Chromatic processing in the zebrafish (Danio rerio) inner retina: bipolar cell physiology and open hardware designs for spectrally accurate stimulation under two-photon

Colour vision describes the ability of animals to differentiate objects based on their spectral reflectance properties independent of light intensity. It is an essential evolutionary trait that allows species to efficiently forage for food, avoid predation, break camouflage, communicate with conspecifics, or to find mates. Zebrafish is a powerful model for studying colour vision as it possesses four cone-photoreceptor types which can be categorised as Red-, Green-, Blue- and UV-sensitive. From first principles, its retina therefore holds the potential to process diverse chromatic computations. In the presented work, the focus was on retinal bipolar cells (BC). These are the retina’s first projection neurons. They receive inputs from the photoreceptors in the outer retina, and send their axon terminals to the inner retina, the inner plexiform layer (IPL). Diverse types within this class of interneuron shape light responses collected by the photoreceptor array into parallel channels with diverse spectral properties. BCs also make connections with all other neuron types within the retina, including horizontal cells in the outer retina, and amacrine as well as retinal ganglion cells in the inner retina. This makes them a central hub for spectral processing within the retina. By combining genetically encoded calcium indicator and two-photon microscopy, light-driven activity from larval zebrafish BC synaptic terminals was systematically recorded in vivo. Synaptic responses to tetrachromatic light stimulation unveiled an unprecedented degree of visual specialisation, including retinal regions dedicated to distinct light-guided behaviours. These regional characteristics were further correlated to functional BC types which were strongly associated with specific retinal positions and axonal stratification depths. Overall, BC projections to the inner plexiform layer displayed a sophisticated level of organisation, structured into chromatic and achromatic functional layers which systematically adjusted their response profiles across the eye to match natural spectral input statistics. Together, these findings bolster our understanding of “colour-processing” in this animal’s inner retina and suggest that unlike in mammals, teleost fish BCs already encode complex chromatic responses in the inner plexiform layer before driving retinal ganglion cells. Additionally, the study of colour vision from an organism requires precise control over the light stimuli’s temporal, spatial and spectral features. Therefore, chromatic stimulators, designed to be combined with two-photon microscopy, were developed throughout this work. These devices allowed circumventing experimental limitations, such as spectral crosstalk between the microscope and the stimulus light. Furthermore, they were conceived as open source projects to be easily replicated and adapted to any organism’s retina with different spectral sensitivities through the free control over the number and spectra of stimulation light sources. These open source projects originated from the desire to set up a stimulation standard for the field of visual neuroscience.