We recently identified a novel retinoid binding protein, PapilioRBP, in the soluble fraction of the eye homogenate of the butterfly Papilio xuthus, and demonstrated that the protein is involved in the visual cycle. We now have localized the protein in the Papilio eye by light and electron microscopic immunohistochemistry using a monospecific antiserum produced against artificially expressed Papilio RBP. We found strong immunoreactivity in the primary as well as secondary pigment cells and in the tracheal cells. The pigment cells have long been regarded as an important site of the visual cycle, and this view is further supported by the present result. Interestingly, the cytoplasm and nuclei of these cells were equally labeled, indicating that the protein exists in both the cytoplasm and the nucleus. We conducted a survey for the existence of the Papilio RBP-like proteins in other insects including several species of butterflies, dragonflies, cicadas, grasshoppers and honeybees. Anti-Papilio RBP immunoreactivity was confirmed in the proteins isolated only from butterflies belonging to the superfamily Papilionoidea and not from other species. In all insects tested, however, fluorescing proteins were clearly detected, suggesting that these insects also have similar retinol-binding proteins.


A visual pigment molecule is composed of a protein opsin with an 11-cis retinal attached to it. The 11-cis retinal absorbs light and changes its configuration to the all-trans form, which subsequently causes a conformational change of the protein into an active form, metarhodopsin. The activated visual pigment triggers the phototransduction cascade, whose end result is the change in membrane potential.

Visual pigments will be depleted by prolonged illumination unless they are regenerated and/or newly synthesized. A key process of the visual pigment regeneration, the visual cycle, is to generate 11-cis retinal from all-trans retinal, retinol and other retinoids. As a result of extensive studies of this process in recent decades, a number of retinoid binding proteins have been identified in both invertebrates and vertebrates(Hara and Hara, 1991; McBee et al., 2001; Ozaki et al., 1987; Saari, 1999; Stavenga et al., 1991). We recently identified a retinol binding protein, Papilio RBP, whose ligand is 3-hydroxyretinol, in the compound eye of the Japanese yellow swallowtail butterfly Papilio xuthus(Wakakuwa et al., 2003). The Papilio RBP is a novel protein, for it has little homology with any other retinoid binding proteins so far reported. We found that the isomer composition of the ligand of Papilio RBP changes between light- and dark-adaptation: the content of 11-cis isoform increases in eyes in the light. In addition, illumination significantly increases the amount of 11-cis isoform, especially in the distal part of the retinal layer(Wakakuwa et al., 2003).

This observation fully agrees with the results described in a previous report, where retinoid composition was measured under various conditions of adaptation (Shimazaki and Eguchi,1995). We thus concluded that the protein is involved in the regeneration of Papilio visual pigment whose chromophore is 11-cis 3-hydroxyretinal (Wakakuwa et al., 2003).

Where in the eye does the Papilio RBP function? Is this novel binding protein strictly specific to Papilio, or shared by other species? To answer these questions, we raised a specific antiserum against Papilio RBP. We then carried out light and electron microscopic immunohistochemistry in the Papilio retina to localize the protein in the retinal tissue. We also performed combined native PAGE and immunoblot analyses on several other insect species to evaluate the distribution of Papilio RBP-like protein among insects.

Materials and methods


Japanese yellow swallowtail butterfly Papilio xuthus L., was taken from a cultured population. The population was derived from eggs laid by females caught in the field. The hatched larvae were fed with fresh citrus leaves at 25°C under a light regime of 10 h:14 h light:dark. The pupae were stored at 4°C for at least 3 months and then allowed to emerge at 25°C.

For the comparative study, we used four papilionid species, Papilio machaon, Papilio protenor, Papilio helenus and Graphium sarpedon, a nymphalid Vanessa indica, a pierid Pieris rapae, and a hesperid Parnara guttata, captured in the field around the campus of Yokohama City University. We also investigated a dragonfly Orthetrum albistylum speciosum (Odonata, Libellulidae), a cicada Graptopsaltria nigrofuscata (Hemiptera, Cicadidae), a locust Oedaleus infemalis (Orthoptera, Acrididae) and a honeybee Apis mellifera (Hymenoptera, Apidae). These insects were captured in the field around the campus of Yokohama City University, except for Apis mellifera, which was taken from a hive culture.

