Efferent modulation of spontaneous lateral line activity during and after zebrafish motor commands

Abstract

Accurate sensory processing during movement requires the animal to distinguish between external (exafferent) and self-generated (reafferent) stimuli to maintain sensitivity to biologically relevant cues. The lateral line system in fishes is a mechanosensory organ that experiences reafferent sensory feedback, via detection of fluid motion relative to the body generated during behaviors such as swimming. For the first time in larval zebrafish ( Danio rerio), we employed simultaneous recordings of lateral line and motor activity to reveal the activity of afferent neurons arising from endogenous feedback from hindbrain efferent neurons during locomotion. Frequency of spontaneous spiking in posterior lateral line afferent neurons decreased during motor activity and was absent for more than half of swimming trials. Targeted photoablation of efferent neurons abolished the afferent inhibition that was correlated to swimming, indicating that inhibitory efferent neurons are necessary for modulating lateral line sensitivity during locomotion. We monitored calcium activity with Tg(elav13:GCaMP6s) fish and found synchronous activity between putative cholinergic efferent neurons and motor neurons. We examined correlates of motor activity to determine which may best predict the attenuation of afferent activity and therefore what components of the motor signal are translated through the corollary discharge. Swim duration was most strongly correlated to the change in afferent spike frequency. Attenuated spike frequency persisted past the end of the fictive swim bout, suggesting that corollary discharge also affects the glide phase of burst and glide locomotion. The duration of the glide in which spike frequency was attenuated increased with swim duration but decreased with motor frequency. Our results detail a neuromodulatory mechanism in larval zebrafish that adaptively filters self-generated flow stimuli during both the active and passive phases of locomotion. NEW & NOTEWORTHY For the first time in vivo, we quantify the endogenous effect of efferent activity on afferent gain control in a vertebrate hair cell system during and after locomotion. We believe that this pervasive effect has been underestimated when afferent activity of octavolateralis systems is characterized in the current literature. We further identify a refractory period out of phase with efferent control and place this gain mechanism in the context of gliding behavior of freely moving animals.

Corollary discharge prevents signal distortion and enhances sensing during locomotion

“…Swimming activity was identified in immobilized larvae using extracellular recordings from ventral motor nerve roots, while lateral line activity was determined from patch clamp recordings (loose-patch configuration) of single afferent neurons (hereafter referred to as units) in the posterior lateral line afferent ganglion. Single neuromasts were sinusoidally stimulated using a glass bead affixed to a piezoelectric transducer at 5, 20, or 40 Hz to simulate the normal tail beat frequencies of freely swimming larvae (37, 42). Afferent fiber responses rapidly habituated to stimulation (exemplar trace at 20 Hz and peristimulus time histogram (PSTH), Figure 1 Ai, ii; 5 and 40 Hz in Supplementary Fig.…”

“…Thus the effect of intense habituation, like CD ON, should be to create a more faithful representation of the body wave, as implied by our model in Figure 1. However, the change in gain due to habituation will depend on highly variable intercellular differences (Figure 2 B,C) as well as variable effects of motor frequency and duration (see also, (37, 42)). Thus, habituation should be an intrinsically unreliable mechanism for increasing the fidelity of afferent responses with respect to the input stimulus.…”

“…Swim bouts and a subsequent post-swim period of equal time were normalized to unit duration. These periods correspond to CD ON (swimming, red) followed by CD OFF (white), although residual effects of CD ON persist into the post-swim period (37). All effects normalised to start of trace.…”

“…The CD to the lateral line tonically modulates the sensitivity of the lateral line during swim bouts (25, 35–37). Efferent projections from the hindbrain synapse on hair cells and evoke an inhibitory hyperpolarizing current while swimming, reducing the amount of presynaptic glutamate vesicle release (38–40).…”

“…Hyperpolarization increases the current required to evoke exocytosis, which demands a larger cupular deflection to achieve the same afferent fiber activity. In addition to the reduced number of spontaneous and evoked afferent fiber spikes (25, 37, 40), the higher deflection threshold will also narrow the response interval (which in turn increases vector strength, Figure 1 Bii). The narrower interval results from delayed early spikes and advanced late spikes (as measured by the quantiles of the spike phase distribution), and affects sustained responses in two important ways.…”