Cloning and Characterization of a Novel all-trans Retinol Short-Chain Dehydrogenase/Reductase from the RPE

Abstract

purpose. In the photic visual cycle, retinal G protein-coupled receptor (RGR) isomerizes all-trans retinal to 11-cis retinal in the retinal pigment epithelium (RPE) after illumination. It is unclear, however, how all-trans retinal, the substrate for RGR, is generated in the RPE, because no all-trans retinol dehydrogenase (atRDH) has been identified in the RPE. This study was conducted to identify the atRDH that generates all-trans retinal in the RPE.

methods. The full-length cDNA encoding a novel atRDH, RDH10, was cloned by PCR based on an expressed sequence tag (EST). Cellular localization was determined at the mRNA level by Northern blot analysis, RT-PCR, and in situ hybridization and at the protein level by immunohistochemistry with an antibody specific to RDH10. The activity was measured by an RDH activity assay with recombinant RDH10 expressed in COS cells.

results. The full-length RDH10 was cloned from the human, cow, and mouse. These cDNAs encode a protein of 341 amino acids and have significant sequence homology with other short-chain dehydrogenases/reductases (SDRs). The human RDH10 shares 100% and 98.6% amino acid sequence identity with the bovine and mouse proteins, respectively, suggesting a highly conserved sequence during evolution. RDH10 is predominantly expressed in the microsomal fraction of the RPE. Human RDH10 expressed in COS cells oxidized all-trans retinol to all-trans retinal. RDH10 displayed substrate specificity for all-trans retinol and preferred nicotinamide adenine dinucleotide phosphate (NADP) as the cofactor.

conclusions. RDH10 is a novel retinol oxidase expressed in the RPE. This enzyme can generate all-trans retinal from all-trans retinol and may play an important role in the photic visual cycle.

In vertebrates, the visual cycle starts with photoisomerization of the chromophore of visual pigments, 11-cis retinal, to all-trans retinal. ^{1 2 3} This photoisomerization triggers conformational changes in these visual pigments that subsequently activate G protein transducins and signal transduction cascades. ^{4 5} The all-trans retinal reductases in photoreceptor cells then reduce all-trans retinal to all-trans retinol, which is then transported to the retinal pigment epithelium (RPE). ^{1 3 6} The RPE is the site for esterification, storage, and isomerization of retinoids. A fundamental question regarding the visual cycle is how all-trans retinoids are isomerized into 11-cis isomers. ⁷ Two visual cycles are known to isomerize all-trans retinoids: the isomerohydrolase cycle in the dark and the photic cycle in the light. ^{8 9} It has been suggested that in the dark, all-trans retinol is isomerized to 11-cis retinol through an all-trans retinyl ester by a yet to be identified enzyme, isomerohydrolase. ^{1 8} The oxidation of 11-cis retinol is catalyzed by an 11-cis retinol dehydrogenase (RDH), RDH5, in the RPE to generate 11-cis retinal for the regeneration of visual pigments. ^{3 10 11}  

Recently, a light-dependent visual cycle that generates 11-cis retinal has been identified in the RPE in addition to the isomerohydrolase pathway. ⁹ This photic visual cycle depends on the RPE retinal G protein-coupled receptor (RGR), a photoisomerase that uses all-trans retinal as a substrate. ⁹ RGR is preferentially expressed in the RPE and Müller cells. ¹² Irradiation of RGR with 470 nm blue and near UV light triggers the stereospecific conversion of the bound all-trans isomer into 11-cis retinal. ¹³ The RGR knockout mouse showed decreased steady state levels of 11-cis retinal and rhodopsin and reduced ERG kinetics in light-adapted eyes, suggesting that the RGR pathway is essential for the regeneration of 11-cis retinal in the light. ⁹ In this RGR visual cycle, however, it is not clear how the all-trans retinal is generated in the RPE to serve as a substrate for RGR. ¹⁴  

Previous studies by other groups have identified more than 10 short-chain dehydrogenases/reductases (SDRs) with cis- and/or trans-RDH activities. Among them, RoDH1, ¹⁵ hRoDH4 or RDH-E, ¹⁶ m17βHSD9, ¹⁷ RDH-TBE, ¹⁸ and mRDH1 ¹⁹ have shown all-trans retinol oxidase activities. However, most of these enzymes are very tissue specific and are not expressed in the RPE. ^{15 16 17 18 19} The mRDH1 is expressed in multiple tissues, but has been identified only in the mouse. Although recent evidence has shown that there is all-trans RDH (atRDH) activity in the RPE, which oxidizes all-trans retinol to all-trans retinal, which can serve as the substrate for RGR, ¹⁴ the enzyme responsible for this atRDH activity has not been identified in the RPE. 

