Animals

Nautilus pompilius were captured at a depth of about 500 m off the main reef at Suva on the island of Fiji. They were adapted overnight to darkness at the University of the South Pacific, and their excised heads, kept in the dark on dry-ice, were sent to Osaka by air, and stored at −25°C until use.

Preparation of rhodopsin and retinochrome

All the operations for preparing photopigment extracts were carried out under a deep red light at temperatures as low as 4°C. About 40 frozen eyes were used for each preparation. They were hemisected to remove the anterior part, and retinal tissues were collected from the eyecoups by scraping. The tissues were homogenized with a teflon homogenizer in 67 mM phosphate buffer, pH 6.5, diluted with the same buffer, and centrifuged at 15,000 rpm for 30 min to yield dark tissue fragments.

The fragments were suspended in a 36% solution of sucrose in phosphate buffer, and centrifuged at 12,000 rpm for 20 min to float the rhabdomal portions of the visual cells. The supernatant was then diluted 2-fold with phosphate buffer, and centrifuged to collect a slightly dark residue. It was washed more than three times with 0.1 M Na₂HPO₄, pH 9.2, to remove unwanted black pigments, and further washed with phosphate buffer to obtain a microvillus-containing fraction (MV-fraction) which was used to extract rhodopsin. This fraction was finally extracted by gentle shaking with 2% digitonin for 20 min.

The black debris which sedimented in 36% sucrose was used to prepare retinochrome. For this preparation, it was most important to remove the remaining microvilli and black pigments to the greatest degree possible. The debris was suspended in 60% sucrose containing phosphate buffer, and centrifuged at 12,000 rpm for 20 min to discard the floated microvilli. After three such flotations, the dark precipitates were resuspended in 75% sucrose, and centrifuged to obtain a supernatant. The supernatant was withdrawn, diluted 3-fold with phosphate buffer and centrifuged. In order to remove the large amount of black pigment, the precipitate was washed once with 0.5% Na₂CO₃ by stirring and centrifuging, and washed with water as many times as necessary until susccessive supernatants were colorless. The residue was immersed in Weber-Edsall's solution for 30 min at 25°C to extrude soluble proteins, washed with water, and lyophilized. The dried material was treated with three successive portions of petroleum ether to remove lipids, and washed with buffer to obtain a myeloid body-containing fraction (MB-fraction), and mixed with 2% digitonin for extracting retinochrome.

Irradiation and spectrophotometry

A 100 W tungsten filament projector lamp shielded by an 8 cm water cell was used for a light source, and irradiation of photopigment samples was carried out at 0°C, unless otherwise specified. Orange lights (>560 and >540 nm) were created by using Toshiba V-O 56 and V-O 54 cutoff glass filters (which transmit about 50% at 560 or 540 nm and about 90% at wavelengths exceeding 600 or 580 nm), respectively, while blue light (450 nm) was isolated using a combination of Toshiba KL-45 interference and Toshiba V-Y 43 cutoff glass filters. Absorption measurements were performed with a Hitachi model 320 or 323 recording spectrophotometer.

Analysis of retinal isomers by HPLC

In order to analyze geometrical isomers of retinal in photopigment samples, high-performance liquid chromatography (HPLC) was carried out using a Hitachi model 635 liquid chromatography system according to the method described previously [26]. Absorbance at 340 nm (A_340nm) was monitored with a JASCO UVIDEC 100-V detector, and the peak area was determined by integrating the absorbance. Each isomer was quantified from the peak area by the use of the following values for extinction coefficients: all-trans syn = 51,000, all-trans anti = 43,000, 13-cis syn = 51,000, 13-cis anti = 47,500, 11-cis syn = 34,300, 11-cis anti = 26,600, and 9-cis syn and anti = 52,000 (provisional value derived from 9-cis syn and anti mixture).

SDS-polyacrylamide gel electrophoresis (SDS-PAGE)

SDS-PAGE was performed to estimate the molecular weight of the photopigments using the following method. The membrane fraction containing rhodopsin or retinochrome (MV- and MB-fraction) was mixed with 200 μl of 1 N HCl and 2 mg of borane dimethylamine [15], and incubated at 4°C for 20 min. The mixture was neutralized with 0.1 N NaOH, and the membranes were washed with phosphate buffer and dissolved in SDS sample buffer. Ali-quots were subjected to electrophoresis on polyacrylamide slab gels (7.5% gel used for rhodopsin and 12% gel for retinochrome) with protein standards (Bio-Rad Instruments).

