Discharge patterns of rostral pontine respiratory-modulated neurons in relation to vagal and superior laryngeal afferent input

Abstract

In midcollicular decerebrate, thoracotomized, paralyzed cats, phrenic nerve discharge served as the indicator of central respiratory rhythm and was used to control ventilation by a cycle triggered respiratory pump, which inflated the lungs in phase with central inspiration. Extracellular microelectrode recordings were taken from neurons of the pontine respiratory group (PRG). Unit activities were analyzed by means of cycle triggered histograms, spectral and coherence histograms, crosscorrelation histograms, and peristimulus histograms. Different types of respiratory modulated discharge activities, inspiratory (I), expiratory (E), and phase spanning (IE and EI) types with or without background activities in the complementary phase (subtyped T- and P-, respectively), were recorded from PRG neurons. About 80% of the neurons were of tonic type; and only about 10% had primarily E-related discharge (T-E and P-E types). No clear anatomical segregation of different types of neuron was found. The great majority of I-related neurons were strongly inhibited by lung inflation, i.e. their respiratory modulation increased during withholding of inflation. Electrical stimulation of vagal (V) afferents, which shortened the I phase, also inhibited the activity of I-related neurons. However, the latency of inhibition (median 120 ms) was much less than the latency to the off-switch of I (350-400 ms). Stimulation of superior laryngeal (SL) afferents, applied during the last third of the I phase, produced a short latency ({dollar}<{dollar}40ms) termination of I and excitation of I-related neuron activity (median latency 45 ms); the unit excitation always followed the abrupt termination of I. Spectral analysis showed that only a small minority (12/242) of I-modulated neurons had high frequency oscillation (HFO) coherent with phrenic HFO; and a very small fraction of PRG neurons had firing locked to individual stimuli (1/50 neurons with V stimulation and 2/36 neurons with SL stimulation). These results suggest that PRG I-related neurons are connected to medullary I neuron populations and to V and SL afferents only by diffuse pathways involving many synapses and that PRG neurons do not play a direct role in promoting the I off-switch produced by afferent in-puts. These conclusions are based on the rarity of short-term locking to HFO and stimuli as well as the direction and timing of the changes of unit activity produced by V and SL afferent inputs.