THE JOURNAL OF BIOLOGICAL CHEMISTRY 8 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 23, Issue of August 15, pp. 15520-15524,1991 Printed in U.S.A.

Purification of Retinol Dehydrogenase from Bovine Retinal Rod Outer Segments*

(Received for publication, January 23, 1991)

Sei-ichi IshiguroS, Yuko Suzuki, Makoto Tamai, and Katsuyoshi Mizuno From the Department of Ophthalmology, Tohoku Uniuersity School of Medicine, 1-1 Seiryo-machi, Sendai, Miyagi 980, Japan

We purified retinol dehydrogenase from bovine rod outer segments using polyethylene glycol precipitation and hydroxylapatite, concanavalin A-Sepharose CL- 4B, and Sepharose CL-GB column chromatography in the presence of NADP. We obtained 13-fold purifica- tion of retinol dehydrogenase with specific activity of 61.8 nmol/min/mg and 3.8% recovery. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed that retinol dehydrogenase had a molecular mass of 37,000 daltons. The K,,, values of purified retinol de- hydrogenase for all-trans retinol and all-trans retinal were 6.6 mM and 0.085 mM, respectively. The purified enzyme reacted with the all-trans retinal but not with 13-, 11-, and 9-cis compounds.

In addition, we prepared antibody to retinol dehy- drogenase using rat. The anti-retinol dehydrogenase antibody precipitated retinol dehydrogenase activity and was confirmed to bind to 37-kDa protein by West- ern blotting. We also found that anti-retinol dehydro- genase antibody bound to bovine rod outer segments specifically by immunohistochemical technique. The molar ratio of retinol dehydrogenase to opsin in rod outer segments estimated by enzyme-linked immuno- sorbent assay was 1:140.

In the visual cycle, all-trans retinal from bleaching rhodop- sin is converted to all-trans retinol by all-trans-specific retinol dehydrogenase in the rod outer segments of photoreceptor cells (1). All-trans retinol derived from the bleaching photo- receptor cells (2), and the choroidal blood supply (3) is deliv- ered to the retinal pigment epithelial (RPE)’ cells and is converted to 11-cis retinal through esterification (4-6), de- esterification (7), isomerization (8-lo), and oxidation by 11- cis-specific retinol dehydrogenase (1, 11). The 11-cis retinal formed in the RPE cells is delivered to rod outer segments (ROS) to regenerate rhodopsin (2, 12-14).

Retinol dehydrogenase in ROS catalyzes the interconver- sion of all-trans retinol and all-trans retinal. The reaction, however, is directed to form all-trans retinol using NADPH in the ROS (15) during light adaptation. The enzyme binds to membrane and is labile when dissolved by detergents (1,

* This work was supported by Grant-in-Aid 63480390 for Scientific Research of The Ministry of Education, Science and Culture, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

f To whom correspondence should be addressed Dept. of Oph- thalmology, Tohoku University School of Medicine, 1-1 Seiryo-ma- chi, Sendai, Miyagi 980, Japan.

The abbreviations used are: RPE, retinal pigment epithelial; ROS, rod outer segment; SDS, sodium dodecyl sulfate; PAGE, polyacryl- amide gel electrophoresis.

16, 17). Therefore, purification of the enzyme has not yet been accomplished. We found that retinol dehydrogenase, even solubilized with detergents, is not so labile in the pres- ence of NADP. We purified the enzyme from bovine ROS and partially characterized it.

MATERIALS AND METHODS

Preparation of Retinal Isomers-Photoisomerization of retinal was performed according to the method of Bridges and Alvarez (18). Retinal isomers were isolated with high pressure liquid chromatog- raphy using Chemcosorb 5Si column (4.6 X 250 mm) according to Tsukida et al. (19). Mobile phase is diethyl etherln-hexane (892) at 1 ml/min. Peaks were detected at 365 nm. All-trans, 9-cis, 11-cis, and 13-cis compounds were collected separately and were kept under argon at -30 “C. Experiments involving the retinoids were carried out in dim red light.

