Mechanisms underlying the production of multiple respiratory patterns by a single neural network in vitro

Are different forms of breathing derived from one or multiple neural networks? We demonstrate that brainstem slices containing the pre-Boetzinger complex (PBC) generate two rhythms in normoxia, with striking similarities to defined breathing patterns known as eupneic ("normal") respiration and sighs. In anoxia, sighs ceased and eupneic activity was reconfigured into a third pattern consistent with the description of gasping in vivo . We conclude that a single medullary network underlies all three breathing patterns.; Sighs were phase-locked to the eupneic rhythm in control conditions, but not in the presence of strychnine (1 muM). A large majority (∼96%) of neurons active during the sigh were also active during eupnea, and vice versa, yet sighs were differentially modulated (by Substance P, and by the muscarinic acetylcholine receptor agonist oxotremorine), and could be selectively abolished by low concentrations of P/Q-type calcium channel blockers. Activation of the group III metabotropic glutamate receptor mGluR8 also abolished sighs, suggesting that it may be a presynaptic P/Q-type channel that is essential. Subsequent experiments tested this hypothesis by investigating the effects of calcium channel toxins and metabotropic glutamate receptor agonists on evoked excitatory postsynaptic potentials (EPSPs), as well as the effects of ionotropic glutamate receptor antagonists on network-level activity. The results of these studies suggest that the glutamatergic synapses subserving eupneic rhythmogenesis comprise two distinct subsets, with different pharmacology, of which only one is essential to the production of sighs.; These results provide a resolution to several longstanding questions within the respiratory field: whether eupnea and gasping are generated by separate networks (no); whether the pattern studied in vitro is really representative of the eupneic pattern in vivo (yes); whether gasps and sighs originate from similar brainstem mechanisms (no). The conventional understanding of sighs as a reflexive behavior is also shown to be incorrect. Perhaps more importantly, these results also constitute the first demonstration of reconfiguration in a mammalian pattern-generating network. Although investigation into the mechanisms responsible for this reconfiguration is only beginning, and many issues remain unresolved, these results provide a necessary framework for any future investigation into the neural origin of different respiratory patterns.