Phasic Inhibition Differentially Modulates Synaptic Dynamics Among Excitatory preBötzinger Complex Populations to Promote Rapid Breathing In Vivo

Abstract

The preBötzinger Complex (preBötC), located in the medulla, generates the inspiratory phase of breathing. A specific lineage of excitatory neurons, derived from Dbx1 expressing precursors, seems to underlie rhythmogenesis in the preBötC. In vitro and modelling studies suggest excitatory synaptic dynamics (e.g. presynaptic facilitation and depression) among Dbx1 neurons result in neuronal synchrony, population burst termination, and a refractory period for neural synchrony following a preBötC population burst. However, Dbx1 neurons are estimated to comprise only ~50% of the total glutamatergic population within the preBötC, and an additional ~50% of all preBötC neurons are inhibitory. Moreover, in intact animals, preBötC activity is influenced by sensory feedback, including vagally‐mediated lung stretch afferents that activate preBötC inhibitory neurons in phase with inspiration and increase respiratory frequency (rf). The mechanism by which sensory feedback increases rf and the functional implications of preBötC heterogeneity remain unclear. To investigate how these populations of preBötC neurons influence rf in vivo, we employed an optogenetic approach using transgenic mice expressing channelrhodopsin‐2 specifically within Dbx1 (Dbx1ERT2‐Cre), glutamatergic (VgluT2‐Cre), and gabaergic (Gad2‐Cre) neurons. In spontaneously breathing anesthetized mice with intact or bilaterally transected vagal nerves, the ventral medulla was exposed and an optical fiber was placed over the preBötC. Rhythmic XII activity was recorded during continuous (10 sec) light stimulation, repeated light stimulation occurring specifically during the inspiratory or expiratory phase, and random 100–200 ms light pulses. In vagus‐intact Dbx1‐Cre and VgluT2‐Cre mice, rf slowed when light stimulation occurred during inspiration. When light stimulation occurred during expiration, rf was dramatically increased in VgluT2‐Cre mice, but only modestly increased in Dbx1‐Cre mice. During continuous light stimulation, the inspiratory and expiratory effects of Dbx1 stimulation offset such that rf was unchanged, whereas VgluT2 stimulation increased rf. Phase shift plots revealed a refractory period for Dbx1 stimulation that was substantially longer than that for VgluT2 stimulation. In contrast, in vagotomized mice, stimulation during inspiration did not slow rf, the ability of Dbx1 and VgluT2 stimulation to reset the rhythm was no longer different, and refractory periods were dramatically increased. In vagus‐intact or vagotomized Gad2‐Cre mice, continuous light stimulation or stimulation during expiration slowed rf. However, if isolated to the inspiratory phase, Gad2 stimulation dramatically increased rf. This was confirmed by phase shift plots revealing that Gad2 stimulation during inspiration advances the subsequent inspiration, has little effect during early expiration (refractory period), and delays the subsequent inspiration during late expiration. All changes in frequency were primarily a result of changes in expiratory time (Te). Given these results, we propose a network structure in which phasic inhibition modulates excitatory synaptic dynamics. Specifically, sensory feedback inhibition differentially influences Dbx1 and non‐Dbx1 glutamatergic preBötC populations and promotes rapid breathing frequencies observed in vivo by reducing presynaptic depression of excitatory synapses.Support or Funding InformationNIH NRSA: F32HL134207NIH RO1: R01HL126523NIH PPG: P01 HL090554