We report the identification and characterization of a novel SDR, RDH10, from the RPE. This enzyme has specific atRDH activity that may be responsible for the generation of all-trans retinal in the RGR visual cycle. 

Methods

Cloning of Human RDH10 cDNA

The amino acid sequences of human retSDR1 and RDH5 were used to screen the expressed sequence tag (EST) database by tFastA software (provided without charge to researchers by the Biomolecular Computing Resource, a part of the Medical University of South Carolina research resource, Internet address at http://bcr.musc.edu), as described previously. ²⁰ Based on the sequences of two EST clones that showed significant homologies with human retSDR1, two oligonucleotides (T30: 5′-AGCCTTAGTGGTCCAG AAGTGTGC-3′, F10: 5′-GGAGAACGTCTACCTGACGGCTG-3′) were synthesized. Rapid amplification of cDNA ends (RACE) was used to amplify the 5′ and 3′ ends of the selected EST cDNAs from the human retina (Marathon cDNA; Clontech, Palo Alto, CA) using a commercial system (High Fidelity PCR; Roche, Indianapolis, IN). The primer T30 and the adapter primer (AP1; Clontech) were used for the 5′ RACE, and primers F10 and AP1 were used for the 3′-RACE. A 1.3-kb cDNA containing the full-length coding region was amplified with a sense primer (F1: 5′-GCAGGAGGCGCCGAGCC-3′) and an antisense primer (T3: 5′-TTAGAATCCATAAAATGTCAGC-3′). PCR mixtures were first incubated at 94°C for 3 minutes and then amplified for 45 cycles at 94°C for 45 seconds, at 64°C for 1 minute, and at 72°C for 3 minutes. The PCR products were cloned into a vector (pCRII; Invitrogen, Carlsbad, CA), and positive clones were identified by colony hybridization, as described previously. ²¹ The positive colonies were selected and sequenced. All sequences were confirmed by sequencing the complementary strand and were verified in another clone from an independent PCR. 

Cloning of Bovine RDH10 cDNA

The bovine RDH10 cDNA was identified by screening a bovine RPE cDNA library (obtained from T. Michael Redmond and Ana Boulanger, National Eye Institute). The 5′ end was amplified with a gene-specific primer (B-T40: 5′-TCATGGTTCTCTCAATGAGC-3′) and a vector primer (V5P: 5′-AGGGATGTTTAATACCACTAC-3′). The 3′ end was amplified with a gene-specific sense primer (B-F10: 5′-GAGCAGACAAGTGTATGTACC-3′) and a vector primer (V3P: 5′-GCACAGTTGAAGTGAACTTGC-3′). The cDNA containing the full-length coding region was amplified by two primers (B-F1: 5′-ccagcaggaggcgccg agcc-3′ and B-T10: 5′-TTGAGACTTCCTGTTCATTCC-3′). The PCR products were cloned and analyzed as described earlier. 

Cloning of Mouse RDH10

Total RNA was purified from mouse eyecups with extraction reagent (TRIzol; Gibco-BRL Life Technologies, Rockville, MD). The first strand of cDNA was synthesized by the reverse transcription (RT) from 1 μg of the eyecup RNA with a random hexamer (Amersham Pharmacia Biotech, Piscataway, NJ). The full-length coding region was amplified by PCR, cloned, and sequenced as described earlier. 

