SEARCH

SEARCH BY CITATION

Keywords:

  • opsin 5;
  • chicken;
  • development;
  • retina;
  • amacrine cells;
  • ganglion cells

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The opsin gene family encodes G protein–coupled seven-transmembrane proteins that bind to a retinaldehyde chromophore for photoreception. It has been reported that opsin 5 is expressed in mammalian neural tissue, but its function has been elusive. As a first step to understand the function for opsin 5 in the developing eye, we searched for chicken opsin 5-related genes in the genome by a bioinformatic approach and isolated opsin 5 cDNA fragments from the embryonic retina by RT-PCR. We found that there are three opsin 5–related genes, designated cOpn5m (chicken opsin 5, mammalian type), cOpn5L1 (chicken opsin 5-like 1), and cOpn5L2 (chicken opsin 5-like 2), in the chicken genome. Quantitative PCR analysis has revealed that cOpn5m is the most abundant in the developing and early posthatching neural retina. In situ hybridization analysis has shown that cOpn5m is specifically expressed in subsets of differentiating ganglion cells and amacrine cells. These results suggest that the mammalian type opsin 5 may contribute to the development of these retinal cells in the chicken. Developmental Dynamics 237:1910–1922, 2008. © 2008 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Opsins are a family of membrane-bound, heptahelical G protein–coupled receptors (GPCRs) characterized by their ability to covalently bind retinaldehyde chromophore via a Schiff base linkage (for a review, see Terakita,2005). There are seven major opsin subfamilies in vertebrates: melanopsin, encephalopsin, ciliary photoreceptor opsins, Go-coupled opsins, opsin 5, peropsin, and photoisomerase subfamilies. Within the past decades, non-canonical opsins other than rod- and cone-opsins, have been identified and shown to be expressed in the eye, mediating non-visual effects of light on physiology. Furthermore, it has been shown that some of the non-canonical opsin genes are expressed even in tissues classically considered non-photosensitive, and their expression is observed in the developing eye of mammals and birds at early stages. For example, encephalopsin (opsin 3), melanopsin (opsin 4), and peropsin are expressed in the mouse eye tissue at embryonic day 10.5 when the earliest cell type in the retina, retinal ganglion cells, begins to differentiate as revealed by RT-PCR analysis (Tarttelin et al.,2003a). In the developing chicken retina, on the other hand, two melanopsin genes are differentially expressed by distinct populations of developing retinal cells in the horizontal, ganglion, and amacrine cell layers (Tomonari et al.,2005,2007). This suggests that the expression of these non-canonical opsin genes might be associated with genesis and early differentiation of certain retinal cell types.

Opsin 5 or Gpr136 was identified by a bioinformatic approach and reported to be expressed in the human and mouse eye, brain, and testis (Tarttelin et al.,2003b; Vanti et al.,2003; Fredriksson et al.,2003). Opsin 5 is most related to peropsin among the Opsin family as deduced from the phylogenetic relationship and their shared intron structure (Tarttelin et al.,2003b). Although the function for Opsin 5 has not been elucidated so far, given that its putative counterion, which balances the Schiff base, is a tyrosine residue as is found for peropsin, along with its phylogenetic position, it has been suggested that Opsin 5 may be functioning as a photoisomerase (Tarttelin et al.,2003b). In addition, it is intriguing to know whether opsin 5 is expressed during retinal development as is the case for other non-canonical opsins, but it has not been studied in any species.

As a first step to understand the function for opsin 5 in the developing eye, here we studied the expression pattern of the chicken opsin 5 genes. We found that there are three opsin 5–related genes in the chicken genome, designated cOpn5m (chicken opsin 5, mammalian type), cOpn5L1 (chicken opsin 5-like 1), and cOpn5L2 (chicken opsin 5-like 2). We examined their expression patterns during chicken retinal development, finding that cOpn5m is predominantly expressed in the developing retina, specifically expressed by small subsets of differentiating ganglion and amacrine cells.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Identification of Three Opsin 5–Related Gene Loci in Chicken

To identify chicken opsin 5–related genes, the GenBank database was screened with the amino acid sequence for mouse Opsin 5 using BLAST algorithm (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi). Three chicken nucleotide sequences encoding a protein similar to Opsin 5 were found as follows: NCBI gene name LOC421093 for nucleotide accession number XM_419178, LOC428670 or opsin 5 for XM_426228, and LOC422050 for XM_420056. Partial cDNA fragments for these three opsin 5–related genes were obtained by RT-PCR from embryonic day 16 (E16) retina using primers designed according to these nucleotide sequences (see Supplementary Table 1, which can be viewed at www.interscience.wiley.com/jpages/1058-8388/suppmat). Thus, it seemed that these opsin 5–related genes were certainly transcribed in the developing chicken retina.

