To Sigh Or Not To Sigh: Role Of Glia In Sigh Generation

The respiratory network generates two distinct rhythmic activities: normal respiration (eupnea), which is relatively fast and sighs that occur at a very slow, but regular frequency. Understanding how the slow periodicity of sighs is generated has been challenging. Here, we propose that glia cells play a critical role in setting up the periodicity of sighs. In mouse transverse brain slice preparations containing the respiratory network, specifically the pre‐Bötzinger complex (PreBötC) we recorded Ca2+ signals over a time interval of 2 to 5 min to capture sighs, and subsequently applied aCSF solution containing 0KCl to differentiate glia from neurons. Neurons drastically decrease their Ca2+ signal, while glia increase intracellular Ca2+ in response to 0KCl. Among the glia cells we observed Ca2+ signals that were synchronized with the periodic sighs. In fact, the phase shift analysis between sighs and Ca2+ signal from non‐neuronal cells showed that on average Ca2+ spikes preceded sigh activity. Furthermore, glia toxins inhibit sighs while eupneic activity was preserved. Based on our pharmacological manipulations in vitro and in vivo we propose the hypothesis that release of glial ATP through activation of P2Y receptors activates glial and neuronal Gq/11 protein. Gq/11 in turn increases glial and neuronal intracellular Ca2+, which in turn leads to the release of glial ATP, thus locking the cycle. Our experimental data are consistent with our computational approach that models the neuronal‐glial interaction. The network model predicts that the presence of IP3‐like channels is necessary to produce “sigh” behavior and to mimic transitions between activity states caused by neuromodulators. We analyze the model dynamics and describe dose‐response curve to neuromodulator such as norepinephrine. Our theoretical data are not only similar to our in vivo and in vitro experimental data, but the model also predicts that depending on the modulatory state of the respiratory center, the network can exhibit bi‐stable behavior, meaning that sighs can be triggered between “on” and “off” states by short perturbation. We detected two different mechanisms that support sigh bi‐stability. The first mechanism is based on the presence of a saddle equilibrium and the range of Ca2+ parameters where the observed bistability is limited by sub‐critical Andronov‐Hopf bifurcation and homoclinic bifurcation for periodic cycle. The second mechanism is based on the presence of a saddle periodic orbit and the parameter range is defined by sub‐critical Andronov‐Hopf bifurcation and saddle‐node for periodic orbits bifurcation. This theoretical prediction is consistent with our experiments in in vivo anesthetized mice: Freely breathing anesthetized mice that did not spontaneously produce sighs were stimulated by repeated, brief pulses of hypoxia, which elicited sustained periodic sighs. Support or Funding Information NIH: F32 HL121939‐01