receptor agonist, inhibits L6 neuronal firing (F_(1,80) = 24, p < 0.0001, repeated-measures two-way ANOVA), also more prominently at higher input steps (125- to 200-pA steps, p < 0.05, Sidak’s multiple comparisons test).

Characterization of L6 neurons in medial prefrontal cortex (mPFC) expressing synaptotagmin 6 and epiphycan. A, Neurons expressing eGFP driven by the synaptotagmin-6 (Syt6) BAC promoter are localized to L6 pyramidal neurons in the prelimbic region of mPFC. B, EGFP is also seen in L6 pyramidal neurons in prelimbic mPFC by anti-EGFP immunohistochemistry in Epyc-Cre mice crossed to a Cre-dependent eGFP reporter. Image is adapted from www.gensat.org. C, Quantification (mean ± SEM) by qRT-PCR of the expression of selected genes in mPFC Epyc-vTRAP IP samples compared with whole PFC input. Positive values indicate enrichment in the IP, and negative values indicate depletion. Dotted lines indicate a twofold difference in either direction. Green bars, genes that are >2-fold enriched in the Epyc cells; red bars, genes that are >2-fold depleted; gray bars, genes expressed at levels similar to the rest of PFC. Of note, mRNAs for serotonin receptors Htr1a and Htr2a were expressed, but not enriched, in Epyc cells, a not-unexpected finding given the expression of 5-HT_1A and 5-HT_2A in other populations of neurons in mouse mPFC beyond L6. *p < 0.05; **p < 0.01 by Student’s t test.

Optogenetic activation of L6 pyramidal neurons of medial prefrontal cortex excites L5 interneurons. A, Schematic representation of light activation of L6 pyramidal neurons of medial prefrontal cortex in Epyc-ChR2 mice with axons projecting to L5 interneurons. B, Channelrhodopsin-expressing pyramidal L6 neurons were robustly excited by targeted light stimulation over L6. The effects of increasing L6 light power are shown for one example L6 pyramidal neuron. C, Light activation of L6 robustly excited L5 interneurons. The effects of increasing L6 light power are shown for three different L5 interneurons. D, Top, close-up of the initial light-evoked action potential in L6 to show the timing from onset of light (blue dotted line) to peak of the spike (black dotted line). Scale bar: 20 mV, 1 ms. Note: In L6 the onset of depolarization from light is <1 ms. Bottom, voltage-clamp recording showing the light-evoked postsynaptic response in a L5 interneuron to demonstrate response latency. Scale bar: 40 pA, 1 ms. Postsynaptic responses in L5 interneurons were initiated 1.1 ± 0.3 ms after initial spike of L6 pyramidal neurons, as indicated by an arrow.

Two distinct groups of interneurons are found in L5 and are activated by light stimulation of L6 pyramidal neurons. A, FS interneurons of L5 characterized by injection of current steps. B, Representative trace of an L5 FS interneuron activated by L6. Note the rapidly depressing response to L6 activation. C, Activation of FS interneurons by L6 elicited action potential firing primarily during the initial phase of activation that rapidly depressed over the duration of the stimulation (number of elicited action potentials in first half of stimulation vs. second half: t₉ = 7.2, p < 0.0001, unpaired t test). D, nFS interneurons of L5 characterized by injection of current steps. E, Representative trace of a L5 nFS interneuron activated by L6, demonstrating a more regular firing pattern. F, L5 nFS interneurons were activated by L6 and fired in a regular pattern over the course of the stimulation (number of elicited action potentials in first half of stimulation vs. second half: t₇ = 1.9, p = 0.1, unpaired t test).