Assay of Retinol Dehydrogenase Actiuity-Retinol dehydrogenase activity was measured by the modification of Lion et al. (1). The oxidation reaction was started by mixing 50 pl of enzyme solution and 50 p1 of substrate solution containing 10 mM all-trans retinol, 6 mM NADP, 1.2% Tween 80, 12% acetone, and 0.4 M Tris-HC1 buffer, pH 8.0. Incubation was carried out at 37 “C. The reaction was stopped with the addition of 0.5 ml of ethanol, and the retinal formed was assayed by adding 0.2 ml of thiourea reagent and 0.2 ml of thiobar- bituric acid reagent according to Futterman and Saslaw (20). After 30 min, samples were centrifuged at 10,000 rpm for 1 min. Color developed was measured at 530 nm. To avoid precipitation of salts, 50 pl of distilled water was added to the supernatant before the measurement was performed.

The reduction reaction was started by mixing 50 pl of enzyme solution and 50 p1 of substrate solution containing 10 nmol of all- trans retinal, 5 mM NADPH, 1.2% Tween 80, 12% acetone, and 0.2 M sodium acetate buffer, pH 5.0. Incubation was carried out at 37 “C, and the reaction was stopped with the addition of 0.5 ml of ethanol. Retinal that was decreased in 2 min was assayed by the thioharbituric acid method, as described.

Purijication of Retinol Dehydrogenase-Bovine eyes were obtained from a local slaughterhouse. The ROS from 300 bovine retinas were prepared in the light, according to the method of Papermaster and Dreyer (21). The purified ROS were suspended with 24 ml of 5 mM phosphate buffer, pH 7.0. To the ROS suspension, 10.72 ml of distilled water, 0.24 ml of 1 M phosphate buffer, pH 7.0, 0.8 ml of 60 mM NADP, 0.24 ml of 200 mM dithiothreitol, 9.6 ml of glycerol, and 2.4 ml of 20% sodium cholate were added, in that order. Final concentra- tion of the suspension was 1 mM NADP, 1 mM dithiothreitol, 20% glycerol, 10 mM phosphate buffer, and 1% sodium cholate. Ammo- nium sulfate (6.91 g) was added to the suspension (48 ml) to make 25% saturated solution for dissolving retinol dehydrogenase from ROS. The pH of the crude enzyme solution was adjusted to 7.0 with NaHC03 powder (crude extract).

The same volume of 20% (w/v) polyethylene glycol 6000 (about 51 ml) was added to the crude extract. After 10 min, precipitated proteins were removed by centrifugation at 14,000 rpm for 20 min (polyeth- ylene glycol supernatant).

The same volume (about 96 ml) of buffer A (0.5% sodium cholate, 0.3 mM NADP, 1 mM dithiothreitol, 20% glycerol, 10 mM phosphate buffer, pH 7.0) containing 10% saturated ammonium sulfate was added to the supernatant. The sample was applied onto hydroxylapa- tite column (50 X 20 cm, 15 g of hydroxylapatite, Bio-Gel HTP, Bio- Rad) equilibrated with buffer A containing 10% saturation of am-

15520

Rod Outer Segment Retinol Dehydrogenase 15521

monium sulfate (buffer B). Then retinol dehydrogenase was eluted with a linear gradient of 0-40% ammonium sulfate in buffer A a t a flow rate of 1.5 ml/min. Total gradient volume was 600 ml. Fraction size was 10 g/tube. Aliquots (50 pl) of the fractions were used to assay retinol dehydrogenase activity. The oxidation reaction was performed for 10 min a t 37 "C. The fractions containing retinol dehydrogenase were concentrated with Amicon Diaflo apparatus using PM-10 mem- brane to approximately 10 ml. The concentrated sample was dialyzed overnight against 200 ml of buffer A containing 0.15 M NaCl (buffer C) at 4 "C.