Photoreaction of nautilus rhodopsin

The absorption spectrum of a digitonin extract from the MV-fraction at pH 6.5 is shown in Figure 1. The peak of the visual absorption appeared at 465 nm (spectrum 1), and on the long-wave side, the absorbance fell to zero near 560 nm. When the extract was irradiated with blue light for 15 min, the peak was shifted markedly toward longer wavelengths, accompanied by an increase in absorbance (spectrum 2). On further irradiation with orange light for 10 min, spectrum 2 was changed into spectrum 3 (broken-like spectrum). Such a conversion between spectra 2 and 3 was photoreversible. The corresponding changes in the isomeric form of the retinal chromophore with this extract are presented in Table 1. In the original (dark), 11-cis-retinal was dominant (98%), indicating that most of the photopigment was rhodopsin. On irradiation with blue light, the 11-cis-retinal decreased to 36% and all-trans isomer increased to 62%, showing the formation of a photoequilibrium mixture of rhodopsin and metarhodopsin. On further irradiation with orange light, the proportion of 11-cis isomer increased from 36% to 79%, indicating that rhodopsin was markedly regenerated from metarhodopsin by light (photoreversal). At the same time, 9-cis isomer increased slightly, suggesting a formation of isorhodopsin in spectrum 3. Actually, spectrum 3 overlapped extensively with spectrum 1, probably because isorhodopsin resembles rhodopsin in the shape of spectrum.

Fig. 1

Photochemical properties of rhodopsin extracted from the MV-fraction of a Nautilus retina. See text for explanation.

Table 1

Configurational change in the retinal chromophore on irradiation of an extract of the MV-fraction (rhodopsin)

Effect of hydroxylamine on rhodopsin and metarhodopsin

The following experiment was aimed at examining the stability of rhodopsin and metarhodopsin in the presence of hydroxylamine (NH₂OH), and at determining their abosrption maxima (λ_max). As shown in Figure 2, rhodopsin (spectrum 1, pH 6.5) was irradiated with blue light for 15 min to produce a photoequilibrium mixture of rhodopsin and metarhodopsin (spectrum 2). When 1 ml of it was mixed with 50 μl of freshly neutralized 2 M NH₂OH (final concentration, ∼100 mM; spectrum 2′ corrected for dilution) and incubated in the dark at 16°C, the absorbance gradually decreased in the visible and increased in the NUV range, as shown by spectra 3 and 4 measured after 1 and 15 min, respectively. This reaction of metarhodopsin to all-trans-retinaloxime was complete after 120 min (spectrum 5), leaving rhodopsin which was stable in the presence of 100 mM NH₂OH. As shown by the difference spectrum between spectra 2′ and 5 (spectrum 7), the λ_max of metarhodopsin lies at 510 nm. The remaining rhodopsin was then irradiatd with blue light at 25°C for 120 min until the photoproduct metarhodopsin was entirely destroyed by the NH₂OH which was present (spectrum 6). The difference spectrum before and after irradiation (spectra 5–6) showed that the λ_max of rhodopsin lies at 465 nm (spectrum 8). This values of λ_max was consistent with that of the original rhodopsin extract shown in spectrum 1. Estimating from the results shown in Figure 2, the extinction coefficient at λ_max of metarhodopsin is about 1.7 times higher than that of rhodopsin (the ratio of A_510nm between spectra 2 and 5 to A_465nm between spectra 1 and 5).

Fig. 2

Effect of hydroxylamine on rhodopsin and metarhodopsin. Since spectra 3 to 6 were derived from the extract diluted with NH₂OH, spectrum 2 was corrected for comparison (broken-line curve, spectrum 2′). See text for further explanation.

Photoreaction of nautilus retinochrome

The absorption spectrum of a digitonin extract from the MB-fraction at pH 6.5 is presented in Figure 3. This extract showed an absorption peak around 500 nm (spectrum 1), and the pigment greatly differed from rhodopsin in photolytic behavior in the following manner. When irradiated with orange light, it was readily bleaced (∼85% for 5 min) so that spectrum 1 was changed into spectrum 2 after 15 min. The λ_max of the difference spectrum before and after bleaching was at 513 nm (spectrum 3). The corresponding changes in the isomeric form of the retinal chromophore are presented in Table 2. In the original (dark), all-trans-retinal was dominant (71%), suggesting that most of the photopigment was retinochrome. On irradiation with orange light, the all-trans isomer decreased to 19% and a large amount of 11-cis isomer appeared (78%), suggesting the conversion of retinochrome to metaretinochrome (or more specifically, a photoequilibrium mixture of retinochrome and metaretinochrome). The original extract (dark) also showed a fairly high content of 11-cis-retinal (26%), meaning a contamination by rhodopsin. However, since rhodopsin can not absorb the orange light (cf. spectrum 1 in Fig. 1), there occurs no photoreaction or isomerization of rhodopsin in this experiment. Accordingly, after the photobleaching of retinochrome, the net molar ratio of retinochrome to metaretinochrome is estimated to be 19: 52 (78–26). In any case, even though rhodopsin is contained in the extract, the spectral change of retinochrome shown in Figure 3 is scarcely affected by the presence of rhodopsin.

