Diversification of non-visual photopigment parapinopsin in spectral sensitivity for diverse pineal functions
Recent genome projects of various animals have uncovered an unexpectedly large number of opsin genes, which encode protein moieties of photoreceptor molecules, in most animals. In visual systems, the biological meanings of this diversification are clear; multiple types of visual opsins with different spectral sensitivities are responsible for color vision. However, the significance of the diversification of non-visual opsins remains uncertain, in spite of the importance of understanding the molecular mechanism and evolution of varied non-visual photoreceptions.
Here, we investigated the diversification of the pineal photopigment parapinopsin, which serves as the UV-sensitive photopigment for the pineal wavelength discrimination in the lamprey, linking it with other pineal photoreception. Spectroscopic analyses of the recombinant pigments of the two teleost parapinopsins PP1 and PP2 revealed that PP1 is a UV-sensitive pigment, similar to lamprey parapinopsin, but PP2 is a blue-sensitive pigment, with an absorption maximum at 460–480 nm, showing the diversification of non-visual pigment with respect to spectral sensitivity. We also found that PP1 and PP2 exhibit mutually exclusive expressions in the pineal organs of three teleost species. By using transgenic zebrafish in which these parapinopsin-expressing cells are labeled, we found that PP1-expressing cells basically possess neuronal processes, which is consistent with their involvement in wavelength discrimination. Interestingly, however, PP2-expressing cells rarely possess neuronal processes, raising the possibility that PP2 could be involved in non-neural responses rather than neural responses. Furthermore, we found that PP2-expressing cells contain serotonin and aanat2, the key enzyme involved in melatonin synthesis from serotonin, whereas PP1-expressing cells do not contain either, suggesting that blue-sensitive PP2 is instead involved in light-regulation of melatonin secretion.
In this paper, we have clearly shown the different molecular properties of duplicated non-visual opsins by demonstrating the diversification of parapinopsin with respect to spectral sensitivity. Moreover, we have shown a plausible link between the diversification and its physiological impact by discovering a strong candidate for the underlying pigment in light-regulated melatonin secretion in zebrafish; the diversification could generate a new contribution of parapinopsin to pineal photoreception. Current findings could also provide an opportunity to understand the “color” preference of non-visual photoreception.
KeywordsAnimal photoreception Color vision Gene duplication Rhodopsin Spectral tuning UV-sensitive pigment
Recent genome-wide sequences of various animals have uncovered unexpected diversifications of opsin genes, which encode protein moieties of photoreceptor proteins, in most animals. The biological meaning of the diversification of visual opsins has been well studied; multiple visual opsins with different spectral sensitivities are responsible for color vision and other visual functions [1, 2, 3]. Most animals also possess multiple opsins that underlie non-visual functions such as circadian photoentrainment, suggesting their possible diverse and complicated involvement in varied types of non-visual photoreception [4, 5, 6, 7]. However, the significance of the diversification of non-visual opsins remains uncertain, although comparative studies of duplicated non-visual opsins such as melanopsin and vertebrate ancient-long (VAL)-opsin have revealed differences in their expression patterns [8, 9, 10, 11, 12]. The pineal and related organs of non-mammalian vertebrates are a good model for examining this issue because they are photosensitive organs with diverse functions. One such function is light-dependent regulation of melatonin secretion, which is involved in various physiological functions such as regulation of locomotor activity and sleep [13, 14]. In addition, two types of neural sensing, achromatic responses for luminescence detection and chromatic responses for wavelength discrimination, are well known .
Some opsins that serve as so-called pineal photopigments have previously been identified. Pinopsin was the first pineal opsin to be identified in vertebrates, and was originally isolated from chicken pineal organs [16, 17, 18, 19]. It has been suggested that pinopsin is involved in light-dependent regulation of melatonin secretion in chicken [16, 20, 21]. Since this discovery, parapinopsin has been identified in the teleost and lamprey pineal and parapineal organs [22, 23], and exo-rhodopsin in the teleost pineal organs . In lizard parietal eyes (a pineal-related organ), the expression of parietopsin as well as pinopsin or parapinopsin has been revealed [25, 26]. The genome-wide sequencing of zebrafish revealed the existence of two genomic DNAs encoding opsins that are highly similar to parapinopsins, indicating that zebrafish have another parapinopsin in addition to the ortholog of parapinopsin originally identified in catfish and rainbow trout [22, 23].
