Loss of the Persistent Sodium Current Elicits Homeostatic Plasticity in Respiratory Rhythm Generation

The neural networks underlying the generation of breathing rhythms must remain stable and dynamic throughout the lifetime of an organism. In many neural networks, it is hypothesized that mechanisms of homeostatic plasticity regulate synaptic and/or intrinsic properties of neurons to maintain stability. The neural network giving rise to the inspiratory phase of breathing, termed the preBötzinger Complex (preBötC), utilizes a combination of intrinsic membrane conductances, such as the persistent sodium current (I NaP ) and the nonspecific cation current (I CAN ), as well as synaptic properties to generate and shape the inspiratory rhythm. When isolated in brainstem slices, the preBötC continues to produce remarkably stable rhythmic population activity in phase with inspiratory motor output. We hypothesized that homeostatic mechanisms within the preBötC maintain stability of the inspiratory rhythm. Slices containing the preBötC were incubated overnight (~16hrs) in 10μM DNQX or 20μM riluzole to block excitatory synaptic transmission or the persistent sodium current, respectively. Upon washout, we observed that DNQX did not significantly alter the inspiratory rhythm. However, in slices incubated with riluzole, inspiratory burst amplitude was increased and frequency was decreased compared to controls (ACSF only). To explore this finding further, rhythmic activity was recorded for the duration of the riluzole exposure. Following ~30–40min, riluzole perturbed the rhythm such that inspiratory frequency dramatically slowed or stopped. Surprisingly, after ~7hrs the rhythm came back spontaneously, indicating plasticity had occurred to overcome the loss of I NaP . The rhythm and burst shape were altered such that amplitude and burst rise slope were increased (~300% and 220% baseline, respectively) and frequency was slowed (~30% baseline). Following thorough washout of riluzole (1–3hrs), the new large and slow inspiratory pattern remained, suggesting long lasting changes in network properties had occurred. In a subset of preBotC slices, riluzole was reapplied following washout. In contrast to naïve slices, there was no effect on inspiratory frequency following >1hrs in riluzole, indicating that the new rhythm was insensitive to loss of I NaP . However, burst amplitude was reduced suggesting I NaP remained present in preBötC neurons. These studies demonstrate that the preBötC can compensate for long‐lasting changes in network properties with mechanisms of homeostatic plasticity and begin to provide a basis for understanding how the neural networks underlying breathing rhythms may adapt to maintain stability in health and disease. Support or Funding Information NIH RO1 HL 126523