Abstract
STRA6 is a multi-transmembrane domain protein not homologous to any other proteins with known function. It functions as the high-affinity receptor for plasma retinol binding protein (RBP) and mediates cellular vitamin A uptake from the vitamin A/RBP complex. Consistent with the diverse roles of vitamin A and the wide tissue expression pattern of STRA6, mutations in STRA6 are associated with severe pathological phenotypes in human. The structural basis for STRA6’s biochemical function is unknown. Although computer programs predict 11 transmembrane domains for STRA6, its topology has never been studied experimentally. Elucidating the transmembrane topology of STRA6 is critical for understanding its structure and function. By inserting an epitope tag into all possible extracellular and intracellular domains of STRA6, we systematically analyzed the accessibility of each tag on the surface of live cells, the accessibility of each tag in permeabilized cells, the effect of each tag on RBP binding and STRA6-mediated vitamin A uptake from the vitamin A/RBP complex. In addition, we used a new lysine accessibility technique combining cell-surface biotinylation and tandem-affinity purification to study a region of the protein not revealed by the epitope-tagging method. These studies not only revealed STRA6’s extracellular, transmembrane and intracellular domains, but also implicated extracellular regions of STRA6 in RBP binding.
Vitamin A and its derivatives (retinoids) are essential for diverse aspects of vertebrate physiology (1–3). Due to the hydrophobic nature of retinoids, it has been assumed that random diffusion is the primary if not the only means of transmembrane transport. However, biochemical evidence suggests that retinol uptake from the small intestine is mediated by a membrane transporter (4). There is also strong evidence for the existence of a specific mechanism to transport 11-cis retinal in the retinal pigment epithelium (RPE) that depends on interphotoreceptor retinoid-binding protein (IRBP). Apo-IRBP is much more effective in promoting the release of 11-cis retinal from the RPE than the apo-forms of other retinoid binding proteins (5). In addition, apo-IRBP is only effective when it is present on the apical, but not basal, side of the RPE (6). Another finding that challenges the assumptions about random diffusion is the identification of an ATP-dependent transporter (ABCR or ABCA4) that transports all-trans retinal released from bleached rhodopsin across membranes (7–9). Mutations in ABCR cause a wide spectrum of human vision diseases from retinitis pigmentosa to macular degeneration. Prior to the surprising discovery of ABCR’s role in retinoid transport, there was no biochemical or physiological evidence for the existence of such a transporter.
Retinol is the main transport form of vitamin A in the blood. Although free retinol can also diffuse through membranes, it seldom exists in its free form. Retinol binding protein (RBP) is the specific carrier of vitamin A in the blood (10, 11). During transport in the blood, virtually all retinol is bound to RBP. RBP solubilizes retinol, and the complex of retinol/RBP cannot freely diffuse through membrane. Unlike ABCR, which was not predicted to exist, evidence has accumulated for more than 30 years for the existence of a membrane receptor for RBP that mediates cellular vitamin A uptake (12–25). Using an unbiased strategy, the membrane receptor for RBP has been identified as a multitransmembrane domain protein STRA6. STRA6 binds to RBP with high-affinity and mediates cellular uptake of vitamin A from the retinol/RBP complex (holo-RBP) (26). STRA6 represents a rare example of a eukaryotic membrane transport system that depends on an extracellular carrier protein but does not rely on endocytosis. Consistent with the essential roles of vitamin A in human development, mutations in human STRA6 cause severe pathological phenotypes such as anophthalmia, mental retardation, congenital heart defects, and lung hyperplasia (27, 28).
Since STRA6 is a novel membrane transport protein not homologous to any other protein with known function, one difficulty in studying STRA6’s structure and function is that it has no obvious functional domains (e.g., ATP binding domain). The transmembrane topology of STRA6 has never been studied experimentally. At the basic level, it is not even known which terminus of STRA6 faces the outside or inside of the cell. Determining the transmembrane topology of STRA6 is of critical importance in understanding its detailed molecular mechanism. Transmembrane topology of a membrane protein contains information regarding extracellular domains, transmembrane domains and intracellular domains. For example, the membrane topology of a channel is essential to elucidate functional domains within it, such as the ligand binding region and the pore.