Gel electrophoresis

For native PAGE, Laemmli's buffer system was used(Laemmli, 1970) but without SDS and 2-mercaptoethanol in the gel, running and sample buffer solutions. Unless otherwise stated, whole compound eyes were used for extraction. The compound eyes were homogenized in 63 mmol l–1 Tris-Cl buffer (pH 6.8),and the homogenate was centrifuged at 15 000 g for 30 min at 4°C. The supernatant was loaded onto a 10% polyacrylamide gel, and soluble proteins were electrophoretically separated. After electrophoresis, the gel was illuminated with UV light, which visualizes the RBP as a single fluorescing band. The RBPs were recovered from cut pieces of the gel containing the fluorescing band (purified RBP). When necessary, the gel was stained with Coomassie Brilliant Blue (CBB). Regular SDS-PAGE was also carried out using a 12% polyacrylamide gel(Laemmli, 1970). The gel was then stained with CBB.

Antiserum production

To produce antigen by expression, we first carried out overexpression of Papilio RBP. We prepared a pair of oligo nucleotide primer(ROLBP-forward, 5′-GTGAAGACATATGTCTTCACGAATATATCC-3′;ROLBP-reverse, 5′-GAACTCGAGTTCAACTTTTGCCCCAAATATTTTG-3′) based on the full-length cDNA sequence of Papilio RBP(Wakakuwa et al., 2003). Using these primers, the entire coding region of Papilio RBP (714 bp) was amplified. We subcloned the polymerase chain reaction (PCR) product into the pET-21 expression vector; the vector was designed to enable recombinant expression of the Papilio RBP as a fusion protein to which a 6× His-tag was added at the C-terminal end for one-step purification by nickel chelate affinity chromatography.

Female Wistar rats were immunized intradermally with 0.1 mg of purified recombinant Papilio RBP in 200 μl of phosphate-buffered saline emulsified 1:1 with Freund's complete adjuvant (Difco Laboratories, Detroit,USA). The rats were boosted every 2 weeks in a similar manner using Papilio RBP in incomplete Freund's adjuvant (Difco Laboratories): the rats were injected antigen six times in total. Immune serum was obtained 7 days after the final boost. Monospecificity of the antiserum was confirmed by immunoblot analysis (see below).


Water-soluble extracts of retinal homogenates or the RBP purified from the native PAGE gel were separated by 12% SDS-PAGE. The protein samples were then blotted onto a polyvinylidene difluoride (PVDF) membrane. Nonspecific binding sites were blocked with 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS), followed by overnight incubation with the antiserum. After washing with PBS, the PVDF membrane was incubated with alkaline phosphatase-conjugated secondary antibody. After washing with PBS, the PVDF membrane was incubated in 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT) in alkaline phosphatase buffer (100 mmol l–1 Tris-HCl, pH 9.5, 100 mmol l–1 NaCl and 5 mmol l–1 MgCl2), until adequate stain intensity was obtained.


For light microscopic immunohistochemistry, isolated compound eyes were fixed in 2% paraformaldehyde and 0.2% picric acid in 0.1 mol l–1 phosphate buffer, pH 7.4 (PB) for 30 min at room temperature. After a brief wash with 0.1 mol l–1 PB, the eyes were then dehydrated in a graded ethanol series, infiltrated with xylene and embedded in paraffin. Thin sections (8 μm) mounted on slides were incubated in 0.3% H2O2 in water for 5 min to quench endogenous peroxidase activity (Larsson,1988). Non-specific binding sites for antibodies were blocked by treating the sections with 10% normal goat serum in PBS for 30 min, and then the sections were incubated in the anti-Papilio RBP in 1% BSA in PBS overnight at 4°C. The sections were subsequently reacted with biotinylated secondary antibody for 30 min, and further incubated with Vectastain ABC Reagent (Vector Laboratories, Burlingame, USA). After washing with PBS, the sections were incubated in peroxidase substrate solution (0.2%3,3′-diaminobenzidine in 50 mmol l–1 Tris-HCl, pH 7.4)until adequate stain intensity was obtained.

