β-Carotene Conversion into Vitamin A in Human Retinal Pigment Epithelial Cells

Abstract

purpose. Vitamin A is essential for vision. The key step in the vitamin A biosynthetic pathway is the oxidative cleavage of β-carotene into retinal by the enzyme β,β-carotene-15,15′-monooxygenase (BCO). The purpose of the study was to investigate β-carotene metabolism and its effects on BCO expression in the human retinal pigment epithelial (RPE) cell line D407.

methods. BCO mRNA and protein expression were analyzed by real-time quantitative PCR and Western blot analysis, respectively. BCO activity was assayed in protein extracts isolated from D407 cells. The conversion of β-carotene to retinoids was determined by measuring retinol levels in D407 cells on β-carotene supplementation.

results. By RT-PCR, BCO mRNA was detected in D407 cells, bovine RPE, and retina. Western blot analyses revealed the presence of BCO at the protein level in D407 cells. Exogenous β-carotene application to D407 cells resulted in a concentration (75% at 0.5 μM and 96% at 5 μM; P < 0.05)- and time (127% at 2 hours and 97% at 4 hours in 5 μM β-carotene, P < 0.05)-dependent upregulation of BCO mRNA expression. Application of exogenous retinoic acid downregulated BCO mRNA levels at higher concentrations (1 μM; −96%, P < 0.0005) and upregulated it at a lower concentration (0.01 μM; 399%, P < 0.005). The RAR-a-specific antagonist upregulated BCO expression by sixfold (P < 0.005). Tests for enzymatic activity demonstrated that the mRNA upregulation resulted in enzymatically active BCO protein (7.3 ng all-trans-retinal/h per milligram of protein). Furthermore, D407 cells took up β-carotene in a time-dependent manner and converted it to retinol.

conclusions. The results suggest that BCO is expressed in the RPE and that β-carotene can be metabolized into retinol. β-Carotene cleavage in the RPE may be an alternative pathway that would ensure the retinoid supply of photoreceptor cells.

Retinoids (vitamin A and its derivatives) are essential components in vision and are also essential in the maintenance of normal growth and development, immunity, and reproduction. ¹ Animals, in general, are unable to synthesize retinoids de novo but rely on a dietary supply of these compounds in the form of vitamin A, mainly as retinyl esters from animal food sources or in the form of provitamin A carotenoids from plants. ² Of all known carotenoids, β-carotene is believed to be the most important in human nutrition. For strict herbivores, dietary carotenoids are the sole source of vitamin A. ³ Benefits of β-carotene supplementation in combating vitamin A deficiency have been demonstrated in several studies, ^{4 5 6} including the reversal of abnormal dark adaptation in humans. ⁷  

Visual pigments (rhodopsins) are retinal photoreceptor proteins of bipartite structure consisting of the transmembrane protein (opsin) and a light-sensitive chromophore (11-cis retinal). ^{8 9} Photoisomerization of the 11-cis to the all-trans-retinylidene group triggers a sequence of reactions that eventually result in the conversion of light energy into a photoreceptor electrical response. This reaction is terminated by the release of all-trans-retinal from the opsin molecule. The regeneration of rhodopsin requires a constant supply of 11-cis-retinal. The supply is obtained through a multistep pathway called the visual cycle, via the photoreceptors and the adjacent RPE. ⁹  

The eyes depend on blood circulation for vitamin A. Alternatively, vitamin A may be generated in the RPE from provitamin A carotenoids. The key step in vitamin A formation is the oxidative cleavage of provitamin A carotenoids by β,β-carotene 15,15′-monooxygenase (BCO), formerly known as β-carotene dioxygenase (EC 1.14.99.36). BCO cleaves β-carotene symmetrically into two molecules of all-trans-retinal. The activity of BCO was reported in cytosolic preparations of rat liver and intestine as early as the 1960s. ^{10 11} In recent years, the BCO gene was cloned from several vertebrate species, including humans. BCO is expressed in a wide range of human tissues, including the intestinal tract, liver, kidney, prostate, testis, ovary, and skeletal muscle. ¹² It is also expressed in monkey retina and human retinal pigment epithelium (RPE). ¹³ Yan et al. ¹⁴ reported the human BCO gene structure and chromosomal localization and also demonstrated that BCO expression is higher in the RPE than in liver, kidney, intestine, or testes. ¹⁴ A Drosophila BCO mutant exhibits significantly reduced rhodopsin content and strongly reduced visual sensitivity that can be rescued exclusively by consumption of retinal, but not by consumption of β-carotene. ¹⁵ In zebrafish (Danio rerio), BCO plays a crucial role in the development of the eyes, the craniofacial skeleton, and the pectoral fin. ^{16 17} These findings suggest that β-carotene conversion by BCO may constitute an important step in the pathway leading to retinoic acid in local tissue environments of vertebrates. 