Northern Blot Analysis

Total RNA was isolated from bovine tissues. The mRNA was purified from the total RNA with an mRNA isolation kit (Amresco, Solon, OH). The same amount of mRNA (1.5 μg) from each tissue was loaded for Northern blot analysis. A 1.6-kb bovine RDH10 cDNA probe was labeled with ³²P-α-dCTP using a kit (Nick Translation; Gibco-BRL Life Technologies). Hybridization solution (Ultrahyb; Ambion, Austin, TX) was used according to the manufacturer’s protocol. After hybridization with the bovine RDH10 probe, the RNA blot was stripped and reprobed with the cDNA of β-actin. 

Real-Time RT-PCR

Bovine cDNA was reverse transcribed from the same amount of mRNA of various tissues as described earlier. Real-time PCR was performed with a commercial system (The Smart Cycler; Cepheid, Sunnyvale, CA) using a nucleic acid gel staining PCR kit (SYBR Green; Perkin Elmer, Warrington, UK). PCR was performed with a sense primer F6 (5′-ATGAACATCGTGGTGGAGTTC-3′) and an antisense primer T40 (5′-TCATGGTTCTCTC AATGAGC-3′) for 33 cycles. The β-actin cDNA was also amplified from the same RT product by real-time PCR. 

Expression of Human RDH10 in COS Cells

The coding region of RDH10 was amplified and subcloned into the pCDNA6/V5/His vector at the EcoRI and XhoI sites (Invitrogen) in frame with the V5 and His tag sequence. The RDH10/pCDNA6/V5/His expression construct and empty pCDNA6/V5/His vector were transfected separately into COS cells with a lipophilic transfection reagent (Lipofectamine-2000; Gibco-BRL Life Technologies). The transfected cells were cultured for one additional day and then harvested for the in vitro activity assays. 

Western Blot Analysis

A peptide QRKQATNNNEAKNGI corresponding to amino acids 327-341 of the human, bovine, and mouse RDH10 sequences were selected for its apparent antigenicity predicted by the Kyte-Doolittle hydropathy program and synthesized. Rabbits were injected subcutaneously with an emulsion of 0.3 mg of the peptide and complete Freund’s adjuvant (CFA; Gibco-BRL, Grand Island, NY), with intramuscular booster injections of 0.3 mg of the same emulsion at 3-week intervals. After significant immune responses had developed, the rabbits were killed, and the whole serum was collected. 

The transfected COS cells were lysed in a lysis buffer (1% SDS, 100 mM NaCl, 10 mM Tris [pH 8.0], and 1 mM EDTA). Protein concentration was measured by a commercial assay (Protein Assay kit; Bio-Rad Laboratories, Hercules, CA), according to the protocol recommended by the manufacturer. Equal amounts of protein (45 μg) from each sample were used for Western blot analysis with the enhanced chemiluminescence (ECL) reagents, as described previously. ²²  

In Situ Hybridization and Immunohistochemistry

Eyes from C57/BL6 mice (2 months old) were fixed and embedded in optimal cutting temperature (OCT) compound, as described previously. ²³ Sagittal cryosections of 6 μm in thickness were cut. All sections included the optic nerve. In situ hybridization was performed as described previously, ²³ with modifications. The sections were incubated at 37°C for 1 hour with prehybridization buffer (5× SSC, 40% deionized formamide). The biotin-labeled antisense (CAGCAGCGTCGGCCGCCTCCAGGTCCCGGTAGATGTGGCGAACCA) and sense (TGGTTCGCCACATCTACCGGGACCTGGAGGCGGCCGACGCTGCTG, positions 524-480 in mouse RDH10) oligonucleotide probes were diluted in hybridization buffer (10% dextran sulfate, 1% SDS, 4× SSC, 30% deionized formamide, 1× Denhardt’s solution, 10 mM dithiothreitol [DTT], 50 μg/mL yeast tRNA, 1 μg/mL denatured salmon sperm DNA, 25 μg/mL heparin, 100 μg/mL polyA, and 500 pM oligonucleotide probes) and then added onto the sections and the sections sealed. 

