Biochemical Properties of Purified Recombinant Human β-Carotene 15,15′-Monooxygenase*

β-Carotene 15,15′-monooxygenase (BCO), formerly known as β-carotene 15,15′-dioxygenase, catalyzes the first step in the synthesis of vitamin A from dietary carotenoids. We have biochemically and enzymologically characterized the purified recombinant human BCO enzyme. A highly active BCO enzyme was expressed and purified to homogeneity from baculovirus-infectedSpodoptera frugiperda 9 insect cells. TheK m and V max of the enzyme for β-carotene were 7 μm and 10 nmol retinal/mg × min, respectively, values that corresponded to a turnover number (k cat) of 0.66 min−1 and a catalytic efficiency (k cat/K m ) of ∼105 m −1·min−1. The enzyme existed as a tetramer in solution, and substrate specificity analyses suggested that at least one unsubstituted β-ionone ring half-site was imperative for efficient cleavage of the carbon 15,15′-double bond in carotenoid substrates. High levels of BCO mRNA were observed along the whole intestinal tract, in the liver, and in the kidney, whereas lower levels were present in the prostate, testis, ovary, and skeletal muscle. The current data suggest that the human BCO enzyme may, in addition to its well established role in the digestive system, also play a role in peripheral vitamin A synthesis from plasma-borne provitamin A carotenoids.

Retinol, also referred to as vitamin A, is a fat-soluble polyisoprenoid essential for tissue development, growth, and vision. Retinol can be either ingested or synthesized within the body from dietary carotenoids. Preformed retinol is found almost exclusively in animals in the form of esters of fatty acids. It is hydrolyzed during the process of digestion, absorbed in the free form, re-esterified with fatty acids within the intestinal mucosa, and transported to the liver via the lymphatic route associated with chylomicrons (1). ⌻he major substrate for the in vivo synthesis of retinol is the plant carotenoid ␤-carotene. Of the more than 600 different carotenoids isolated from nature, ϳ50 possess biological activity; hence, these compounds are termed provitamin A carotenoids (2)(3)(4).
In vivo studies in humans show that the majority of the ingested ␤-carotene is cleaved at the central carbon 15,15Јdouble bond to form two molecules of retinal (retinaldehyde) (5,6), and consequently, the enzyme that catalyzes this first step in vitamin A synthesis in the intestinal mucosa was named ␤-carotene 15,15Ј-dioxygenase (BCO) 1 (7). Although the enzyme activity was first described in the mid-1950s, it was not until 1965 that the research groups of Goodman (8) and Olson (7) independently demonstrated that the cleavage of ␤-carotene to retinal could be studied in vitro by using soluble enzyme preparations from the intestine and liver. Characterization of the native enzyme was performed with a 100,000 ϫ g supernatant fraction or an ammonium sulfate precipitate thereof. It was shown that the reaction is dependent on molecular oxygen and that nicotinamide dinucleotide cofactors are not required for catalysis. Also, the facts that the reaction was inhibited by the addition of iron chelating agents and that cyanide, which inhibits ferric protoporphyrin enzymes, did not attenuate the conversion of ␤-carotene to retinaldehyde suggested that BCO was a nonheme iron-containing enzyme. Furthermore, the native enzyme had the ability to cleave the 15,15Ј-double bond of a variety of carotenoids other than ␤-carotene, including ␣-carotene, ␤-apocarotenols, and ␤-apocarotenals (9 -11). The Michaelis-Menten constant (K m ) for the native enzyme of a variety of animal species relative to ␤-carotene has been determined to be in the 1-10 M range. The pH optimum for the reaction was in the slightly alkaline range, and the enzyme was inhibited by sulfhydryl alkylating reagents such as p-chloromercuribenzoate and N-ethylmaleimide. BCO was either protected or activated by sulfhydryl reducing agents such as cysteine and glutathione (7, 10 -13).