Recently, we used spectroscopic, immunohistochemical and electrophysiological analyses to demonstrate that parapinopsin is the ultraviolet (UV)-sensitive photopigment underlying pineal UV reception in the lamprey . The pineal organs of lampreys and teleosts, as well as the pineal-related organs of frogs (the frontal organ) and lizards (parietal eye), detect the ratio of UV to visible light in the environment [27, 28, 29, 30, 31, 32, 33], suggesting that these non-mammalian vertebrates can discriminate between different wavelengths of light using the pineal and related organs, independently of image-forming color vision in the eyes . In previous studies, the electrophysiological response most sensitive to UV light was only observed in chromatic responses for wavelength discrimination and not in achromatic responses for luminescence detection in lamprey and teleost pineal organs [28, 29, 30]. We isolated parapinopsin from rainbow trout and clawed frogs, in which the pineal and related organs have been reported to exhibit antagonistic chromatic responses to UV and visible light [29, 32]. In addition, parapinopsin expression in the pineal organ of a closely related species of rainbow trout was recently reported . These facts suggest that parapinopsin may be the common molecular basis for pineal UV reception in wavelength discrimination. This idea is supported by our recent finding that parapinopsin is expressed in the photoreceptor cells of the iguana parietal eye, where the ratio of UV to visible light is detected .
Here, we investigated and compared two types of parapinopsins in teleosts, PP1 and PP2, in order to clarify the functional meanings of the diversification. Our spectroscopic, immunohistochemical and transgenic analyses of teleost parapinopsins revealed that PP1 is a UV-sensitive pigment, similar to the lamprey parapinopsin involved in wavelength discrimination, while PP2 is a blue-sensitive pigment that could be involved in melatonin secretion rather than wavelength discrimination. Current findings indicate that the diversification of parapinopsin with respect to spectral sensitivity contributes to other pineal function besides wavelength discrimination, which could be a link between the diversification of non-visual pigments and diverse photoreception in animals.
Two parapinopsins in teleosts
Diversification of teleost parapinopsin with respect to spectral sensitivity
Mutually exclusive expression of the two parapinopsins in the teleost pineal organ
Different characteristics of PP2-expressing cells from PP1-expressing cells
The finding that PP2 is not a UV-sensitive pigment but a blue-sensitive pigment raised questions about the potential involvement of PP2 in pineal wavelength discrimination. In addition, the mutually exclusive expression of two parapinopsins, which is conserved in three teleost species, allows PP1-expressing and PP2-expressing cells to capture light independently. In order to characterize PP1-expressing and PP2-expressing cells and examine their involvement in pineal wavelength discrimination, we generated transgenic zebrafish in which either PP1-expressing cells or PP2-expressing cells were labeled, because antibodies to opsins essentially label only the outer segments where photopigments are localized, but not the cell bodies or neuronal processes of the photoreceptor cells. We isolated the upstream genomic DNA of zebrafish PP1 (~5.3 kb) and PP2 (~6.7 kb) and introduced green fluorescent protein (GFP) and red fluorescent protein (RFP) into zebrafish under the upstream sequence, respectively. We obtained transgenic zebrafish where GFP and RFP were specifically expressed in the pineal organ (Additional file 6: Figure S6A, B). The GFP and RFP exhibited mutually exclusive expression patterns in the transgenic zebrafish pineal organ (Additional file 6: Figure S6C), which was consistent with the result of the immunohistochemical analysis using antibodies specific for PP1 and PP2 (Fig. 4). The expression patterns of GFP and RFP closely matched those of PP1 and PP2 mRNA revealed by in situ hybridization in the transgenic zebrafish pineal organ, respectively (Additional file 7: Figure S7), indicating that these upstream sequences promote specific expression of reporter genes into PP1-expressing and PP2-expresing cells. In the transgenic zebrafish where PP1-expressing cells or PP2-expressing cells were labeled, we detected neuronal processes of PP1-expressing cells, which is consistent with their involvement in pineal wavelength discrimination (Additional file 8: Figure S8A). However, we rarely detected neuronal processes of PP2-expressing cells (Additional file 8: Figure S8B), suggesting the possibility that PP2-expressing cells are mainly involved in non-neural responses rather than neural responses like wavelength discrimination.