After 8 hours’ hybridization at 37°C, the sections were washed extensively and then treated with RNase A (10 mM Tris [pH 7.5], 1 mM EDTA, 0.5 M NaCl, and 0.02 mg/mL RNase A). The sections were blocked with 3% BSA and 10% goat serum in TBST for 30 minutes, and the signal was detected by immunohistochemistry (ELF97 Kit; Molecular Probes, Eugene, OR). The reaction was stopped by a rapid rinse 12 to 15 times with the stopping buffer (25 mM EDTA and 0.05% Triton X-100 in PBS [pH 7.2]). The sections were mounted with mounting medium. Immunohistochemistry was performed on sagittal cryosections from BALB/c mice, as described previously. ²⁴  

Preparation of Microsomal and Cytosolic Fractions

The transfected cells and bovine RPE were homogenized in a lysis buffer (15 mM Tris [pH 7.6], 1 mM DTT, 0.32 M sucrose) followed by three freeze-thaw cycles. The cell debris and nuclei were removed by low-speed centrifugation at 300g for 10 minutes. The supernatant was then spun at 30,000g for 20 minutes, and the supernatant was collected. To separate microsomal and cytosolic proteins, the supernatant from the last centrifugation was centrifuged again at 412,000g for 1 hour to separate the microsomes from cytosol. The supernatant (cytosolic fraction) was saved for an RDH activity assay. The pellet was washed five times. The final pellet (microsomal fraction) was subjected to Western blot analysis and an RDH activity assay. 

RDH Activity Assay

All the following procedures were performed under dim red light. To measure the production of all-trans retinal in RDH10-transfected cells, [³H] all-trans retinol (2.5 μCi, specific activity 52.5 Ci/mmol; Perkin Elmer-NEN, Boston, MA) was added to the culture medium of the COS cells 24 hours after transfection. The cells were incubated at 37°C in 5% CO₂ for 5 hours. Retinoids were extracted from the cells with methanol and hexane for HPLC analysis. For spiking, an internal standard of [³H] all-trans retinal was mixed with a fraction of the retinoid extract from the transfected cells, and the resultant retinoid profile was analyzed by HPLC. 

For the in vitro RDH activity assay, the aforementioned cytosolic and microsomal fractions were used. The microsomal pellets were resuspended in the reaction buffer containing 15 mM Tris, (pH 7.6), 1 mM DTT, 0.5% BSA, and 200 μM nicotinamide adenine dinucleotide (NAD) or nicotinamide adenine dinucleotide phosphate (NADP; Sigma, St. Louis, MO). The mixture was sonicated for 1 minute on ice to resuspend the microsomes. A total of 200 μg of the cytosolic or microsomal proteins was mixed with the reaction buffer and 1 μCi [³H] all-trans retinol to a final volume of 200 μL. The reaction mixture was incubated at 37°C for 1 hour with agitation. The reaction was terminated by the addition of 300 μL methanol and the retinoids extracted for HPLC analysis. 

HPLC Analysis

The retinoids were extracted by addition of 300 μL hexanes into the reaction mixture. After vortexing, the aqueous and organic phases were separated by centrifugation. Retinoids in the organic phase were dried under argon and redissolved in the mobile phase of HPLC (85.4% hexane, 11.2% ethyl acetate, 2% dioxane, and 1.4% octanol). The extracted retinoids were separated on a normal-phase column (250 mm × 4.6 mm; Lichrosphere SI-60 5U; Alltech, Deerfield, IL), by using the aforementioned mobile phase. Retinoids were analyzed with a commercial HPLC system (Waters, Milford, MA, accompanied by Millennium software and a 996 Photodiode Array Detector). An in-line flow scintillation analyzer (Packard Radiomatic 500TR; Perkin Elmer, Wellesley, MA) was used to monitor the radioactivity of the retinoids. The retinoids were also monitored by UV absorbance at 320, 360, and 370 nm. ²⁵  

The human, bovine, and mouse RDH10 cDNA sequences have been deposited into GenBank with accession numbers AF456765, AF456766, and AF456767, respectively (GenBank is provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD, and is available at http://www.ncbi.nlm.nih.gov/Genbank). 

