Fig. 3

Macropatch recordings confirm tight synchronization of inspiratory onset along VRC networks. A Anatomical distribution of paired macropatch recordings from regions along the ventral respiratory column. Paired recordings of preBötC/pFRG (orange), VRG/pFRG (yellow), and preBotC/Pico (blue) were carried out in 7 preparations (indicated by distinct symbols).B.i Pseudo-spikes identified by thresholding the macropatch signal (criterion + -3 SD of the whole signal with a refractory period of 1.2 ms) are indicated by black points. B.ii Pseudo-spike times are convolved with a Gaussian window (SD = 3.7 ms, equivalent to a 270 Hz sampling rate) to obtain a pseudo-firing rate time signal which models a local summed potential of macropatch activity. C Cross-correlation of pseudo firing rate between paired recordings, whose anatomical sources are color-coded as in A. Peaks to the left of zero on the x-axis indicate that caudal structures led. Dashed green box spans ± 0.02 s. D.i We tested relative onset times by applying a threshold (set to 33% of the maximum pseudo firing rate for that recording) to detect burst onset using successively smaller Gaussian windows to ensure that threshold would not be trapped in a local minimum at the smallest 3.7 ms window. Both methods produced the same result: in preBötC/pFRG pairings, preBötC led pFRG in 2/5 experiments; in VRG/pFRG recordings, VRG led in 1/2 experiments, and in the PBC/PiCo pair onset times matched. In all cases but one, onset time differences led or lagged by less than 0.02 s (dashed green box, C). These findings are congruent with synchronous onset times detected using optical recording methods