Possible involvement of PP2 in melatonin-secretion
Spectral tuning mechanism of teleost parapinopsins
In this study, we identified two types of parapinopsins, PP1 and PP2, in three kinds of teleosts, and characterized their functions. We found clear differences between PP1 and PP2; PP1 is a UV-sensitive pigment, like the lamprey and iguana parapinopsins [23, 26], but unexpectedly, PP2 is not UV-sensitive and is a novel blue-sensitive parapinopsin (Fig. 3), which shows the diversification of non-visual opsins with respect to spectral sensitivity. Our phylogenetic and synteny analyses suggested that PP1 and PP2 are derived from TSGD [37, 38]. Furthermore, we have clearly shown that the parapinopsin of spotted gar, which diverged before TSGD [37, 38], is a UV-sensitive pigment (Fig. 3c). In addition, based on the fact that the parapinopsins of other vertebrates are also UV-sensitive pigments, we can infer that the blue-sensitive parapinopsin evolved from a UV-sensitive parapinopsin. It is of interest to discuss how the functional differentiations between PP1 and PP2 occurred. In the phylogenetic tree, the branch length between the gene duplication and the deepest node in teleost PP2 (0.291 amino acid substitutions per site) is more than three-fold longer than that between the gene duplication and the deepest node in teleost PP1 (0.086), suggesting a higher accumulation rate of amino acid substitutions in the course of PP2 evolution. On the other hand, sequence comparison of transmembrane helices, which are involved in spectral tuning, between the UV-sensitive parapinopsin group, including the spotted gar parapinopsin, teleost PP1s and other vertebrate parapinopsins, and the teleost PP2 group did not reveal any positions occupied by group-specific amino acid residues with different physicochemical properties other than site 86 (Fig. 6e). Judging from the observation that no significant spectral changes occurred owing to substitutions of amino acid residue at site 86 in pufferfish and zebrafish parapinopsins, multiple amino acid residues, which might be species specific, could be involved in the spectral tuning of parapinopsins. Further mutational analyses including other teleost parapinopsins could reveal a novel spectral tuning mechanism between UV-sensitive and visible-light-sensitive pigments.
We also found that PP1 and PP2 were expressed in different pineal cells in three teleost species belonging to phylogenetically distant teleost orders (Fig. 4, Additional file 4: Figure S4, Additional file 5: Figure S5). These results strongly suggest that the differences in both the molecular properties and expression profiles between PP1 and PP2 are common in teleosts. Furthermore, our characterization of PP1-expressing and PP2-expressing cells in the zebrafish pineal organs revealed that PP1 can be considered a UV-sensitive pigment involved in pineal wavelength discrimination, but, conversely, PP2 is likely involved in light-regulated melatonin secretion (Fig. 5).
We emphasize that the teleost PP1 is a UV-sensitive photopigment (Fig. 3) and its pineal expression is conserved in three kinds of teleosts, including rainbow trout, which has the pineal wavelength discrimination ability  (Fig. 4, Additional file 4: Figure S4, Additional file 5: Figure S5). Accordingly, it is strongly suggested that PP1 underlies UV reception for pineal wavelength discrimination in teleosts. Taken together with the involvement of parapinopsins in UV reception in the lamprey pineal organ and iguana parietal eye [23, 26], current findings strengthen the idea that UV-sensitive parapinopsin is a fundamental molecule for wavelength discrimination involving UV reception in the pineal and related organs of non-mammalian vertebrates.
