The latencies of antidromic soma spikes of the respiratory bulbo-spinal neurons are not constant, but vary slightly during the respiratory cycle. Based on studies made on spinal motoneurons antidromic latency (ADL) variations give a good indication of synaptic events which drive cell activity. The ADL variation of 65 respiratory neurons, the anatomical and functional characteristics of which were inferred from their discharge patterns and from the possibility of getting an antidromic invasion after spinal cord stimulation (bulbo-spinal neurons), vagus nerve stimulation (laryngeal motoneurons), or stimulation of other medullary regions (propriobulbar neurons), was studied in cats. Our results deal with 46 inspiratory bulbo-spinal neurons (IBSN), 9 expiratory bulbo-spinal ones (EBSN), 7 laryngeal motoneurons (LM, inspiratory and expiratory) and 2 inspiratory propriobulbar neurons (IPBN). IBSN (20) were located in the ventro-lateral region of the nucleus solitarius (the dorsal respiratory nucleus); the other neurons were found in the nucleus ambiguus or its vicinity (the ventral respiratory nucleus). For each neuronal type the ADL reached a maximal value in inspiration for expiratory neurons, and in expiration for inspiratory neurons, while the minimal latency occurred during the period of cell discharge; the latency difference varied between 0.60-0.04 ms, depending on the neurons, with a mean value of 0.16 .+-. 0.04 ms for the nonvagal inspiratory neurons, 0.22 .+-. 0.11 ms for the expiratory ones and 0.29 .+-. 0.25 ms for the laryngeal motoneurons (P = 0.05). For each inspiratory neuron (except a few inspiratory LM), ADL decreased more or less rapidly when phrenic activity appeared (E [expiration] .fwdarw. I [inspiration] transition). This observation holds true for early discharging neurons or late discharging ones. All inspiratory neurons receive a synchronous excitatory input corresponding to a general uprising of excitability in the inspiratory pool. During the silent period (or minimal activity), of inspiratory neurons, 3 patterns of ADL evolution were observed: a plateau (type I), a progressively decreasing curve (type II), or an intermediary curve, i.e., 1st plateau-like, and then decreasing at a moment which varied from one respiratory cycle to another cycle (type III). It is significant that IBSN of type II were exclusively located in the dorsal respiratory nucleus, where they represented 35% of the whole IBSN population (vs. 4% in the ventral nucleus). Such neurons may be involved in the genesis of respiratory rhythm. During maintained inflation, the most consistent changes in ADL were exhibited during expiration by IBSN belonging to the dorsal nucleus. An increase or a decrease of ADL was observed. It was not possible to obtain an antidromic invasion of 50% of EBSN during their silent period, vs. 2% of IBSN. Apparently, inhibitory mechanisms are predominant in expiratory off-switching; from this view point, the removal of inhibition during inspiratory phase would occur either suddenly, at the I .fwdarw. E transition or progressively. For 3 inspiratory LM (out of 6), a strong diminution of ADL could occur several hundred milliseconds before E .fwdarw. I transition (type IV). This decrease has probably the significance of a central respiratory drive potential, for the intercostal motoneurons. Such motoneurons may be involved in driving other laryngeal neurons.
