Click here to close
Hello! We notice that you are using Internet Explorer, which is not supported by Xenbase and may cause the site to display incorrectly.
We suggest using a current version of Chrome,
FireFox, or Safari.
???displayArticle.abstract???
Information processing in the vertebrate brain is thought to be mediated through distributed neural networks, but it is still unclear how sensory stimuli are encoded and detected by these networks, and what role synaptic inhibition plays in this process. Here we used a collision avoidance behavior in Xenopus tadpoles as a model for stimulus discrimination and recognition. We showed that the visual system of the tadpole is selective for behaviorally relevant looming stimuli, and that the detection of these stimuli first occurs in the optic tectum. By comparing visually guided behavior, optic nerve recordings, excitatory and inhibitory synaptic currents, and the spike output of tectal neurons, we showed that collision detection in the tadpole relies on the emergent properties of distributed recurrent networks within the tectum. We found that synaptic inhibition was temporally correlated with excitation, and did not actively sculpt stimulus selectivity, but rather it regulated the amount of integration between direct inputs from the retina and recurrent inputs from the tectum. Both pharmacological suppression and enhancement of synaptic inhibition disrupted emergent selectivity for looming stimuli. Taken together these findings suggested that, by regulating the amount of network activity, inhibition plays a critical role in maintaining selective sensitivity to behaviorally-relevant visual stimuli.
Figure 2. Behavioral experiments in freely-swimming tadpoles. (A) Trajectories of the tadpole and the visual stimulus over the course of one experiment; small dots represent points of avoidance maneuver initiation. Scaled images of a circle and a tadpole are superimposed on their respective trajectories. (B) A summary of the parameters measured for every collision (see 2 for a detailed description; R.-C. offset stands for rostro-caudal offset). (C) Escape probability as a function of visual stimulus maximal angular size during the collision event; each square represents an average of 29 events and the red line is a fourth degree polynomial fit. (D) One of the pair-wise comparisons between escapes in response to stimuli of different size. Gray circles represent data from individual collision events, boxes and whiskers show medians and interquartile ranges, and the asterisk marks statistical significance (P < 0.05). Tadpoles escaped from smaller circles at a faster speed than from larger ones. (E) Relative efficacy (probability of triggering an avoidance maneuver, normalised over the individual responsiveness of each animal), as observed in freely-swimming tadpoles, for dark circles either colliding with the animal, or instantaneously appearing near it.
Figure 3. Experiments in partially immobilised animals. (A) General view of the preparation, with the animal partially immersed in agarose gel, and an optic fiber placed near the eye. (B) Simplified representation of stimuli presented to the animal. Here ‘flash’ or ‘f’ represents an abrupt change in lightness, ‘ramp’ or ‘r’ represents a gradual change, ‘grid’ or ‘g’ stands for an array of linearly expanding squares, ‘crash’ or ‘c’ stands for a linearly expanding circle, and ‘control’ stands for the absence of a visual stimulus. (C) Typical responses (tail deflections) to these stimuli. (D) Sample swimming trajectory reconstructions for different stimuli; black circles mark the response onset. (E) Average ‘distances virtually traveled’ for every animal, in response to different stimuli (see 2); boxes represent medians, black whiskers show the interquartile range, and gray lines and asterisks indicate comparisons with ‘flash’ with PW < 0.05.
Figure 4. Typical responses to visual stimuli recorded at different points in the visual system, and their relative cumulative response amplitudes. (A) Recordings from the optic nerve. (B and C) Whole-cell recordings in the OT for inhibitory and excitatory currents, respectively. (D) Spiking output of the OT cells. In every panel ‘f’ stands for ‘flash’, ‘r’ for ‘ramp’, ‘g’ for ‘grid’, and ‘c’ for ‘crash’. For relative response amplitudes (right column) all measurements are shown as gray circles, together with medians represented by boxes, and interquartile ranges. Gray lines and asterisks mark significance levels of P < 0.05 compared with response to ‘flash’. Outliers were brought into axes limits. EPSC, excitatory post-synaptic current; IPSC, inhibitory post-synaptic current.