The dialysate was applied onto concanavalin A-Sepharose CL-4B column (0.9 X 12 cm) equilibrated with buffer C a t a flow rate of 0.5 ml/min. Unbound retinol dehydrogenase was eluted with 100 ml of buffer C. Fraction size was 10 g/tube. After eluting retinol dehydro- genase, 100 ml of buffer C containing 0.25 M methyl-a-D-mannopyr- anoside was used to elute the bound retinol dehydrogenase. Aliquots (50 pl) of the fractions were used to assay retinol dehydrogenase activity. The oxidation reaction was performed for 10 min a t 37 "C. The fractions containing unbound retinol dehydrogenase were con- centrated to approximately 4 ml. The concentrated sample was di- alyzed overnight against 200 ml of buffer C a t 4 "C. Concanavalin A- Sepharose CL-4B column chromatography removes the remaining opsin. This step sometimes decreases recovery of retinol dehydrogen- ase, but is necessary to obtain the highly purified enzyme.

The dialysate was applied onto Sepharose CL-GB column (1.9 X 100 cm) equilibrated with buffer C. Retinol dehydrogenase was eluted with buffer C a t a flow rate of 0.75 ml/min. Fraction size was 3 g/ tube. Aliquots (50 pl) of the fractions were used to assay retinol dehydrogenase activity. The oxidation reaction was performed for 1 h a t 37 "C. The fractions containing retinol dehydrogenase were concentrated to approximately 4 ml.

Final purity of the retinol dehydrogenase was determined by SDS- polyacrylamide gel electrophoresis (SDS-PAGE), according to Laem- mli (22), with 10% acrylamide gel. After electrophoresis, proteins were detected with a silver staining kit (Bio-Rad).

Protein concentrations were measured according to Bensadoun and Weinstein (23) using sodium cholate with bovine serum albumin as the standard.

Preparation of Antibodies-Rabbit anti-bovine opsin antibody was prepared, as described previously (24).

We prepared anti-bovine retinol dehydrogenase antibody using rat. Retinol dehydrogenase was partially purified by polyethylene glycol, hydroxylapatite, and Sepharose CL-GB column chromatography. As the column chromatography of concanavalin A-Sepharose CL-4B decreased the recovery, we removed this step. Retinol dehydrogenase was further purified by SDS-PAGE with 10% gel from the partially purified sample. We stained the gel with Coomassie Brilliant Blue R- 250 (25) and eluted the 37-kDa protein electrophoretically using the electrode buffer of Laemmli (22). The eluted sample was concentrated by lyophilization.

We mixed 20 pg (200 pl) of the electrophoretically purified retinol dehydrogenase with the same volume (200 pl) of complete adjuvant and injected the emulsified sample intraperitoneally into rat two times. Incomplete adjuvant and 50 pg of protein were used for the third injection. Plasma was collected 3 days after the final injection.

Western Blotting-Western blotting was performed according to Towbin et al. (26). After electrophoretic transfer of proteins to nitro- cellulose membrane, the membrane was treated with 3% gelatin for 1 h at 37 "C. We used 1:lOO diluted rat anti-retinol dehydrogenase plasma and 1:50 diluted goat anti-rat IgG antibody conjugated with peroxidase. The antibodies were diluted with phosphate-buffered saline (0.14 M NaCl and 10 mM phosphate buffer, pH 7.4) containing 1% bovine serum albumin and 0.05% Tween 20. Incubation time was 1 h a t room temperature. Washing was performed after each step with phosphate-buffered saline containing 0.05% Tween 20. Color development with o-dianisidine was carried out for 20 min at room temperature.

Light Microscopic Immunohistochemistry of Retinol Dehydrogen- ase-Bovine retinas were prepared with periodate-lysine-paraform- aldehyde fixative (27) a t 4 "C overnight and embedded in paraffin. Indirect staining was performed, as reported previously (24), except that 1:200 diluted rat anti-retinol dehydrogenase plasma as first antibody and peroxidase-conjugated goat anti-rat IgG antibody as second antibody were used. We prepared control sections treated with anti-retinol dehydrogenase plasma absorbed with excess antigen (10 pg) or omitted the step of first antibody reaction.