For electron microscopic immunohistochemistry, isolated eyes were prefixed by immersing in 2% paraformaldehyde and 2% glutaraldehyde in 0.1 mol l–1 cacodylate buffer, pH 7.3 (CB) for 30 min on ice. After a 10 min wash with 0.1 mol l–1 CB, the eyes were postfixed in 2% OsO4 in 0.1 mol l–1 CB for 30 min on ice. After dehydration with graded methanol series, the eyes were embedded in LR White resin. Ultrathin sections, mounted on nickel grids, were treated with 4% BSA in PBSG (0.25% fish gelatin in PBS) for 20 min to block non-specific antibody binding sites, and then incubated with anti-Papilio RBP in PBSG overnight at 4°C. After washing with PBSG, the sections were reacted with secondary antibody-conjugated 15 nm colloidal gold particles for 1 h. After washing with PBSG, the sections were stained with 4% uranyl acetate in distilled water, and observed in a transmission electron microscope (JEM 1200EX; Tokyo, Japan).


Antibody specificity (immunoblot)

Monospecificity of the produced polyclonal antiserum against purified recombinant Papilio RBP was confirmed by immunoblot analysis. Fig. 1 shows the results of SDS-PAGE and the immunoblot. Immunoblot labeled with the antiserum revealed a single band of about 31 kDa, in both the retinal homogenate and the purified Papilio RBP, which was immunoreactive to the antiserum. The band was not detected by non-immune serum (data not shown). These results indicate that the antiserum detects Papilio RBP monospecifically. We therefore refer to the antiserum as anti-Papilio RBP.

Localization of Papilio RBP in the eye of Papilio xuthus

We studied the possible localization of Papilio RBP viathe distribution of anti-Papilio RBP immunoreactivity in the compound eye (Fig. 2). An ommatidium of a Papilio eye consists of a dioptric apparatus, the corneal facet lens and crystalline cone, and a retinula, containing nine photoreceptor cells(Fig. 2A). The photoreceptor cells together construct a photoreceptive rhabdom, which is a long, slender cylinder in the center of the ommatidium. Neighboring ommatidia are optically separated by the primary and secondary pigment cells, both containing dark-brown pigment granules. The primary pigment cells wrap the crystalline cones, whereas the secondary pigment cells are located between the ommatidia along the entire length of the retinal layer. We found that anti-Papilio RBP labeled these pigment cells but not the photoreceptor cells (Fig. 2B,C). The anti-Papilio RBP strongly labeled the cytoplasm of the tracheal cells, proximal of the basement membrane, forming the fenestrated layer between the retina and the lamina, i.e. the first optic ganglion (Fig. 3D). Non-immune serum gave no labeling (Fig. 3Binset).

The subcellular localization of Papilio RBP was further studied by electron microscopic immunohistochemistry. The results fully confirmed the light microscopical findings. Fig. 3 shows five pairs of electron micrographs, each consisting of one at low magnification and one at high magnification. In the retinal layer the gold particles, which indicate the localization of anti-Papilio RBP,were exclusively found in the pigment cells and not in the photoreceptor cells(Fig. 3A–H). Interestingly, the anti-Papilio RBP labeled the nuclei as well as the cytoplasm of the pigment cells at similar density(Fig. 3E–H). Electron microscopy revealed that the strongly labeled structures proximal to the basement membrane (Fig. 2D) are the nuclei and the cytoplasm of the tracheal cells(Fig. 3I,J). The photoreceptor axons, which pass through the basement membrane, were not labeled.