Results

Cloning and Sequence Analysis of Human, Bovine, and Mouse RDH10 cDNAs

The cloned full-length human RDH10 cDNA consists of 1445 bp, including 260 bp of the 5′ untranslated region (UTR), 1023 bp of the coding region, and 162 bp of the 3′ UTR (with GenBank accession number AF456765). A polyadenylation signal AATAAA is located at position 1420-1425 in the 3′ UTR. A translation initiation codon ATG is located at position 261-263, according to the consensus sequence of Kozak. ²⁶ The termination codon is at position 1286-1288. This open reading frame (ORF) encodes a protein of 341 amino acid residues (Fig. 1A) . The deduced protein has a calculated molecular mass of 38,087 Da and an isoelectric point of 7.35. The global GenBank search showed that human RDH10 does not match any known gene but has high sequence homologies with a number of EST clones. 

To determine the sequence homology of this novel enzyme across species, we also cloned the bovine and mouse RDH10 cDNAs containing the complete coding region (GenBank accession numbers AF456766 and AF456767, respectively; Fig. 1A ). The mouse and bovine RDH10 cDNAs also encode a protein of 341 amino acids. The deduced bovine and mouse proteins have 100% and 98.6% amino acid sequence identity, respectively, with human RDH10 (Table 1 , Fig. 1 ). At the nucleic acid level, the bovine and mouse sequences showed 90% and 87% sequence identity, respectively, with the human RDH10 cDNA. 

Amino Acid Sequences of RDH10

The N-terminal region of RDH10, amino acids 1-24, is highly hydrophobic and is predicted to form a transmembrane helix. RDH10 retains all the motifs of the SDR superfamily members, including the invariant TGXXXGXG (position 42-49) in the nucleotide binding domain and YXXXK (position 210-214) in the substrate domain. It also contains the LXNNAG motif that is conserved among most of the SDRs ²⁰ (Fig. 1) . RDH10 shares 34% to 52% sequence similarity with other RDH SDRs at the amino acid level (Table 1) , indicating that RDH10 belongs to the SDR superfamily. 

Tissue Distribution of the RDH10 mRNA

As shown by Northern blot analysis, the RDH10 mRNA was detected as a single transcript of approximately 3 kb in length in the bovine RPE, but not in the other 10 bovine tissues analyzed. When the highly sensitive real-time RT-PCR was used, the RDH10 mRNA was detected at a high level in the RPE, but at relatively lower levels in most tissues analyzed, including the retina, kidney, pancreas, liver, lung, skeletal muscle, and brain (Fig. 2) . 

Cellullar Localization of RDH10

The cell type-specific expression of RDH10 in the mouse eye was determined at the mRNA level by in situ hybridization with an antisense oligonucleotide probe specific for RDH10. The antisense probe showed a strong hybridization signal in the mouse RPE, whereas the sense oligonucleotide probe from the same sequence region did not show any specific hybridization, demonstrating the specificity of the hybridization (Fig. 3A) . 

The cellular localization of RDH10 was confirmed at the protein level by immunohistochemistry. The specificity of the antibody against RDH10 was first demonstrated by Western blot analysis with COS cells transfected with RDH10/pcDNA6/V5/His, an expression construct of human RDH10 attached by a V5 epitope and a 6xHis tag. The RDH10 antibody recognized a single band with an apparent molecular mass of approximately 42 kDa, which matched the calculated molecular mass of recombinant RDH10 (RDH10 fused with the V5 epitope and the His tag). Moreover, the same band was recognized by an antibody against His-tag which was attached to the recombinant RDH10 (Fig. 3B) . The preimmune serum did not recognize any specific band in the same transfected cells (data not shown). No band was detected in untransfected COS cells under the same conditions by either antibody. 

The antibody against RDH10 showed intensive immunostaining in the mouse RPE. The specificity of the immunostaining was demonstrated by the negative control, which showed no immunostaining when the primary antibody was omitted (Fig. 3C) . 

Conversion of all-trans Retinol to all-trans Retinal in COS Cells Expressing RDH10

After incubation with [³H] all-trans retinol at 37°C for 5 hours, the HPLC spectrum of radioactive retinoids showed a novel peak in the RDH10-transfected COS cells, but not in the cells transfected with the empty pCDNA6 vector under the same conditions. This peak showed a retention time of 6.5 minutes, which matched the standard of [³H] all-trans retinal (Fig. 4) . 

