In vision, it is ultimately the properties of photoreceptors that determine what visual features can be detected, and what features will be missed. Accordingly, it is critical to tune photoreceptors to the visual tasks they support. However, these tasks can differ dramatically across visual space, meaning that the properties of photoreceptors must vary depending on their position in the eye. For example, recent work showed that cones of the primate fovea are slower than their peripheral counterparts, likely in a trade-off to support our extreme spatial acuity vision (Sinha et al. 2017 Cell; Baudin et al. 2019 eLife). Similarly, mouse cones surveying the upper visual field preferentially respond to dark contrasts on a bright background, which may be linked to the appearance of predatory birds in the open sky (Baden et al. 2013, Neuron). However, in each of these cases, the extent and mechanistic basis of such within-type “photoreceptor tuning” remains far from understood, and the links to specific behavioural requirements remain largely anecdotal.
Here, we show that larval zebrafish use an extreme form of photoreceptor tuning to enable visual guided prey capture, one of the best-studied visual behaviours in vertebrates. The performance of this system is remarkable. We show that:
- Both prey detection and subsequent capture behaviours are driven by single UV-cones.
- For this, zebrafish use highly specialised UV-cones in their area temporalis, an “evolutionary forerunner” of a fovea.
- In the area temporalis, UV-sensitivity is boosted up to 42-fold.
- To better resolve the hyperpolarising response to UV-bright prey, these cones in addition feature elevated synaptic calcium levels. In showing this, we also provide the first ever in vivo recordings of individual photoreceptor-synapse light-responses in any vertebrate.
- Transcriptomic analysis reveals that this elevated calcium baseline stems from molecular tuning of phototransduction.
- Further photoreceptor optimisation occurs at the level of glutamate release from the synapse.
Our findings tally with hypothesised mechanisms of photoreceptor tuning in the primate fovea, and hint at a common set of molecular hotspots that can be targeted during convergent evolution. For example, primate foveal cones express increased levels of the “slow” rod version of transducin, the opsin’s G-protein that kicks off the phototransduction cascade. We show that zebrafish have tweaked the same target, yet in a different way: They rather express reduced levels of the “fast” cone transducin. Here, our dissection of the extreme photoreceptor specialisations across zebrafish UV-cones opens the way for a nuanced understanding of this fundamental property of vision. After all, of the three vertebrates that have ever been surveyed for region-specific functional tuning of photoreceptors (primate, mice, and now zebrafish), all three use regionally tuned photoreceptors, yet in each case for a different set of purposes.