The 1D4 Antibody Labels Outer Segments of Long Double Cone But Not Rod Photoreceptors in Zebrafish

Abstract

Purpose.: In experimental eye research, zebrafish has become a powerful model for human retina disorders. The purpose of the present study is the characterization of antibodies commonly employed in zebrafish models for rod photoreceptor degeneration.

Methods.: The 1D4 monoclonal antibody, developed against bovine rhodopsin, has been widely used in studies addressing structural and functional features of rhodopsin and was reported as an informative marker to stain rod outer segments in both mice and zebrafish. We have used transgenic reporter lines and histologic analysis to determine the photoreceptor types identified by 1D4 and other antibodies in zebrafish.

Results.: We demonstrate that 1D4, in contrast to what has been reported previously, does not recognize rod outer segments in zebrafish, but instead labels long double cone outer segments consistent with sequence conservation of the respective epitope. As an alternative marker for zebrafish rods, we characterized the monoclonal antibody zpr-3, which was found to stain outer segments of both rods, as well as double cones.

Conclusions.: Our findings highlight the importance to confirm specificity of antibodies in cross-species experiments for correct interpretation of experimental data. Our findings clarify conflicting published information arising from studies using 1D4 and zpr-3 antibodies in zebrafish.

Introduction

The zebrafish has become an established model for eye research and for recapitulating aspects of human retinopathies. ^1–4 Its retina is organized similar to that of humans and many key factors controlling formation and maintenance of photoreceptors are highly conserved. ⁵ Furthermore and different to mammals, the teleost retina has an enormous potential to regenerate, making zebrafish an excellent model to study mechanisms underlying photoreceptor recovery. ³ Antibodies have been extensively used to analyze photoreceptor morphology and intracellular composition. Most commercially available antibodies are specific to proteins from humans, mice, and other mammals and the demand for zebrafish specific antibodies is rapidly increasing. Given the high conservation of photoreceptor proteins, monoclonal as well as polyclonal antibodies established in mammals are frequently used in zebrafish to characterize degenerative processes. Noteworthy, a detailed characterization of species-specificity is often lacking, opening the possibility for misinterpretation of results. 

Rhodopsin is a rod-specific opsin that is sensitive under low light. Mutations in this gene result in human eye diseases, such as congenital stationary night blindness ⁶ and retinitis pigmentosa (RP). ^7,8 The detailed molecular mechanisms underlying these diseases remain unclear and significant efforts are taken to establish suitable cell culture and animal models. Hence, rhodopsin specific antibodies are valuable tools to study retinal degeneration, which is the basis for future drug discovery and therapy designs. The mouse monoclonal antibody 1D4 was initially developed against bovine rhodopsin by immunizing with rod outer segment disc membranes. ⁹ A peptide comprising the C-terminal eight amino acids E-T-S-Q-V-A-P-A was used to map the binding sites of 1D4. ^10,11 Given its high specificity, it has been widely used in studies addressing structure and function of rhodopsin, ¹² as well as a reliable fusion tag for purification of nonrelated receptors. ^13,14 In mice, it has been used for the detection of rhodopsin ¹⁵ and its subcellular distribution ¹⁶ during retinal degeneration. This antibody is commercially available (ab5417; Abcam, Cambridge, UK; R5403; Sigma-Aldrich, St. Louis, MO; sc-57,432; SCBT, Santa Cruz, CA; MAB5356; Millipore, Billerica, MA) and according to the supplier is used to detect rhodopsin also in zebrafish. Accordingly, two recent zebrafish publications used the 1D4 antibody as a marker to examine rhodopsin localization and to describe rod deformations in studies characterizing two RP related genes, RPGR ¹⁷ and ZNF513. ¹⁸  

In our study, we used histologic analysis of a rod-specific transgenic reporter line and surprisingly found that 1D4 does not stain rod outer segments in zebrafish. Instead, we present evidence that 1D4 recognizes long double cone outer segments. We characterize the antibody zpr-3 as an alternative zebrafish rod marker, as it stains outer segments of rods and double cones. Our studies highlight the necessity to determine antibody binding specificities, and suggest revisiting earlier reports that have used the 1D4 antibody. 