Fig. 3

Photochemical properties of retinochrome extracted from the MB-fraction of a Nautilus retina. See text for explanation.

Table 2

Configurational change in the retinal chromophore on irradiation of an extract of the MB-fraction (retinochrome)

Effect of hydroxylamine on retinochrome and metaretinochrome

Figure 4 shows the effect of NH₂OH on the decomposition of retinochrome and metaretinochrome. In the experiment shown in Figure 4A, 1 ml of retinochrome (spectrum 1) was mixed with 10 μl of 2 M NH₂OH (final concentration, ∼20 mM; both rhodopsin and metarhodopsin are stable in 20 mM NH₂OH), and incubated in the dark at 18°C. As shwon by the spectrum measured after 12 min (spectrum 2), retinochrome was only slightly decomposed into all-trans-retinaloxime, increasing in the NUV absorption. When irradiated with orange light for 10 min, all remaining retinochrome (spectrum 2) was bleached to spectrum 3, suggesting that metaretinochrome is readily decomposed by NH₂OH. In another experiment shown in Figure 4B, the same retinochrome (spectrum 1) was converted to metareti nochrome by irradiation (spectrum 2) and then mixed with ∼20 mM NH₂OH in the dark. The metaretinochrome was completely decomposed into 11-cis-retinaloxime within 3 min (spectrum 3), proving again that it is far more susceptible to NH₂OH than retinochrome. In addition, the difference spectrum between spectra 1 and 3 in Figure 4A or 4B revealed that retinochrome has a λ_max at 510 nm.

Fig. 4

Effect of hydroxylamine on retinochrome and metaretinochrome. Addition of NH₂OH to retinochrome (A) and metaretinochrome (B). See text for explanation.

Dark regeneration of retinochrome from metaretinochrome

In the beginning, it was examined whether or not retinochrome extract contains aporetinochrome (chromophore-free retinochrome) which may be liable to react with retinal. When 0.5 ml of retinochrome (spectrum 1 in Fig. 5A) was kept in the dark with 15 μl of all-trans-retinal (A_390nm 2.88) for 20 min at 18°C, there was no marked increase in absorbance around 500 nm (spectrum 2). Spectrum 2 showed an absorption due to the added retinal in the NUV range, but overlapped to a great extent with spectrum 1 in the visible range, indicating the absence of aporetinochrome. In the experiment shown in Figure 5B, such retinochrome (spectrum 1) was converted to metaretinochrome by irradiation (spectrum 2), and then 0.5 ml of it was similarly mixed with all-trans-retinal. During dark incubation, the absorbance of the mixture gradually increased in the visible until the change was complete after 20 min (spectrum 3). This mixture was readily bleached by reirradiation with orange light (spectrum 4), indicating that the regenerated pigment was as light-sensitive as the original retinochrome. These results clearly showed that the 11-cis chromophore of metaretinochrome is easily replaced by all-trans-retinal to regenerate retinochrome. In consequence, the experimental result shown in Figure 5A additionally reveals the absence of metaretinochrome as well as aporetinochrome in the original extract of retinochrome.

Fig. 5

Dark regeneration of retinochrome from metaretinochrome in the presence of all-trans-retinal. Addition of the retinal to retinochrome (A) and metaretinochrome (B). See text for further explanation.

