Serotonin inhibits low-threshold spike interneurons in the striatum

Abstract

Low-threshold spike interneurons (LTSIs) are important elements of the striatal architecture and the only known source of nitric oxide in this nucleus, but their rarity has so far prevented systematic studies. Here, we used transgenic mice in which green fluorescent protein is expressed under control of the neuropeptide Y (NPY) promoter and striatal NPY-expressing LTSIs can be easily identified, to investigate the effects of serotonin on these neurons. In sharp contrast with its excitatory action on other striatal interneurons, serotonin (30 μM) strongly inhibited LTSIs, reducing or abolishing their spontaneous firing activity and causing membrane hyperpolarisations.These hyperpolarisations persisted in the presence of tetrodotoxin, were mimicked by 5-HT(2C) receptor agonists and reversed by 5-HT(2C) antagonists. Voltage-clamp slow-ramp experiments showed that serotonin caused a strong increase in an outward current activated by depolarisations that was blocked by the specific M current blocker XE 991. In current-clamp experiments,XE 991 per se caused membrane depolarisations in LTSIs and subsequent application of serotonin (in the presence of XE 991) failed to affect these neurons.We concluded that serotonin strongly inhibits striatal LTSIs acting through postsynaptic 5-HT(2C) receptors and increasing an M type current.

Figure 1. Serotonin hyperpolarises striatal LTS interneurons

A, typical responses of an LTSI to positive and negative current injections which produce low-threshold calcium spikes during or after the pulse, respectively. B, cell-attached recording from an LTSI showing reversible depressing effects of serotonin (5-HT) on spontaneous firing; reapplication of serotonin after washout elicited similar effects as the first application. C, dose–response curve for the inhibitory effects of serotonin on spontaneous firing frequency of LTSIs. Results were obtained with cell-attached recordings. For each neuron, the effects of each concentration of serotonin were expressed as percentage of the firing frequency observed in the absence of serotonin. Percentage of inhibition was then defined as (100 – firing frequency in serotonin expressed as percentage of control). A sigmoidal curve was fitted to the data using a Matlab routine. This fit yielded a half-maximal response dose of 16 μm. Each concentration of serotonin was tested in at least 4 LTSIs. D, serotonin (30 μm) reversibly hyperpolarised a spontaneously active LTSI and fully blocked action potential generation. E, effects of serotonin on spontaneous firing frequency in individual LTSIs from rats or BAC NPY-GFP transgenic mice.F, average effects of serotonin on the spontaneous firing frequency of LTSIs from Sprague–Dawley rats or BAC NPY-GFP mice. Asterisk denotes statistical significance (*P < 0.05; assessed with Mann–Whitney U test; n = 8 for mice and 6 for rats). G, in the presence of TTX, serotonin reversibly hyperpolarises LTSIs; a typical example is shown. H, cell-attached recording from an LTSI in which the serotonin reuptake blocker citalopram reversibly abolished spontaneous firing.

Figure 2. The effects of serotonin on LTSIs are mediated by 5-HT_2C receptors

A, in a representative experiment, serotonin (30 μm) hyperpolarised an LTSI (recorded without any current injection) in the presence of TTX (1 μm). Subsequent addition of ketanserin (10 μm) caused the LTSI to depolarise to control level. B, another LTSI (recorded without current injection), with a resting membrane potential of –74 mV, was reversibly hyperpolarised by the 5-HT_2C receptor agonist WAY 161503 (10 μm) in the presence of TTX. C, in a different LTSI, in the presence of TTX, the hyperpolarising effects of WAY 161503 (10 μm) were reversed by subsequent application of ketanserin (10 μm; still in the presence of WAY 161503). D, in a cell-attached experiment, the 5-HT_2C receptor antagonist RS 102221 (1 μm) increased spontaneous firing frequency in a LTSI; in the presence of RS 102221, serotonin failed to affect the LTSI firing activity. E, in this experiment, an LTSI (recorded without current injection) displayed spontaneous firing activity in control solution. Serotonin induced a hyperpolarisation accompanied by a reduction in spontaneous firing frequency. Subsequent application of the 5-HT_2C receptor antagonist RS 102221 (1 μm) repolarised the LTSI and increased spontaneous firing activity. F, bar chart showing the average changes in firing frequency (with respect to control solution) caused by combinations of serotonin receptor ligands. Only LTSIs that were spontaneously active in control solution were included in this analysis. Asterisks denote statistical significance (*P < 0.05; **P < 0.001, assessed with Mann–Whitney U test; n = 14 for serotonin; n = 6 for WAY 161503; n = 5 for 5-HT + ketanserin; n = 5 for 5-HT + RS 102221). Firing frequency in the presence of either serotonin or WAY 161503 was significantly (P < 0.001) lower than in control.

Figure 3. The effects of serotonin on LTSIs are mediated by a reduction in an M type outward current

A, serotonin increases a voltage-dependent outward current in LTSIs. An example of the voltage-clamp experiments carried out in the presence of TTX. Voltage ramps (from –110 mV to –10 mV, 10 mV s⁻¹) were applied to an LTSI, before and after serotonin application. Membrane currents are plotted vs. voltage in control solution (black) and in the presence of serotonin (grey). B, voltage dependence of the serotonin-induced current in the same LTSI. The steady-state current induced by serotonin was calculated for each voltage by subtracting the membrane current measured in the presence of serotonin from that measured in control solution, before serotonin application. The zero level is indicated by the grey dashed line. C, the 5-HT_2C receptor agonist WAY 161503 increased a similar voltage-dependent outward current in an LTSI. Subsequent application of the 5-HT₂ receptor antagonist ketanserin reversed the effect of WAY 161503. D, in a representative experiment, XE 991 per se reduced a voltage-dependent outward current in LTSIs similar to the one induced by serotonin. Membrane currents are plotted vs. voltage in control solution (black) and in the presence of XE 991 (grey). E, voltage dependence of the current blocked by XE 991 in the same LTSI. The steady-state current blocked by XE 991 was calculated for each voltage by subtracting the membrane current measured in control solution from that measured after application of XE 991. F, in the presence of M current blocker XE 991 (20 μm), serotonin failed to induce an outward current. The zero level is indicated by the grey dashed line.

Figure 4. The inhibitory effects of serotonin on LTSIs are blocked by XE 991

A, in this LTSI, application of serotonin caused a reversible hyperpolarisation that abolished spontaneous firing activity. After washout of serotonin, application of the M current blocker XE 991 (20 μm) caused a depolarisation accompanied by an increase in spontaneous action potential frequency. In the continuous presence of XE 991, serotonin was reapplied but failed to affect the LTSI membrane potential or its firing frequency. Expanded traces show the changes in spontaneous firing activity in response to serotonin alone and in the presence of XE 991. B, in another experiment, in the presence of TTX, application of XE 991 depolarised an LTSI by 4 mV. Subsequent application of serotonin in the presence of XE 991 had no effects on the membrane potential of the LTSI.

Figure 5. Wortmannin prevents the inhibitory effects of serotonin

In this representative experiment, an LTSI's spontaneous firing rate increased significantly (P < 0.001 for the ISI) after wortmannin application (10 μm). In the presence of wortmannin, application of serotonin failed to affect the LTSI firing activity or membrane potential.