Light-regulated melatonin secretion is the primary function of pineal organs and has therefore been well investigated thus far. However, the underlying photopigments has been largely uncertain in teleosts. We have demonstrated that zebrafish PP2-expressing pineal cells contain serotonin and aanat2 (Fig. 5) and some of them also express LWS opsin, which has been shown to be involved in light-regulated melatonin secretion in the lamprey and chicken [21, 45] (Additional file 9: Figure S9). These results suggest that PP2 and LWS opsin are strong candidates for the light sensor for regulation of melatonin secretion in teleosts. In fact, the combination of PP2 and LWS opsin is consistent with a previous observation, showing that a broad visible light ranging from 400 to 600 nm influences melatonin secretion in zebrafish pineal organs . This is similar to the case of the chicken pineal organ, where the blue-sensitive pigment pinopsin and LWS opsin are suggested to be involved in melatonin secretion [16, 20, 21]. Interestingly, teleosts with PP2 do not possess pinopsin in their genomes whereas the spotted gar, which does not contain PP2, has pinopsin. These observations lead us to speculate that blue light might be suitable for the regulation of melatonin secretion, with the evolution of PP2 potentially being related to the loss of pinopsin in teleosts. Accordingly, our findings provide a clear-cut example for functional differentiation among paralogs; the teleost PP1 could have retained the putative ancestral function, whereas teleost PP2 acquired a novel functional role via a neofunctionalization event.
Recent genome projects of many animals have revealed that non-visual opsins have undergone a number of diversifications. However, the functional meaning of these diversifications remain unclear, in contrast to those of visual opsins, which diversified in spectral sensitivity to enable color vision and other visual function [1, 2, 3]. In this paper, we have clearly shown the differences in the molecular properties of duplicated non-visual opsins, by demonstrating the diversification of parapinopsin with respect to spectral sensitivity, which generated a blue-sensitive parapinopsin from the original UV-sensitive parapinopsin. Moreover, we also showed a plausible link between this diversification and its physiological consequences; this diversification could allow parapinopsin to contribute to the light regulation of melatonin secretion, in addition to wavelength discrimination. Current findings that two parapinopsins with different spectral sensitivities could contribute to different pineal functions (wavelength discrimination and melatonin secretion) also provide an opportunity to improve our understanding of the “color” preferences of non-visual photoreception.
This experiment was approved by the Osaka City University animal experiment committee (#S0032, #514) and complied with the Regulations on Animal Experiments from Osaka City University.
Zebrafish (Danio rerio) were obtained from the Zebrafish International Resource Center (ZIRC) and National Bio Resource Project (NBRP) Zebrafish. Pufferfish (Takifugu rubripes), rainbow trout (Oncorhynchus mykiss), and spotted gar (Lepisosteus oculatus) were commercially obtained.
Partial cDNAs of parapinopsins of the pufferfish, zebrafish, and spotted gar were isolated from the RNA of pineal organs by RT-PCR using gene-specific primers based on gene sequences found in genome databases. A partial cDNA of the rainbow trout PP2 was isolated from the RNA of pineal organs by RT-PCR using the following degenerate primers: 5′-TGYACiGTiGCiYTNATHGCNGT-3′ as the sense primer and 5′-GAATTCAIIGCRTADATIAINGGRTT-3′ as the antisense primer. These primers were based on CTVALIAV and NPIIYAL, respectively. PCR amplifications using degenerate primers were carried out at an annealing temperature of 40 °C as described . Full-length cDNAs of the parapinopsins were obtained using the 3′ RACE and 5′ RACE systems (Invitrogen, Life Technologies, Carlsbad, California, USA).
Amino acid sequences of vertebrate visual and non-visual pigments, including parapinopsins, were aligned with MAFFT version 7.221 employing the L-INS-i method . Unambiguously aligned sites were selected with trimAl version 1.4 using the automated1 option  followed by trimming gapped sites. A molecular phylogenetic tree was inferred with PhyloBayes 4.1b, which employs the Bayesian framework, assuming the CAT-GTR model . The PhyloBayes run performed 10,900 cycles of MCMC chain from which the first 1,000 cycles were discarded as burn-in. A phylogenetic tree based on the maximum-likelihood approach was inferred with RAxML version 8.1.22 assuming the PROTCATGTR model . One thousand bootstrap replicates were performed for maximum-likelihood tree inference (“-f a” option for RAxML). The accession numbers of the sequences used for the phylogenetic tree inference are provided in Additional file 11.