Figure 5. Role of excitation in inhibition in looming stimulus selectivity. (A) Effect size (difference between responses to ‘crash’ and ‘flash’, divided by trial-to-trial variability) observed at different stages of information processing in the tadpolebrain. Left to right: recordings in the optic nerve (Eye), GABAergic inputs to OT cells (GABA), glutamatergic inputs to OT cells (Glu), and OT cell spike output (Spikes). Asterisks mark effects that are significantly larger than zero. (B) Predictions of cell spike output that are based on amplitudes of synaptic excitation and inhibition (circles) match actual median normalised spike output (black squares) better than those based on excitatory inputs alone (triangles). Stimuli were normalised by the average response across all stimulus types (similarly to Fig. 4). See details in the text. (C) Average synaptic responses to different visual stimuli, normalised by total synaptic charge, with inhibitory (pale wide line) and inverted excitatory (darker narrow line) curves superimposed. Inhibitory currents are lagging behind excitatory currents for all stimuli. (D) Average positions and values of cross-correlogram peaks for excitatory and inhibitory synaptic currents recorded in OT cells (horizontal axis represents the lag between excitation and inhibition; n = 28; whiskers represent SEM). Responses to ‘flash’ and ‘crash’ are not significantly different (PMW > 0.05). f, ‘flash’; r, ‘ramp’; g, ‘grid; c, ‘crash’.
Figure 6. Pharmacological modulation of OT network activity. In every panel, columns of data represent different experimental conditions [left to right: control experiments, high K+ ACSF, GABAergic transmission blocked with picrotoxin (PTX), and GABAergic transmission enhanced by diazepam (BDZ)]. Gray circles show individual cell data, red boxes indicate medians, whiskers show interquartile ranges, and gray lines and asterisks indicate significance levels of P < 0.05 compared with control; outliers are brought into axes limits. (A) Average number of spikes per stimulus (averaged across all five stimulus types from our assay) generated by OT cells in different solutions. (B) Effect size for OT looming stimulus selectivity (difference between spike outputs in response to ‘crash’ and ‘flash’, normalised by response variability); dashed line marks effect reversal. (C) Index of individual selectivity of OT cells (see definition in 2). Dashed line indicates the value (2.4), which with our sample sizes corresponded to PANOVA of 0.05 (significance threshold for individual OT cell selectivity, without a false discovery rate adjustment).
Figure 7. Temporal coding in the retina does not contribute to the looming stimulus selectivity in the tectum. (A) Average periodograms of optic nerve responses to different visual stimuli. The 10–30 Hz frequency band was prominent in responses to ‘flash’, but was absent from spectra of responses to both ‘grid’ and ‘crash’. (B) Relative spectral power of the γ-band in responses to different stimuli; the γ-band is more prominent in responses to ‘flash’ than to all other stimuli (PW < 2e−5; N = 22). f, ‘flash’; r, ‘ramp’; g, ‘grid’; c, ‘crash’.
Figure 8. Conceptual model of a reverberation network acting as a collision detector (see 4 for details). In every panel the leftgrid represents a one-dimensional ‘retina’, and the rightgrid represents a one-dimensional ‘tectum’; rows correspond to discrete time steps (time running from top to bottom), and columns correspond to ‘RGCs’, or retinotopically arranged ‘tectal cells’, accordingly. Cells that spike during the current time frame are shown as dark and silent cells are pale. (A) Response to instantaneous ‘flash’. (B) Response to reversed looming, or receding stimulus. (C) Response to looming stimulus, or ‘crash’. Total cumulative spiking output in the ‘tectum’ is stronger for ‘crash’ (10 spikes) than for either ‘flash’ or receding stimulus (seven spikes each).