Enzyme-linked Zmmunosorbent Assay of Opsin and Retinol Dehy- drogenase-Crude extracts of rod outer segments were prepared, as

0.1- . I . . I . . I . . t 0

0 2 0 4 0 6 0 8 0

fraction number

1 .o- e .

0 . 6

0.4-

0 . 8

0 . 6 0

0 . 4 g v

cy

0.2

0 . 0

4:

0 . 2 a - 0 . 2 0 1 0

fraction number 2 0

0.1 - 0

OD cy s

0 . 0 a

-0.1 3 0 4 0 5 0 6 0 7 0 8 0 9 0 100

fraction number FIG. 1. Results of column chromatographies. Panel a, hydrox-

ylapatite column chromatography. Arrow indicates start of a linear gradient of 0-40% ammonium sulfate in buffer A. Oxidation reaction of retinol dehydrogenase is shown as absorbance a t 530 nm (see description under "Materials and Methods"). Panel b, concanavalin A-Sepharose CL-4B column chromatography. Oxidation reaction was performed for 10 min a t 37 "C. Arrow indicates start of elution with buffer C containing 0.25 M methyl-a-D-mannopyranoside. Panel c, Sepharose CL-GB column chromatography. Elution positions of the molecular weight standards are as follows: 1) Blue Dextran 2000; 2) thyroglobulin (669,000); 3) catalase (232,000); 4) ovalbumin (43,000); and 5) pyridoxal 5-phosphate (247).

described before, and diluted with 1% Triton X-100 containing 1% bovine serum albumin, 0.14 M NaCl, and 10 mM phosphate buffer, pH 7.4. Opsin content and retinol dehydrogenase content were meas- ured by competitive enzyme-linked immunosorbent assay, as reported previously (24).

RESULTS

The stabilizing effect of NADP on retinol dehydrogenase was determined using 1, 0.1, 0.01, and 0.001 mM in the presence of 20% glycerol, 0.1 M NaCl, 10 mM phosphate buffer, pH 7.0, and 1% sodium cholate. After 30 min, we measured the oxidation reaction of retinol dehydrogenase in a linear range of activity and obtained 73, 69, 59, and 21% activity compared with undissolved ROS, respectively. When glycerol was removed, the remaining activity decreased markedly. We used at least 0.3 mM NADP for the purification of the enzyme. We performed column chromatography at pH 7.0 because retinol dehydrogenase is rather stable from pH 5.0 to 7.5.

We found that retinol dehydrogenase dissolved with 1%

15522 Rod Outer Segment Retinol Dehydrogenase TABLE I

Purification of retinol dehydrogenase from bovine rod outer segments ~~~ ~ ~

Purification step Volume

mi Crude extract 51 Polyethyleneglycol 98 Hydroxylapatite 10.4 Concanavalin A-Sepharose 3.9 SeDharose CL-GB 4.0

Protein Activity Specific activity Purification Yield rng nmollrnin nrnol/rnin/rng protein -fold %

417 1990 4.77 1 100 56.8 1320 23.3 4.9 66 16.6 496 29.9 6.3 25 2.83 78.4 27.7 5.8 3.9 1.22 75.4 61.8 13.0 3.8

97- 66-

43-

31-

22- 14-

a b C d e FIG. 2. SDS-PAGE of purified retinol dehydrogenase and

characterization of rat anti-bovine retinol dehydrogenase an- tibody by Western blotting and immunostaining techniques. Lanes a d , SDS-PAGE silver staining. Lane a, molecular weight standards. Lane b, purified retinol dehydrogenase (1 pg, arrow). Lane c, electrophoretically purified retinol dehydrogenase (0.5 p g , arrow). Lune d, crude extract (7.5 pg). Lane e, Western blot of crude extract (7.5 pg) reacted with anti-bovine retinol dehydrogenase (described under “Materials and Methods”). The antibody reacts with a single molecular species that corresponds with the migration of the retinol dehydrogenase.