We quantified the labeling density in different regions of the eye(Fig. 4). The labeling of the pigment cells is significantly higher than the background labeling of the photoreceptor cells (*, P<0.01, one-way ANOVA, Tukey test). In the primary and the secondary pigment cells, both nuclei and the cytoplasm were equally labeled. The strongest labeling was found in the tracheal cells, where the calculated particle density was even higher in the nucleus than in the cytoplasm. This is in fact due to the uneven labeling pattern in the cytoplasm: labeling was rather confined to the region where more electron-dense materials are concentrated (data not shown). Local labeling densities in the nucleus and the cytoplasm were comparable (see for example Fig. 3J). Although the labeling density is higher in tracheal cells than in all other cells observed(Fig. 4), this does not contradict our previous results that the Papilio RBP is mainly distributed in the distal retinal layer(Wakakuwa et al., 2003): the total volume of the pigment cells is overwhelmingly large compared to that of the tracheal cells.

Papilio RBP-like proteins in other insect species

Fig. 5 shows the results of the native PAGE analysis of crude retinal extracts from seven butterfly species belonging to the superfamily Papilionoidea. We identified a single band emitting whitish fluorescence under UV illumination in all species tested(Fig. 5). The mobility of the fluorescing proteins in the native gel seems genus dependent. Mobilities in species of the genus Papilio are similar, whereas those of Graphium (Papilionidae), Vanessa (Nymphalidae) and Pieris (Pieridae) are distinctly lower.

Fig. 6 shows results of immunoblot analysis of these proteins. We first cut out the gel piece containing the fluorescing band. The proteins were then extracted from the gel pieces and separated by SDS-PAGE. The anti-Papilio RBP revealed a single band in all species. In the Papilionid species the molecular mass of these proteins was about 31 kDa and, in Vanessa and Pierisit was 25 kDa. Although in Vanessa three bands were evident in the SDS-PAGE, the anti-Papilio RBP detected only one of them at 25 kDa.

We also searched for Papilio RBP-like proteins in the eye of insects other than Papilionoidea, namely a skipper Parnara guttata(Lepidoptera, Hesperiidae), a dragonfly Orthetrum albistylum speciosum (Odonata, Libellulidae), a cicada Graptopsaltria nigrofuscata (Hemiptera, Cicadidae) and a grasshopper Oedaleus infemalis (Orthoptera, Acrididae). We identified a fluorescing band in all of these species under UV illumination on native PAGE(Fig. 7). However, none of them cross-reacted with the anti-Papilio RBP on immunoblots. We also checked whether the honeybee Apis mellifera (Hymenoptera, Apidae) has a corresponding protein, but we could not detect any fluorescing band on the native gel and any protein reacting with the anti-Papilio RBP on the immunoblot (data not shown).

Localization of Papilio RBP-like protein in the eye of Pieris rapae

Fig. 8 shows the localization of the anti-Papilio RBP imunoreactivity in the eye of Pieris rapae. As in Papilio xuthus, specific labeling was confined to the primary pigment cells, the secondary pigment cells and the tracheal cells proximal to the basement membrane. No labeling was detected in the photoreceptor cells.


Localization and function of Papilio RBP in the eye of butterflies

We found a strong Papilio RBP-like immunoreactivity in the retinal primary and secondary pigment cells (Fig. 2). This finding supports the presently accepted view that the primary pigment cells are somehow involved in the visual cycle (Schwemer, 1989, 1993; Smith and Goldsmith, 1991). Of course, so far we have no direct evidence to indicate that the primary pigment cells are the site where the visual cycle is taking place in Papilio. However, the distal portion of the Papilio retina functions in the light-dependent oxidation of 3-hydroxyretinol and isomerization of 3-hydroxyretinal (Shimazaki and Eguchi,1993). We have also demonstrated that Papilio RBP is involved in retinoid metabolism in the distal portion of the eye(Wakakuwa et al., 2003). Taken together, the primary as well as the secondary pigment cells are probably also involved in the visual cycle in the Papilio eye. Identification and localization of respective enzymes for the visual cycle, as in honeybees(Smith and Goldsmith, 1991),would provide further support for the view.

