Unexpected diversity and photoperiod dependence of the zebrafish melanopsin system

Abstract

Animals have evolved specialized photoreceptors in the retina and in extraocular tissues that allow them to measure light changes in their environment. In mammals, the retina is the only structure that detects light and relays this information to the brain. The classical photoreceptors, rods and cones, are responsible for vision through activation of rhodopsin and cone opsins. Melanopsin, another photopigment first discovered in Xenopus melanophores (Opn4x), is expressed in a small subset of retinal ganglion cells (RGCs) in the mammalian retina, where it mediates non-image forming functions such as circadian photoentrainment and sleep. While mammals have a single melanopsin gene (opn4), zebrafish show remarkable diversity with two opn4x-related and three opn4-related genes expressed in distinct patterns in multiple neuronal cell types of the developing retina, including bipolar interneurons. The intronless opn4.1 gene is transcribed in photoreceptors as well as in horizontal cells and produces functional photopigment. Four genes are also expressed in the zebrafish embryonic brain, but not in the photoreceptive pineal gland. We discovered that photoperiod length influences expression of two of the opn4-related genes in retinal layers involved in signaling light information to RGCs. Moreover, both genes are expressed in a robust diurnal rhythm but with different phases in relation to the light-dark cycle. The results suggest that melanopsin has an expanded role in modulating the retinal circuitry of fish.

Figure 1. Five zebrafish opn4-related genes arose by duplication and retrotransposition.

Schematic diagrams of the chromosomal regions surrounding mouse (Mus musculus), chicken (Gallus gallus), frog (Xenopus tropicalis) and zebrafish (Danio rerio) opn4-related loci. Gray boxes represent melanopsin exons and arrows indicate direction of transcription. The opn4a and opn4.1 genes were previously identified , . Orange and green boxes represent syntenic genes located upstream and downstream of the chicken opn4 locus. Blue and yellow boxes represent syntenic genes located upstream or downstream of the mouse Opn4 locus. Pink boxes represent conserved genes upstream of duplicated zebrafish opn4m loci. The single exon opn4.1 locus shows no synteny with other opn4 chromosomal regions and likely arose by retrotransposition.

Figure 2. Zebrafish proteins share features of mammalian melanopsin.

(A) Alignment of the core region of the predicted zebrafish melanopsin-related proteins with the Xenopus Opn4x and mouse Opn4 proteins. The core region includes seven transmembrane domains (TM1-7, underlined) and associated intracellular and extracellular loops. The DRY tripeptide motif of G-protein coupled receptors and the two signature motifs of rhabdomeric opsins are indicated by brackets (a, b and c, respectively). Glu (*) or Tyr (∧) is a possible counter ion for Schiff base linkage and Lys (+) is the site of chromophore binding. (B) Phylogenetic analysis separates the five zebrafish proteins into the Opn4m and Opn4x groups as indicated in a neighbor-joining tree rooted to Amphioxus Opn4 with five hundred bootstrap values. (C) Percentage of amino acid similarity between zebrafish Opn4-related and mouse Opn4 proteins (blue) or core regions (orange). (D) Normalized absorbance difference spectrum for zebrafish Opn4.1 with a maximum of 403 nm (dashed line). (E) Time course in seconds (s) of the response of HEK-293 cells (green) or transiently transfected with zebrafish Opn4.1 (blue) or mouse Opn4 (grey), as measured by fluorescent calcium imaging.

Figure 3. Diverse expression of zebrafish opn4-related genes in the developing retina.

Profile of opn4-related gene expression in the larval retina from 3 to 5 dpf. Following whole-mount RNA in situ hybridization, larvae were embedded in LR gold media and 4 µm sections prepared. In A, the lens and retinal cell layers are indicated (GCL, ganglion cell layer, INL, inner nuclear layer, PCL, photoreceptor cell layer). (A–C) opn4xa is expressed in a small subset of cells in the ganglion cell layer. (D–F) opn4xb is transcribed in bipolar cells in a broad domain of the INL. (G–I) opn4a is expressed in clusters of bipolar cells scattered throughout the INL. (J–L) opn4b transcripts are found in three domains within the INL, where amacrine (arrowhead in L), bipolar (arrow in K) and horizontal cells (open arrowhead in K) are located. (M–O) opn4.1 expression is weakly detected at 3 dpf but, one day later, strong expression is observed in horizontal cells in the outer shell of the INL. Sparse opn4.1 expression is also found in the photoreceptor cell layer.

Figure 4. Photoperiod length influences melanopsin expression.

(A, C) Larvae housed in 14∶10 LD or (B,D) 18∶6 LD cycles were fixed at ZT1 at 96 hpf, and assayed for opn4-related gene expression. opn4a and opn4.1 expression in the inner nuclear layer is greatly reduced in the prolonged light conditions. Additionally, opn4.1 transcripts are only detected in the photoreceptor cells (arrowhead) of larvae raised in the 14∶10 cycle. Over 90 larvae were assayed in 3 independent experiments.

Figure 5. Rhythmic expression of opn4a and opn4.1.

(A, B) Larvae maintained in 14∶10 LD or 18∶6 LD cycles were collected every 4 hours starting at ZT1. Expression of (A) opn4a and (B) opn4.1 show diurnal rhythms with different phases for each light cycle. The rhythm of opn4a expression is a single wide waveform in 14∶10 LD, and either a dual waveform or an ultradian rhythm at 18∶6 LD. Under the same LD conditions, expression of opn4.1 shows a reversal in the waveforms with respect to opn4a. (C) Summary graph representing the relative expression levels of opn4a and opn4.1 under the two photoperiods depicted as white (light phase) and black (dark phase) bars.