The all-trans retinal peak generated in the RDH10-transfected cells displayed an absorbance spectrum with a λ_max of 368.2 nm, matching that of all-trans retinal. The identity of the formed all-trans retinal was further confirmed by spiking with the standard [³H] all-trans retinal (Fig. 4) . The results demonstrated that the RDH10-transfected cells converted all-trans retinol to all-trans retinal. 

Subcellular Localization of RDH10 Protein and Activity

The cytosolic and microsomal fractions from bovine RPE and transfected COS cells were subjected to Western blot analysis with the anti-RDH10 antibody. The antibody recognized a single band in the microsomal fractions from both the RPE and the RDH10-transfected COS cells (Fig. 5) . The apparent molecular mass of the band was approximately 38 kDa in the RPE and 42 kDa in the transfected cells, matching the calculated molecular masses of the native and recombinant RDH10. No band was detected in the cytosolic fraction of either the RPE or the transfected COS cells (Fig. 5) . 

RDH activity was measured in the cytosolic and microsomal fractions of the RDH10-transfected COS cells. The result showed that the RDH10 enzymatic activity exists only in the microsomal but not in the cytosolic fraction (data not shown), suggesting that RDH10 is a microsomal protein. 

Analysis of the atRDH Activity of RDH10

To determine the cofactor preference, RDH10 activity was analyzed in the microsomal fraction from the RDH10-transfected cells after extensive washes to remove the intrinsic NAD and NADP cofactors. The activity in microsomes of the untransfected COS cells was measured in parallel and subtracted from the results as the background. Only basal levels of all-trans retinal were generated in the microsomal fraction without the addition of exogenous NAD or NADP. The addition of NADP resulted in more efficient oxidation of all-trans retinol into all-trans retinal, when compared with the addition of NAD (Fig. 6A) , suggesting that RDH10 prefers NADP as the cofactor. 

Under the same assay conditions, RDH10 did not oxidize 11-cis retinol, 9-cis retinol, or 13-cis retinol into the respective retinal (pH 7.6, in the presence of NAD or NADP), indicating the substrate specificity of RDH10 (Fig. 6B) . When all-trans retinal was used as a substrate (pH 5.5 and pH 7.6), and NADH or NADPH was used as the cofactor under the assay conditions described in the Methods section, only a low level of all-trans retinol was generated by RDH10 (Fig. 6B) . 

Discussion

In the present study, we identified a novel enzyme, RDH10, belonging to the SDR superfamily. RDH10 is predominantly expressed in the RPE and specifically converts all-trans retinol into all-trans retinal. Therefore, RDH10 is the first all-trans retinol oxidase identified from the RPE and may be an important component of the RGR visual cycle. 

The SDR superfamily consists of multiple enzymes with low sequence homology and widely variable functions among the family members. The members of the SDR superfamily share several highly conserved, characteristic sequence motifs. ^{27 28} The RDH10 sequences contain all the strictly conserved motifs of the SDR superfamily, including the TGXXXGXG motif and the SDR-specific YXXXK active site, which are absent in the medium-chain dehydrogenase superfamily. ²⁹ Although its predicted molecular mass of 38 kDa is slightly higher than those of most SDR members (25–35 kDa), RDH10 aligns well with other SDRs (Table 1 , Fig. 1 ). The higher molecular mass can be ascribed to the potential N-terminal transmembrane domain containing 24 highly hydrophobic amino acids. Some known enzymes in the SDR family have molecular masses similar to RDH10. ²⁸ Moreover, activity assays showed that it catalyzed the oxidation of all-trans retinol to all-trans retinal. Therefore, based on sequence comparison and its activity, we conclude that RDH10 belongs to the SDR superfamily. 

RDH10 shares much higher sequence homologies with retSDR1 and retSDR2 than with other members in the SDR superfamily (Table 1) . This close evolutionary relationship suggests that there is a novel subfamily with RDH activities within the SDR superfamily (Fig. 1B) . Similar to retSDR1 (98.3% amino acid sequence identity between human and bovine retSDR1), ²⁰ the RDH10 sequence has been highly conserved during evolution (Table 1) . The 99% to 100% amino acid sequence identities between bovine, mouse, and human RDH10 are not common in the SDR family. For example, the human and bovine RDH5 have 90% amino acid sequence identity, and human and bovine prRDH have 85.5% identity at the amino acid level. ^{10 30} The highly conserved sequence during evolution suggests that RDH10 may have important physiological functions. 

