In this study, we aimed to investigate the functional role of EAATs expressed on ON-bipolar cells in the retina. These transporters are responsible for clearing glutamate from the synaptic cleft while simultaneously inducing an anion conductance that hyperpolarizes the expressing cell in the presence of glutamate. As a result, they serve a dual function as both glutamate transporters and inhibitory glutamate receptors (REF). Our previous studies demonstrated that the loss of EAAT5b and EAAT7 reduces ON-bipolar cell responses, with eaat5b mutants being less affected than eaat5b/eaat7 double mutants, which also exhibit altered response kinetics (Niklaus et al., 2024).
In order to gain more insight into the modulatory role of these transporters and their effect on vision, we first assessed the electroretinographic responses to a variety of spectral illumination.
We found that a lack of EAAT5b and EAAT7 leads to modest reductions in the ON-response under UV-blue and red light. Interestingly, we did not observe any reduced responses under green illumination, suggesting a spectral-specific role in modulation by these proteins. However, due to the strong overlap of UV and blue cones’ light sensitivities, our whole retina recordings with a wide-field stimulus cannot separate these two channels (Yoshimatsu et al., 2021).
Recent studies have uncovered a complex anisotropy of the larval zebrafish eye, matched to the visual ecology that they live in Zimmermann et al., 2018. This adds a complexity of visual processing that cannot be assessed by the whole-field stimulation of our experiments.
In order to assess anisotropic distribution of EAAT5b and EAAT7 in the retina, we performed immunohistochemistry on larval dissected retinas. In line with previous data, we observed that EAAT5b and EAAT7 are largely co-expressed at most dendritic tips (Niklaus et al., 2024). More intriguingly, we found an unequal density of immunoreactive across the retina. EAAT7 is highest expressed in ventral regions of the retina with an intensity drop at the area temporalis or the strike zone (SZ). Conversely, EAAT5b’s signal is elevated in this specific area, where the highest concentration of UV cones is also found. This zone is aptly named for its importance in prey detection (Zimmermann et al., 2018; Yoshimatsu et al., 2020).
Previous work investigating the importance of UV cones for prey-capture circuits proposed that while UV-sensitive photoreceptors in the SZ are important for hunting, morphologically similar cones in the nasal patch might be used to monitor the periphery for predators (Yoshimatsu et al., 2020).
The peculiar localization of Eaat5b in the SZ, along with the observed defect in the UV/Blue range upon KO, suggests an involvement of Eaat5b in prey detection and capture. Hunting behavior has been extensively studied in the past years, both using live paramecia (Semmelhack et al., 2014; Yoshimatsu et al., 2020) and virtual assays (Bianco et al., 2011; Semmelhack et al., 2014), where generally a moving white dot is projected on a screen and displayed to a larva. In order to test UV-specific sensitivity in our mutants, we built a behavioral setup to deliver UV-bright artificial stimuli simulating swimming paramecia to head-mounted larvae. The setup delivers programmable stimuli at different light intensities and can be used to determine wavelength detection thresholds. We validated the setup by testing WT fish reactions to yellow and UV light, first confirming that responses to UV light were significantly more frequent than those to yellow light, and second, that the UV response rate was directly proportional to light intensity. Surprisingly, we did not observe a significant effect on prey-detection and capture behavior in the Eaat5b mutants, while Eaat7 mutants showed an increased response rate and a lower threshold than WT fish. These unexpected results may be accounted for by the different dynamics of Eaat5b and Eaat7. Namely, removal of Eaat5b leads to a shortening of the time to peak of the measured b-waves, while Eaat7 causes a lengthening of the same, indicating opposing roles in regulating the ON-BCs depolarization timing (Niklaus et al., 2024). UV-cones in the SZ have longer signal integration compared to the other photoreceptors, supposedly to better track a moving prey (Yoshimatsu et al., 2020). The concentration of Eaat5b in the BC terminals of this region might be another mechanism to slow down the signal and facilitate this behavior. Loss of Eaat5b might then result in a reduced prey-tracking ability; while conversely, the absence of a functional Eaat7 could lead to slower integration times throughout the whole retina, consequently making stimulus sensitivity higher. A possible way to test his hypothesis would be to modify our existing set-up to display the moving dot approaching from the periphery of the larval eye and observing if the animal’s ‘detection field’ in absence of Eaat7 has expanded outside of the SZ.
Since our monochromatic ERG revealed long-wavelengths sensitivity defects in both Eaat5b and Eaat7 mutants, we designed a behavioral setup to expose our mutant larvae to OMR-eliciting moving gratings, featuring black and red bands at varying contrasts. Our tests revealed that while Eaat5b mutants exhibited a reduced response to the stimulus, Eaat7 mutants showed no significant difference compared to WT. Since OMR is driven by both ON- and OFF-stimuli (Orger and Baier, 2005; Kist and Portugues, 2019) and zebrafish larvae typically respond equally to both contrast changes when their field of view is restricted (as in our assay), it is possible that any impairment in the ON-response was compensated by the larvae’s strong reaction to light-dark transitions.
Behavioral assays can be easily structured to deliver high-throughput results with intuitively measurable effects. However, these results are the outcome of the whole neural circuit, making it difficult to assess complex mechanisms without more targeted assays. Both Eaat5b and Eaat7 transcripts are expressed throughout the brain (Niklaus et al., 2024), and both localization and effects of their loss in the zebrafish brain are yet to be investigated. Combining monochromatic ERGs with full retinal immunohistochemistry and behavioral experiments finally helped us understand more about the functions and differences of these two postsynaptic transporters. Taken together, our data suggests that while mGluR6b remains the main regulator of ON-BCs depolarization throughout the retina, Eaat5b and Eaat7 modulate response speeds in different regions of the retina to optimize behavior-specific integration dynamics, particularly in relation to UV light. In the future, analysis of published single-cell transcriptomics databases could help identify whether eaat5b and eaat7 are co-localized in all ON-bipolar cells or if they are, at least in some cases, segregated in different cell populations.