Syntenic regions near parapinopsin genes in bony fish genomes were identified by the Genomicus database version 80.01 . Three gene families, CACNA2D3/cacna2d3, TKT/tkta, and ERC2/erc2, in which the paralogs generated in teleost-specific genome duplication were both conserved in at least one teleost species, were subject to phylogenetic analysis. Amino acid sequences of these genes were obtained with the aid of a web tool, aLeaves . Bayesian phylogenetic trees of these genes were inferred with the procedures described above. Ambiguous sites of the ERC2/erc2 amino acid sequence alignment were removed by visual inspection as well as with trimAl. The PhyloBayes runs were performed for 13,300, 19,800, and 4,200 cycles of MCMC chain for CACNA2D3/cacna2d3, TKT/tkta, and ERC2/erc2 genes, respectively, from which the first 1,000 cycles were discarded as burn-in.
Expression of the opsin-based pigment and spectroscopy
The cDNAs of the parapinopsins and their various mutants were tagged with the monoclonal antibody Rho 1D4 epitope sequence (ETSQVAPA). The tagged cDNA was inserted into a pcDNA3.1 plasmid vector (Invitrogen). Chimeric mutants with respect to the transmembrane helix between PP1 and PP2 were generated by combining two cDNA fragments using PCR with primers at the ends of the combined sequence. R55/Q56, M89/G90, N130/P131, D168/L169, A226/E227, and Y270/I271 in the pufferfish PP1, and K56/Q57, Q90/G91, K131/P132, E169/L170, M227/E228, and K271/I272 in the pufferfish PP2 are boundaries of helices I/II, II/III, III/IV, IV/V, V/VI, and VI/VII, respectively. A point mutation was introduced into the cDNA with a commercial site-directed mutagenesis kit, QuikChange (Stratagene, Agilent Technologies, Santa Clara, California, USA), according to the manufacturer’s instructions. Pigment expression in HEK293S cells and purification were carried out as previously described . Briefly, to reconstitute the pigment, the expressed proteins were incubated with excess 11-cis retinal overnight. The pigments were then extracted with 1 % dodecyl β-D-maltoside (DM) in 50 mM HEPES buffer (pH 6.5) containing 140 mM NaCl (buffer A). To purify the pigment, pigments in the crude extract were bound to 1D4-agarose, washed with 0.02 % DM in buffer A (buffer B) and eluted with buffer B that contained the 1D4 peptide. The absorption spectra of pigments were recorded at 4 °C with a Shimadzu UV2450 spectrophotometer. UV, blue, and orange light were supplied by a 1-kW halogen lamp (Philips, Eindhoven, Netherlands) with a UV glass filter, UTVAF-50S-36U (Sigma Koki, Saitama, Japan), a 460-nm interference filter, and an O56 glass cutoff filter (Toshiba, Tokyo, Japan), respectively.
mRNA from pineal organ, brain, and eye of zebrafish was purified with an Oligotex -dT30 Super mRNA Purification Kit (TaKaRa, Otsu, Japan), and the mRNAs were reverse transcribed to cDNAs with a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, Massachusetts, USA). The β-actin genes were used as internal references. The primer sequences used for amplification of each gene and the length of amplicons are as follows. 5′-TGCGGCAGGTGAGTCGTCTG-3′ and 5′-GATCCCTGAACTGTCTGTTC-3′ for amplification of a 270-bp fragment of PP1, 5′-GCTGAGACAAGTTGCTAAGG-3′ and 5′-ACCTCTGGAACTGTTTGTTC-3′ for amplification of a 225-bp fragment of PP2, 5′-TCGATTGCAGGTCTTGTGAC-3′ and 5′- TGGGTGGACTCTGACTCGGC-3′ for amplification of a 380-bp fragment of SWS1 opsin, and 5′-TGGAGAAGAGCTATGAGCTG-3′ and 5′-ACTCATCGTACTCCTGCTTG-3′ for amplification of a 386-bp fragment of β-actin.