The human BCO cDNA encodes a hydrophilic protein of 547 amino acids with a predicted molecular weight of 62,637 (14). Amino acid comparison of the human BCO with the mouse (15), rat (GenBank TM accession number NM_053648), chicken (16), zebrafish (GenBank TM accession number AJ290390), and Drosophila (17) enzymes show sequence identities of 85, 84, 67, 56, and 22%, respectively. Based on biochemical and amino acid sequence data, it has been proposed that BCO belongs to the nonheme iron-containing dioxygenase family. However, recently it was demonstrated that the reaction mechanism of enzymatic cleavage of the central carbon 15,15Ј-double bond in ␤-carotene involves a monooxygenase-type mechanism (18). By using a partially purified chicken intestine BCO preparation and isomerically pure ␣-carotene as substrate, it was shown that both 17 O 2 and H 2 18 O were incorporated into the two retinal products. Thus, the BCO enzyme was renamed ␤-carotene 15,15Ј-monooxygenase (19). The human BCO gene structure and chromosomal localization were also recently reported (14). However, no biochemical characterization of a purified human enzyme has been described.
In the current study, we describe a detailed biochemical and enzymological characterization of the purified recombinant human BCO enzyme. Furthermore, a comprehensive analysis of tissue-specific expression is described.
Plasmid Constructions-Standard molecular biology techniques were used (20). The complete coding sequence of human BCO cDNA was obtained from a liver cDNA pool (CLONTECH Laboratories, Inc., Palo Alto, CA) by PCR using the following primers: hBCOf2 (sense) 5Ј-AAATCGATCTCCCTCGGCACC-3Ј and BacBCOr (antisense) 5Ј-CG-GAATTCTTATCAGGTCAGAGGAGCCCCGTG-3Ј, corresponding to nucleotides 191-211 and 1863-1842, respectively, in the human cDNA sequence (GenBank TM accession number AK001592). The PCR product was first subcloned into the pCR2.1-TOPO vector (Invitrogen) using the extra deoxyadenosine added by the Taq polymerase and then subcloned in the pCMV-Sport2.0 vector (Invitrogen) using EcoRI.
The insert in the pCMV-hBCO vector was subcloned into pBluescript SK Ϫ (Stratagene, La Jolla, CA) using KpnI and XbaI, and the 3Ј part of the hBCO coding insert was removed by digestion with SalI and AccI. A 494-bp fragment containing the corresponding coding region part but with six histidine codons inserted before the stop codon was obtained by PCR using primers CHis5 (sense) 5Ј-AAGTCTACTGCCAGCCG-GAATTTCTTTATGAAGGC-3Ј and CHis3 (antisense) 5Ј-TTGTCGAC-TATCAATGATGGTGATGATGGTGGGTCAGAGGAGCCCCGTGGC-3Ј. The fragment was then digested with SalI and AccI and subcloned into the SalI/AccI-digested pBluescript-hBCO vector, resulting in pBluescript-hBCO-His. The pFastBac1-hBCO-His donor plasmid for subsequent production of recombinant baculovirus was constructed by transfer of the pBluescript vector insert into pFastBac1 (Invitrogen) using EcoRI and SalI.
Insect Cell Culture, Expression, and Purification of Recombinant Human BCO Protein-Spodoptera frugiperda 9 cells were maintained in monolayer cultures with Sf-900 II SFM (catalog number 10902-096; Invitrogen), or Grace's insect medium (catalog number 11605-094; Invitrogen) supplemented with 10% fetal calf serum. Using the Bac-To-Bac TM baculovirus expression system (Invitrogen) and the pFastBac1-hBCO-His plasmid (above), recombinant baculovirus was obtained. Expression of the C-terminally histidine-tagged recombinant human BCO protein was performed by infecting S. frugiperda 9 cells at a multiplicity of infection of 10 and harvesting the cells after 3 days culture. The cells from three 225-cm 2 flasks were detached and collected by centrifugation 5 min at 500 ϫ g, then resuspended in 2.7 ml of chilled 50 mM sodium phosphate buffer (pH 7.0), kept on ice 5 min, and subsequently homogenized by 20 strokes in a glass tissue grinder. After centrifugation 30 min at 10,000 ϫ g at 4°C, the supernatant (S10) was collected, and sodium chloride was added to a final concentration of 100 mM. The S10 was then immediately incubated with 0.5 ml of Talon CellThru resin (CLONTECH) with gentle agitation for 30 min at 4°C. The resin was washed three times with 10 column volumes of 50 mM sodium phosphate buffer (pH 7.0) containing 100 mM sodium and 5 mM imidazole (wash/extraction buffer), and then bound protein was eluted by two 15-min incubations with one column volume of wash/extraction buffer containing 300 mM sodium chloride and 150 mM imidazole each time. The slurry was centrifuged at 500 ϫ g for 2 min after each wash and elution step, and the supernatant was collected. The purified hBCO-His enzyme was stored on ice until used.