Phylogenetic analysis of the deduced amino acids, obtained by search in NCBI (http://www.ncbi.nlm.nih.gov/sites/entrez) and Ensembl (http://www.ensembl.org/index.html) databases, has shown that LOC428670 is most related to mammalian Opsin 5 (Opn5), while LOC421093 is related to a Xenopus Opn5 and a teleost Opn5, and LOC422050 is close to other teleost Opn5-related proteins (Fig. 1A and Table 1). Alignment of the chicken Opsin 5–related proteins with representatives of those from other species reveals that several key features are conserved (Fig. 1B). These include seven putative transmembrane α-helices and a lysine residue at position 296 (K296 in human Opsin 5) required to form the Schiff base with the chromophore, which is considered diagnostic for the opsin family. The positively charged Schiff base is balanced by a counterion in the third transmembrane domain, a tyrosine residue at position 109 in the case of mammalian Opsin 5 (Tarttelin et al.,2003b), which is all conserved in the Opsin 5–related proteins aligned in Figure 1B. Two cysteine residues for disulphide bond formation (C106 and C183 in human Opsin 5) are also conserved. A site at the C-terminus for palmitoylation (C315, C316 in human Opsin 5) is retained in the chicken mammalian-type Opsin 5 encoded by LOC428670 (Fig S1B). In contrast, it seems that the cysteine residues are relatively shifted to the C-terminal tail in the presumptive LOC421093 and LOC422050 proteins. It is noticeable that a highly conserved tripeptide (D/E)R(Y/W) motif found in GPCRs at the end of transmembrane domain 3 (DRY 123-125 in LOC428670), which is important for G-protein binding (Franke et al.,1992), is not present in the presumptive LOC421093 protein as is the case for OlOpn5L1, TnOpn5L1, DrOpn5L1, and XtOpn5L1 (Fig. 1B). The presumptive LOC421093 protein has cysteine residues instead, suggesting that it could not activate G-proteins and might function as a dominant negative form, which should be verified by further studies.

thumbnail image

Figure 1. Phylogenetic tree of vertebrate Opsin 5 proteins (A), and their amino acid alignment (B). A: The Opsin 5–related proteins are depicted by the NJ (neighbor-joining) method. The chicken candidate Opsin 5–related proteins are shown in yellow. Branchiostoma belcheri Amphiop3, which belongs to the peropsin group (Koyanagi et al.,2002), was used as an outgroup. The scale bar is calibrated in substitutions per site. Numbers show bootstrap confidence values. Amino acid sequences used in tree construction are deduced from the nucleotide sequences listed in Table 1. B: The predicted amino acid sequences of the chicken candidate opsin 5–related genes (LOC428670, LOC421093, and LOC422050) are aligned with other Opsin 5–related proteins. Portions of amino acid residues corresponding to seven transmembrane domains (TM) are lined according to Tarttelin et al (2003b). Key features of Opsin 5 mentioned in the text are shown by red asterisks. The color codes for amino acid residues and rules to indicate conserved residues (in black asterisks and dots) are according to the instruction of ClustalX (http://bips.u-strasbg.fr/fr/Documentation/ClustalX/). The deduced amino acid sequences for OlOpn5L1, TIOpn5L1, and DrOpn5L2 are partial and truncated at N-termini.

Download figure to PowerPoint

Table 1. Sequence Sources With Associated Nucleotide Database Accession and ID Numbers Used to Construct the Phylogenetic Tree Shown in Fig. 1A.
Opsin nameGene name/symbolProposed nameAnimalReference sequenceOther reference sequence
  • *

    Ensembl Gene ID is indicated.

  • **

    Ensembl Transcript ID is indicated.

Amphiop3Amphiop3 Japanese lanceletAB050610 
DrOpn5L1LOC570397 ZebrafishXM_693874 
DrOpn5L2LOC100002138 ZebrafishENSDART00000074613** 
DrOpn5msi:ch211-105n9.2 ZebrafishXM_687530 
HsOpn5OPN5 HumanNM_181744 
LOC421093LOC421093cOpn5L1ChickenXM_419178AB368181
LOC422050LOC422050cOpn5L2ChickenXM_420056AB368183
LOC428670OPN5cOpn5mChickenAB368182XM_426228
MmOpn5Opn5 MouseNM_181753 
OlOpn5L1ENSORLG00000017226* MedakaENSORLT00000021557** 
OlOpn5L2ENSORLG00000010136* MedakaENSORLT00000012711** 
OlOpn5mENSORLG00000015647* (OPN5) MedakaENSORLT00000019587** 
TnOpn5L1GSTENG00006437001* TetraodonGSTENT00006437001** 
TnOpn5L2GSTENG00018381001* TetraodonGSTENT00018381001** 
TnOpn5mGSTENG00035513001* TetraodonGSTENT00035513001** 
XtOpn5L1LOC780307 (MGC147314) Xenopus tropicalisNM_001079378 
XtOpn5mENSXETG00000011322* (OPN5) Xenopus tropicalisENSXETT00000024724** 

Chromosomal Localization of the Three Opsin 5–Related Genes: Mammals Have Lost the Two Opsin 5–Related Genes

To further investigate the possibility that there are three distinct opsin 5–related genes in the chicken genome, we examined chromosomal localization of the genes and compared them with those of their orthlogous genes from other species by searching the genome databases of NCBI and Ensembl. Searches in the chicken genome database have revealed that LOC421093 maps to the chromosome 2 (Fig. 2B), while both LOC428670 and LOC422050 map to chromosome 3 at distinct loci (Fig. 2A,C). Since we found that LOC428670 is located between Gpr115 and a Patched-related gene as OPN5/Opn5 is in the mammalian genomes (Fig. 2A), it is reasonable to think that LOC428670 is orthologous to mammalian opsin 5. In zebrafish, si:ch211-105n9.2 on chromosome 20 seems to be orthlogous to chicken LOC428670 by comparison of their chromosomal regions; but gene order near Gpr115 has not been conserved between the two species (see Suppl. Fig. 1A).

thumbnail image

Figure 2. Schema of the chromosomal regions surrounding the three chicken opsin 5–related gene loci and those for their orthologous genes (shown in red) from other species. A: Comparison of syntenic regions encompassing the LOC428670 (cOpn5m) locus on chicken Chromosome 3 with the Opn5 loci on human Chromosome 6p12.3 and mouse Chromosome 17. Gene order and distance (∼190 kb) have been conserved. B: Comparison of syntenic regions encompassing the LOC421093 (cOpn5L1) locus on chicken Chromosome 2 with the orthologous loci on Xenopus and zebrafish chromosomes. Gene order has been almost conserved. C: Comparison of syntenic regions encompassing the LOC422050 (cOpn5L2) locus on chicken Chromosome 3 with the candidate orthologous loci on Tetraodon and Medaka chromosomes. Gene order has been almost conserved.

