9 mM and ∼6 mM, respectively. When applied Galunisertib mouse over the rostral lumbar segments, an aCSF composed of 0.9 mM [Ca2+]o and 6 mM [K+]o triggered an episode of locomotor-like activity in all preparations tested (n = 7; Figure 1H). Blind whole-cell recordings were performed from ventromedial neurons in the L1-L2 region to further investigate the relationship between changes in ionic concentrations and firing properties. Interneurons were identified by their high input resistance (604 ± 75 MΩ, n = 18) and the absence of antidromic response

to ventral root stimulation. A few minutes after NMA and 5-HT were applied, and long before the locomotor-like activity emerged, all interneurons (n = 18) were spiking (Figure 2A). Simultaneous recordings with ion-sensitive microelectrodes and intracellular pipettes enabled linking changes in ionic

concentrations to the cellular activity (Figures S2A and S2B). Half of the recorded neurons switched their firing pattern from spiking to bursting, either at the onset of locomotor-like activity (3/8 neurons; Figure 2B) or during ongoing locomotion (5/8 neurons). Superfusion of riluzole (5 μM) to block INaP progressively reduced the amplitude of membrane oscillations, which then became undetectable ( Figures 2D and 2E). As described previously ( Tazerart et al., 2007; Zhong et al., 2007), the ventral root burst progressively decreased in amplitude with little effect on the cycle frequency until locomotor-like activity disappeared ( Figure S2C). Furthermore, when preincubated for 45 min before the application of NMA/5-HT, riluzole (5 μM) Selleck BVD-523 prevented the emergence of locomotion (n = 3,

Figure S2D). The following set of experiments was performed to assess whether locomotor-related changes in [Ca2+]o and [K+]o may initiate intrinsic bursting properties. Whole-cell recordings were performed in neonatal rat slice preparations from spinal Ribonucleotide reductase interneurons (n = 187) located in the area of the locomotor CPG (ventromedial part of L1-L2). To discriminate the effect of ionic changes from that of neurotransmitters in generating membrane oscillations, we omitted the application of NMA and 5-HT. Reducing [Ca2+]o to 0.9 mM while keeping [K+]o at 3 mM did not affect the firing pattern (Figure 2F, left). Bursting could be induced only when reducing [Ca2+]o further to 0.3 or 0.0 mM (Figure 2G). At a constant [Ca2+]o (1.2 mM), pacemaker activities could not be evoked by increasing [K+]o to near 6 mM (Figure 2F, right) and appeared only at values above 9 mM (Figure 2G). A striking observation was the synergistic effect of reducing [Ca2+]o and increasing [K+]o on the generation of bursts. A concomitant reduction of [Ca2+]o to 0.9 mM and increase of [K+]o to 6 mM induced bursts in 25% of neurons (Figures 2G and 2H). These bursts were attributable to INaP as they were reversibly abolished by low concentrations of TTX (0.