Antibodies and immunohistochemistry
The anti-zebrafish PP1, anti-zebrafish PP2, anti-pufferfish PP1, and anti-pufferfish PP2 antibodies were generated against the C-terminal 50-amino-acid region of the protein. The anti-rainbow trout PP1 and anti-rainbow trout PP2 antibodies were generated against the C-terminal 40 amino acids and 44 amino acids, respectively. The anti-zebrafish LWS opsin was generated against the C-terminal 41-amino-acid region, which is conserved in two zebrafish LWS opsins, LWS1 and LWS2 . The antigens were prepared using the pMAL Protein Fusion and Purification System (New England Biolabs, Ipswich, Massachusetts, USA), according to a previously reported method . The dissected pineal organs with brain were immersion-fixed in 4 % paraformaldehyde, cryoprotected in 0.1 M phosphate buffer containing 30 % sucrose, frozen with OCT Compound (Sakura Finetechnical, Tokyo, Japan), and sectioned at 12–20 μm. The sections were incubated with 1:500 diluted antisera, followed by incubation with Alexa Fluor 488 or 594 anti-mouse IgG or Alexa Fluor 488 or 594 anti-rabbit IgG (Molecular Probes, Life technologies, Carlsbad, California, USA) for immunofluorescent detection. For immunohistochemical labeling of serotonin, the anti-serotonin antibody (ImmunoStar, Hudson, Wisconsin, USA) was used. The anti-GFP antibody (Nacalai Tesque, Kyoto, Japan) and anti-DsRed antibody (Clontech, Mountain View, California, USA) were used to visualize the GFP and RFP, respectively. Confocal images were obtained with confocal laser scanning microscopes C1 siReady (Nikon, Tokyo, Japan) or Leica TCS SP8 (Leica, Wetzlar, Germany).
Generating and analyzing transgenic zebrafish
The 5,293-bp and 6,719-bp upstream sequence from the ATG initiation codon of zebrafish PP1 and PP2 genes, respectively, were obtained from zebrafish genomic DNA by PCR. The promoter sequence of aanat2 was obtained according to a previous report . Each DNA fragment was inserted into the Tol2 vector pT2AL200R150G, which contains the GFP expression cassette . To introduce RFP, the GFP cDNA in the vector was replaced by mRFP cDNA obtained from pT2ZUASRFP . Microinjection of the construct DNA and transposase mRNA to zebrafish zygotes was carried out as previously reported . We also generated the transgenic line where Gal4 is introduced under the PP1-promoter and the F1 fish from the Gal4 line and UAS-GFP line  was used to examine the presence of serotonin and aanat2 in the PP1-expressing cells. For labeling LWS1-expressing and LWS2-expressing cells, Tg(LWS1up2.6 kb:GFP)#1509 and Tg(LAR:LWS2up1.8 kb:GFP)#1499 were used .
In situ hybridization
Digoxigenin (DIG)-labeled antisense and sense RNA probes for zebrafish PP1 and PP2 mRNAs were synthesized using the DIG RNA labeling kit (Roche, Basel, Switzerland). Sections were pretreated with proteinase K and hybridized with each RNA probe in ULTRAhyb Ultrasensitive Hybridisation Buffer (Ambion, Life Technologies, Carlsbad, California, USA). The probe on the sections was detected by incubation with alkaline phosphatase-conjugated anti-DIG antibody (Roche) followed by a blue 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium color reaction.
The sequences reported in this paper have been deposited in the DDBJ database [accession nos. AB626964 - AB626967, AB675727].
We thank Robert S Molday (University of British Columbia) for his kind supply of rho 1D4-producing hybridoma. We thank Hiromi Hirata (Aoyama Gakuin University) for his helpful advice on zebrafish experiments. We also thank Shoji Kawamura (The University of Tokyo) for providing transgenic line Tg(LWS1up2.6 kb:GFP)#1509 and Tg(LAR:LWS2up1.8 kb:GFP)#1499 and the National BioResource Project, Zebrafish, for zebrafish supply. This work was supported in part by grants-in-aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science and Technology (nos 19207006, 23247011 to AT and nos 20770057, 22687005 to MK), the Yamada Science Foundation (to AT), and the Naito Foundation (to AT and MK). SW and EK-Y are supported by Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists.
- 28.Morita Y, Dodt E. Slow Photic Responses of the Isolated Pineal Organ of Lamprey. Nova Acta Leopoldina. 1973;38:331–9.Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.