Download figure to PowerPoint

As for LOC421093, its fully syntenic region has been found in the chromosome of Xenopus tropicalis scaffold_84 (Fig. 2B). In zebrafish, the chromosomal region encompassing LOC570397 has been almost syntenic to that of LOC421093 in the chicken (Fig. 2B). In Tetraodon and Medaka genomes (Fig. S1B), the orthologous genes to LOC421093, GSTENG00006437001, and ENSORLG00000017226 have been found, respectively; however, many gene loci between the LOC421093 orthologs and RIPK1, and two loci near MEP1B have been lost in Tetraodon and Medaka. In contrast to the syntenic conservation exhibited by the chicken LOC428670 and mammalian Opn5 loci (Fig. 2A), it appears that a mammalian ortholog of LOC421093 is absent as revealed by searching of genome databases. To further explore this apparent lack of a mammalian LOC428670 ortholog, we compared genomic localization for LOC421093 with syntenic regions of the human and mouse genome (Fig. S1C). In the chicken genome, LOC421093 localizes between RCJMB04_19h3 (B4galt6) and a pseudogene for KIAA1012 (Fig. S1C). There are no genes between the two loci in the mouse syntenic region while there are other distinct genes in the human syntenic region (Fig. S1C), suggesting that there is no “LOC421093” locus in the mammalian genome.

As for LOC422050, it localizes to chicken Chromosome 3 in a region of approximately 80 kb that emcompasses the following immediately adjacent gene cluster, MUT-LOC422050-LOC422051-CDC5L-LOC771603-RCJMB04_23d3-Runx2-CLIC5(Fig. 2C). The Tetraodon orthlogs of MUT, Runx2, and CLIC5 localize to Tetraodon Chromosome 14 and there are two LOC422050 orthologs between MUT and Runx2 (Fig. 2C). Similarly, in Medaka, the following gene cluster, MUT-ENSORLG00000010136 (LOC422050 ortholog)-Runx2-CLIC5 is found in the chromosome 24 (Fig. 2C). In zebrafish, gene loci for MUT, Runx2, and CLIC5 were found in the chromosome 20 and LOC100002138, the zebrafish ortholog of chicken LOC422050 localized to the adjacent region (Fig. S1D). In contrast, the human orthologs of MUT, CDC5L, and CLIC5 localize to human Chromosome 6p21.1 (CDC5L and CLIC5) and 6p12.3 (MUT) and are separated by approximately 5.1 Mb (Fig. S1E). In the chicken, Tetraodon, and Medaka genomes, the orthologs of LOC422050 localize adjacent to MUT, Runx2, and CLIC5; however, there is no “LOC422050” locus in this portion of human Chromosome 6.

Thus, it seems most likely that there are three Opsin 5–related genes in the chicken, but mammals have only one Opsin 5 gene. To better indicate the gene identity, we designated LOC428670 or Opsin 5 as cOpn5m (chicken opsin 5, mammalian type), LOC421093 as cOpn5L1 (chicken opsin 5-like gene 1), and LOC422050 as cOpn5L2 (chicken opsin 5-like gene 2) for further description in this report.

Comparison of Amino Acid Identity: The Opn5m Group Is Most Conserved Among Representative Vertebrate Species

A global comparison of deduced amino acid sequences over the seven transmembrane domains and associated intracellular and extracellular loops but excluding N- and C-termini, i.e., the core region (Bellingham et al.,2006) of Opsin 5–related proteins from representative vertebrate species further confirms the distinction between the three Opsin 5–related genes (Table 2 and Suppl. Table 2). cOpn5m shows 85% amino acid identity with human Opn5 and more than 75% identity with other Opn5m proteins, but only 37% and 34% with cOpn5L1 and cOpn5L2, respectively (Table 2). As for cOpn5L1, it shows the highest amino acid identity and similarity with zebrafish Opn5L1 (70% identity and 87% similarity) (Tables 2 and S2). As for cOpn5L2, it shows the highest amino acid identity and similarity with Tetraodon Opn5L2 (54% identity and 72% similarity) (Tables 2 and S2). It is known that there are two melanopsin (ospin 4) genes, Opn4m and Opn4x, in the chicken and other nonmammalian vertebrates: Within each of the Opn4m and Opn4x groups, identity is 70% or greater and greater than 66%, respectively, while between the two groups, identity is lower at approximately 55% (Bellingham et al.,2006). Compared with the melanopsin genes, a greater percent of amino acid identity is found within the Opn5m group, comparable within the Opn5L1 group, and lower identity within the Opn5L2 group. However, amino acid identity is much lower between the three groups (30% identity and 52% similarity between cOpn5L1 and cOpn5L2) (Tables 2 and S2).

Table 2. Amino Acid Identity (Percent) Across the Core Region of Opsin 5-Related Proteins
Opsin nameReference sequencecOpn5mHsOpn5MmOpn5DrOpn5mcOpn5L1cOpn5L2DrOpn5L2Amphiop3XtOpn5L1DrOpn5L1OlOpn5L1OlOpn5L2OlOpn5mTnOpn5L1TnOpn5L2TnOpn5mXtOpn5m
  • a

    AB368182 was used to calculate amino acid identity shown here.