Materials and Methods

Generation of Transgenic Reporter Fish

The Tg(Rho:EGFP) transgenic line was generated according to information previously provided. ^1,19,20 The Tg(LCR:RH2-1:EYFP) transgenic line was generated by using a 0.5-kb locus control region and 1.8-kb RH2-1 promoter region of zebrafish Rh2-1 as described. ²¹ In this study, homozygous Tg(Rho:EGFP) and heterozygous Tg(LCR:RH2-1:EYFP) specimens were used for analysis. All experiments were performed in accordance with Institutional Animal Care and Use Committee protocols of the National University of Singapore (approval numbers 075/07). All animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Immunohistochemistry

Cryosectioning and immunostaining of fixed zebrafish eyes was done as described before (Yin et al., 2011). Briefly, all sections were taken at the same angle and at the level of the lens to allow comparative analyses. Sections were incubated overnight at 4°C with primary monoclonal antibody anti-Rhodopsin 1D4 (ab5417; 1:50 dilution; Abcam), a polyclonal serum obtained after immunization with 1D4 peptide (ab3424; 1:900; Abcam), the zpr-1 or zpr-3 antibody (1:200; both Zebrafish International Resource Center [ZIRC], Eugene, OR) or polyclonal sera directed against zebrafish opsins that were generously provided by David Hyde (as described in Vihtelic et al., 1999). Sections were treated with secondary antibody Alexa-Fluor 488-, 568-, or 633-conjugated anti-mouse or anti-rabbit antibodies (Life Technologies, Grand Island, NY) for 1 hour at room temperature in the dark. Sections were incubated with Bodipy-FL or Bodipy-TR (0.33 mg/mL; Molecular Probes/Life Technologies, Grand Island, NY) for 1 hour at 37°C. Following this, the sections were treated with 4′,6-diamidino-2-phenylindole (DAPI 1:500) for 12 minutes at room temperature in the dark. After three washes with phosphate-buffered saline/tween, samples were mounted with antifade (Mowiol; Sigma-Aldrich). Images were acquired by separate sequential detection of the channels using the 405 (DAPI), 488 (1D4), 543 (zpr-1 and all opsin antibodies) or 633 (zpr-3) nm laser lines of a Zeiss LSM 510 confocal microscope (Zeiss, Jena, Germany). 

Western Blot Analysis

For Western blot analysis, adult zebrafish and bovine retinas were dissected and solubilized in lysis buffer (100 mM NaH²PO⁴; 10 mM Tris-HCl pH 4.5) containing 8M urea. Efficient tissue homogenization was achieved by sonication. After removal of cellular debris by centrifugation, the concentration of total protein in the clear supernatant was determined by Bradford analysis. After adjusting the protein concentration to 1 μg / μL, equal amounts of retinal lysate were subjected to standard SDS-PAGE and semi-dry transfer to nitrocellulose membrane. Blocking was performed for 30 minutes with 5% non-fat dry milk in tris-buffered saline tween-20 (TBS-T). Membranes were incubated with primary antibodies 1D4 (ab5417; Abcam) or zpr-3 (zpr-3 hybridoma culture supernatant; ZIRC) diluted 1:1000 in NET-Gelatine (150 mM NaCl; 5 mM EDTA; 50 mM Tris pH 7.5; 0.05% Triton-X 100; 2.5 g/L gelatine) containing 0.02% sodium-azide. For detection, horseradish peroxidase conjugated secondary antibodies diluted in TBS-T (1:5000) and standard electrochemiluminescence (ECL) reaction mix was used. 

Sequence Alignment

Sequences were obtained from Ensembl (www.ensembl.org). Multiple sequence alignments were done with the CLUSTAL-W computer program. ²²  

Results

Characterization of 1D4 in a Tg(Rho:EGFP) Transgenic Zebrafish Line

We generated a rod-specific transgenic line Tg(Rho:EGFP) in zebrafish using a 1.2 kb zebrafish rhodopsin promoter as published previously. ^1,19,20 In this line, green fluorescent protein (GFP) is found in the rod cytoplasm, including the inner segments, outer segments and cell body (Figs. 1A–C). During its development, the rod gradually changes its morphology in different aspects, such as overall length, shape and density. At 14 days, the rod outer segment (ROS) is still short and extends only slightly above the rod inner segment (RIS) and rod cell body (RCB; Fig. 1A). At 21 days, rods distribute throughout the complete photoreceptor cell layer. Their inner segments are elongated and outer segments of neighboring rods are tightly packed (Fig. 1B). In 4-month-old fish (Fig. 1C) the number of rods has increased. There are five to seven distinct layers of rod nuclei located in the ventral part of the outer nuclear layer. In these matured rods, only the base and tip of the spindle shaped inner segments are visible by GFP emission, while the middle part is too thin and GFP expression too faint to be detectable (Fig. 1C). At this stage, the outer segments have become long and rod-like, with a compact distribution in the outer layer of the retina. 

