2B) which were each followed by a large mAHP (Fig. 2C), as previously described (Beck et al., 2004; Scuvee-Moreau et al., 2004). No AHP was observed at the end of positive current injection, contrary to what is seen in cortical pyramidal neurons. Action potentials were broad (duration at half-amplitude was 1.13 ± 0.25 ms; n = 90), with a typical shoulder on the

falling phase (Fig. 2B). It has been suggested that this shoulder is due to an influx of Ca2+ (Aghajanian & Vandermaelen, 1982; Vandermaelen & Aghajanian, 1983; Penington et al., 1992). The amplitude of the spikes as measured from the threshold was 67 ± 6 mV (n = 90). In intracellular experiments, presumed 5-HT neurons were PCI32765 defined according to the following criteria: they were either silent, with a resting membrane potential between −55 and −70 mV, or had a slow spontaneous firing rate. The action potential and mAHP were strictly similar to those recorded in young animals. Their input resistance was 280 ± 41 MΩ and the membrane time constant, τ, was 35 ± 3 ms (n = 22). These neurons were depolarized by the α1 agonist phenylephrine (3–10 μm; not shown), as already described (Vandermaelen

& Aghajanian, 1983). Taken together, these features identify them as presumed 5-HT neurons. Phenylephrine (10 μm) was added to the superfusion solution in all extracellular recordings. Under these conditions, presumed 5-HT neurons had a slow, regular firing rate of 0.4–3 spikes/s consisting of broad (> 2 ms) triphasic action potentials. These neurons were most probably serotonergic because their firing was inhibited by the 5HT1A Lumacaftor molecular weight agonist 8-OH-DPAT (30 nm) and this effect was blocked by the 5HT1A antagonist WAY100635 (100 nm; not shown). We have previously

shown that the potency of WAY100635 as an antagonist of 5HT1A receptors in these neurons (pKB, 9.57) is similar to its affinity at identified 5HT1A receptors (Defraiteur et al., 2007). In order to prevent any effect of the activation of 5HT1A receptors in the extracellular recordings reported in this paper, 100 nm WAY100635 was superfused together with the blockers mentioned above. In order to characterize the outward current underlying the mAHP observed in current clamp, we performed voltage-clamp experiments. As a first step, we used the same protocol as the one used previously in dopaminergic and other neurons Coproporphyrinogen III oxidase (Wolfart & Roeper, 2002). Neurons were held at −60 mV and 20-ms depolarizing pulses (referred to below as long pulses) were given to a range of voltages between −10 and +100 mV. This type of voltage step induced a subsequent outward current peaking immediately and lasting ~400 ms (Fig. 3A). In order to isolate the SK current from voltage-dependent K+ currents and/or synaptic currents, we used synaptic blockers and 5 mm TEA (Fig. 3A), as explained in detail in ‘Materials and methods’. The remaining outward current had a peak amplitude of 47 ± 21 pA (n = 69). Its mean time to peak was 75 ± 15 ms.