  • b

    Ensembl Transcript ID is indicated.

cOpn5mAB368182, XM_426228a-85847837343229373435337736347583
HsOpn5NM_181744 -987438343129363437327538347380
MmOpn5NM_181753  -7437343229363437327538347381
DrOpn5mXM_687530   -38353228373637338737348576
cOpn5L1XM_419178    -302826577066304064294038
cOpn5L2XM_420056     -5226313130513530543335
DrOpn5L2ENSDART00000074613b      -22282829433230583132
Amphiop3AB050610       -222222262822252928
XtOpn5L1NM_001079378        -6056323755313639
DrOpn5L1XM_693874         -71323668293635
OlOpn5L1ENSORLT00000021557b          -273782313738
OlOpn5L2ENSORLT00000012711b           -3328423234
OlOpn5mENSORLT00000019587b            -37338975
TnOpn5L1GSTENT00006437001b             -313739
TnOpn5L2GSTENT00018381001b              -3334
TnOpn5mGSTENT00035513001b               -72
XtOpn5mENSXETT00000024724b                -

These analyses have indicated that on the basis of phylogeny, sequence similarity, and chromosomal localization, cOpn5m group genes are present in birds, frogs, fish, and mammals, whereas cOpn5L1 group genes are found in birds, frogs, and fish, and cOpn5L2 group genes are limited to birds and fish. Compared with the Opn5m group, the cOpn5L1 group is less conserved and the cOpn5L2 group is much less conserved.

cOpn5m Is Predominantly Expressed in the Developing Chicken Retina, While the Expression Level of cOpn5L1 and cOpn5L2 Is Much Lower

To examine the relative abundance of each transcript, we performed quantitative PCR analysis in the retina at E10, E17, and posthatching day 6 (P6). We measured the amount of cOpn5L1, cOpn5m, and cOpn5L2 transcripts relative to the ubiquitous β-actin mRNA, taken as an internal standard, using total RNA from the neural retina (NR) and retinal pigmented epithelium (RPE) (Fig. 3 and Suppl. Table 3). With the level of cOpn5L1 in E10 NR referred to as 1, the relative mRNA level of cOpn5L1 in the NR increased to 7.1 at E17 and 17.9 at P6. Under the same standard, the relative mRNA level of cOpn5m in the NR was 12.7 at E10, increased to 46.0 at E17, and further increased to 102 at P6. Similarly, the relative mRNA level of cOpn5L2 in the NR was 0.66 at E10, 3.4 at E17, and 9.9 at P6. In the RPE, on the other hand, with the level of cOpn5L1 in E10 NR referred to as 1, the relative mRNA level of cOpn5m was 3.52 at E10 and decreased to 0.356 at E17 and 0.415 at P6. As for cOpn5L1 and cOpn5L2, under the same standard, the relative mRNA levels were much lower than that of cOpn5m (Fig. 3 and Table S3). These results indicate that the expression levels of cOpn5m in the neural retina at E10, E17, and P6 are much higher than those of the other two cOpn5L genes and that cOpn5m is distinctly expressed in the RPE at E10.

thumbnail image

Figure 3. Relative mRNA levels of the three opsin 5–related genes in the neural retina (NR) and retinal pigmented epithelium (RPE) at E10, E17, and P6 as revealed by quantitative PCR analysis. The level of cOpn5L1 in E10 NR was referred to as 1. Black column (1), cOpn5L1; gray column (2), cOpn5m; white column (3), cOpn5L2. cOpn5m is predominantly expressed in the developing neural retina, while cOpn5L1 and cOpn5L2 are expressed at much lower levels. cOpn5m is also expressed in the RPE at E10.

Download figure to PowerPoint

cOpn5m Is Expressed by Certain Cell Types of Developing Retina But cOpn5L1 or cOpn5L2 Is Not

To verify the relative expression levels and scan through the expressing cells for the three genes, we performed in situ hybridization on retinal sections at E3.5, E10, E17, and P5 with their RNA probes using the sense probes for control. At E3.5, no specific signals for cOpn5L1, cOpn5m, or cOpn5L2 mRNA were detected in the eye or surrounding head region (Suppl. Fig. 2A–F), while Chx10 was distinctly expressed in the neural retina as shown previously (Fig. S2G) (Chen and Cepko,2000). At later stages, only cOpn5m was expressed by a subset of cells in the neural retina (Fig. S2H–M, O–T, V–AA) and Chx10 was expressed by the developing bipolar cells (Fig. S2N, U, AB). In the pineal gland at E17, which is also a photosensitive organ, none of the three opsin 5–related genes exhibited specific expression patterns (not shown). These results indicate that although there are three opsin 5–related genes in the chicken and PCR analysis could detect their expression in the developing retina, only cOpn5m-expressing cells could be identified by a standard in situ hybridization method. Therefore, we focused on cOpn5m and examined the detailed spatio-temporal expression pattern in the developing retina.

At E5, specific expression of cOpn5m was not yet observed in the retina (Fig. 4A, B). By E7, cOpn5m mRNA was detected in a small subset of cells in the vitreal neuroepithelium (Fig. 4C,D). At E10, when three nuclear laminar structures, the ganglion cell (GC) layer, inner nuclear layer (INL), and outer nuclear layer (ONL) are formed, the signals for cOpn5m mRNA became more intense in the GC layer (Fig. 4E, F). By E12, the distinct expression was found in subsets of cells in the GC layer and the inner layer of INL, or the amacrine cell (AC) layer (Fig. 4G, H). At E15 and E17, cOpn5m continued to be expressed by subsets of cells in the GC and AC layers (Fig. 4I, J, K–M). At this stage, cOpn5m was also expressed in a subset of cells in the developing trigeminal ganglion (Fig. 4N). At P5, cOpn5m continued to be expressed by subsets of cells in the GC and AC layers (Fig. 4O; Fig. S2V). It is interesting to note that cOpn5m mRNA is often detected on the vitreal side of the cell body in the AC layer (Fig. 4P, Q). These results show that the onset of cOpn5m expression in the developing retina is around E7 and it seems that Opn5m continues to be expressed by subsets of cells in the GC and AC layers until at least P5, the latest stage examined in this study.

