Ca2+-binding proteins play important roles in neuronal function by transducing Ca2+ signals and thereby regulating crucial processes like synaptic signaling and neuronal development, growth and survival. The Ca2+-binding protein (CaBP) subfamily is part of the vast EF-hand containing calmodulin superfamily. Eight genes encoding CaBPs have been identified in zebrafish, and many of them are expressed in specific subpopulations of retinal neurons during development. Among them, cabp2a and cabp5b have been shown to be expressed in the retinal inner nuclear layer (INL). Here, we demonstrate that their paralogues, cabp2b and cabp5a, are also specifically expressed in the INL of the developing retina. By extending expression analysis of cabp2a, cabp2b, cabp5a and cabp5b to the adult retina, we reveal exclusive expression of all four genes in the INL after retinal development is completed. Thus, our findings suggest functions of cabp2a, cabp2b, cabp5a and cabp5b in Ca2+ signaling in mature retinal neurons, besides a role in the developing retina.

Ca2+ signaling is required for many fundamental aspects of neuronal function, such as signaling at pre- and postsynaptic sites, gene activation, growth, development and survival^[1]. Ca2+-binding proteins sense and transduce Ca2+ signals by undergoing conformational changes upon Ca2+ binding, and consequently regulating target proteins. The largest group of Ca2+-binding proteins are structurally conserved EF-hand containing proteins belonging to the calmodulin superfamily. Within this superfamily, the vertebrate-specific CaBP subfamily comprises CaBP1, CaBP2, CaBP4 and CaBP5 in mammals^[2][3][4]. In zebrafish, the CaBP subfamily is expanded to 8 members encoded by cabp1a, cabp1b, cabp2a, cabp2b, cabp4a, cabp4b, cabp5a and cabp5b^[5], most likely due to duplicate gene retention after teleost-specific whole genome duplication^[6].

CaBPs are specifically expressed in parts of the nervous system, including the retina^[2][5][7]. The role of CaBP4 in retinal function is currently best understood. CaBP4, being within the retina specifically expressed in photoreceptors, has an established role in retinal disease and photoreceptor synaptic function^{[7][8][9][10][11][12][13][14]}. Functions of CaBPs expressed in second-order retinal neurons (i.e. bipolar and horizontal cells) are less well understood^[15][16]. Bipolar cells of the mouse retina have been reported to express CaBP5^[2], and cabp2 transcripts have also been detected in this cell type^[17]. Characterization of the CaBP family in zebrafish has revealed two orthologues for each mammalian CaBP-encoding gene, and expression analysis was performed in the developing embryo^[18]. Similar to mammals zebrafish cabp4b is expressed in photoreceptors, while no specific expression pattern was detected for cabp4a. cabp1a and cabp1b are both expressed in amacrine cells. Like the mammalian orthologues, cabp2a and cabp5b are expressed in retinal bipolar cells during development, while cabp2b expression was reported exclusively in hair cells. cabp5a expression has not been detected at these stages^[5].

Here, we evaluate expression of cabp5a in the developing zebrafish retina and extend expression analysis of cabp2a, cabp2b, cabp5a and cabp5b to the adult retina.

In order to study expression of cabp2a, cabp2b, cabp5a and cabp5b in the retina, we performed RNA in situ hybridization on developing whole-mount zebrafish and on adult retinal sections. In 3 day old zebrafish larvae, all four genes studied show clear expression in the retinal inner nuclear layer (INL) (Fig. 1A-D). While staining for cabp2a, cabp5a and cabp5b results in strong labeling of the INL (Fig. 1A,B1,C1,D1), cabp2b staining is only weak in the retina (Fig. 1B1), also when compared to strong cabp2b staining observed in the inner ear (Fig. 1B2). Expression in the inner ear can also be detected for cabp5a and cabp5b (Fig. 1C2,D2). In the adult retina, all four cabps examined show exclusive and strong staining in the INL (Fig. 1G-H). Based on the shape and location of labeled cell bodies, all four genes are expressed in bipolar cells (central and distal INL), and possibly in horizontal cells (distal INL). The proximal INL, harboring amacrine and displaced ganglion cells, is devoid of staining.

Expression patterns of cabp2a and cabp5b in developing retina and ear are consistent with the findings by Di Donato et al.^[5]. While expression of cabp2b has previously been reported only in sensory hair cells, we detected weak expression in the retinal INL, likely reflecting low abundance of cabp2b transcripts in this region. Here, we show for the first time expression of cabp5a expression, which is strongly expressed in the retinal INL and in the inner ear. Our finding that expression of cabp2a, cabp2b, cabp5a and cabp5b is maintained up to adulthood strongly suggests functions in the mature retinal INL.