BCO Enzyme Assay and HPLC Analysis of Reaction Products-All of the reactions were performed in a volume of 100 l in an assay buffer consisting of 100 mM Tricine-KOH (pH 8.0), 125 mM sodium chloride, 10 M Fe 2 SO 4 , 5 mM Tris(2-carboxylethyl)phosphine hydrochloride (TCEP) (Pierce), and 1% (w/v) 1-S-octyl-␤-D-thioglucopyranoside (OTG) (Pierce). The carotenoid concentration in the reactions varied between 0.125 and 128 M, and the hBCO-His enzyme amount varied between 60 and 1000 ng/reaction. The assays were set up and run under subdued light as follows; carotenoid in hexane was added to 25 l of 4% (w/v) OTG in 2-ml Eppendorf tubes, and the solvent was evaporated. The assay buffer and enzyme, in some cases preincubated with various potentially inhibitory compounds for 30 min on ice, were then added to the carotenoid/detergent mix, and the tubes were incubated for 5 min to 1 h at 37°C with gentle agitation (70 rpm). HPLC analyses of reaction products were performed essentially as described by During et al. (21).
Briefly, 25 l of 37% (v/v) formaldehyde was added, and the incubation was continued for 10 min at 37°C. For extraction of the products, 250 l of acetonitrile was added, and the tube was vortexed and kept on ice for 5 min. After centrifugation for 10 min at 10,000 ϫ g at 4°C, the supernatant was separated on a 4.6 ϫ 150-mm Phenomenex LUNA 3 C18 column (Phenomenex, Torrance, CA) or a 4.6 ϫ 150-mm XTerra MS C18 3.5 m column (Waters, Milford, MA) in a mobile phase consisting either of 90% acetonitrile, 10% water, 0.1% (w/v) ammonium acetate or of 85% acetonitrile, 15% water, 0.1% (w/v) ammonium acetate with a flow rate of 1 ml/min (Waters 501 HPLC pump) and UV detection (Waters 484 tunable absorbance detector) at 380 nm. The enzyme kinetics were calculated using GraFit Version 5.0.1 (Erithacus Software Limited, Horley, UK). This program fits the data to the Michaelis-Menten equation using nonlinear regression analysis. For spectral scan analyses of the products, the 4.6 ϫ 150-mm XTerra MS C18 3.5 m column was used under the same conditions as above, but with a LC-10AT liquid chromatograph, SPD-M10AVP diode array detector, and Class-VP chromatograph data system, version 4.2 (Shimadzu, Columbia, MD).
Retinal formed during the reactions was quantified from its peak height by using a standard curve obtained by incubating 1-50 pmol of all-trans-retinal in a 100 mM Tricine-KOH (pH 8.0) buffer, containing 125 mM NaCl, 10 M Fe 2 SO 4 , 5 mM TCEP, 1% (w/v) OTG, and 250 ng of heat-inactivated BCO for 15 min at 37°C with gentle agitation (70 rpm). The samples were then treated the same way as described above for the BCO assay samples and were separated on a 4.6 ϫ 150-mm Phenomenex LUNA 3 C18 column in a mobile phase consisting of 90% acetonitrile, 10% water, and 0.1% (w/v) ammonium acetate.