thumbnail image

Figure 4. Expression patterns of the cOpn5m gene during chicken retinal development from E5 to P5. Large magnifications of A, C, E, G, I, and K are shown in B, D, F, H, J, and M, respectively. AM, OQ: The posterior retina at each stage is shown. Vitreal side is up. At E5, no specific signals are observed in the developing retina (B). At E7 (D) and lateron, cOpn5m is expressed by subsets of cells in the ganglion and amacrine cell layers. L: The cOpn5m sense probe gives no signals. N: cOpn5m mRNA is also detected in subsets of cells in the trigeminal ganglion at E17. O: cOpn5m expression in the retina near the optic nerve. P, Q: In later stages (E17 and P5), cOpn5m mRNA is often detected on the vitreal side (up) of the cell body in the amacrine cell layer. ac, amacrine cell layer; gc, ganglion cell layer; inl, inner nuclear layer; ipl, inner plexiform layer, nr, neural retina; on, optic nerve; onl, outer nuclear layer; pe, pigmented epithelium; vit, vitreous body. Scale bars = 100 μm in A, C, E, G, I; 100 μm in B, D, F, H, J; 100 μm in K, L; 100 μm in M; 100 μm in N; 50 μm in O; 20 μm in P, Q.

Download figure to PowerPoint

cOpn5m Is Expressed by Subsets of Large Ganglion Cells and GABAergic Amacrine Cells

To further characterize cOpn5m-expressing cells, we examined whether Pax6 or Islet1 was co-localized with Opn5m in the E17 retina by immunofluorescence after in situ hybridization. These transcription factors are known to be expressed in the differentiating and mature retinal cells of the chicken, as summarized in Table 3 (Belecky-Adams et al.,1997; Edqvist et al.,2006; Fischer et al.,2007). We found that cOpn5m-expressing cells in the AC layer were middle-sized, Pax6-positive (Fig. 5A, A′) and Islet1-negative (Fig. 5C–D′). On the other hand, cOpn5m-expressing cells in the GC layer were relatively large and positive for Pax6 (Fig. 5B, B′). As for Islet1, there were two distinct populations in the cOpn5m-expressing cells of the GC layer: Islet1-positive (Fig. 5E) and Islet1-negative cells (Fig. 5F, F′). It is known that the Islet1-positive cells in the GC layer abutting the inner plexiform layer are likely displaced amacrine cells (Edqvist et al.,2006). However, as Islet1-positive, cOpn5m-expressing cells in the GC layer are not located in that lamina (Fig. 5E), it is unlikely that they might be displaced amacrine cells.

Table 3. Retinal Marker Genes and Neurotransmitter-Related Molecules Used In This Study
Protein nameExpressing cells/neuronsAntibody/DilutionReference
Pax6GC, all amacrine, HCDSHB, PAX6/1:100Edqvist et al. (2006)
Islet1GC, cholinergic amacrine, bipolar, HCDHSB, 39,4D5/1:50Edqvist et al. (2006)
Tyrosine hydroxylase (TH)Dopaminergic, adrenergic, norudrenergicChemicon, MAB318/1:100Ohyama et al., (2005)
Glutamic acid decarboxylase 65/67 (GAD65/67)GABAcrgicAbcam, ab11070/1:200This study
Choline acetyltransferase (ChAT)CholinergicChemicon, AB144P/1:100Yamaguta and Sunes (1995)
thumbnail image

Figure 5. Double staining of cOpn5m and retinal markers in E17 retinae. cOpn5m-expressing amacrine cells are GABAergic. AB′: Co-labeling for cOpn5m mRNA (arrow in A, B) with Pax6 protein. Large magnifications of A and B are shown in A′ and B′, respectively. CF′: Co-labeling for cOpn5m mRNA (arrows in C, D) with Islet1 protein. E: cOpn5m-expressing Islet1-positive ganglion cell (GC). F, F′: cOpn5m-expressing Islet1-negative GC. GG″: Co-labeling for cOpn5m mRNA with GAD65/67 protein. Arrows show the cOpn5m-expressing cells positive for GAD65/67. The merged image of G and G′ is shown in G″. H, H′: Co-labeling for cOpn5m mRNA with TH protein. Large magnification of H is shown in H′. Vitreal side is up. Nuclei are stained with Hoechst in D′ F′, G, H, H′. ac, amacrine cell layer; bc, bipolar cell layer; gc, ganglion cell layer; hc, horizontal cell layer; inl, inner nuclear layer; ipl, inner plexiform layer; onl, outer nuclear layer. Scale bars = 100 μm in A–C, H; 10 μm in A′, B′, D–F′; 30 μm in G, G′, H′.

Download figure to PowerPoint

Retinal amacrine cells are subdivided into more than 29 distinct cell types in mammals and a variety of neurotransmitters are produced from amacrine cells (Masland,2001; Dowling,1987). We, therefore, examined what types of neurotransmitter-related molecules are present in the cOpn5m-expressing amacrine cells at E17 (Table 3). Since cOpn5m-expressing amacrine cells were negative for Islet1, they are not cholinergic (Edqvist et al.,2006; see Experimental Procedures section). We found that they are GABAergic as glutamic acid decarboxylase 65/67 (GAD65/67) was present in the cytoplasm of cOpn5m-expressing cells (Fig. 5G–G″). In contrast, cOpn5m-expressing cells were negative for tyrosine hydroxylase (TH), a biosynthetic enzyme to generate dopamine and other amines, in the E17 retina. (Fig. 5H, H′).

