Our results show a segregation in location of Shox2 neurons based on their connectivity or lack of connectivity to motor neurons (Figure 6H). Therefore, we next determined

if the Shox2 INs connecting with flexor GSK1349572 manufacturer and extensor motor neurons are also segregated anatomically. In experiments tracing monosynaptic rabies virus spread separately from GS and TA motor neurons, we found that the percentage of Shox2 INs labeled from TA was three-fold greater than from GS (Figures 6I–6L). Whereas Shox2 INs constituted 5% of last order neurons labeled from the TA motor neurons, they only made up 1.5% of GS premotor neurons (Figure 6L), confirming the clear flexor bias of these connections observed also by anterograde Everolimus datasheet tracing (Figures 6D and 6L). This flexor dominance was evident at the level of all Shox2 premotor INs, regardless of rostral-caudal location. Both GS and TA injections labeled Shox2 INs in overlapping areas of the most

lateral area of lamina VII (Figures 6I–6K) demonstrating that the Shox2 INs projecting to flexor and extensor motor neurons are intermingled. In summary, our findings suggest that Shox2 INs segregate into a laterally located premotor population and a more medially-positioned population, which corresponds to the location of the Shox2+ nonpremotor INs. Additionally, within the premotor Shox2 IN population, there is a connectivity bias toward flexor motor neurons. Based on findings in other locomotor networks, rhythm-generating neurons are interconnected and provide excitation to several other identifiable CPG neurons. We therefore further evaluated the connectivity of Shox2 INs (Figure 7A). Rhythm-generating neural networks in Xenopus tadpole and lamprey are thought to be excitatory neurons that are recurrently, although sparsely, interconnected ( Roberts et al., 1998 and Grillner, 2003). To probe recurrent connectivity within the Shox2 population, we performed dual recordings from fluorescently labeled Shox2 INs in Shox2::Cre; Z/EG mice in dorsal-horn-removed preparations.

Depolarizing synaptic connections were detected in 4 of 41 pairs of Shox2 INs ( Figures 7B–7D). In all four cases, coupled pairs were found mafosfamide in close proximity and connections were unidirectional: spiking in one neuron resulted in EPSPs in the second neuron, but there was no reciprocal activation. In two of the connected pairs, EPSPs built up with each successive spike ( Figure 7C). The amplitude of the EPSPs ranged from 0.05 to 1 mV. Thus, Shox2 INs are sparsely interconnected, without direct monosynaptic feedback. Connectivity among neurons in excitatory populations may be expected and has been examined in a similar manner in other populations ( Dougherty and Kiehn, 2010a, Zhong et al., 2010, Wilson et al.