The principal function of the pigment cells is to optically isolate the ommatidia by absorbing off-axis incident light, so to optimize spatial resolving power. The light-driven isomerase, which may exist in the primary pigment cells (e.g. Smith and Goldsmith,1991), therefore, occupies an ideal location for receiving light,so to serve the visual pigment regeneration cycle. The secondary pigment cells are elongated in shape, lying between photoreceptor cells along the entire length of the retinal layer (Fig. 2). Assuming that these cells are also involved in the visual cycle, a possible function of the Papilio RBP then is to remove retinoid in the all-trans form from the photoreceptors and/or to supply retinoid in the 11-cis form back to the photoreceptors, to replenish visual pigment molecules.

The function of Papilio RBP in the tracheal cells is difficult to understand at present. Of course, the tracheae are not restricted to the eye,but exist throughout the body. We carried out a preliminary immunoblot analysis on the abdominal tracheae and found slight immunoreactivity (data not shown): maybe the Papilio RBP has some function specific to the tracheal system, or it may function in transporting retinoids from and to the haemolymph in turnover and de novo synthesis.

Even more conspicuously, Papilio RBP-like immunoreactivity was found in the nuclei of both the pigment cells and the tracheal cells. To the best of our knowledge, this is the first example of retinoid-binding proteins in cell nuclei.

Comparative aspects

Is Papilio RBP specific to Papilio, or is it shared by other species? This point is particularly important for elucidating a general scheme of the function of Papilio RBP in the visual cycle. With this question in mind, we carried out a comparative biochemical and immunohistochemical analysis in several other insect species.

In native PAGE, we detected a fluorescing band in all tested butterfly species. According to Seki et al.(1987), butterflies use 3-hydroxyretinal as the visual pigment chromophore and most of them also contain an excess amount of 3-hydroxyretinol. Presumably therefore, the fluorescing substance in tested butterfly species is 3-hydroxyretinol. The mobility of the fluorescing proteins in the native gel differs considerably,indicating that their surface charge, size and/or the three-dimensional structure rather vary. The proteins with lower mobilities may exist in certain polymerized forms. In addition to the fluorescence, the proteins were found to be immunoreactive to the anti-Papilio RBP in all butterflies, except for the skipper Parnara guttata (Hesperiidae). The molecular mass of the protein revealed by SDS-PAGE appeared to be similar in seven other butterflies (25–31 kDa; Fig. 6). In Vanessa indica, however, three bands were evident on SDS-PAGE (Fig. 6), probably due to overlap of three proteins in the location of the fluorescing protein(Fig. 5). Immunohistochemical localization of the Papilio RBP-like protein in the Pieriseye, using the anti-Papilio RBP, revealed that the distribution pattern is similar in Pieris and Papilio, indicating a similar function in both species (Figs 2 and 8).

Our data on the Papilio RBP-like protein in other insects is still preliminary. For example although we found a fluorescing protein band in the gel (Fig. 7), we have not yet identified the fluorescing materials themselves. These insects have actually been shown to have significant amounts of retinol and/or 3-hydroxyretinol in the eye (Seki et al., 1987, 1989). Therefore the proteins detected here probably have the ability to bind retinols, although they do not bind to the anti-Papilio RBP.

We conclude that the Papilio RBP-like protein is shared by butterflies belonging to the superfamily Papilionoidea. In addition, native PAGE indicated that other insects possess their own RBPs. These findings suggest that insects may share a basic pathway of visual cycle to regenerate rhodopsin. Nevertheless, the molecular characteristics of RBP could not be identical between species (or at least beween genuses) (Figs 5 and 7). Such variability in the protein properties may reflect structural variability in the rhabdom among species, because the visual cycle involves the removal and incorporation of visual pigments from and to the rhabdom. In fact many butterflies have apposition eyes with photoreceptors with small rhabdoms, whereas the eyes of nocturnal species are of the superposition type containing photoreceptors with large rhabdoms (Eguchi, 1978). Different rhabdoms would require a somewhat different mechanism for removal and incorporation of the visual pigment molecules. At any rate, elucidation of the function of the Papilio RBP and Papilio RBP-like proteins in the visual cycle requires further study, including immunohistochemical localization and the uncovering of other enzymes involved in the process.