Figure 1. 

 

Monoclonal 1D4 immunohistochemistry does not detect rod outer segments in Tg(Rho:EGFP) transgenic fish retinas. (A) Transverse section across lens at 14 days post-fertilization (dpf). White box shows high magnification view of single rod. GFP is expressed in rod cytoplasm, including outer segment, inner segment, and cell body. Second column shows 1D4 staining, present in the same layer as GFP positive rods. Overlay in fourth column shows no overlap between 1D4 and GFP. DAPI stained nuclei in third column. (B) At 21 dpf, 1D4 signal distinct from GFP. (C) At 4 months, 1D4 positive segments are located below layer of GFP expressing rod outer segments. Note that DAPI signal is weak in the outer nuclear layer due to less efficient staining. ROS, rod outer segment; RIS, rod inner segment; RCB, rod cell body; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer.

As previously reported ^17,18 the mouse anti-rhodopsin 1D4 monoclonal antibody has been used in order to detect rhodopsin protein in the zebrafish rod outer segments. We tested 1D4 in our Tg(Rho:EGFP) line and found that at 14 days (Fig. 1A) the 1D4 signal was evident in the same layer where GFP positive rod outer segments are located. Importantly, however, the 1D4 signal did not overlap with the GFP signal but instead was presenting more basal to GFP positive rod outer segments. At 21 days (Fig. 1B), this separation became even more obvious and we found the 1D4 signal nonoverlapping with GFP positive rod outer segments. Also at 4 months (Fig. 1C), the 1D4 positive domains were clearly distinct from GFP expressing rod outer segments and were positioned basal to rod outer segments in the region of cone outer segments. Our results indicate that the monoclonal 1D4 antibody does not label rod outer segments in zebrafish, which suggests that it does not detect rhodopsin present exclusively in rod receptors. Based on the distribution pattern we instead hypothesize that 1D4 stains outer segments of a subset of zebrafish cones. 

1D4 Epitope Sequence Is Not Conserved between Mammals and Fish

The 1D4 monoclonal antibody was initially purified from mouse hybridoma cell lines generated by immunization with bovine rod outer segment disc membranes. ⁹ The epitope was determined to be the last eight amino acid residues of rhodopsin. ¹⁰ 1D4 is commercially available as monoclonal as well as polyclonal antibodies. We used both for immunostaining of larval and adult zebrafish retinas and found that both antibodies yield overlapping signals, with the polyclonal 1D4 antibody staining additional cells (Fig. 2A). We then aligned the 1D4 epitope from bovine rhodopsin with corresponding sequences in human, mouse and zebrafish rhodopsin and found that human, mouse and bovine sequences are identical, while the zebrafish sequence contains a five amino acid insertion and shares only six identical residues with mammals (Fig. 2B). This significant epitope variation makes it unlikely that 1D4 recognizes zebrafish rhodopsin. To test this experimentally, Western blot analysis was performed using bovine and zebrafish retinal extracts (Fig. 2D). As predicted, 1D4 failed to detect zebrafish rhodopsin (lanes 1, 3, and 5) while in the bovine extracts a robust signal was observed that corresponded to the different oligomeric and glycosylated forms of rhodopsin (lanes 2, 4, and 6). Importantly, when the zpr-3 antibody was used the opposite was observed: No signal was detected in the bovine extracts (lanes 8, 10, and 12) while a staining pattern corresponding to that of rhodopsin was revealed in the zebrafish extracts (lanes 7, 9, and 11). Together, these data suggest that the 1D4 epitope is not present in zebrafish rhodopsin. 

Figure 2. 

 

(A) Comparative immunostaining of zebrafish retinas at 4 dpf (top; larval) and at 12 months (bottom; adult) with 1D4 monoclonal and polyclonal antibodies. Note partial overlap of signals obtained by the two antibodies, with the polyclonal staining additional cells. (B) Amino acid alignment of 1D4 epitope sequence from bovine, human, mouse, and zebrafish rhodopsin. (C) Amino acid alignment of bovine 1D4 epitope sequence with the eight zebrafish cone opsin sequences of LWS1 (accession: Q9W6A7.1), LWS2 (Q8AYN0.1), RH2-1 (Q9W6A5.2), RH2-2 (Q8AYM8.1), RH2-3 (Q8AYM7.1), RH2-4 (Q9W6A6.2), SWS1 (Q9W6A9.2), and SWS2 (Q9W6A8.1). (D) Western blot analysis of monoclonal 1D4 and zpr-3 antibodies on adult zebrafish and bovine retina extracts. Three different amounts (5, 10, and 20 μg) of extracts are loaded.