Retinal isomerase activity of retinochrome

In Figure 5B, spectrum 4 did not intersect spectrum 3 of the regenerated retinochrome, the absorbance being lowered in the NUV range. The reason is that, when the final irradiation is carried out for a long time in the presence of excess all-trans-retinal, the bleaching and regeneration of retinochrome are quickly repeated in the mixture to accumulate 11-cis-retinal having a far lower extinction coefficient. It was thus suggested that nautilus retinochrome acts as a catalyst in the light to convert the all-trans isomer of retinal into the 11-cis form [6, 8]. The following experiment shown in Figure 6 was an attempt to examine such retinal photo-isomerase activity. 0.4 ml of retinochrome was added to 0.7 ml of all-trans-retinal (A_390nm 1.30), and the mixture was steadily irradiated with orange light (>540 nm, not absorbed by retinal). During irradiation, the absorbance of all-trans-retinal around 390 nm gradually decreased and λ_max shifted toward shorter wavelengths as the trans-to-cis photo-isomerization advanced with time, as revealed by transient spectra measured after 60, 140 and 260 min. The reaction was completed after 500 min. The time-course of the fall in absorbance at 390 nm apparently followed first-order kinetics, as shown by a straight line on the inserted graph with the semilogarithmic scale. The cahgnes in retinal chromophore composition were analyzed by HPLC before and after irradiation. The molar ratios of all-trans, 13-cis, and 11-cis forms were changed from 99:0:1 to 17:5:78, indicating that much of the all-trans-retinal was isomerized to 11-cis. The reason that the proportion of the 11-cis was limited to 78% was probably because some of the all-trans-retinal had been trapped by lipids and proteins other than retinochrome and were not accessible for photoisomerization. This isomerization practically excluded any other isomers except for the above three forms. Nautilus retinochrome is thus capable of converting all-trans-retinal to the 11-cis form in the light, which is required for rhodopsin formation in the retina.

Fig. 6

Light isomerization of all-trans-retinal with retinochrome. Retinochrome was added to all-trans-retinal, and steadily irradiated with orange light. The inset shows the time-course of the fall in A_390nm during irradiation (isomerization), where A is the absorbance at various times and × is its minimal value after prolonged irrdiation. See text for further details.

Molecular weight of rhodopsin and retinochrome

The molecular weights (M.W.) of nautilus rhodopsin and retinochrome were estimated by SDS-PAGE. Each of the photopigment membrane fractions was reduced by borane dimethylamine to produce N-retinyl protein, solubilized with SDS and electrophoresed. Figure 7 presents the SDS-PAGE pattern photographed under NUV light (a) and that after staining with Coomassie brilliant blue (b). In each case, there are many protein bands on the CBB-stained gel, but the desired photopigment band could readily be selected by consulting the position of fluorescence due to N-retinyl protein on the fluorogram. The M.W. of nautilus rhodopsin was thus estimated to be 84,000, far larger than those of squid and octopus rhodopsins [12]. In the absorption spectrum of rhodopsin shown in Figure 1, the γ-band at 280 nm was very high, probably associated with such a large molecular weight of opsin. On the other hand, the M.W. of nautilus retinochrome was 26,000, the same as that of squid retinochrome [3].

Fig. 7

SDS-PAGE patterns of rhodopsin and retinochrome. The photopigment membrane was reduced by boranae dimethylamine and subjected to electrophoresis. (a) photographed under NUV light. (b) stained with Coomassie brilliant blue.

Contents of rhodopsin and Retinochrome

Our previous work with epifluorescence microscopy suggested that the nautilus retina contains far less retinochrome than rhodopsin [11]. In order to estimate the relative amount of these photopigments, dark-adapted retinas were homogenized and analyzed for retinoids by HPLC. On the average of six samples derived from two separate retinas, the molar ratio of all-trans-, 13-cis-, 11-cis retinal and all-trans-retinol are presented in Table 3. Since the preceding experiments showed the absence of metarhodopsin and metaretinochrome in the extracts from dark-adapted retinas, the 11-cis- and the all-trans-retinal must be derived from rhodopsin and retinochrome, respectively. Consequently, from the data in Table 3, the ratio of rhodopsin to retinochrome in the retina is calculated to be approximately 96:4 (93.0/97.2:4.2/97.2). In the nautilus retina, retinochrome is surely so scanty compared with rhodopsin.

Table 3

Composition of isomeric forms of retinoids found in the nautilus retina by HPLC

Rhodopsin

The λ_max of rhodopsin was at 465 nm, which is very close to the value of ∼467 nm which was reported by Muntz [16]. The light absorption of rhodopsin yields a photoequilibrium mixture with metarhodopsin, which shows a higher absorbance at the λ_max of 510 nm. Since the distance in wavelength between the λ_max of rhodopsin and metarhodopsin is very large (45 nm), the metarhodopsin regenerates greatly into rhodopsin when reirradiated with light of long wavelengths (photoreversal), resembling the cases of the octopus [2] and the conch [22]. As seen in Figure 2, metarhodopsin was slowly decomposed in the presence of 100 mM NH₂OH, whereas rhodopsin was completely stable. This metarhodopsin seemed a little sensitive to NH₂OH compared with squid metarhodopsin [25]. As well known in dibranchiate cephalopods [7], nautilus metarhodopsin was also an acidbase indicator whose absorption peak shifts toward 380 nm at alkaline pH. The molecular weight of nautilus rhodopsin was estimated to be 84,000 by SDS-PAGE (Fig. 7), but it was reduced to 66,000, when the membrane fraction had been treated with Peptid-N-glycosidase F (Boehringer Mannheim Biochemica) to remove carbohydrate from the rhodopsin molecule. Even this value was still far larger than that of squid rhodopsin at 51,000 [19] or 49,800 [12], but rather comparable with that of conch rhodopsin at 59,000 (Terakita, personal communication). Unlike squid rhodopsin [19], nautilus rhodopsin did not combine with concanavalin A, so that Con A-Sepharose affinity chromatography was ineffective in purifying nautilus rhodopsin.