Monoclonal Antibody Production-The synthetic peptide [C]RN-RKEQLEPVRAKVTGK, corresponding to amino acid residues 7-23 in human BCO, was coupled to keyhole limpet hemocyanin using mmaleimidobenzoyl-N-hydroxysulfosuccinimide ester (Pierce) (22), and used for immunization of mice as described previously (23). Hybridomas were established and screened by an enzyme-linked immunosorbent assay using the synthetic peptide as antigen as described previously (23). The anti-BCO monoclonal antibody, designated mAb-1-11, was determined to belong to the IgG 1 / subclass as revealed by an enzymelinked immunosorbent assay-based mouse typer subisotyping kit (Bio-Rad).
Gel Electrophoresis, Immunoblotting, and Protein Quantitation-SDS-polyacrylamide electrophoresis was performed according to the method of Laemmli (24), and the proteins were either detected in the gel with Coomassie Brilliant Blue R or with silver staining using a Silver Stain Plus kit (Bio-Rad). Alternatively, the proteins were transferred onto an Immobilon-P polyvinylidene difluoride membrane (Millipore, Bedford, MA) using a previously published method (25). For immunodetection of BCO proteins, mAb-1-11 containing tissue culture medium was incubated 1 h at room temperature, followed by a secondary horseradish peroxidase-conjugated goat anti-mouse IgG polyclonal antibody (catalog number 170-6516; Bio-Rad) diluted 1:10,000, and incubated for 1 h at room temperature. Antibody binding was detected by chemiluminescence using an ECL kit (Amersham Biosciences). The protein concentrations were determined using a Coomassie Plus protein assay reagent (Pierce).
Immunocytochemistry-The human hepatocellular carcinoma cell line, HepG2 (ATCC HB-8065), maintained in Dulbecco's modified Eagle's medium (catalog number 11995-065; Invitrogen) supplemented with 10% fetal calf serum, 10 mM HEPES, and 1% penicillin/streptomycin was cultured on 12-mm diameter microscope glass covers. The cells were transiently transfected with pCMV-hBCO using Fugene 6 (Roche Diagnostics Corporation, Indianapolis, IN) according to the manufacturer's instructions. 36 h after transfection, the cells were rinsed twice with cold PBS (10 mM sodium phosphate, pH 7.4, 150 mM sodium chloride), fixated with cold methanol for 10 min at Ϫ20°C, then rinsed twice with cold PBS, and blocked 1 h at room temperature with PBS containing 1% (w/v) bovine serum albumin. The cells were then incubated 1 h at room temperature with undiluted mAb-1-11 monoclonal antibody containing tissue culture medium, rinsed three times with PBS, and then incubated 40 min at room temperature with fluo-rescein isothiocyanate-conjugated donkey anti-mouse IgG (Jackson Im-munoResearch, West Grove, PA) diluted 100 times in PBS containing 1% (w/v) bovine serum albumin. The samples were analyzed by confocal microscopy using a Zeiss axioplan research microscope with a Bio-Rad MRC600 laser scanning head.
RNA Blotting-A human digestive system 12-lane MTN ® blot (catalog number 7782-1), a human 12-lane MTN ® blot (catalog number 7780-1) containing ϳ1 g of poly(A) ϩ RNA/lane, and a human MTN ® blot II (catalog number 7759-1) containing ϳ2 g of poly(A) ϩ RNA/lane were purchased from CLONTECH. A 552-bp fragment, corresponding to residues 714 -1265 in the human BCO cDNA (GenBank TM accession number AK001592) was obtained by PCR and radiolabeled with [␣-32 P]dCTP (Amersham Biosciences) using a Rediprime TM II kit (Amersham Biosciences). ExpressHyb TM hybridization solution (CLON-TECH) was used for 2 h before hybridization at 68°C and for overnight hybridization at 68°C with 2 ϫ 10 6 cpm probe/ml. Washing was performed according to the blot manufacturer. Autoradiography was carried out at Ϫ80°C with Hyperfilm TM MP (Amersham Biosciences) and Quanta Rapid intensifying screens (DuPont) for 3 days. The same filters were subsequently hybridized with a human ␤-actin probe and exposed at room temperature for 1 h.