BCO shares sequence homology with RPE65, a retinal pigment epithelium (RPE)–specific protein essential for the maintenance of normal vision. ¹⁸ In humans, mutations in RPE65 result in severe forms of childhood retinal dystrophies, including Leber congenital amaurosis (LCA). ^{19 20}  

In the present study, we elucidated the possible role of BCO in the conversion in the RPE of provitamin A, which plays an essential role in the retinoid metabolism of the eyes. We demonstrated BCO mRNA, protein, and enzyme activity in the human RPE cell line D407. Application of exogenous β-carotene to D407 cells resulted in an upregulation of BCO mRNA and protein levels, accompanied by a conversion of β-carotene to retinoids. Together, our analyses suggest that β-carotene cleavage, via BCO in the RPE, contributes to the function and survival of the adjacent photoreceptor cells. 

Materials and Methods

Chemicals

All-trans-β-carotene in water-soluble beadlet form and the identical formulation without β-carotene (placebo) were generously provided by BASF (Ludwigshafen, Germany). All cell culture media and reagents were obtained from Biochrom (Berlin, Germany). The RAR-α antagonist Ro 41-5253 was a gift from Hoffmann-La Roche (Basel, Switzerland). All-trans-retinoic acid was purchased from Sigma-Aldrich (Munich, Germany). 

Cell Culture

The D407 (human RPE) cell line was a gift from Richard Hunt (Department of Pathology and Microbiology, University of South Carolina Medical School, Columbia, SC), ²¹ and the Caco-2 TC7 clone was kindly provided by Monique Rousset (INSERM U178, Unite de Recherches sur la Differenciation Cellulaire Intestinale, Villejuif cedex, France). ²² The D407 and Caco-2 TC7 cells were grown in DMEM supplemented with 10% FBS. 

Chemical Treatment

Water-soluble all-trans-β-carotene was dissolved in sterile distilled water by incubating at 37°C for 30 minutes and was stored at −70°C in aliquots. An aliquot of the suspension was extracted as described for the cells, and β-carotene concentration was determined by measuring the absorbance at λ = 450 nm in a UV/vis-spectrophotometer. When stored at −70°C, the β-carotene concentration was unchanged in the suspension for a period of 6 months. The cell cultures were maintained in serum-free medium for 48 hours before incubation in supplemented medium. All the preparations involving β-carotene were made under safe dim yellow light. The cells were incubated with all-trans-β-carotene-supplemented medium. Stocks of all-trans-retinoic acid and Ro 41-5253 (1 mM) were prepared in absolute ethanol and used for supplementation. The controls received the same amount of placebo (water-soluble formulation without β-carotene) or ethanol. 

Reverse Transcription–Polymerase Chain Reaction

Total RNA was isolated from the cells in affinity columns (RNeasy; Qiagen, Hilden, Germany). DNase I (Qiagen) digestion was performed to avoid any DNA contamination. RNA was quantified in a spectrophotometer (Smartspec; Bio-Rad, Munich, Germany). Total RNA (2 μg) was reverse transcribed with an oligo(dT)₂₀ primer (Thermo Hybaid, Ulm, Germany) using reverse transcriptase (Omniscript; Qiagen) according to the manufacturer’s instructions. 

Gene specific primers were designed from the respective gene mRNA sequences available in the GenBank sequence database (http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) using the Primer 3 program (http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) ²³ available online. Primer sequences used: BCO (GenBank accession no. NM_017429) forward, 5′-CATCTTCCTTGAGCAGCCTTTC-3′, reverse 5′-GCAGCCGTCCTCTTCGTAG-3′; GAPDH (NM_002046) forward, 5′-ATGACATCAAGAAG GTGGTGA-3′, reverse, 5′-CTGTAGCCAAATTCGTTGTCA-3′; CYP26 A1 (NM_000783) forward, 5′-GGAGGACACGAAACCAC-3′, reverse, 5′-TCAGAGCAACCCGAAACC-3′; RAR-a (X56057) forward, 5′-GGGGAAGGAGTGTAGGATACC-3′, reverse 5′-CTGGGAAGGGCGAGTCTTA-3′; and RXR-a (NM_002957) forward, 5′-GCCTGAGTCTTCTCCTTGCT-3′, reverse, 5′-AGTTCCTGAGCCCCTCTCTC-3′. PCR amplification conditions were: 95°C for 4 minutes, followed by 35 cycles of amplification (95°C 30 seconds, 56°C 30 seconds, 72°C 30 seconds), followed by final an extension at 72°C for 5 minutes. The amplification reaction was performed in a thermal cycler (iCycler; Bio-Rad, Munich, Germany). The presence of the corresponding PCR products was determined by agarose gel electrophoresis. 

