, 2012, Soc. Neurosci., abstract). Therefore, the Shox2+ dI5 INs and/or the V2d neurons are likely responsible for decreased locomotor LY294002 datasheet frequency seen in this study. Another hallmark of vertebrate excitatory rhythm generating neurons is their recurrent connectivity (Li et al., 2006 and Parker and Grillner, 2000). Although connectivity was seen among Shox2 INs, it was sparse and we cannot ascribe this connectivity directly to Shox2+ non-V2a INs. It is notable that synaptic connectivity was not observed in previous studies of V2a neurons in the rodent spinal cord in Chx10-GFP

mice (Dougherty and Kiehn, 2010a and Zhong et al., 2010), nor has it been seen among excitatory Hb9 neurons (Wilson et al., 2005 and Hinckley and Ziskind-Conhaim, 2006). The rostrocaudal distribution of rhythmicity found in Shox2 INs may match with the subsets of Shox2 INs having a role in rhythm generation. Thus, the rhythm generating capability in the spinal cord is distributed (Kiehn and Kjaerulff, 1998) throughout the lumbar cord but with a rostral (L1–L3) dominance (Cazalets, 2005 and Kiehn and Kjaerulff,

1998). Notably, this rostral-caudal difference in rhythmicity was not seen in V2a neurons, as Chx10-GFP rhythmic neurons were equally distributed along the lumbar spinal cord (Dougherty and Kiehn, 2010a, Dougherty and Kiehn, 2010b and Zhong et al., 2010). Could there be an alternate explanation for the decrease in frequency observed in this study? Shox2 neurons could provide drive to the rhythm generating neurons—in which case a reduction in the glutamatergic drive AZD2281 order to rhythm generating neurons would account for the decrease in locomotor frequency. Metalloexopeptidase We do not favor this possibility, since the majority of Shox2 neurons, particularly in more rostral segments, are rhythmically active during locomotion, thereby placing them either as part of the rhythm generator or downstream from it. If Shox2 neurons provide tonic drive to rhythm generating neurons, they would have to be located locally as Shox2-halorhodopisin experiments involved application

of yellow light to an area of approximately three lumbar segments—with a consequent reduction in locomotor frequency. Another possibility is that the non-V2a Shox2 neurons are not rhythm generating but the effect seen is due to a nonspecific decrease in the number of excitatory neurons required for rhythm generation. Essentially when a critical mass of excitatory cells is eliminated, the frequency will drop. However, the Chx10 neurons outnumber the Shox2 neurons by at least 20%–25%. Therefore, if the critical excitatory cell mass hypothesis was correct, we would expect there to be a pronounced reduction in frequency in Chx10DTA experiments (that Crone et al., 2008 did not see), an intermediate reduction in the Shox2-Chx10DTA experiments (that we did not see), and a reduction in the frequency in Isl1-vGlut2Δ/Δ experiments (which Bui et al., 2012, Soc. Neurosci., abstract did not see).