We thank to Dr Doekele G. Stavenga for critical reading of the manuscript. This work was supported by the Sasagawa Grant for Scientific Research from the Japan Science Society to M.W., Grants-in-Aid for Scientific Research from Japan Society for the Promotion of Science (JSPS) to K.A. and to K.O.


Eguchi, E. (
). Comparative fine structure of lepidopteran compound eyes, especially skippers (Hesperioidea).
Zool. Mag.
Hara, T. and Hara, R. (
). Retinal-binding protein: function in a chromophore exchange system in the squid visual cell. In
Progress in Retinal Research
, vol.
(ed. N. Osborne and G. J. Chader), pp.
-206. Oxford-New York: Pergamon Press.
Laemmli, U. K. (
). Cleavage of structural proteins during the assembly of the head of bacteriophage T4.
Larsson, L. (
Immunocytochemistry: Theory and Practice
. Boca Raton-Florida: CRC Press.
McBee, J. K., Palczewski, K., Baehr, W. and Pepperberg, D. R. (
). Confronting complexity: the interlink of phototransduction and retinoid metabolism in the vertebrate retina.
Prog. Ret. Eye Res.
Ozaki, K., Terakita, A., Hara, R. and Hara, T.(
). Isolation and characterization of a retinal-binding protein from the squid retina.
Vision Res.
Saari, J. C. (
). Retinoids in mammalian vision. In
Retinoids: The Biochemical and Molecular Basis of Vitamin A and Retinoids Action
, vol.
(ed. H. Nau and W. S. Blaner), pp.
-588. Berlin:Springer-Verlag.
Schwemer, J. (
). Visual pigments of compound eyes–structure, photochemistry, and regeneration. In
Facets of Vision
(ed. D. G. Stavenga and R. C. Hardie), pp.
-133. Berlin-Heidelberg:Springer-Verlag.
Schwemer, J. (
). Visual pigment renewal and the cycle of the chromohphore in the compound eye of the blowfly. In
Sensory Systems of Arthropods
(ed. K. Wiese and F. G. Gribakin). Berlin-Heidelberg: Springer-Verlag.
Seki, T., Fujishita, S., Ito, M., Matsuoka, N. and Tsukida,K. (
). Retinoid composition in the compound eyes of insects.
Exp. Biol.
Seki, T., Fujishita, S. and Obana, S. (
). Composition and distribution of retinal and 3-hydroxyretinal in the compound eye of the dragonfly.
Exp. Biol.
Shimazaki, Y. and Eguchi, E. (
). Synthesis of 3-hydroxyretinal in the cytosol of the butterfly compound eye.
Vision Res.
Shimazaki, Y. and Eguchi, E. (
). Light-dependent metabolic pathway of 3-hydroxyretinoids in the eye of a butterfly, Papilio xuthus.
J. Comp. Physiol. A
Smith, W. C. and Goldsmith, T. H. (
). Localization of retinal photoisomerase in the compound eye of the honeybee.
Vis. Neurosci.
Stavenga, D. G., Schwemer, J. and Hellingwerf, K. J.(
). Visual pigments, bacterial rhodopsins, and related retinoid-binding proteins. In
Photoreceptor Evolution and Function
, vol.
(ed. M. G. Holmes), pp.
-349. London: Academic Press.
Wakakuwa, M., Arikawa, K. and Ozaki, K. (
). A novel retinol-binding protein in the retina of the swallowtail butterfly, Papilio xuthus.
Eur. J. Biochem.