Real-Time Quantitative PCR

The authenticity of the PCR amplified fragments was determined by DNA sequencing. The PCR products were cloned into the PCR II TOPO vector using a cloning kit (TOPO TA; Invitrogen, Karlsruhe, Germany). The plasmids containing amplicons were linearized with BamHI and used as standards for absolute quantification. 

The cDNA from a 20-μL reverse transcription reaction was diluted to 100 μL with water and 5 μL was used for each PCR reaction. The amplification was performed with PCR master mix (QuantiTect SYBR Green; Qiagen) in triplicate reactions. The reaction mixture consisted of 50 μM of each forward and reverse primer, 1× master mix, and 10 nM fluorescein calibrating dye. The amplification was performed in a thermal cycler with an optical detection system (iCycler; Bio-Rad). The amplification conditions are 15 minutes at 94°C for the enzyme activation followed by 40 cycles of denaturation at 94°C for 30 seconds, primer annealing at 56°C for 30 seconds, and extension at 72°C for 30 seconds. The fluorescence data were collected at the extension step of each cycle. At the end of every PCR run, a melting curve was generated by increasing the temperature very slowly (0.5°C every 10 seconds) from 65°C to 95°C. The postrun data were analyzed with the real-time detection system software of the thermal cycler (iCycler iQ ver. 3.0; Bio-Rad). Briefly, the threshold value (C _T) was selected in the log phase of the amplification curve. A standard curve (plot of C _T values/crossing points of different standard dilutions against the log of amount of standard) was generated by using serial 10-fold dilutions (10⁰–10⁶ copies) of the standards with known quantities of DNA (linearized plasmid containing the PCR fragment). The copy numbers of the samples were deduced by plotting the C _T values of unknown samples on the standard curve. Melting curve analysis was included in every run, to confirm the specificity of the PCR reaction. The changes in the mRNA levels were quantified by calculating the average of triplicate reactions and normalizing with mRNA levels of the GAPDH gene. 

Western Blot Analysis and Immunodetection

After the treatment, cells were washed with PBS, scraped into PBS, and sedimented by centrifugation. The cell pellets were dissolved in lysis buffer (1% NP-40, 0.1% SDS, 0.5% sodium deoxycholate, and protease inhibitors in PBS) and incubated on ice for 30 minutes. The lysates were centrifuged at 10,000g for 15 minutes at 4°C, and the supernatants were saved. Protein concentration was then determined (Dc Protein assay system; Bio-Rad). Protein from each preparation (100 μg) was separated by 10% SDS-PAGE, transferred onto a polyvinylidene difluoride (PVDF) membrane (Roche Diagnostics, Mannheim, Germany), blocked with 3% BSA, and probed with a murine BCO antiserum ¹⁶ overnight. Peroxidase-conjugated anti-mouse IgG was used as the secondary antibody. The blots were visualized by a chemiluminescence system (Cell Signaling Technology, Beverly, MA). Afterward, the detection blots were stripped and probed with actin antibody as an internal control. 

BCO Activity Assay

Enzyme preparation and BCO activity assays were performed as described by During et al. ²⁴ with minor modifications. Protein isolated from Caco-2 TC7 cells was used as a positive control in assay reactions. The reaction was initiated by the addition of 45 μL of 75 μM water-soluble β-carotene. The reaction mixture contained 40 μM tricine-KOH buffer (pH 8.0), 0.2 μM dithiothreitol (DTT), 1.6 μM sodium cholate, and 1 mg protein preparation from the cells. The enzyme reaction was performed at 37°C in a water bath with gentle shaking for 60 minutes and was stopped by adding 50 μL of 37% formaldehyde. The reaction mixture was extracted three times with n-hexane. Solvent was evaporated under a nitrogen gas stream, and the residues were dissolved in 100 μL acetonitrile and analyzed by reversed-phase HPLC. Results are expressed as nanograms of all-trans-retinal per milligram of protein per hour. 