0 . 0 8

0 . 0 6

. > y 0.04

0 . 0 2

0 Id - 1 0 0 1 0 2

1 i s

0.1

0 -1 0.0

1 - 1

1 I s FIG. 3. Determination of VmaX and K,,, by Lineweaver-Burk

plot. Reduction (panel a, 30 pg of fully purified enzyme, 5 min of incubation) and oxidation (panel b, 25 pg of fully purified enzyme, 5 min of incubation) reactions were performed in a linear range of time (during first 6 min of both oxidation and reduction reactions) and enzyme concentration (at least 30 pg) (see description under “Mate- rials and Methods”). The lines were obtained by a least-squares approximation.

sodium cholate, sodium deoxycholate, Triton X-100, and Nonidet P-40 retained its enzyme activity, but dodecyltrime- thylammonium bromide inactivated the enzyme even in the presence of NADP. As the enzyme bound to hydroxylapatite firmly and eluted with high concentration of ammonium sulfate (Fig. la), we selected sodium cholate.

As shown in Table I, concanavalin A-Sepharose column chromatography (Fig. l b ) sometimes decreased recovery of

1 8

Ilme(min) Iime(mln)

FIG. 4. Stereospecificity of retinol dehydrogenase. Panel a, rod outer segments (250 pglreaction mixture). Panel b, fully purified retinol dehydrogenase (25 pglreaction mixture). Substrates used are all-trans (X), 9-cis (O), 11-cis (O), and 13-cis (B) retinal. Reduction reaction was carried out (see description under “Materials and Meth- ods”). All-trans retinal is a specific substrate both for retinol dehy- drogenase in rod outer segments and purified retinol dehydrogenase.

0.1

Z I [r 0.0 0 10 0

antibody added( pl)

FIG. 5. Immunoprecipitation of retinol dehydrogenase (RDH) with rat anti-bovine retinol dehydrogenase antibody. We mixed 2 pg of partially purified retinol dehydrogenase by polyeth- ylene glycol, hydroxylapatite, and Sepharose CL-GB with indicated amounts of anti-retinol dehydrogenase (0) or unimmunized rat plasma (0) in buffer C. The mixtures were incubated at 4 “C for 2 h in a volume of 45 pl. Then, 55 pl of goat anti-rat IgG serum was added to the mixtures, and the final mixtures were incubated at 4 “C for 18 h. The samples were centrifuged at 12,000 rpm for 10 min, and 50 pl of the supernatant was used for oxidation reaction of retinol dehydrogenase. Nonenzymatic color development of plasma was sub- tracted.

the enzyme, and we obtained only 3.8% in final recovery. However, without this step, we could not obtain highly puri- fied retinol dehydrogenase.

Retinol dehydrogenase was larger than thyroglobulin by Sepharose CL-GB column chromatography (Fig. IC), whereas its molecular mass was 37,000 daltons estimated by SDS- PAGE (Fig. 2). Dimer molecules of retinol dehydrogenase were noted (Fig. 2e).

We carried out nondenaturing polyacrylamide gel electro- phoresis to obtain further information about the active form

Rod Outer Segment Retinol Dehydrogenase 15523

r: 1

FIG. 6. Immunohistochemical demonstration of retinol de- hydrogenase in bovine retina. Marked immunoreactivity was present in the rod outer segments (arrow in panel b) , whereas no immunoreactivity was observed in the control sections, for which the step of first antibody reaction was omitted (panel a). OS, outer segment; IS , inner segment; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Bar indicates 50 pm.