Eight types of cone opsins expressed in four distinct zebrafish cone types share similarity with rhodopsin. ^23,24 We thus compared the bovine 1D4 epitope sequence to corresponding sequences of these eight cone opsins in zebrafish (Fig. 2C). This showed that red opsins LWS1 and LWS2, expressed in long double cones (red cones), have seven residues in common with the bovine epitope. The four RH2 green opsins, expressed in short double cones (green cones), share six while the blue opsin SWS2, expressed in long single cones (blue cones) shares six but has one amino acid residue missing at the C-terminus. The UV opsin SWS1, expressed in short single cones (UV cones) shares five identical amino acid residues with the bovine epitope. This reveals that the respective red and green opsin sequences in zebrafish are more similar to the 1D4 epitope than rhodopsin, thus opening the possibility that 1D4 may recognize cone opsins instead of rhodopsin. 

1D4 Antibody Labels Long Double Cone Outer Segments in Zebrafish But Not Green Cones

To determine which type of cone is detected by 1D4 in zebrafish, we used Bodipy ceramide (Molecular Probes/Life Technologies) to label all cell membranes in the adult eye. Rod and cone outer segments are composed of distinct membrane arrays and organized in a well-defined pattern in the retina, so that all outer segments can be efficiently discriminated by Bodipy ceramide staining. As shown in Figures 3A, 3B (first column), outer segments of rods, double cones, long single cones, and short single cones can be identified after staining with Bodipy TR (red color) and Bodipy FL (green color). We first analyzed Bodipy TR staining in the retina of adult Tg(Rho:EGFP) transgenic fish (Fig. 3A). As expected, EGFP is located in the rod outer segment layer confirming that the used zebrafish rhodopsin promoter drives GFP expression exclusively in rods. Next, we simultaneously stained wild type adult fish retinas with Bodipy FL and monoclonal 1D4 antibody. The 1D4 signal co-localized with green Bodipy signal in double cone outer segments (Supplementary Material, Fig. 3B; Fig. S1). This suggests that 1D4 binds to opsins in zebrafish double cones. For further confirmation that 1D4 stains double cones instead of rods, we next performed double staining of polyclonal 1D4 with zpr-1, which detects an unknown antigen specific to double cones ²⁶ (Fig. 3C). We found partially overlapping signals with both antibodies in the outer segments suggesting that 1D4 stains outer segments in a subset of double cones. 

Figure 3. 

 

1D4 stains outer segments of double cones in zebrafish. (A) Bodipy TR staining (red) of adult Tg(Rho:EGFP) transgenic fish retina. EGFP under control of rhodopsin promoter is expressed in rod outer segment layer. DAPI stains nuclear layers. (B) Bodipy FL (green) staining in wild-type adult fish retina. Monoclonal 1D4 signal is in double cone outer segment layer. (C) zpr-1 and polyclonal 1D4 immunohistochemistry in wild-type adult fish retina. (D) Monoclonal 1D4 immunohistochemistry in Tg(LCR:RH2-1:EYFP) transgenic fish retina. EYFP is expressed in cytoplasm of RH2-1 opsin expressing short double cones. 1D4 signal does not overlap with YFP in short double cone outer segments. CB, cell body; CNL, cone nuclear layer; DOS, double cone outer segment; LSCOS, long single cone outer segment; IF, inner fiber; INL, inner nuclear layer; IS, inner segment; ONL, outer nuclear layer; OS, outer segment; P, pedicle; ROS, rod outer segment; RNL, rod nuclear layer; SSCOS, short single cone outer segment.

There are two types of double cones, short double cones expressing RH2 green opsin and long double cones expressing LWS red opsin. These cones always appear in pairs and are located in the same retina layer, showing only subtle differences in length. In general, long double cones appeared slightly more extended than short double cones, but this difference was difficult to score reliably, even after Bodipy staining. 

Therefore, in order to determine which type of double cone is stained by 1D4, we used the Tg(LCR:RH2-1:EYFP) transgenic line, in which EYFP is driven by the green opsin RH2-1 promoter/long control region (LCR). ²¹ In this line, enhanced yellow fluorescent protein (EYFP) is exclusively expressed in the cytoplasm of RH2-1 positive short double cones, including outer and inner segments, cell bodies, inner fibers and pedicles (Fig. 3D). After immunostaining, we found that 1D4 signal did not overlap with EYFP in short double cone outer segments (Fig. 3D). Instead, it was located directly above the short double cone outer segments. This suggests that 1D4 does not label RH2-1 short double cone outer segments, but probably outer segments of long double cones. 

