For all frequencies, there was a significant left-right and flexor-extensor alternation (preferred direction around 0.5 in the circular plots; Figures 3D and 3E). The increased variability in burst parameters (see above) likely contributes to the slightly weaker left-right coupling in Shox2-vGluT2Δ/Δ cords (seen as shorter r-vectors

at lower frequencies of locomotion Figures 3D and 3E). At higher speeds of locomotion, the variability decreased but coupling was unaltered. Flexor-extensor coupling was not significantly different at any locomotor frequency. These findings underscore a role for Chx10off Shox2 INs in rhythm generation (a change in frequency) but with little to no effect on left-right and flexor-extensor coordination. However, in addition to influencing locomotor frequency, elimination of Shox2 INs also affected the stability of the rhythm, in a manner similar selleck chemical to that seen when all V2a neurons were ablated (Crone et al., 2008), and when the Shox2+ GW786034 price V2a neurons are eliminated (this study). The reduction in locomotor frequency is seen in Shox2-vGluT2Δ/Δ mice, where both Shox2+ V2a INs and Shox2+ non-V2a INs are silenced, but not in Shox2-Chx10DTA mice, where Shox2+ V2a INs alone

are ablated, suggesting a specific role for Shox2+ non-V2a INs in rhythm generation. Bath application of locomotor drugs exposes all spinal neurons to neural excitants uniformly and tonically, which likely is not the case in vivo. To test the locomotor phenotype of Shox2-vGluT2Δ/Δ mice in a more physiological context, we evaluated locomotor-like activity induced by brainstem stimulation, an efficient way of evoking bouts of locomotor activity. Rhythmic ventral root bursting was elicited by descending fiber stimulation in both control and Shox2-vGluT2Δ/Δ spinal cords ( Figures 3G and 3H). In controls, heptaminol burst frequency increased with stimulus strength ( Figure 3I). However, locomotor frequency in Shox2- vGluT2Δ/Δ cords was slower than controls at all stimulus intensities, apart from the lowest ( Figure 3I). Additionally, in Shox2-vGluT2Δ/Δ spinal cords locomotor frequency failed to increase with increasing stimulus

intensity ( Figure 3I). Left-right and flexor-extensor activities were in alternation (offset ∼180°) in both control and Shox2-vGluT2Δ/Δ mice at all stimulation intensities tested ( Figures 3J and 3K). These experiments show that Shox2 INs may be involved in mediating the descending locomotor drive and/or generating the rhythmic activity in the spinal cord. Although some of the reduction in frequency seen in neural-evoked locomotion may be due to a loss of Shox2-related descending drive, the drug-induced method to evoke locomotor-like activity bypasses the neural-evoked pathways for initiating locomotor-like activity. Together, the lower drug-evoked and stimulus-evoked locomotor frequencies seen in Shox2-vGluT2Δ/Δ cords as compared to controls suggest that spinal Shox2 INs play a significant role in rhythm generation.