Several SDR members have displayed atRDH activity; however, none of them is expressed in the RPE. ²⁰ Previous evidence has shown atRDH activity in the RPE. ³¹ This observation was supported by a recent study that confirmed atRDH activity in cultured RPE cells. ¹⁴ This atRDH activity appears membrane bound and is specific for all-trans retinol, but not for 11-cis retinol. ¹⁴ Our results showed that RDH10 is predominantly expressed in the RPE and located in the microsomes, consistent with the atRDH activity reported previously. ¹⁴ The oxidation of all-trans retinol to all-trans retinal by RDH10 was substrate specific—RDH10 did not oxidize 11-cis retinol, 13-cis retinol, or 9-cis retinol. Moreover, RDH10 preferred NADP as a cofactor. These properties of RDH10 are identical with those of the atRDH activity reported by Yang and Fong, ¹⁴ suggesting that RDH10 is responsible for the atRDH activity in the RPE. 

Regeneration of 11-cis retinal in the eye is essential for formation of visual pigment and normal vision. It has been shown that the isomerohydrolase isomerizes all-trans retinoids to 11-cis retinal in the dark, ^{1 8} whereas the RGR-dependent photic visual cycle generates 11-cis retinal after illumination. ^{9 12 13} Gene knockout of RGR results in reduced 11-cis retinal levels and abnormal ERG in the light-adapted eye, suggesting that this photic visual cycle is important for maintaining the steady state levels of 11-cis retinal in the light. ⁹ Moreover, this photic visual cycle is essential for normal photoreceptor development and function, because mutations in RGR are associated with retinitis pigmentosa. ³² The function of RDH10 may be to provide all-trans retinal, the substrate for RGR in the RPE, ¹⁴ and thus may play an essential role in the photic visual cycle. Alternatively, RDH10 may play a role in the generation of all-trans retinoid acid, because conversion of all-trans retinol to all-trans retinal is the rate-limiting step for generating all-trans retinoic acid, which is an important regulator of development and cell differentiation. ³³ This possible function may explain the low-level expression of RDH10 in several nonocular tissues. Thus, RDH10 may also play a role in the retinoic acid regulation of development and differentiation. 

²  Contributed equally to the study.

 Supported by National Eye Institute Grants EY12231, EY12600, and EY04939, a grant from Foundation Fighting Blindness, and a grant from the Research to Prevent Blindness, Inc.

 Submitted for publication May 20, 2002; revised July 10, 2002; accepted July 23, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Alignment of the deduced amino acid sequences of human (h), bovine (b), and mouse (m) RDH10. Important sequence patterns of SDR enzymes are labeled with consensus sequences. Underscored bold letters: unconserved residues among the three RDH10 sequences.

Table 1.

 

Comparisons between Human RDH10 and Other Retinoids Related to SDR

SDR	Genbank Accession Number	Amino Acid homology with hRDH10			Major Retinoid Substrate	Major Activity
				Similarity (%)			Identity (%)
bRDH10	AF456766	100	100	all-trans-retinol	Oxidation
mRDH10	AF456767	99.4	98.6	all-trans-retinol	Oxidation
hretSDR2	AF126780	51.9	42.8	unknown	Unknown
hretSDR1	AF061741	49.3	38.5	all-trans-retinal	Reduction
mRDH1	AY028928	39.6	28.8	all-trans-retinol	Oxidation
hRDH-E/hRoDH4	AF057034/AF086735	38.8	28.7	all-trans-retinol	Oxidation
hRDH5	U43559	38.7	29.3	11-cis-retinol	Oxidation
				9-cis-retinol
hRDH-TBE	AY017349	38.2	28.1	all-trans-retinol	Oxidation
m17βHSD9/mRDH8	AF103797	36.7	28.6	all-trans-retinol, 11-cis-retinol	Oxidation
rRoDH1	U18762	36.0	26.5	all-trans-retinol	Oxidation
rRoDH2	U33500	36.0	25.5	all-trans-retinol	Oxidation
rRoDH3	U33501	35.6	26.1	Unknown	Unknown
mCRAD2	AF056194	35.3	26.9	11-cis-retinol	Oxidation
mCRAD1	AF030513	34.4	24.5	11-cis-retinol	Oxidation
				9-cis-retinol
hPR-RDH	AF229845	34.2	26.6	all-trans-retinal	Reduction
mRDH4	AF013288	38.4	27.6	11-cis-retinol	Oxidation
				9-cis-retinol