To directly determine which cones are identified by the 1D4 antibody, we stained adult retina sections with monoclonal 1D4 together with antibodies directed against UV, blue, green, and red opsin, which label short single cones, long single cones, short double cones and long double cones, respectively. ²⁵ This showed that 1D4 signal overlaps with red opsin in long double cones (Supplementary Material, Fig. 4A; Fig. S2A), while no overlap was observed for the other antibodies (Fig. 4B–D; Fig. S2B–D). 

Figure 4. 

 

Monoclonal 1D4 signal co-localizes with red opsin in long double cones. Immunostaining of 1D4 together with red opsin (A), green opsin (B), blue opsin (C), and UV opsin (D). (E) Control staining (secondary antibody only) for Alexa 568 (E; anti-rabbit), Alexa 488 (E′; anti-mouse) and overlay of Alexa 568, Alexa 488, and DAPI (E″).

zpr-3, An Alternative Antibody for Rod Staining in Zebrafish

As our data indicated that 1D4 does not stain rod outer segments, we searched for alternative rod markers in zebrafish. Zpr-3, a monoclonal antibody developed by ZIRC, University of Oregon, was originally obtained after immunization with total retina extracts. It is widely used as a rod photoreceptor marker in zebrafish. ^26,27 We used zpr-3 for immunostaining of retina sections from Tg(Rho:EGFP) transgenic fish. Zpr-3 overlapped with GFP in rod outer segments (Fig. 5A). However, it also stained a layer directly below the rod outer segment layer (Fig. 5A; marked by asterisk). To determine which other cell layers are stained by zpr-3, we double stained wild type fish retinas with zpr-3 antibody and Bodipy TR (Figs. 5B, 5C). This revealed that zpr-3 signal is found in the rod as well as double cone outer segment layer. We next stained retina sections from the Tg(LCR:RH2-1:EYFP) transgenic line and found partially overlapping signals for zpr-3 and YFP (Fig. 5D). All these results suggest that zpr-3, while detecting the rod outer segments, is not an exclusive rod marker. 

Figure 5. 

 

zpr-3 antibody stains rod outer segments as well as short double cones in zebrafish. (A) zpr-3 immunohistochemistry in adult Tg(Rho:EGFP) transgenic fish retina. zpr-3 signal is located in EGFP positive rod outer segment layer but also in the layer below (white bracket with asterisk). (B) Staining of wild-type adult fish retina with zpr-3 and Bodipy TR. zpr-3 signal is seen in rod and double cone outer segment layer. (C) Higher magnification view of images in B shows that zpr-3 signal is co-localized with Bodipy TR in double cone outer segment layer. (D) zpr-3 immunohistochemistry in adult Tg(LCR:RH2-1:EYFP) transgenic fish retina. zpr-3 signal overlaps with YFP in short double cone outer segments. CB, cell body; CNL, cone nuclear layer; DCOS, double cone outer segment; LSCOS, long single cone outer segment; IF, inner fiber; INL, inner nuclear layer; IS, inner segment; ONL, outer nuclear layer; OS, outer segment; P, pedicle; RNL, rod nuclear layer; ROS, rod outer segment; SSCOS, short single cone outer segment.

Discussion

The mouse monoclonal antibody 1D4 was originally developed by immunizing mice with bovine rod outer segment discs. ⁹ It binds to the C-terminal eight amino acids of rhodopsin and labels rod outer segments in mice, cows, and humans. ^15,16 Recently, two zebrafish reports used this antibody as a predicted rod outer segment marker for functional studies on retinitis pigmentosa (RP) related genes. In the first study, the protein ZFPRGP was examined in the presumptive rod cilium region using 1D4 as a co-localizing marker. ¹⁷ In the second report, authors used 1D4 staining to claim that rhodopsin expression was almost absent and rod numbers largely reduced in zebrafish larvae after znf-513 knock-down. ¹⁸  

In our study, we provide experimental evidence that the monoclonal 1D4 antibody does not recognize zebrafish rhodopsin and hence does not label rod outer segments. First, 1D4 did not detect rhodopsin in Western blots from zebrafish retinal extracts. Second, 1D4 also failed to stain rod outer segments in retinal cryosections from adult zebrafish. Instead, by using photoreceptor specific reporter transgenes, Bodipy ceramide and staining with a double cone marker as well as various opsin markers, we localized the 1D4 signal in the layer comprising long double cone outer segments. We did not find overlap of 1D4 staining with reporter expression in RH2-1 opsin positive short double cones. As the epitope sequence in RH2-1 is identical to sequences in other green opsins (RH2-2, RH2-3, and RH2-4), this strongly suggested that 1D4 binds to red opsins in the outer segment of long double cones. This was confirmed by double labeling of long double cones with a red opsin antibody and 1D4 (Fig. 4). 