HPLC Analysis

D407 cells were incubated with different concentrations of β-carotene. After the incubation, the cells were washed three times with phosphate-buffered saline (PBS), scraped with a cell scraper, and pelleted down by centrifuging at 1500g for 5 minutes. Vitamin A metabolites were extracted as described elsewhere. ²⁵ Briefly, the cell pellet was dissolved in 100 μL of PBS, and the cells were lysed by three freeze–thaw cycles in liquid nitrogen. Ethanol (100 μL) was added, and the lysate was shaken with 2 mL of n-hexane for 30 minutes. The hexane phase was separated by centrifuging at 3000g for 1 minute and transferred into a fresh tube. The hexane extraction was repeated two times, and the phases were pooled and dried under a stream of nitrogen gas. The dried residues were reconstituted in 100 μL acetonitrile. A portion (50 μL) of the sample was injected into a reversed-phase C18 3-μm column with the dimensions of 300 × 4 mm (Nucleosil; Grom, Herrenberg, Germany). β-Carotene was eluted with an isocratic solvent containing 82% acetonitrile, 15% dioxane, and 3% methanol with 0.1% triethylamine at the flow rate of 1 mL/min and detected by UV absorption (λ = 450 nm). Retinol and retinal were eluted with the gradient of (A) acetonitrile/methanol (90/10, vol/vol) and (B) tetrahydrofuran, in the following program: 0 minutes: 98% A + 2% B; 6 minutes: 94% A + 6% B; 7 minutes: 80% A + 20% B; 16 minutes: 70% A + 30% B; 17 minutes: 98% A + 2% B; and 25 minutes: end of run and detected by UV absorption (λ = 325 nm). Chromatograms were recorded and analyzed (Star Chromatography Workstation Software, ver. 5.31; Varian Deutschland GmbH, Darmstadt, Germany). 

Statistical Analysis

The statistical significance at each time point, comparing the control with the treated conditions, was determined with a paired two-tailed Student’s t-test. P < 0.05 was considered significant. 

Results

BCO Expression in the RPE Cell Line and Bovine Ocular Tissues

RT-PCR analysis of total RNA from the D407 cell line, Caco-2 TC7 cell line, bovine RPE/choroids, and retina yielded a PCR product of the expected size (237 bp), indicating that BCO mRNA is expressed in these cell lines and tissues (Fig. 1A) . Relative BCO expression, normalized with the GAPDH expression in corresponding tissues, is shown in Figure 1B . Caco-2 TC7 is a clone of Caco-2, which is a human colon cancer cell line. The TC7 clone has been reported to possess BCO activity ²⁴ and was chosen as a standard for BCO expression. 

Total protein was isolated from confluent cultures of D407 and Caco-2 TC7 cell lines. Protein extract (100 μg) was separated by SDS-PAGE, and BCO protein was detected by Western blot analyses. A single band of 65 kDa was detected in D407 cells and Caco-2 TC7 cells (Fig. 2) . The BCO antibody used in this study shows cross-reaction with a second member of the BCO gene family, β-carotene 9′,10′-oxygenase (BCO2). ²⁶ However, BCO2 mRNA was not detectable in this cell line (Fig. 2B) , thus excluding the possibility that BCO2 protein was detected in the Western blot analysis, as well as BCO. In addition, BCO heterologously expressed in COS7 cells is specifically detected by the antiserum (Von Lintig J, personal communication, June 2004). 

Regulation of BCO Expression by β-Carotene

D407 cells possessed only low-abundance BCO mRNA levels, which made it difficult to assess quantitatively the possible changes in its mRNA levels after treatment of the cells with β-carotene. To overcome this problem, the cDNA preparation was initially amplified for 15 cycles and subsequently used for the quantification of BCO mRNA levels by real-time PCR. This preamplification step minimized inter- and intra-assay variations and increased the efficiency of the PCR method, in accordance with other studies in which a comparable protocol was used for accurate quantification of changes in the expression of genes with low-abundance mRNA levels. ²⁷  

D407 cells were incubated with 0.5 and 5 μM β-carotene in the medium for 4 hours to investigate the effect of β-carotene on BCO expression. Upregulation of BCO mRNA was observed in the cells incubated with 0.5 μM (75%, P < 0.05) and 5 μM β-carotene (96%, P < 0.05) compared with control cells (Fig. 3A) . BCO protein expression increased 40% at concentrations of 0.5 and 5 μM, as judged by quantitative Western blot analysis (Fig. 3C) . To establish the time course of β-carotene action, we incubated cells in the medium supplemented with 5 μM β-carotene for 2, 4, 6, 24, and 48 hours. BCO mRNA was upregulated by 127% (P < 0.05) at 2 hours and 97% at 4 hours and was downregulated by 37% at 24 hours and 40% at 48 hours (Fig. 3B) . BCO protein expression had increased twofold at 2 and 4 hours but then decreased after 24 and 48 hours (Fig. 3C) . 