RESULTS
Identification and Analysis of Human BCO cDNA and Protein-Using the Drosophila BCO as query sequence (17), we performed BLAST searches of the GenBank TM data base and identified a human protein (GenBank TM accession number AK001592) with a 22% sequence identity to the Drosophila protein.
To investigate whether this was the human BCO homologue, we isolated the corresponding cDNA from human liver cDNA by the PCR. DNA sequencing of multiple independently amplified cDNAs revealed an aspartate instead of a glycine in position 302 of the BCO protein. These results were in concordance with the human genomic BCO sequences present in the Celera data base, and with the recently published gene sequence (GenBank TM accession number NP_059125) (14). The cDNA was subcloned into a FastBac vector with a hexahistidine tag added to the C terminus of the protein. The recombinant protein was expressed in S. frugiperda 9 insect cells and purified to homogeneity by Co 2ϩ column chromatography. Fig.  1A shows a Coomassie-stained polyacrylamide gel of uninfected (lane 1) and infected (lane 2) insect cell homogenates, and 100 ng of the purified histidine-tagged protein of ϳ64 kDa (lane 3), consistent with the predicted molecular weight of 63,460 (including the six histidines). To further examine the purified protein, 650 ng of the protein was analyzed by SDS-polyacrylamide gel electrophoresis followed by silver staining. The purified protein showed a major band of ϳ64 kDa (lane 4). Fig 1B  shows an immunoblot with a monoclonal antibody (mAb-1-11) against a synthetic peptide derived from the putative human BCO protein. Taken together, these data show that isolation of a highly purified recombinant protein was achieved. Fig. 2 shows the results of an in vitro time course experiment in which 2.5 M ␤-carotene was used as substrate with 60 ng of purified BCO protein. The products at the different time points were analyzed by reverse-phase HPLC and found to migrate with the same retention time as an all-trans-retinal standard, and the formation of product was linear up to 20 min. Preincubation of the purified protein at 95°C for 5 min eliminated enzyme activity (data not shown). Fig. 3 shows the spectral properties of the enzyme reaction products by photodiode array detector on-line analysis after separation by HPLC. The enzyme dependent product after incubation with ␤-carotene (Fig.  3, solid line) has an absorbance spectrum virtually identical to that of authentic all-trans-retinal (Fig. 3, dashed line C). Identical absorbance spectrum and retention time were seen for one of the enzyme-dependent reaction products when ␤-cryptoxanthin was used as substrate, indicating that all-trans-retinal is one of the two reaction products produced from this substrate (data not shown). Taken together, these data showed that the purified recombinant protein catalyzed the central cleavage of ␤-ionone ring containing carotenoids, suggesting that the PCRamplified cDNA encoded a human homologue of the Drosophila BCO.
Characterization of Purified Recombinant BCO Enzyme-The biochemical properties of the purified protein were investigated by performing enzyme assays, gel filtration chromatography, and immunocytochemistry. Table I summarizes data on the effect of seven different detergents on BCO enzyme activity. The data show that maximal enzyme activity was achieved when OTG was used at a final concentration of 1% (w/v). Table  II shows the effect of four different sulfhydryl reducing agents on enzyme activity. The data are consistent with the observations that sulfhydryl reducing agents are imperative for maximal enzyme activity in vitro (11)(12)(13) and that TCEP was the most effective agent when used at concentrations above 0.5 mM. In contrast to the native enzyme (11), inclusion of ferrous iron in the enzyme assay buffer was not essential for maximal activity of the recombinant human enzyme (data not shown).
The data shown in Fig. 4 indicate that the enzyme demonstrates a slightly alkaline pH optimum, similar to the activity present in intestinal homogenates of rat (12), rabbit (10), hog  R (lanes 1-3) and silver staining (lane 4). B, proteins were transferred to a nitrocellulose membrane and incubated with 10 g/ml of mAb-1-11, and the antibody-antigen complexes were visualized by a chemiluminescence method as described under "Experimental Procedures." The film was exposed for 10 s. The positions of prestained molecular size markers are shown on the left.