, -I ,

9 0 . a

0 ' 4 0 ' 30 2 .o

1 0

5 , . 1 0 0.01 0.1 1 . 1 0 10-4 lb-3 ' 1'0-2. 1 b - 3 1

pmollwell sample added(p1)

FIG. 7. Competitive enzyme-linked immunosorbent assay of opsin and retinol dehydrogenase. Panel a, standard curves of retinol dehydrogenase (0) and opsin (0) show similar results. Con- centration of purified retinol dehydrogenase is calculated based on M, = 37,000. Panel b, content of retinol dehydrogenase (0) and opsin (0) in the crude extract is calculated by its % bound. Lines were obtained by a least-squares approximation. Molar ratio of retinol dehydrogenase to opsin was 1:140.

of retinol dehydrogenase according to the method of Davis (28) and Weber and Osborn (25) using fully purified retinol dehydrogenase (26 pg/lane), various detergents (sodium cho- late, Triton X-100, Emulphogene BC-720), varying detergent concentration (0.5, 1.0, 2.0%), and varying separation gel concentration (5.0, 7.5, 10, 12%) in the presence of 0.3 mM NADP. We, however, failed to detect retinol dehydrogenase activity in the gel by tetrazolium salt staining (5 mM all-trans retinol, 6% acetone, 0.6% Tween 80, 3 mM NADP, 0.3 mM phenazine methosulfate, 1 mM nitrotetrazolium blue, 0.2 M phosphate buffer, pH 7.0) after electrophoresis.

We obtained K, and V,,, values by Lineweaver-Burk plot using fully purified material (Fig. 3), ROS (350 pg of protein/ mixture), and ROS crude extract (350 pg of protein/mixture) in a linear range of time (during first 6 min) and enzyme concentration. The K,,, values of retinol dehydrogenase in the ROS, crude extract, and purified sample for all-trans retinal

were 0.080, 0.042, 0.085 mM, respectively. Apparent K,,, value of retinol dehydrogenase in the ROS was not changed after solubilization and purification. The results show that opsin and other ROS proteins do not affect the K,,, value of retinol dehydrogenase. In determining the K,,, values of these prepa- rations for all-trans retinol, we obtained 2.9, 1.3, and 6.6 mM, respectively. The Vmax values of the purified retinol dehydro- genase were 38.9 nmol/min/mg for all-trans retinol and 37.1 nmol/min/mg for all-trans retinal, respectively. Thus, it ap- pears clear that purified retinol dehydrogenase has oxidore- ductase activity.

Stereospecificity of purified retinol dehydrogenase was de- termined using reduction reaction with all-trans, 9-cis, ll-cis, and 13-cis retinal as substrates (Fig. 4). Retinol dehydrogen- ase both in ROS and in purified sample had substrate pref- erence toward all-trans compound.

We prepared rat anti-bovine retinol dehydrogenase anti- body. The antibody bound to 37-kDa protein (Fig. 2e) and precipitated retinol dehydrogenase activity (Fig. 5). It was also confirmed that the antibody bound specifically to bovine ROS (Fig. 6). In control sections, in which the step of first antibody reaction was omitted or the sections were treated with first antibody absorbed with excess antigen, no immu- noreactive product was observed. This immunohistochemical result was consistent with the histochemical result reported previously by Kissun et al. (29).

The molar ratio of retinol dehydrogenase to opsin was determined by enzyme-linked immunosorbent assay (Fig. 7). Similar standard curves were obtained for retinol dehydrogen- ase and opsin. In crude extract, the ratio of retinol dehydro- genase to opsin was 1:140 k 22 (mean & SD, n = 4).

DISCUSSION

The results obtained herein constitute the first purification of retinol dehydrogenase to homogeneity from bovine ROS. The purified enzyme has stereospecificity for all-trans retinal. The results show that ROS retinol dehydrogenase does not convert 11-cis retinal delivered from RPE cells to 11-cis retinol. The 11-cis retinol is present in the interphotoreceptor matrix (14), but its origin is unknown.