 h, human; b, bovine; r, rat; m, mouse.

Figure 2.

 

Distribution of the RDH10 mRNA in bovine tissues. (A) Northern blot analysis detected the RDH10 mRNA only in the RPE, but not in other bovine tissues analyzed. (B) Real-time RT-PCR analysis of RDH10 and β-actin. The RDH10 mRNA was amplified from various tissues in addition to the RPE. Lane 1, lung; lane 2, brain; lane 3, heart; lane 4, liver; lane 5, kidney; lane 6, pancreas; lane 7, skeletal muscle; lane 8, spleen; lane 9, retina; lane 10, RPE; and lane 11, cornea.

Figure 3.

 

Cellular localization of RDH10. (A) In situ hybridization: the antisense probe detected the RDH10 mRNA in the RPE (400×) and the sense probe showed no hybridization. (B) Western blot analysis using 45 μg of microsomal proteins: both the anti-RDH10 and anti-His-tag antibodies detected RDH10 in cells transfected with the RDH10 expression vector (lane 1) but not in the untransfected cells (lane 2), demonstrating the specificity of the anti-RDH10 antibody. (C) Immunohistochemistry. The anti-RDH10 antibody showed strong immunostaining in the RPE (400×).

Figure 4.

 

Generation of all-trans retinal from all-trans retinol in the RDH10-transfected cells. [³H] all-trans retinol was added to the culture medium of COS cells transfected with the RDH10 expression construct or with empty vector and incubated at 37°C in the dark for 5 hours. Retinoids were extracted from the cells and analyzed by HPLC. (A) Control cells transfected with the empty pCDNA6 vector. Inset: absorbance spectrum of peak 2 with a λ_max of 324.5 nm (in the mobile phase of HPLC) which matches the λ_max of all-trans retinol. ³⁴ (B) Cells transfected with the RDH10 expression construct. Inset: absorbance spectrum of peak 1 with a λ_max of 368.2 nm (in the mobile phase of HPLC) which matches the λ_max of all-trans retinal. ³⁴ (C) [³H] all-trans retinal standard. (D) One fourth of the retinoids from (B) were spiked with the [³H] all-trans retinal standard. Peaks: 1, all-trans retinal; 2, all-trans retinol.

Figure 5.

 

Subcellular localization of RDH10. Western blot analysis detected RDH10 in the microsomal fractions of the COS cells expressing RDH10 and of the bovine RPE but not in the cytosolic fractions when 45 μg protein from transfected cells and 15 μg protein from bovine RPE were used. Lane 1: cytosol of bovine RPE; lane 2: microsomes of bovine RPE; lane 3: cytosol of the transfected COS cells; lane 4: microsomes of transfected cells; and lane 5: microsomes of untransfected COS cells.

Figure 6.

 

Characterization of RDH10 activity. The microsomal fraction from COS cells expressing RDH10 was used for the atRDH activity assay (200 μg microsomal protein per assay). (A) Cofactor preference of RDH10 activity. RDH10 activity was measured after the addition of NADP or NAD to different concentrations. All-trans Retinol was used as the substrate for this assay. The values represent the percentages of all-trans retinol converted to all-trans retinal (mean ± SEM, n = 2). (B) Substrate specificity of RDH10. Microsomal fractions from cells expressing RDH10 were incubated with all-trans, 9-cis, 11-cis, and 13-cis retinol in the presence of NADP and NAD or with all-trans retinal in the presence of NADPH and NADH. The resultant retinoids were analyzed by HPLC and the respective products quantified and expressed as percentages of total retinoids.