So far, no reports on zebrafish CaBP function exist. Studies of CaBP5 in mammalian cell culture and mouse retina suggest an important modulatory role in retinal signaling^[16], possibly by regulation of voltage-gated Ca2+ channels^[17] and a function in synaptic vesicle recruitment^[18]. Since expression of cabp5 orthologues is conserved between mammals and zebrafish, it is well conceivable that cabp5a and cabp5b have similar tasks.

The diversity of Ca2+ binding proteins expressed in the nervous system allows finely-tuned Ca2+ signaling through their different Ca2+ affinities, subcellular and cellular localizations and different binding partners, making them non-redundant regulators of neuronal signaling^[1]. This and other studies^[2][5][15] suggest that bipolar cells exploit a repertoire of CaBPs for decoding Ca2+ signals, laying out the framework for further functional studies.

We demonstrate expression of four genes encoding CaBP subfamily members in the retinal INL of zebrafish: cabp2a, cabp2b, cabp5a and cabp5b. Since all four genes are expressed in the INL of both developing and adult zebrafish, our findings suggest functions of cabp2a, cabp2b, cabp5a and cabp5b in Ca2+ signaling in mature retinal neurons, besides possible developmental roles.

We show expression that strongly implies, but does not prove, function. The position of the cell bodies in the retina strongly suggests expression in bipolar and horizontal cells, but is not a formal proof.

The study implies a functional of these CaBPs in retinal processing. The obvious next step is to genetically manipulate these genes (e.g. by CRISPR/Cas9 genome editing) and assess altered retinal function in these fish. Due to overlapping expression patterns of paralogues in the INL, it might be necessary to generate double knockouts to observe a phenotype.

Fish husbandry

Zebrafish were kept at 26°C under a 14/10 h light/dark cycle as previously described^[19]. Embryos were raised at 28°C in E3 medium and staged according to development in days post fertilization^[20]. All fish used for this study were from the Tü wildtype strain.

Generation of DIG labeled RNA probes

Fragments of cabp2a, cabp2b, cabp5a and cabp5b coding sequence were amplified from cDNA of 5dpf zebrafish by PCR and cloned into the pCRII-TOPO vector. For primer sequences, see supplementary table 1. Identity of the cloned fragments was confirmed and orientation was checked by restriction digestion and sequencing. Plasmids were linearized using appropriate restriction enzymes, and in vitro transcription of antisense and sense RNA probes was performed using the Roche DIG-RNA Labeling Kit (Roche).

RNA in situ hybridization on whole-mount larvae and adult retinal sections

For whole mount in situ hybridization, embryos were treated with 3 µM PTU (1-phenyl-2-thiourea, Sigma-Aldrich) to suppress pigmentation. 3 day old larvae were fixed in 4% paraformaldehyde overnight at 4°C. Whole-mount in situ hybridization was carried out as described^[21], with modifications^[22].

For in situ hybridization on adult retinal sections, zebrafish were euthanized using tricaine (MS-222, Sigma-Aldrich). Eyes were excised and fixed in 4% paraformaldehyde overnight at 4°C. Tissue was washed twice in PBS for 5 min at room temperature, incubated in 30% sucrose overnight at 4°C, embedded in in Richard-Allan Scientific Neg-50 Frozen Section Medium (Thermo Fisher Scientific) and cryosectioned at 16 μm thickness. In situ hybridization on sections was performed as described^[22]. Sections were coverslipped with Kaiser’s glycerol gelatin (Merck). Whole-mount and section in situ hybridization experiments were carried out using antisense probes. Control experiments using sense probes resulted in no signal.

Image acquisition and processing

Images were acquired on a light microscope (Olympus Bx61) equipped with a Color Camera (ColorViewIII, Soft Imaging System, Olympus). Whole-mount larvae were imaged in the bright field mode, and adult retinal sections in the DIC mode. Images were adjusted for brightness and contrast and figure was composed using Adobe Photoshop.

This work was supported by the Swiss Science Foundation (31003A_153289/1).

We thank Ramon Maegert for assistance during an internship, and Kara Dannenhauer and Martin Walther for excellent animal care.

The authors declare no conflicts of interest.

All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the local authorities (Veterinäramt Zürich TV4206).

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.