Substrate specificity also distinguishes retinol dehydrogen- ase purified in the present study from another retinol dehy- drogenase in RPE. Stereospecificity of retinol dehydrogenase in RPE is directed toward 11-cis compound (1, 30), while retinol dehydrogenase purified in this study has specificity for all-trans compound.

We found that anti-retinol dehydrogenase antibody bound to ROS specifically and had no cross-reactivity toward retinol dehydrogenase in RPE. Also of immunohistochemical interest was whether all-trans-specific retinol dehydrogenase is pres- ent in Muller cells or not. Muller cells have cellular retinal- binding protein (31, 32) and cellular retinol-binding protein (33), and it has been suggested that the glial cells play some role in vitamin A metabolism in the retina. All-trans-specific retinol dehydrogenase, however, was not found by our im- munohistochemical technique. The reduction reaction of all- trans retinal to all-trans retinol seems to be a specific reaction in the ROS in the eye.

Opsin and retinol dehydrogenase, being membrane pro- teins, have a marked tendency to exist as a large micellar complex with many molecules of the detergent. Nicotra and Livrea (17) showed that a major peak of retinol dehydrogenase solubilized with nonionic detergent Lubrol 12A9 corresponds to Stokes radius of 8.5 nm which is close to thyroglobulin molecule on Bio-Gel A-5 m. In the present study, we also

15524 Rod Outer Segment Retinol Dehydrogenase

observed a large micellar complex on a gel filtration column chromatography (Fig. IC).

We obtained the molecular mass of 37,000 daltons for retinol dehydrogenase by SDS-PAGE. Nicotra and Livrea (17) showed that the molecular weight of retinol dehydrogen- ase was 70,000 by SDS-PAGE using [14C]retinol binding capacity. A molecular weight of 70,000 is almost equivalent t o the dimer of our molecule. At present, it is unknown whether the active form of retinol dehydrogenase requires the dimer structure or not.

Retinol dehydrogenase constitutes about 0.6% of ROS pro- tein since the molar ratio of retinol dehydrogenase to opsin is 1:140 (Fig. 7), and rhodopsin comprises over 80% of the protein in ROS (21). Highest V,,, value of our fully purified retinol dehydrogenase for all-trans retinal was 58.1 nmol/ min/mg. The turnover number (molecular activity) can be calculated based on this specific activity and 37,000 Da as the molecular mass, i e . (58.1 nmol/6O s)/(l,OOO,OOO ng/37,000) = 0.036 s-'. This means that the conversion of 140 molecules of all-trans retinal released from 140 molecules of photobleached rhodopsin by one molocule of retinol dehydrogenase in ROS requires about 65 min to end at constant velocity. If the fully purified material contains inactive enzyme to some extent, the turnover number increases and the conversion finishes earlier. Retinal content of the rat retina practically reaches a level 10-15% after 30 min in the light (2). This result was obtained using different species, but the time required for the conversion seems to be not markedly different from the result of our calculation (15% after 55 min).

Finally, we purified labile, membrane-bound retinol dehy- drogenase from bovine ROS. The enzyme plays an important role in the visual cycle in the eye. A deficiency of the enzyme may cause the accumulation of a potentially toxic all-trans retinal derived from photolysis of rhodopsin in the ROS of photoreceptor cells and also may result in the deterioration of dark adaptation by losing the supply of all-trans retinol from bleached photoreceptor cells to RPE cells. Thus, it seems likely that a deficiency of ROS retinol dehydrogenase causes photoreceptor degeneration. Further molecular analysis of retinol dehydrogenase may provide an important clue in un- derstanding the disorders caused by abnormality in vitamin A metabolism.

Acknowledgments-We wish to thank Dr. Satoshi Hara for his helpful comments and Eiko Sat0 for the preparation of the histologic sections.

REFERENCES 1. Lion, F., Rotmans, J. P., Daemen, F. J. M., and Bonting, S. L.