CYP26A1 Expression in Cells Incubated with β-Carotene

CYP26A1 is an enzyme involved in retinoic acid catabolism, which converts retinoic acid into polar metabolites. CYP26A1 mRNA levels were measured in cells incubated with β-carotene, which had been used for BCO mRNA quantification. CYP26A1 mRNA levels were upregulated 44% at 0.5 μM and 70% at 5 μM after 4 hours of incubation (Fig. 3A) . In cells incubated with 5 μM β-carotene, CYP26A1 mRNA-concentration increased up to 10-fold with the incubation time and reached maximum peak levels (P < 0.05) after 48 hours (Fig. 3B) . This suggests that, from β-carotene, RA is synthesized in D407 cells, which subsequently induces CYP26A1 expression. 

Regulation of BCO Expression by Retinoic Acid

To elucidate the possible role of RA in the regulation of BCO expression, cells were incubated with all-trans-retinoic acid–supplemented medium for a fixed time of 6 hours. Subsequently, we analyzed BCO mRNA levels by real-time PCR. BCO expression was upregulated by 60% (P < 0.05) at 0.01 μM, 18% (P < 0.05) at 0.1 μM, and downregulated by 56% (P < 0.05) at 1 μM (Fig. 4A) . BCO protein expression was upregulated by onefold at 0.01 and 0.1 μM and downregulated at 1 μM (Fig 4C) . The cells were incubated with 0.01 μM retinoic acid for the indicated time points to establish the time course of the effect of retinoic acid on BCO expression. BCO mRNA levels were upregulated by 130% (P < 0.05), 196%, 250% (P < 0.005) and 399% at 2, 4, 6, and 24 hours, respectively, in cells incubated with 0.01 μM retinoic acid compared with the control (Fig. 4B) . BCO protein was upregulated over time with a maximum fivefold increase at 24 hours in 0.01 μM retinoic acid–incubated cells (Fig. 4D) . BCO mRNA and protein were decreased 10-fold in cells incubated with 1 μM retinoic acid (Figs. 4E 4G) . 

To demonstrate the direct involvement of RA-signaling in the regulation of BCO mRNA expression, cells were incubated with the RAR-α-specific antagonist Ro 41-5253 at 1 μM concentration. After this treatment, BCO expression increased twofold at 2 hours (P < 0.005), threefold (P < 0.05) at 6 hours, and fivefold after 24 hours (P < 0.005; Fig. 4F ). BCO protein also increased with incubation time, to a maximum of fourfold at 24 hours (Fig. 4H) . 

Expression of RAR-a and RXR-a

RAR-a and RXR-a mRNAs were detected in D407 cells. Expression of both mRNAs was strongly induced by retinoic acid in a concentration- and time-dependent manner (data not shown). β-Carotene incubation also induced RAR-a and RXR-a. RAR-a was induced by 50% at 0.5 μM, 80% at 5 μM, and 107% at 40 μM after 4 hours. RXR-a was induced by 50% in 0.5 and 5 μM and was unchanged in 40 μM β-carotene incubated cells (Fig. 5A) . In cells incubated with 5 μM β-carotene, RAR-a expression increased over the incubation time with a maximum of a 10-fold increase after 24 hours. RXR-a expression increased by fourfold at 2 hours and twofold at 4 hours and then declined to basal level at 6 hours and 24 hours (Fig. 5B) . 

BCO Enzyme Activity

So far, we have shown that BCO is expressed at the mRNA and protein level. To demonstrate BCO at the functional level, we performed tests for enzymatic activity in total protein extracts from D407 cells and Caco-2 TC7 cells. Enzyme activity was observed by the formation of 4.5 ng retinal/h per milligram protein in D407 cells. In extracts from Caco-2 TC7 cells, 7.3 ng retinal/h per milligram protein was found (Fig. 6) . All-trans-retinal was the only product detected in the in vitro test system. 

