In studies of the central neural control of breathing, little advantage has been taken of comparative approaches. We have developed an in vitro brainstem preparation using larval Rana catesbeiana which generates two rhythmic neural activities characteristic of lung and gill ventilation. Based on the pattern of the facial (VII) nerve activity both lung and gill rhythm-related respiratory cycles were divided into three distinct phases. The purpose of this study was to characterize and classify membrane potential trajectories of respiratory motoneurons in the VII nucleus at intermediate stages (XII-XVII) of development. 2. Seventy-five respiratory-modulated neurons were recorded intracellularly within the facial motor nucleus region. Their resting membrane potential was between -40 and -80 mV. Sixty of them were identified as VII motoneurons and fifteen were non-antidromically activated. Membrane potentials of fifty-six of the seventy-five neurons were modulated with both lung (5-27 mV) and gill rhythms (3-15 mV) and the remaining nineteen neurons had only a modulation with lung rhythmicity (6-23 mV). No cells with gill modulation alone were observed. 3. All of the cells modulated with lung rhythmicity had only phase-bound depolarizing or hyperpolarizing membrane potential swings which could be categorized into four distinct patterns. In contrast, of the fifty-six cells modulated with gill rhythmicity, thirty-two were phasically depolarized during distinct phases of the gill cycle (four patterns were distinguished), whereas the remaining twenty-four were phase spanning with two distinct patterns. The magnitudes of lung and gill modulations were proportionally related to each other in the cells modulated with both rhythms. 4. In all sixteen neurons studied, a reduction or a reversal of phasic inhibitory inputs during a portion of the lung or gill respiratory cycle was observed following a negative current or chloride ion (Cl-) injection. The phasic membrane resistance modulation in relation to the gill rhythm was analysed in six neurons and a relative decrease in the somatic membrane resistance (0.7-8.1 M omega) was detected during the periods of hyperpolarization. 5. We propose that, at these intermediate stages of development: (a) both gill and lung respiratory oscillations in motoneurons are generated by respiratory premotor neurons having only a few distinct activity patterns; (b) these patterns delineate distinct portions of the centrally generated respiratory cycles; and (c) phasic synaptic inhibition, mediated by Cl-, contributes to shaping the membrane potential trajectories of respiratory motoneurons.