Although such cross-reactivity with red opsin was not detectable in our Western blot analysis (Fig. 2), these data consistently show that 1D4 is unable to detect zebrafish rhodopsin. Thus, our data suggest that the previously reported zebrafish studies on ZFPRGP and znf-513 describe effects on double cones instead of rods as proposed by the authors. ^17,18 While the reported defects could be indicative for so-called bystander effects as described previously, ²⁸ it remains to be tested whether rods are indeed affected in these RP models. Interestingly, the study by Li et al. shows overlap of the staining pattern of 1D4 with that of an antibody directed against red opsin. This supports our findings that 1D4 labels long double cones expressing red opsin from at least 14 days post-fertilization (dpf)onwards. We have not shown that 1D4 recognizes long-wavelength double cones in larvae prior to two weeks of age. 

The different 1D4 binding characteristics can be explained by highly divergent sequences of the epitope in fish versus mammals. When the nine amino acid epitope sequence of 1D4 was compared to mammalian and zebrafish rhodospin sequences, a five-residues insertion was evident and three out of nine residues varied, making it very unlikely that 1D4 detects zebrafish rhodopsin (Fig. 2A). In addition, Xenopus rhodopsin, which shares the last 14 residues with zebrafish has been proposed to be not recognized by 1D4. ²⁹ On the other hand, the 1D4 epitope shares seven out of nine residues with the corresponding sequence in two zebrafish red opsins (LWS1 and LWS2) expressed in long double cones, which makes it possible that 1D4 detects red opsins. Consistent with the idea that 1D4 related epitopes could be present in several opsins, a polyclonal serum (Abcam 3424) obtained after immunization with corresponding peptide labeled more cells in the zebrafish retina than monoclonal 1D4 (Fig. 2A). Importantly, partial overlap was observed between cells labeled with polyclonal 1D4 serum and zpr-1 (Fig. 3C). Phylogenetic studies had suggested that the four cone opsin genes arose from a single ancestral opsin gene through gene duplications with all four cone opsins evolving before rod-specific rhodopsin. ^30–32 Overall, sequence variations in opsins between mammals and teleost fish are not surprising given the rapid evolution of vertebrate opsin as well as other phototransduction sequences and the fact that fish have adapted to different environments. ^30,33  

Our analyses were performed in 14 dpf larval as well as adult transgenic fish retinas, which allowed a much better discrimination of different photoreceptor types than in embryonic and early larval stages as in previously described morpholino knock-down studies. There, a morphologic distinction of individual photoreceptor types is difficult, as photoreceptors have not yet developed their matured shape and different retinal layers are densely packed. In our studies, sequence analysis together with marker stainings and morphologic analysis in rod and double cone transgenic fish strongly suggests that 1D4 labels the outer segments of long double cones and not rods as reported earlier. 

The zpr-3 monoclonal antibody has been used as alternative marker for zebrafish rods in the past. ^26,27,34 Our data confirm that zpr-3 stains rod outer segments; however, we also found evidence for staining in double cone outer segments, which needs to be taken into account for interpretation of past and future data. As the demand for antibodies is rising in the zebrafish community, it is important to determine the epitope of antibodies generated against proteins of other species, especially when proteins of rapidly evolving gene families such as the opsins are targeted. 

Supplementary Materials

pdf S1 

pdf S2 

Acknowledgments

We thank David R. Hyde (University of Notre Dame) for providing zebrafish opsin antibodies. We thank Thuy To for helpful discussions and the NUS-DBS Confocal Imaging Unit for constant support. 

Footnotes

 Supported by AcRF Tier-2 Grant T207B3107 from the Ministry of Education, Singapore, to CW and by SFB581 and RVZ-network grants from the Deutsche Forschungsgemeinschaft (DFG) to UF.

Footnotes

 Disclosure: J. Yin, None; J. Brocher, None; B. Linder, None; A. Hirmer, None; H. Sundaramurthi, None; U. Fischer, None; C. Winkler, None