Figure 6.

V3 neurons evoke a glutamate-mediated PAD, independent of GABA, by direct innervation of afferents. A: V3 neuron contact (VGLUT2⁺) on large myelinated Ia afferent branch in the intermediate spinal cord, at a branch point where the node of Ranvier is located (node also identified by the paranodal taper; afferent filled with neurobiotin). V3 neurons labeled with EYFP in Sim1//ChR2-EYFP mouse. Contact shown in yellow computed from 3-D reconstruction on right. B: PAD evoked by a brief light pulse (10 ms pulse, λ = 447 nm laser, 0.7 mW/mm², 3xT, laser aligned as in Fig. 4B) in Sim1//ChR2-EYFP mouse, and GABA PAD, NMDA PAD, and AMPA PAD components estimated by the change with sequential application of gabazine, APV, and CNQX, respectively (50 µM each). Recorded in S3 DR by grease gap (DRPs). C: same as B, but APV applied first, with sequential application of APV, gabazine, and CNQX. D and E: same as B and C, but for drPAD evoked by DR stimulation (Ca1 DR 0.1 ms, 2xT). F: group averages of V3 PAD early peak (at 15–20 ms), late peak (at 100 ms), and long-lasting NMDA PAD (at 250 ms), with sequential application of gabazine, APV CNQX (gabazine first condition), or APV first (only detailed for late V3 PAD since early PAD unaffected, not shown). *Significant PAD reduction with drug application, P < 0.05, n = 20 gabazine first recordings (of which 6 also had CNQX), and n = 11 APV first recordings, in S4 and S3 DR combined from 5 mice each. G: group averages of drPAD amplitudes at same times as measured for V3 PAD in F, for comparison. *Significant PAD reduction with drug application, P < 0.05, n = 20 gabazine first recordings, and n = 28 APV first recordings from 10 mice each. H: schematic of the putative disynaptic V3 neuron circuit mediating PAD independently of GABA, but with GABA tonically inhibiting V3 neurons, leading to a disinhibition of this glutamate-mediated PAD with gabazine. DR, dorsal root; NMDA, N-methyl-d-aspartate; PAD, primary afferent depolarization.