It was reported that the earliest timepoint for putative amacrine cells to express Islet1 and Pax6 in the chicken retina is around E7 (Edqvist et al.,2006). This stage seems to correspond to the later stages of amacrine cell genesis (Kahn,1974). It is noticeable that E7 is also the approximate timepoint for the earliest expression of cOpn5m, which is expressed in a subset of developing amacrine cells, and that the onset of cOpn5m expression is similar to those of Pax6 and Islet1 expression in the developing amacrine cells.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

It was reported that the four mammalian non-rod, non-cone opsins, i.e., RGR opsin, peropsin, encephalopsin (opsin 3), and melanopsin (opsin 4) are expressed in the developing retina from earlier stages than the classical rod and cone opsin genes are expressed (Tarttelin et al.,2003a). However, cell types of those opsin-expressing cells have not been fully elucidated or a function for these genes in retinal development remains unknown. For example, melanopsin is expressed already at embryonic day 18 (E18) by rat retinal cells of the inner neuroblast layer (Fahrenkrug et al.,2004) and around the same stage in mice (Tomonari et al.,2005). Since melanopsin has proved to be photoreceptive, it was suggested that melanopsin would already play a photoreceptive role in the developing retinohypothalamic tract. In the chicken, on the other hand, two types of melanopsin, Opn4m and Opn4x, are distinctly expressed in subsets of differentiating retinal cells (ganglion, amacrine, horizontal, and bipolar cells) of the embryonic retina (Tomonari et al.,2005,2007) and are more broadly expressed in the GCL, INL, and ONL of the 2-week-old retina (Bellingham et al.,2006), suggesting another role for melanopsin in retinal development other than photoreception. This study has for the first time shown that a chicken opsin 5 gene is expressed in the developing retina; specifically cOpn5m, which exists in mammals as well, is expressed by a subset of ganglion cells with larger cell bodies and a subset of GABAergic amacrine cells. Together with the previous studies on melanopsin, these results imply that the non-canonical opsins may be distinctly expressed by subsets of developing retinal cells and may contribute to retinal differentiation, which should be verified by further studies.

It was reported that Opsin 5 is expressed in mammalian eye, brain, and testis of adult tissues (Tarttelin et al.,2003b). Therefore, although this study has shown that only cOpn5m is predominantly expressed by subsets of developing retinal cells until posthatching day 5, the latest stage examined, we could not exclude the possibility that cOpn5L1 and cOpn5L2 might be mainly expressed in the adult retina or other tissues.

Functions for opsin 5 have not been elucidated until now, but from the phylogenetic position, it is thought to function as a photoisomerase-like peropsin does (Tarttelin et al.,2003b). However, from the amino acid comparison, even peropsin has the possibility to function as a GPCR because it has an NPXXY motif in the seventh helix while this motif is changed to NAINY in the RGR photoisomerase subfamily (Koyanagi et al.,2002). Both cOpn5m and cOpn5L2 (LOC422050) have the NPXXY motif (Fig. 1B) and, therefore, these Opsin 5-related proteins might function as GPCRs in which retinal-like molecules would act as ligands.

Opsin 5 (cOpn5m) is expressed in a subset of developing amacrine cells and what role might be proposed for cOpn5m in these cells? It has been known that retinal amacrine cells are subdivided into more divergent cell types with characteristic arborization patterns of the processes in mammals (Masland,2001). Given that there are such subtypes in the chicken amacrine cells as well, it is intriguing to speculate that the non-canonical opsins such as opsin 5 might be involved in regulation of neurite arborization and mosaic spacing of certain amacrine cell types as recently reported for a cell adhesion molecule in mice (Fuerst et al.,2008). Another notable thing is that cOpn5m mRNA becomes localized on the vitreal (apical) side of the differentiating amacrine cell bodies from which the processes extend to the inner plexiform layer. This is reminiscent of classical opsins, i.e., rod and cone opsins present in the outer segments on the apical side of photoreceptors. Moreover, homozygotes for a targeted null mutation in the rod opsin gene fail to develop retinal rod outer segments and lose their photoreceptors (Lem et al.,1999), showing that rod opsin is required for the development of apical structures (i.e., outer segments) in the photoreceptors. Thus, it is tempting to speculate a role for cOpn5m in the maintenance of apical structures in certain amacrine cell types.

In the animal kingdom, there are two major types of photoreceptors, rabdomeric and ciliary, from the aspect of their cytological origins, microvilli and cilia, respectively, of the outer segments (Smith,2000). In ciliary photoreceptors, light-dependent channels are controlled via cGMP (Gomez and Nasi,1995) resulting in hyperpolarization of the membrane potential. In view of the transduction cascade, however, ciliary photoreceptors are further divided into two subgroups, ciliary vertebrate photoreceptors and ciliary invertebrate photoreceptors, because particular G-protein subtypes that mediate phototransduction differ in each other. In ciliary invertebrate photoreceptors, a G-protein of the Go subtype activates a membrane guanylate cyclase to induce an increase in cGMP (Gomez and Nasi,2000). In contrast, in ciliary vertebrate photoreceptors, the Gt subtype, transducin, activates a phosphodiesterase, leading to hydrolysis of cGMP (for review, see Yau and Baylor,1989). It has been known that the opsins present in rhabdomeric photoreceptors and melanopsin in a subset of photosensitive retinal cells of vertebrates are coupled with the Gq subtype, while rod and cone opsins in ciliary photoreceptors of vertebrates are coupled with the Gt subtype (for review, see Hankins et al.,2008). Thus, it has been thought that the Go-coupled ciliary photoreceptors would represent a novel, separate subgroup among visual receptors (Gomez and Nasi,2000). Until now Go-coupled opsins are not found in vertebrates but only in invertebrates such as scallops and lancelets (for review, see Terakita,2005). Since the phylogenetic analysis reveals that Japanese lancelet (Branchiostoma belcheri), an ancestral vertebrate, possesses the three categories of the G protein–coupled opsins (Koyanagi et al.,2002; Terakita,2005), Go-coupled opsins might be present in vertebrates as well. In the phylogenetic tree, the opsin 5 subfamily is most related to the peropsin subfamily, and subsequently to the Go-coupled opsins (Terakita,2005). Therefore, it would be intriguing to know whether Opsin 5 proteins might be coupled with the Go-proteins or not. If this is the case, it would turn out that the cells retaining the molecular property of the third retinal subtype might exist at least in the developing chicken retina.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Isolation of Chicken Opsin 5–Related cDNAs