β-Carotene Uptake and Metabolism by D407 Cells

Cells were incubated in medium supplemented with 5 μM β-carotene, to determine its uptake and metabolism. Cells took up β-carotene, but the uptake rate slowed with the incubation time (Table 1) . In addition, we performed HPLC profiling for retinol and retinal at different time points. As shown in the Table 1 , retinol was present in cells incubated with 5 μM β-carotene and slightly increased with incubation time, whereas no retinol was found in the control cells or at the starting point of β-carotene incubation. Retinal was below the detection limit in both control and β-carotene-incubated cells, which may have been due to the rapid metabolic conversion to retinol in the cells. 

Discussion

In humans, provitamin A carotenoids are the major dietary precursors for vitamin A. In the photoreceptor outer segments, rhodopsin exists in millimolar concentrations. Therefore, eyes have a very high demand for vitamin A, to maintain the visual process. Bernstein et al. ²⁸ reported the presence of β-carotene in the human RPE/choroid (10 ng/0.2 g tissue). Recently, Yan et al. ¹² demonstrated that BCO expression in the RPE is higher than in liver, kidney, intestine, and testes. In the present study, we investigated provitamin A metabolism in the human RPE cell line D407. This cell line is an immortal human RPE line that retains many of the metabolic and morphologic characteristics of RPE cells found in vivo, even after several passages, and it has a long survival time. ²¹ In the present study, exogenous β-carotene application to D407 cells resulted in an upregulation of BCO mRNA and protein levels in a time- and concentration-dependent manner, resulting in the conversion of β-carotene to retinol. 

Furthermore, β-carotene incubation resulted in an upregulation of CYP26A1 mRNA in a concentration- and time-dependent manner. CYP26A1 is known to be induced by retinoic acid and its receptors through a transactivation of a retinoic acid–responsive element (RARE) located in its promoter. ²⁹ This observation indicates a possible involvement of retinoic acid in the regulation of β-carotene metabolism in the RPE. Indeed, BCO mRNA and protein levels first increased in D407 after β-carotene supplementation, but decreased after longer incubation periods. The primary product of β-carotene cleavage via BCO is retinal, which can be further metabolized either to retinol or retinoic acid. In D407 cells, we found significant amounts of retinol, but retinoic acid was below the detection limit during the initial periods and was detectable only after 24 hours. Nevertheless, traces of retinoic acid are sufficient to activate its nuclear receptors. In addition, an increase in cellular retinoic acid levels rapidly induces its catabolism through CYP26 monooxygenases. Thus, the strong induction of CYP26A1 at the mRNA level and the downregulation of BCO after prolonged β-carotene supplementation may be caused by the increase of retinoic acid in D407 cells after β-carotene supplementation. Retinoic acid’s effects are mediated through its nuclear receptors, RAR and RXR. ^{30 31} We showed that RAR-a and RXR-a are expressed in D407 cells, and their expression is enhanced by β-carotene. 

Retinoic acid regulated BCO expression in a bidirectional manner by inducing expression at lower concentrations and inhibiting it at higher concentrations. A possible involvement of RA signaling in the regulation of BCO expression was further corroborated by applying the RAR-a antagonist Ro 41-5253 to D407 cells. This treatment resulted in the upregulation of BCO mRNA expression, which points to a RAR-a-mediated repression of the BCO gene. Indeed, Ro 41-5253 is exclusively specific to RAR-a at the concentration used in this study (1 μM). ³²  

In a recent report, peroxisome proliferator-activated receptor (PPAR)-gamma/RXR-a heterodimers were shown to regulate BCO gene expression, and 9-cis-retinoic acid treatment induced the expression in mice. ³³ In our study, neither 9-cis-retinoic acid nor 9-cis-β-carotene was detected in the cells after β-carotene incubation. Therefore, a role of 9-cis-retinoic acid in the observed effects on the regulation of BCO expression in D407 cells can most probably be excluded. We also performed a transcription factor binding site search in the 5′ flanking region of the human BCO gene (GenBank sequence database) that revealed a putative RARE (GGGTCActtgAGGTCA, 600 bases upstream to the start site; Chichili GR, unpublished observation, 2004). Therefore, we assume that RAR-a and RXR-a are involved in the regulation of BCO expression in D407 cells and thus most probably are also involved in vivo in the RPE. Bachman et al., ³⁴ demonstrated in rats and chicken that retinoic acid feeding results in a decrease in BCO enzyme activity in intestines but did not affect BCO activity in liver. Our data partially agree with the observation that high RA concentrations inhibit BCO expression. However, we also found that low RA levels have a positive effect on BCO mRNA expression. The fact that RA induces BCO expression at a lower concentration may contribute to efficient the use of β-carotene for retinoid synthesis, whereas the suppression of BCO expression at higher RA concentration may avoid an excess of retinoid synthesis from its provitamin A precursor. Further investigations are necessary to gain a mechanistic understanding of the regulation of BCO expression via RA and its nuclear receptors in different cell types and organisms. 