(1975) Biochim. Biophy~. Acta 384,283-292

2. Dowling, J. E. (1960) Nature 188 , 114-118 3. Bok, D., and Heller, J. (1976) Exp. Eye Res. 22 , 395-402 4. Andrews, J. S., and Futterman, S. (1964) J. Bid. Chem. 239,

5. Saari, J. C., and Bredberg, D. L. (1988) J. Biol. Chem. 263,8084-

6. Saari, J. C., and Bredberg, D. L. (1989) J. Bio2. Chem. 264,8636-

7. Deigner, P. S., Law, W. C., Canada, F. J., and Rando, R. R.

8. Bernstein, P. S., Law, W. C., and Rando, R. R. (1987) Proc. Natl.

9. Bridges, C. D., and Alvarez, R. A. (1987) Science 236,1678-1680 10. Trehan, A., Canada, F. J., and Rando, R. R. (1990) Biochemistry

11. Zimmerman, W. F., Lion, F., Daemen, F. J. M., and Bonting, S.

12. Lai, Y. L., Wiggert, B., Liu, Y. P., and Chader, G. J. (1982)

13. Liou, G. I., Bridges, C. D. B., Fong, S.-L., Alvarez, R. A., and

14. Lin, Z., Fong, S.-L., and Bridges, C. D. B. (1989) Vision Res. 29,

15. Futterman, S., Hendrickson, A., Bishop, P. E., Rollins, M. H.,

16. Blaner, W. S., and Churchich, J. E. (1980) Biochem. Biophys.

17. Nicotra, C., and Livrea, M. A. (1982) J. Biol. Chem. 257,11836-

18. Bridges, C. D. B., and Alvarez, R. A. (1982) Methods Enzymol.

19. Tsukida, K., Kodama, A., Ito, M., Kawamoto, M., and Takahashi,

20. Futterman, S., and Saslaw, L. D. (1961) J. Biol. Chem. 236,

21. Papermaster, D. S., and Dreyer, W. J. (1974) Biochemistry 13 ,

22. Laemmli, U. K. (1970) Nature 227 , 680-685 23. Bensadoun, A., and Weinstein, D. (1976) Anal. Biochem. 70,

24. Ishiguro, S., Fukuda, K., Kanno, C., and Mizuno, K. (1987) Cell

25. Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244,4406-4412 26. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad.

27. McLean, I. W., and Nakane, P. K. (1974) J. Histochem. Cytochem.

28. Davis, B. J. (1964)Ann. N . Y. Acad. Sci. 121,404-427 29. Kissun, R. D., Graymore, C. N., and Newhouse, P. J. (1972) Exp.

30. Zimmerman, W. F., Lion F., Daemen, F. J. M., and Bonting, S.

31. Saari, J. C., Bredberg, D. L., and Garwin, G. G. (1982) J. Biol.

32. Bunt-Milam, A. H., and Saari, J. C. (1983) J. Cell Biol. 97, 703-

33. Bok, D., Ong, D. E., and Chytil, F. (1984) Inuest. Ophthalmol.

4073-4076

8090

8640

(1989) Science 244 , 968-971

Acad. Sci. U. S. A. 8 4 , 1849-1853

29,309-312

L. (1975) Exp. Eye Res. 21,325-332

Nature 298,848-849

Gonzalez- Fernandez, F. (1982) Vision Res. 2 2 , 1457-1467

1699- 1709

and Vacano, E. (1970) J. Neurochem. 17, 149-156

Res. Commun. 94,820-826

11841

81,463- 485

K. (1977) J. Nutr. Sci. Vitaminol. 2 3 , 263-264

1652-1657

2438- 2444

241-250

Struct. Funct. 12 , 141-155

Sci. U. S. A. 76,4350-4354

22,1077-1083

Eye Res. 14,150-153

L. (1975) Exp. Eye Res. 21 , 325-332

Chem. 257,13329-13333

712

Vis. Sci. 25,877-883