Total RNA was extracted from E16 chicken embryonic retina using RNAqueus (Ambion). Reverse transcription was carried out using 1 μg of total RNA, oligo dT primer (Invitrogen), and Superscript II reverse transcriptase (RT) (Invitrogen). The RT reaction mix was treated with ribonuclease H (Invitrogen) at 37°C for 20 min. PCR was performed using 1 μl of the 1/10 attenuate cDNA reaction mix and a thermal cycler (MasterCycler, Eppendorf). Gene-specific primers spanning exon-intron boundaries were designed according to the sequence deposited in GenBank as shown in Supplementary Table 1. In total, 40 cycles were performed at annealing temperatures of 68.6°C (cOpn5L1), 63.8°C (cOpn5m), and 69.0°C (cOpn5L2), respectively. The PCR products were cloned into a plasmid vector (pGEM-T Easy), and multiple clones were sequenced to obtain partial cDNAs for cOpn5L1 (635 bp), cOpn5m (744 bp), and cOpn5L2 (752 bp), which were used for template DNAs to generate RNA probes for in situ hybridization analysis. These partial coding sequences are predicted to encode half of the 2nd transmembrane domain (TM) to half of the 7th TM for cOpn5L1, the 1st TM to half of the 7th TM for cOpn5m, and half of the 1st TM to half of the 7th TM for cOpn5L2. The nucleotide identities between the cloned sequences were 50.2% (cOpn5m and cOpn5L1) and 50.1% (cOpn5m and cOpn5L2), respectively. To identify the 5′ and 3′ end sequences of the cOpn5m cDNA, rapid amplification of cDNA ends (RACE) was performed using cDNAs from the E17 retina and primers listed in Supplementary Table 4, according to the instructions of 5′ and 3′ RACE Systems (Invitrogen). We found that cOpn5m encodes a 357–amino acid protein containing seven-transmembrane domains, which is nearly identical to that predicted by NCBI (364 amino acids). The full-length cDNA sequence for cOpn5m and partial cDNA sequences for cOpn5L1 and cOpn5L2 have been submitted to DDBJ/GenBank under the accession numbers AB368182 (cOpn5m), AB368181 (cOpn5L1), and AB368183 (cOpn5L2), respectively.

Sequence and Phylogenetic Analysis

Assembly of predicted sequences, sequence analysis, and nucleotide identity comparisons were undertaken using GENETYX-SV/RC 12.2.6 (Genetyx). For phylogenetic purposes, amino acid sequences were aligned with ClustalX 1.83 (ftp://ftp-igbmc.u-strasbg.fr/pub/ClustalX/) (Thompson et al.,1997) and neighbor-joining trees were constructed with bootstrap confidence values based on 1,000 replicates using NJplot (Perrière and Gouy,1996) (Fig. 1A, B). Amino acid identity and similarity comparisons were undertaken using MacVector 9.5.4 (MacVector) (Table 2 and Suppl. Table 2).

Quantitative PCR

Total cellular RNA was extracted from NR and RPE at E10, E17, and P6 using an RNAqueus kit (Ambion) and treated with RNase-free DNaseI according to the manufacturer's instructions. To remove RPE from E17 and P6 retinae, the eyes were briefly treated with EDTA and Dispase II (Bruhn and Cepko,1996). One microgram RNA was reverse transcribed to cDNA with Superscript II RT. Quantitative PCR (Q-PCR) was performed using SYBR Green PCR Master Mix (Applied Biosystems) on ABI 7900 Real Time PCR System (Applied Biosystems) under the following conditions: 2 min at 50°C, 10 min at 95°C; 40 cycles of 15 sec at 95°C, 1 min at 60°C; and hold at 25°C. The Q-PCR primer sequences are as follows: cOpn5L1: 5′-TTCTATGCGTTCTGTGGGCTATTCT-3′ (forward) and 5′-ATATTTGTCCATGTTTTCTCCTGA-ACCT-3′ (reverse), cOpn5m: 5′-GGGCTGGCTTCTTCTTTGGCTGT-GG-3′ (forward) and 5′-CAGGCAGATAAAGGCATGGTGT-3′ (reverse) and cOpn5L2: 5′-CTGATGGGTTTCCTTTTTGGTGT-3′ (forward) and 5′-CTTAGAGATTTTGTTACTGTTAGA-TTTG-3′ (reverse). The primer sequences for β-actin were described previously (Tomonari et al.,2005). The amplicon size for cOpn5L1, cOpn5m, and cOpn5L2 was 133, 128, and 111 bp, respectively. Efficiencies of amplification (E) were cOpn5L1 mRNA, E = 0.7662 ± 0.0833; cOpn5m mRNA, E = 0.7346 ± 0.0564; cOpn5L2 mRNA, E = 0.6995 ± 0.1043; β-actin mRNA, E = 0.7906 ± 0.0206. cOpn5L1 to β-actin, cOpn5m to β-actin, and cOpn5L2 to β-actin ratios were determined by the comparative Ct method (Winer et al.,1999). The Q-PCR analysis was performed in triplicate for the E10 and E17 tissues and duplicate for the P6 tissues.