FIG. 2. In vitro time course experiment with purified BCO and
␤-carotene as substrate. Reaction velocity (pmol product formed per reaction) as a function of time (min) is plotted for a reaction with 60 ng of purified BCO and 2.5 M ␤-carotene at 37°C. The assays were performed as described under "Experimental Procedures," and product quantitation was performed by reverse-phase HPLC analysis as described under "Experimental Procedures." (13), and guinea pig (11). The experiments of Fig. 5 show that the enzyme is sensitive to the metal chelating agents ␣,␣bipyridyl and o-phenanthroline, as well as the sulfhydryl alkylating agents N-ethylmaleimide and p-chloromercuribenzoate. These findings are consistent with data regarding the biochemical properties of the intestinal BCO activity from different species. These results suggested that the purified recombinant human BCO enzyme possessed similar biochemical properties as compared with the partially purified native enzyme from a variety of animal species.
To ascertain whether BCO is a monomer or oligomer in solution, gel filtration chromatography experiments were performed with the purified enzyme. Fig. 6 shows a typical elution profile as assessed by enzymatic activity and immunoblotting of different Sephadex S-300 column chromatography fractions. When chromatography was performed under buffer conditions identical to those used in the enzyme assays, a majority of the enzymatically active and immunodetectable protein migrated at ϳ230 kDa. These data suggested that the purified BCO enzyme was a tetramer in its enzymatically active form. It should be pointed out that because the gel filtration experiments were performed with 1% OTG in the elution buffer, it is conceivable that the enzyme migrated on the column in experimentally created detergent-protein micelles; hence, the oligo-    The intracellular localization of BCO was determined by immunocytochemistry. HepG2 cells were transfected with a mammalian expression vector encoding the full-length human BCO, and the expressed enzyme was visualized with the monoclonal antibody mAb-1-11 raised against human BCO and a secondary antibody, fluorescein isothiocyanate anti-mouse IgG. When the cells were examined under a laser confocal microscope, intense cytoplasmic staining was observed (Fig. 7A). There was no apparent reticular or nuclear staining visible.
In Vitro Kinetic Analysis of Purified BCO with Carotenoids as Substrates-To analyze the substrate specificity and kinetic constants for purified recombinant BCO, we performed in vitro assays with the carotenoid substrates ␤-carotene, ␤-cryptoxanthin, zeaxanthin, and lycopene. Incubation conditions were designed to ensure less than 10% substrate conversion with concentrations in the 0.25-32 M range for ␤-carotene and in the 2.5-256 M range for ␤-cryptoxanthin. The formation of product was linear with respect to protein over a 60 -1000 ng range and time over a 20-min period ( Fig. 2 and data not shown). Fig. 8 shows the reaction velocity (nmol product formed/mg protein ϫ min) as a function of substrate concentration (M) plotted for a 15-min reaction with 250 ng of BCO (in duplicate). Product quantitation was performed as described under "Experimental Procedures"; however, when ␤-cryptoxanthin, zeaxanthin, and lycopene were used as substrates, the mobile phase was 15% water and 0.1% ammonium acetate in acetonitrile (for retinal, R t ϭ 12.2 min; for 3-hydroxyretinal, R t ϭ 3 min). The K m and V max of the enzyme with ␤-carotene as substrate were 7.1 Ϯ 1.8 M and 10.4 Ϯ 3.3 nmol retinal/mg ϫ min, respectively (Fig. 8C), values that correspond to a turnover number (k cat ) of 0.660 min Ϫ1 , assuming that each subunit is catalytically active. This output corresponds to a catalytic efficiency (k cat /K m ) of 93,000 M Ϫ1 ⅐min Ϫ1 . ␤-Cryptoxanthin was also accepted as substrate with an apparent K m of 30.0 Ϯ 3.8 M and a V max of 0.9 Ϯ 0.2 nmol/mg ϫ min, values that correspond to a turnover number of 0.057 min Ϫ1 and a catalytic efficiency of 1,900 M Ϫ1 ⅐min Ϫ1 . Zeaxanthin and lycopene were not substrates for the BCO enzyme under the assay conditions used.
