Fig. 7

Both neurons and glia are constituents of the syncytium. A To directly test whether neurons participated in syncytial oscillations, patch-clamp recordings were carried out following synaptic blockade (N = 4). Here, serial patch clamp recordings in the same preparation following synaptic blockade at the level of pFRG (blue trace and square, A.i) and preBötC (orange trace and square, A.ii) are shown. Intracellular recording locations relative to VIIn (dashed oval) reconstructed from images obtained during optical recording are shown in cartoon at top right. Fluctuations in membrane potential phase-locked to syncytial activity detected at ROIs obtained during concurrent optical recording (gray traces) confirms that neurons participate in syncytial activity. Boxes to the right of traces indicate the location of ROIs (green dots) in relation to the patch electrode tips (orange and blue squares), establishes that syncytial rhythm could be detected far from the patch-clamped cell, consistent with gap-junction coupling. B Following bath application of the Na_(V) blocker TTX (1 μM), stationary syncytial rhythmic activity was replaced by transient spindle activity (top), which was uncoupled from neighboring activity, as confirmed by the absence of peaks in cross-correlations between all traces (bottom). C To directly assess the contribution of neurons and glia to syncytial oscillations, we crossed GCaMP6F mice with mice expressing cre- in neurons (Tg(Actl6b-Cre)4092Jiwu/J) and glia (B6;FVB-Tg(Aldh1l1-cre)JD1884Htz/J) respectively. Prior to synaptic blockade, both mice expressing GCaMP in neurons (C.i, left panel; N = 3) and glia (C.ii, left panel, N = 5) showed robust activity during inspiration. Following synaptic blockade, fluctuations in luminance matching syncytial rhythm in germline mice was detected (C.i, C.ii, right panels). Taken together, these findings confirm that neurons participate in syncytial oscillations and suggest that glia do so as well