In Situ Hybridization (ISH) and Immunofluorescence

The chicken embryos and embryonic heads at the desired developmental stages were collected and fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) at 4°C for 3 hr. Digoxigenin (DIG)-labeled antisense and sense control riboprobes were generated using T7 and SP6 RNA polymerase according to standard procedures. The template cDNAs used to generate riboprobes were aforementioned. In situ hybridization to tissue sections with chromogenic substrates (BCIP/NBT) for alkaline phosphatase was performed on 18-μm-thick frozen sections (E3.5 to P5) according to Schaeren-Wiemers and Gerfin-Moser (1993). Briefly, all RNA probes for the three opsin 5–related genes were hybridized at 68°C and sections were treated with proteinase K at 1 μg/ml for 10 min at room temperature. The RNA probe concentration and time for the color reaction to obtain the result shown in Figure S2 is summarized in Supplementary Table 5. For double staining with antibodies, retinal frozen sections were processed without proteinase K digestion during ISH in which the DIG-labeled probes were hybridized at 1 μg/ml, DIG was immunodetected with anti-DIG antibody conjugated with peroxidase, and then the mRNA signals were detected with Fluorescein-Tyramide (Perkin-Elmer) according to the manufacturer's instructions. After washing and blocking with 5% normal goat serum in PBST (0.2% Triton X-100 in PBS), the sections were incubated with the first antibody at 4°C overnight, washed, and stained with Cy3-conjugated anti-mouse secondary antibody (Jackson) (diluted at 1:500) at room temperature for 3 hr. After washing with PBST, nuclei were counterstained with Hoechst 33342. Details for antibodies are shown in Table 3. The Pax6 and Islet1 antibodies were obtained from Developmental Studies Hybridoma Bank, Iowa University. We isolated partial cDNAs for chicken TH, GAD65, and GAD67 to confirm that the protein localization as revealed by anti-TH and anti-GAD65/67 antibodies corresponds well to their mRNA distribution (not shown). We also confirmed that Islet1 protein is co-localized with ChAT protein in the E17 chicken retina as described (Edqvist et al.,2006), but it seemed that the ChAT antibody used in this study also stained cells in the GC layer other than displaced amacrine cells, which needs further elucidation. Sections were analyzed and fluorescence micrographs were acquired using a Nikon C1si confocal imaging system with EZ-C1 software (Nikon, Japan). Color conversion of the fluorescence micrographs was carried out by Adobe Photoshop 7.0 (Adobe Systems).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We are grateful to Yuki Yamahoshi and Shuang Song for technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The Supplementary Material referred to in this article can be viewed at www.interscience.wiley.com/jpages/1058-8388/suppmat

FilenameFormatSizeDescription
dvdy21611-FigS10513Final.tif2412KSupplementary Fig. 1. Schema of the chromosomal regions surrounding the three chickenopsin 5 –related gene loci and those (shown in red) from other species.A:Comparison of syntenic regions encompassing theLOC428670( cOpn5m ) locus on chicken Chromosome 3 with theOpn5loci on zebrafish Chromosome 20. Gene order has been almost conserved.B:Comparison of syntenic regions encompassing theGSTENG00006437001( LOC421093ortholog) locus onTetraodonChromosome un random with that on Medaka Chromosome 17. Gene order has been conserved between the two species.C:Comparison of representative loci between RCJMB04 19h3 (B4galt6 ortholog) andMEP1Bgenes on chicken Chromosome 2 (approximately 220 kb apart) and the orthologous loci on human Chromosome 18q12.1 (approximately 590 kb apart) and on mouse Chromosome 18A2 (approximately 420 kb apart). Note the lack of aLOC421093orthologous locus in mammals.D:Syntenic regions encompassing theLOC100002138( LOC428670ortholog) locus on zebrafish Chromosome 20 according to Ensembl (upper) and NCBI (lower) databases.E:Comparison of representative loci betweenMUTandCLIC5genes on chicken Chromosome 3 (approximately 80 kb apart) and the orthologous loci on human Chromosome 6p21.1-p12.3 (approximately 5 Mb apart). Note the lack of aLOC422050orthologous locus in humans.
dvdy21611-FigS2.tif4191KSupplementary Fig. 2. In situ hybridization analysis of the threeopsin 5 –related genes in the chicken. Eyes at E3.5 ( A–G ), and posterior retinae at E10 ( H–N ), E17 ( O–U ), and P5 ( V–AB ) are shown. Hybridized with thecOpn5mantisense probe (A, H, O, V),cOpn5msense probe (B, I, P, W),cOpn5L1antisense probe (C, J, Q, X),cOpn5L1sense probe (D, K, R, Y),cOpn5L2antisense probe (E, L, S, Z), andcOpn5L2sense probe (F, M, T, AA). Arrowheads in H, O, V indicatecOpn5mexpression in subsets of developing retinal cells. As a positive control, the retinae were hybridized with theChx10antisense probe, showingChx10expression in the retinal neuroepithelium (G) and developing bipolar cells (N, U, AB) as reported previously. le, lens; re, retina; vit, vitreous body.
dvdy21611-SupplementaryTable1.doc134KSupporting Information file dvdy21611-SupplementaryTable1.doc

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.