Tissue Distribution of BCO mRNA-RNA blots of human poly(A) ϩ RNA show that the BCO mRNA is expressed through-out the intestinal tract, with the highest level in jejunum, the segment of the small intestine that has been shown to possess the highest BCO enzyme activity (Fig. 9A) (26). The predominant mRNA is ϳ2.6 kb, and a less abundant mRNA of ϳ5.8 kb is also detected. Fig. 9B shows that in addition to the small intestine, BCO mRNA is also present at high levels in the liver and kidney and that lower levels are in the prostate, testis, ovary, colon, and skeletal muscle. DISCUSSION This paper describes the cloning, expression, purification, and characterization of human BCO, a cytosolic enzyme that catalyzes the first step in the in vivo synthesis of vitamin A by cleaving the central carbon 15,15Ј-double bond in provitamin A carotenoids. The identification of the present BCO enzyme as the human homologue of BCO characterized from a variety of species was based on multiple criteria, including: 1) the amino acid sequence has a high sequence identity with previously cloned BCO proteins belonging to the nonheme iron containing oxygenases; 2) the absorbance spectra of the enzyme dependent reaction product with ␤-carotene as substrate is identical to that of all-trans-retinal; 3) the recombinant enzyme shares biochemical properties with the native BCO enzyme reported from a variety of animal species; and 4) the BCO mRNA is present at high levels along the whole intestinal tract, with the highest levels in jejunum, in concordance with reports on BCO enzyme activities in different segments of the intestine.
Since the initial in vitro characterizations of the intestinal and liver BCO enzymes by the research groups of Goodman and Olson almost 40 years ago, a considerable body of work has focused on the biochemical properties of the native BCO enzyme of partially purified preparations from different animal species such as rat (7,12), rabbit (11), hog (13), guinea pig (11), and chicken (18,19). However, no data on the biochemical properties of the native human isoenzyme have been reported. Furthermore, earlier efforts to purify the native mammalian enzyme to homogeneity have proven unsuccessful. A breakthrough in the area of BCO research came in 2000, when von Lintig and Vogt (17) isolated a cDNA encoding the fruit fly Drosophila melanogaster BCO by employing an expression cloning strategy using Escherichia coli cells genetically engi-neered to synthesize ␤-carotene de novo. This work was followed shortly after by other studies by Wyss et al. (15,16), who partially purified the BCO enzyme from chicken intestine and cloned the chicken and mouse BCO cDNAs. At the same time, we performed GenBank TM data base searches and identified a human BCO cDNA that was found to encode a hydrophilic protein of 547 amino acids with a predicted molecular weight of 62,637 and that has a significant sequence homology with the fruit fly, chicken, and mouse isozymes.
Because expression of enzymatically active BCO in bacteria proved difficult, we chose the baculovirus system as a source of recombinant protein. The human BCO enzyme used in the present study was expressed in insect cells and purified to apparent homogeneity. The purified enzyme possesses a specific activity of 10 nmol retinal formed/mg protein ϫ min when ␤-carotene is used as substrate under optimal reconstitution conditions. This specific activity is 300-fold higher than that of  9. Distribution of BCO mRNA in human tissues. 32 P-Labeled human BCO cDNA probes were hybridized to poly(A) ϩ RNA (1 or 2 g/lane) from the indicated human tissue as described under "Experimental Procedures." The human digestive system filter (A) and human multi tissue filters (B) were exposed to Amersham Biosciences Hyperfilm MP with two intensifying screens at Ϫ80°C for 3 days. The same filters were subsequently hybridized with a human ␤-actin probe and exposed at room temperature for 1 h. The positions of molecular size markers are shown on the left. PBL, peripheral blood leukocytes. the partially purified native enzyme that was enriched 226-fold from chicken intestinal mucosa (15), and 300 -1600-fold higher than that of the bacterially expressed and purified mouse BCO histidine (27) or glutathione S-transferase fusion proteins (28). The recombinant human enzyme catalyzes the cleavage of ␤-carotene and ␤-cryptoxanthin, but not of zeaxanthin or lycopene, suggesting that at least one half-site of an unsubstituted ␤-ionone ring substrate is imperative for cleavage of the central carbon 15,15Ј-double