class="head no_bottom_margin" id="sec1title">IntroductionVisual aftereffects are often considered the by-products of neuronal adaptation processes for the optimization of sensory perception. Typical examples are calibration between movement perception and self-produced locomotion, decorrelation to increase efficiency of sensory coding, and gain control of sensory stimuli to extend the dynamic range of detection (). Therefore, they are useful tools to study the neuronal mechanisms underlying visual perception.A particular example of visual aftereffects is the motion aftereffect (MAE), in which exposure to continuous, coherent, moving visual stimuli induces, following the cessation of the moving stimulus, the illusory perception of motion in the opposite direction. MAE was first described in ∼330 BC by Aristotle in his book Parva Naturalia (trans. ). Since then, many studies have described different psychophysical aspects of the phenomenon (, , , ). In addition to perceptual MAE, continuous, coherent, moving visual stimuli can induce oculomotor MAE (, , ). Despite the vast literature on MAE, only a handful of studies have examined the underlying neuronal mechanisms. MAE was found to be associated with either a decrease or an increase in the response of direction-selective neurons. Direction-selective neurons are specialized for detecting motion along specific axes of the visual field, and they respond to visual stimulus moving in a given direction (the preferred direction) but do not respond or respond less to those moving in the opposite direction (the null direction). Using single-neuron recordings, MAE-associated adaptations have been described in different brain regions of different animal species: the rabbit’s retina (), the owl monkey’s medial temporal lobe (), the cat’s primary visual cortex (), the pigeon’s nucleus lentiformis mesencephali (), and the fly’s lobula plate ().Despite these advances, we lack a comprehensive explanation of the underlying mechanisms and the neuronal correlates of MAE at the circuit level. To that end, and to shed light on the potential mechanisms underlying visual motion perception, we used transgenic zebrafish larvae expressing the genetically encoded calcium indicator GCaMP3. We monitored the dynamics of large neuronal circuits from different brain regions using two-photon microscopy in an intact, non-anesthetized, behaving vertebrate model.In zebrafish, the retinal ganglion cells (RGCs) project to at least ten arborization fields, with the optic tectum (OT) being the largest (, href="#bib29" rid="bib29" class=" bibr popnode">Nevin et al., 2010). The optic tectum is the zebrafish’s most complex layered brain structure, and it is essential for visually guided prey detection and capture (href="#bib15" rid="bib15" class=" bibr popnode">Gahtan et al., 2005, href="#bib47" rid="bib47" class=" bibr popnode">Romano et al., 2015). Direction-selective neurons are found in both the retina (href="#bib30" rid="bib30" class=" bibr popnode">Nikolaou et al., 2012) and the optic tectum (href="#bib14" rid="bib14" class=" bibr popnode">Gabriel et al., 2012, href="#bib16" rid="bib16" class=" bibr popnode">Gebhardt et al., 2013, href="#bib18" rid="bib18" class=" bibr popnode">Grama and Engert, 2012, href="#bib20" rid="bib20" class=" bibr popnode">Hunter et al., 2013, href="#bib47" rid="bib47" class=" bibr popnode">Romano et al., 2015).Using two-photon calcium imaging, it has been shown that the pretectum and the superficial layers of the optic tectum respond to large-field coherent visual motion presented to the contralateral eye (href="#bib42" rid="bib42" class=" bibr popnode">Portugues et al., 2014). Similarly, unilateral stimulation of the pretectal area induced eye movements resembling the optokinetic response (OKR; href="#bib22" rid="bib22" class=" bibr popnode">Kubo et al., 2014).Here, we show that following the presentation of a coherently moving visual pattern (conditioning stimulus, CS) capable of inducing OKR, zebrafish larvae generated, in the absence of sensory stimuli, optokinetic movements in the direction opposite that induced by the CS. Reminiscent of MAE, these results suggest that following the CS, the larvae experienced perception of visual motion in the opposite direction. Using optogenetics to transiently block eye movements during the presentation of the CS, we show that neither muscular fatigue nor eye proprioception feedback plays a role in the generation of optokinetic MAE-like behavior. Moreover, two-photon laser ablation of the optic tectum significantly reduced MAE-like behavior. Using two-photon calcium imaging of transgenic zebrafish larva expressing GCaMP3, we monitored the neuronal activities of the larva’s two main visual centers (retina and optic tectum). We found that following stimulus cessation, direction-selective neurons tuned to the direction of the CS displayed strong habituation in the optic tectum but not in the retina. Furthermore, we observed sustained rhythmic neuronal activity associated with the optokinetic MAE-like behavior among a specific group of direction-selective tectal neurons, thus arguing for a neuronal correlate of the MAE-like behavior. Finally, an empirical mathematical model based on the competition between direction-selective tectal neurons related to their activity could reproduce the OKR, the optokinetic MAE-like behavior, and the unconditioned spontaneous eye movements observed in the absence of moving visual stimulation. Overall, our results propose a functional neuronal circuit in the zebrafish optic tectum that is capable of generating perception of visual motion.
展开▼