Bhatti et al., ¹³ reported that BCO is expressed at variable levels in the RPE of different species but is not expressed in RPE cell lines. This disparity may be due to the sensitivity of the RT-PCR and RNA purification used in our study. RNA isolated with affinity columns was used for RT-PCR in our study, which works better than RNA isolated by phenol-based methods (data not shown). Primers with high PCR efficiency derived from the human BCO sequence were used in the present study. This may explain the differences between the two studies, because PCR sensitivity varies between different primer sets. In addition, our mRNA expression data in RPE cells are further supported by assessing BCO protein levels as well as by tests for BCO enzymatic activity. 

Recently, expression of BCO has been reported in epithelia of several tissues in human including glandular cells in the prostate, endometrium, mammary tissue, kidney tubules, keratinocytes of the squamous epithelium of skin; steroidogenic cells in testis, ovary, and adrenal gland; and skeletal muscle cells. ³⁵ Epithelia in general are structures that are very sensitive to vitamin A deficiency. Local β-carotene cleavage may contribute to the maintenance of a steady state level of retinoids in these tissues, particularly under conditions of high retinoid demand. BCO mRNA expression was also found in bovine RPE and retina, indicating a possible role of this enzyme in the vitamin A metabolism of other species. Thus, our study supports previous findings ^{14 28} and provides functional evidence that β-carotene conversion to vitamin A exists in the RPE and may contribute to ensure retinoid-dependent processes in vision. 

A most interesting finding in two siblings with impaired retinol transport caused by mutations in the RBP gene was that only mild clinical vitamin A deficiency symptoms such as night blindness and modest retinal dystrophy were present. ³⁶ In these patients, cleavage of β-carotene in the RPE may compensate for the impairments in retinol transport. In such situations, high doses of β-carotene supplementation may help to avoid vitamin A deficiency. 

More generally, β-carotene is the major source to ensure the vitamin A demand of humans. Yet, it was thought that β-carotene is converted to vitamin A immediately after its uptake in the small intestine and stored in the liver in the form of retinyl esters, implying that peripheral tissues mainly rely on the supply of preformed vitamin A in the circulation. Our study provides evidence for a tissue-specific vitamin A synthesis in the RPE. This finding is in agreement with the fact that β-carotene is transported in lipoproteins and BCO is expressed in various human tissues. 

Taken together, the evidence in this study showed that BCO is expressed and can be induced by β-carotene in RPE cells. In addition, β-carotene can be converted into vitamin A in RPE cells in a regulated manner, thus providing functional evidence for an eye-specific provitamin A metabolism in humans. 

²  Present affiliations: Molecular Immunogenetics Program, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma; the

³  Department of Biochemistry, Meharry Medical College, Nashville, Tennessee.

 Supported by The European Union (ProVitA).

 Submitted for publication January 25, 2005; revised May 24, 2005; accepted August 25, 2005.

 Disclosure: G.R. Chichili, None; D. Nohr, None; M. Schäffer, None; J. von Lintig, None; H.K. Biesalski, None

 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.

 

BCO mRNA expression in RPE cell lines and bovine retinal tissues. Total RNA from confluent cell cultures and bovine tissues was extracted, 2 μg of total RNA was reverse transcribed into cDNA, and PCR amplification was performed with human BCO gene–specific primers. PCR amplicons were separated by electrophoresis on a 1% agarose gel, stained with ethidium bromide, and documented in a UV gel documentation system. (A) Lane M: 100-bp ladder size marker; lane 1: Caco-2 TC7 cells; lane 2: D407 (human RPE cell line); lane 3: bovine RPE; and lane 4: bovine retina. GAPDH expression in corresponding samples is shown. (B) BCO mRNA levels were quantified by real time quantitative PCR and normalized with GAPDH mRNA expression in corresponding samples. The mean ± SE of two separate experiments is shown.

Figure 2.