bond (Fig. 8C). This requirement coincides with the fact that ␤-carotene and ␤-cryptoxanthin are considered to have nutritional value as provitamin A carotenoids, in contrast to zeaxanthin and lycopene. Interestingly, the BCO enzyme shows a 4-fold higher K m when ␤-cryptoxanthin is used as substrate as compared with ␤-carotene. Because ␤-carotene has two unsubstituted ␤-ionone ring half-sites and ␤-cryptoxanthin consists of one unsubstituted ␤-ionone ring half-site and one 3-hydroxylated ␤-ionone ring, it is conceivable that each half-site may bind to separate BCO subunits. This hypothesis is plausible because our data show that the enzymatically active BCO enzyme is an oligomer in solution. Another explanation could simply be that one molecule of enzyme binds one molecule of substrate in an extended hydrophobic pocket with the active site positioned in the middle. It should be mentioned that our substrate specificity data on the recombinant human isozyme corroborates data obtained with the partially purified chicken enzyme (19). Alignment of the BCO amino acid sequences from the presently cloned species shows that ϳ8% of highly hydrophobic amino acids (tryptophane, tyrosine, and phenylalanine) are conserved in the enzyme, and it thus is likely that at least some of these hydrophobic residues participate in the formation of the substrate-binding pocket.
The recombinant human isozyme was shown to be sensitive to micromolar concentrations of the metal chelating agents ␣,␣-bipyridyl and o-phenanthroline. This inhibition is similar to that of the native enzyme and corroborates earlier suggestions that iron is the metal ligand in the active site of the enzyme. The crystal structures of a number of oxygen-activating mononuclear nonheme iron enzymes have revealed a common structural motif in which two histidines and one carboxylate occupy a face of the iron coordination sphere, termed the 2-His-1-carboxylate facial triad (29,30). Therefore, the study of histidine and aspartate/glutamate residues that are conserved in BCOs across animal species will be imperative to define the active site of the enzyme.
The observation that BCO is dependent on reduced sulfhydryl groups for maximal enzyme activity in vitro is intriguing. We know from denaturing SDS-polyacrylamide gel electrophoreses experiments that the purified BCO migrates as a 64-kDa band regardless of preparation of protein samples in the presence or absence of the sulfhydryl reducing agent 2-mercaptoethanol (data not shown), thus excluding the existence of intramolecular or intermolecular disulfide bonds. Because cysteine residues in general are not involved in enzyme catalysis in nonheme iron containing enzymes, it is instead conceivable that alkylation of one or several free thiols by reagents such as N-ethylmaleimide or p-chloromercuribenzoate prevents substrate binding to the enzyme. It is also possible that substrate bound to the enzyme may block the alkylating reagents from exerting their inhibitory effect and thereby help to define crucial cysteines in the substrate-binding pocket. Interestingly, sequence alignment analysis of the human, mouse, rat, and chicken BCO isozymes reveal six cysteines that are conserved among these family members.
The human BCO gene is highly expressed in the digestive tract and the liver. These tissues have been most commonly used as sources of the native enzyme, and its physiological role in those tissues is well established. However, the expression of the gene in nondigestive tissues including kidney, testis, ovary, prostate, and skeletal muscle is much less clear. Because a majority of provitamin A carotenoids are converted to vitamin A (retinolesters) in epithelial cells of the intestinal mucosa and then transported in chylomicrons via the lymphatic system to the liver for storage, ample amounts of vitamin A are normally available to peripheral tissues via the general circulation. We speculate that the capacity to convert circulating provitamin A carotenoids to vitamin A in peripheral tissues is an alternate pathway for local vitamin A supply in situations of special need or during deficient states. It is well known that testes, for example, require retinoids for spermatogenesis (31,32), and local cleavage of carotenoids could possibly uphold a steady level of retinoids for a time in situations where the organism is nutritionally deprived of vitamin A. Hence, future studies require the study of cell type-specific expression of BCO and the enzymes that constitute the pathway of retinoid biosynthesis from carotenoids.