The ventilatory response to CO 2 changes as a function of neonatal development. In rats, a ventilatory response to CO 2 is present in the first 5 days of life, but this ventilatory response to CO 2 wanes and reaches its lowest point around postnatal day 8. Subsequently, the ventilatory response to CO 2 rises towards adult levels. Similar patterns in the ventilatory response to CO 2 are seen in some other species, although some animals do not exhibit all of these phases. Different developmental patterns of the ventilatory response to CO 2 may be related to the state of development of the animal at birth. The triphasic pattern of responsiveness (early decline, a nadir, and subsequent achievement of adult levels of responsiveness) may arise from the development of several processes, including central neural mechanisms, gas exchange, the neuromuscular junction, respiratory muscles and respiratory mechanics. We only discuss central neural mechanisms here, including altered CO 2 sensitivity of neurons among the various sites of central CO 2 chemosensitivity, changes in astrocytic function during development, the maturation of electrical and chemical synaptic mechanisms (both inhibitory and excitatory mechanisms) or changes in the integration of chemosensory information originating from peripheral and multiple central CO 2 chemosensory sites. Among these central processes, the maturation of synaptic mechanisms seems most important and the relative maturation of synaptic processes may also determine how plastic the response to CO 2 is at any particular age.

We thank D. Phillips for technical assistance, J. Wagner for the 123.7 cells, J. Boulter for providing nAChR cDNA clones, S. Froehner for use of his peristaltic pump, J. Dunlap for use of his densitometry equipment, and J. Margiotta and J. Wagner for comments on the manuscript, and P.D.G. thanks B.P. for inspiration.

No abstract

The pattern of efferent neural activity recorded from the isolated brain stem preparation of the tadpole Rana catesbeiana was examined to characterize fictive gill and lung ventilations during ontogeny. In vitro recordings from cranial nerve (CN) roots V, VII, and X and spinal nerve (SN) root II of premetamorphic tadpoles showed a coordinated sequence of rhythmic bursts occurring in one of two patterns, pattern1, high-frequency, low-amplitude bursts lacking corresponding activity in SN II and pattern 2, low-frequency, high-amplitude bursts with coincident bursts in SN II. These two patterns corresponded to gill and lung ventilatory burst patterns, respectively, recorded from nerve roots of decerebrate, spontaneously breathing tadpoles. Similar patterns were observed in brain stem preparations from postmetamorphic tadpoles except that they showed a greater frequency of lung bursts and they expressed fictive gill ventilation in SN II. The laryngeal branch of the vagus (Xl) displayed efferent bursts in phase with gill and lung activity, suggesting fictive glottal constriction during gill ventilation and glottal dilation during lung ventilation. The fictive gill ventilatory cycle of pre- and postmetamorphic tadpoles was characterized by a rostral to caudal sequence of CN bursts. The fictive lung ventilatory pattern in the premetamorphic animal was initiated by augmenting CN VII discharge followed by synchronous bursts in CN V, X, SN II, and Xl. By contrast, postmetamorphic patterns of fictive lung ventilation were characterized by lung burst activity in SN II that preceded burst onset in CN V and followed the lead burst in CN VII. We conclude that recruitment and timing of pattern 1 and pattern 2 rhythmic bursts recorded in vitro closely resemble that recorded during spontaneous respiratory behavior, indicating that the two patterns are the neural equivalent of gill and lung ventilation, respectively. Further, fictive gill and lung ventilatory patterns in postmetamorphic tadpoles differ in burst onset latency from premetamorphic tadpole patterns and resemble fictive oropharyngeal and pulmonary burst cycles in adult frogs.

Sites of central CO2 chemosensitivity were investigated in isolated brain stems from Rana catesbeiana tadpoles and frogs. Respiratory neurograms were made from cranial nerve (CN) 7 and spinal nerve 2. Superfusion of the brain stem with hypercapnic artificial cerebrospinal fluid elicited increased fictive lung ventilation. The effect of focal perfusion of hypercapnic artificial cerebrospinal fluid on discrete areas of the ventral medulla was assessed. Sites of chemosensitivity, which are active continuously throughout development, were identified adjacent to CN 5 and CN 10 on the ventral surface of the medulla. In early- and middle-stage tadpoles and frogs, unilateral stimulation within either site was sufficient to elicit the hypercapnic response, but simultaneous stimulation within both sites was required in late-stage tadpoles. The chemosensitive sites were individually disrupted by unilateral application of 1 mg/ml protease, and the sensitivity to bath application or focal perfusion of hypercapnia was reassessed. Protease lesions at CN 10 abolished the entire hypercapnic response, but lesions at CN 5 affected only the hypercapnic response originating from the CN 5 site. Neurons within the chemosensitive sites were also destroyed by unilateral application of 1 mM kainic acid, and the sensitivity to bath or focal application of hypercapnia was reassessed. Kainic acid lesions within either site abolished the hypercapnic response. Using a vital dye, we determined that kainic acid destroyed neurons by only within 100 microm of the ventral medullary surface. Thus, regardless of developmental stage, neurons necessary for CO2 sensitivity are located in the ventral medulla adjacent to CN 5 and 10.