 

(A) BCO protein expression in D407 and Caco-2 TC7 cell lines. Protein extracted from confluent untreated cell cultures was analyzed for BCO protein expression by Western blot analysis. Lane +: Positive control (BCO fusion protein with an N-terminal His6-tag); lane 1: D407 cells; and lane 2: Caco-2 TC7 cells. (B) BCO II mRNA expression. Total RNA was extracted from confluent cell cultures, 2 μg of total RNA was reverse transcribed into cDNA, and PCR amplification was performed with human BCO II gene–specific primers. PCR amplicons were separated by electrophoresis on a 1% agarose gel, stained with ethidium bromide, and documented in a UV gel documentation system. Lane M : 100-bp size marker, BCO II; lane 1: negative control; lane 2: D407; lane 3: positive control, GAPH; lane 4: negative control; lane 5: D407; lane 6: positive control.

Figure 3.

 

Concentration- and time-dependent regulation of BCO and CYP26A1 expression by β-carotene. (A) D407 cells were incubated in the medium supplemented with 0.5 and 5 of all-trans-β-carotene for 4 hours. (B) D407 cells were incubated in the medium supplemented with 5 μM β-carotene for 2, 4, 6, 24, and 48 hours. Total RNA was extracted and the changes in mRNA levels of BCO and CYP26A1 were quantified by real-time quantitative PCR. The mean ± SE of results in three separate experiments is shown except when the SE is too small to be seen. *P < 0.05) significant difference from the control. (C) Top: 100 μg of protein extracts from D407 cells treated with β-carotene was analyzed for BCO protein expression indicated the time points by Western blot analysis. Bottom: actin expression in the corresponding samples in the same blot.

Figure 4.

 

Concentration- and time-dependent regulation of BCO expression by all-trans-retinoic acid. (A) D407 cells were incubated in the medium supplemented with 0.01, 0.1, and 1 μM all-trans-retinoic acid for 6 hours and (B) in medium supplemented with 0.01 μM all-trans-retinoic acid for 2, 4, 6, and 24 hours. (E) All-trans-retinoic acid (1 μM). (F) The RAR-a antagonist (1 μM). Total RNA was extracted, and the changes in the BCO mRNA levels were quantified by real-time quantitative PCR. The mean ± SE of results in three separate experiments is shown except when the SE is too small to be seen. **P < 0.005; *P < 0.05, significant difference from the control. BCO protein expression: D407 cells were incubated with (C) different concentrations or (D) 0.01 μM or (G) 1 μM all-trans-retinoic acid for the indicated times. (H) RAR-a antagonist (1 μM Ro 41-5253). BCO protein expression was analyzed by Western blot analysis Bottom: actin expression in corresponding samples in the same blot.

Figure 5.

 

Regulation of RAR-a and RXR-a mRNA expression by β-carotene. (A) D407 cells were incubated in the medium supplemented with 0.5 and 5 μM of all-trans-β-carotene for 4 hours. (B) D407 cells were incubated in the medium supplemented with 5 μM β-carotene for 2, 4, 6, and 24 hours. Total RNA was extracted, and the changes in mRNA levels of RAR-a and RXR-a were quantified by real-time quantitative PCR. The mean ± SE of results in three separate experiments is shown, except when the SE is too small to be seen.

Figure 6.

 

BCO enzymatic activity in D407 and Caco-2 TC7 cells. Protein extraction and a BCO activity assay were performed. Retinal produced from β-carotene in the presence of protein was quantified by HPLC. The mean ± SE of results in three separate experiments is shown.

Table 1.

 

β-Carotene Uptake and Conversion to Retinol by D407 Cells

Time Point (h)	β-Carotene Uptake (ng/μg of DNA)	Conversion to Retinol (ng/μg of DNA)
0 (control cells)	0	0
3	21.26 ± 0.53	0.038 ± 0.00410
6	48.44 ± 1.37	0.048 ± 0.00079
24	62.39 ± 4.24	0.052 ± 0.00380

 D407 cells were incubated in medium supplemented with 5 μM of β-carotene for indicated time periods. β-Carotene and retinol were extracted and analyzed by HPLC as described in the experimental procedures. The mean ± SE of two separate experiments is shown.

The authors thank Christian Köpsel (BASF) for providing water-soluble β-carotene, Willi Hunziker (Hoffmann-la Roche) for providing the Ro 41-5253; Andrea Flaccus for assistance with the PCR assays and Michael Wolter for assistance with the HPLC analysis.