Evidence from motoneurone synchronization for disynaptic pathways in the control of inspiratory motoneurones in the cat.

1. Motoneurone synchronization was measured by cross-correlation between paired inspiratory discharges in external and internal intercostal nerves or their intramuscular branches (T3 to T8) or in the phrenic nerve (C5 root or both C5 and C6 roots independently) in anaesthetized, paralysed cats. 2. All cross-correlation histograms showed central peaks, for which the durations at half-amplitude (half-widths) from internal nerve pairs in adjacent segments were all less than for external nerve pairs in adjacent segments or within a segment (means, 1.6 ms vs. 3.4 ms for adjacent segments). Values for external-internal pairs covered the ranges for both these two. Lowest values came from two phrenic pairs (1.2 and 1.4 ms). 3. The peaks from ipsisegmental external-internal pairs were usually asymmetric and the maximum of the peak was often displaced to a lag of about -1 ms (external nerve providing the reference spikes), whereas peaks from external-external pairs were always symmetrical and centred on zero. Phrenic-internal peaks gave maxima with lags about 1 ms less than for phrenic-external peaks from the same segments. 4. Two explanations were considered possible for the differences in duration and timing: an extra synapse on the pathway to the external nerve motoneurones, or a correlation kernel for a monosynaptic connection to the external nerve motoneurones that had a slower time course than that for the internal or phrenic nerve motoneurones. Computer simulations, assuming the extra synapse, gave a good fit to the observed time courses of the correlation peaks for all categories of nerve pairs using single values of parameters (e.g. EPSP rise time) consistent with those in the literature. This could not be achieved with the different correlation kernel model. The timing of high-frequency oscillation (HFO), which was sometimes present in the correlations, was also better predicted with the extra synapse model. 5. It is concluded that most of the synchronization between external nerve motoneurones is derived from disynaptic common inputs and that any motoneurone synchronization peak with a half-width greater than about 2.2 ms should be assumed to be likely to contain di- or oligosynaptically derived components.