Retinochrome

The λ_max of nautilus retinochrome was at 510 nm, about 45 nm longer in wavelength than that of rhodopsin. Since the spectral peak of the difference spectrum before and after irradiation of retinochrome was observed at 513 nm (Fig. 3), the λ_max of metaretinochrome was inferred to be at 490–500 nm. Metaretinochrome was more susceptible to NH₂OH than retinochrome, showing a lower affinity for retinal, and readily regenerated retinochrome when mixed with all-trans-retinal. Consequently the protein moiety could act as an enzymatic catalyst to convert the all-trans-retinal to 11-cis in the light. We also observed that this metaretinochrome slowly returned to retinochrome when kept alone in the dark (spontaneous regeneration), just like squid metaretinochrome [21]. A noticeable difference from other molluscs was in the content of retinochrome in the retina. It amounted to only about 4% of the rhodopsin content, and the actual yield of retinochrome from one retina was as little as 0.02 in optical density at λ_max (10 mm light path) per milliliter. Such a proportion of retinochrome to rhodopsin was the lowest that we have ever observed among various cephalopod eyes. Usually extracts of retinochrome contain some aporetinochrome, e.g. ∼80% of the retinochrome content in conch photic vesicles [22] and ∼25% in squid myeloid bodies [3]. In contrast, aporetinochrome was not detected at all in the nautilus retina. In general, in a rhodopsin-rich retina, the amount of aporetinochrome tends to decrease to yield retinochrome in retinal-carrying form.

Rhodopsin-retinochrome system

The present study has made clear that the nautilus visual cells have a retinochrome system in myeloid bodies as well as a rhadopsin system in rhabdomes. In order for these systems to cooperate with each other and promote the regeneration of their original photopigments, an additional shuttle which serves to transport retinal between those systems within the visual cell may be necessary. We did not extract it from the nautilus retina, but atempted to ascertain the presence of a retinal-binding protein (RALBP) such as that found in the squid retina [23]. The separated proteins on SDS-PAGE of the retinal homogenates of squid and nautilus were electrophoretically transferred to nitrocellulose paper, and cross-reaction with monoclonal anti-squid RALBP anti-body [4] was tested. The findings confirmed that the nautilus retina contained a RALBP of almost the same M.W. (50,000) as that found in the squid retina (unpublished data). Consequently, the nautilus retina would also possess the rhodopsin-retinochrome conjugate system which contributes to maintaining visual photosensitivity through the recycling of retinal and the regeneration of photopigments, as has been established in the squid retina [5, 10, 26].

Visual life of nautilus

Nautilus rhodopsin with a λ_max at very short wavelengths may be a typical example of the adaptation of an animal to light conditions in a deep-sea environment [14, 16]. In the dark-adapted nautilus, the retina always contains much rhodopsin and little metarhodopsin (Table 1). Such a situation may be attained by the following two mechanisms. One is the photoreversal of rhodopsin from metarhodopsin, and the other is dark regeneration of rhodopsin with the aid of the retinochrome system. In the latter, RALBP serves to send the 11-cis-retinal of metaretinochrome to metarhodopsin and to bring back the all-trans-retinal of metarhodopsin to metaretinochrome. At this time, both rhodopsin and retinochrome are regenerated by the exchange of cis and trans retinals between RALBP and the meta-pigments. At any rate, light absorption by metarhodopsin for photoreversal and by retinochrome for 11-cis-retinal production is essential for promoting the rhodopsin formation in the retina. Since both the λ_max of metarhodopsin and retinochrome are at 510 nm, the light of far longer wavelengths than the λ_max of rhodopsin (465 nm), e.g. green light, should be most desirable. In spite of a pin-hole eye, the nautilus retina is well-developed and abounds in visual cells retaining much rhodopsin. As the animal may utilize only a trace of rhodopsin for catching faint light sources, a capacity of photoreception would be maintained steadily as long as the retina is at times allowed exposure to green or orange daylight. The nautilus eye must be an excellent device for keeping high visual sensitivity for fairly long periods of time in the dark depths of the sea.