Common methods to determine transmembrane topology of membrane proteins on cell surface include epitope tagging (29, 30) and cysteine modification (31). Because STRA6 has a large number of cysteine residues (14 for bovine STRA6), mutating all cysteine residues to create the cysteineless protein is likely to have a large impact on the protein’s structure and function. We therefore chose epitope tagging as the primary method to elucidate STRA6’s transmembrane topology. This technique takes advantage of the fact that STRA6’s natural cellular location is the plasma membrane (32, 33), which allows interaction with RBP, its ligand outside the cell. As a secondary method, we used a lysine accessibility technique to study a region of STRA6 not resolved by the epitope tagging method.
EXPERIMENTAL PROCEDURES
Myc tagging
All experiments in this study were performed on bovine STRA6. For Myc insertions into different regions of STRA6, an Eco RI and Bam HI linker was engineered into distinct sites in STRA6 cDNA by PCR. For the Myc-tagged proteins M1 through M13, the linker was inserted between the following pairs of residues: 16/17, 84/85, 133/134, 170/171, 197/198, 272/273, 323/324, 357/358, 407/408, 461/462, 509/510, 548/549, 593/594. For M14, the linker was appended to the C-terminus. Two linker oligos that have Eco RI and Bam HI sites at each end and encode the Myc sequence (EQKLISEEDLN) flanked by 3 glycines were inserted into these engineered Eco RI and Bam HI sites (the EcoRI and BamHI sites used in the cloning encode amino acids EF and GS, respectively). All final constructs were sequenced to rule out spurious mutations.
Cell transfection
COS-1 cells or HEK293 cells were transfected using FuGENE 6 transfection reagent (Roche). All assays and harvesting of cells were performed 24–30 hours after transfection. All assays based on live cells were done using COS-1 cells due to their strong attachment to the culture dish during washes. These studies include cell-surface protein quantitation studies, vitamin A uptake assays, and RBP binding assays. Since the transfection efficiency is higher for HEK293 cells than COS-1 cells, HEK293 cells were used for localization studies since these protocols are based on fixed cells instead of live cells.
Live cell staining and permeabilized cell staining
STRA6-Myc constructs were transfected into HEK293 cells growing on gelatin-coated coverslips using FuGENE6 transfection reagent (Roche). At 24 hours after transfection, purified anti-Myc monoclonal antibody was added to the media at 2 µg/ml. After 1 hour incubation at 37°C, the cells were washed with Hank’s Buffered Salt Solution (HBSS) 3 times and fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) for 10 minutes at room temperature. After 3 washes with PBS, the fixed cells were incubated with the blocking buffer (5% normal goat serum and 0.3% Triton X-100 in PBS) for 1 hour at room temperature. The cells were then incubated with Alexa Fluor 488-conjugated goat anti-mouse antibody (Molecular Probes) diluted in the blocking buffer for another hour at room temperature. After 3 washes with PBS, the coverslips containing cells were mounted onto slides using VectaShield mounting medium (Vector Laboratories). The cell surface expression of STRA6-Myc proteins was examined by fluorescence microscopy. Live cell staining for Rim-tagged STRA6 proteins was performed identically as described above except that anti-Rim tag monoclonal antibody (Rim3F4) was used. Permeabilized staining was performed similarly to live cell staining procedures except that cells were incubated with the primary antibody after they have been fixed. Briefly, 24 hours after transfection, cells were fixed in 4% paraformaldehyde in PBS for 10 minutes at room temperature. After 3 washes with PBS, the cells were incubated with the blocking buffer for 1 hour. All subsequent procedures were identical as described above for live cell staining.
Quantitation of cell surface expression
For quantitation of cell surface expression, COS-1 cells were transfected with STRA6-Myc constructs. At 24 hours after transfection, purified anti-Myc monoclonal antibody was added to the medium at 2 µg/ml for 1 hour. After 3 washes with HBSS, the cells were incubated in 5% normal rabbit serum in serum free medium (SFM) for 15 minutes. Rabbit anti-mouse antibody conjugated to alkaline phosphatase diluted 1:4000 in 5% normal rabbit serum in SFM was then added to the cells and incubated for 1 hour at 37°C. After 4 washes with HBSS, cells were lysed in 1% Triton X-100 in PBS with protease inhibitors. Insoluble materials were removed by centrifugation at 16,000 g for 5 minutes at 4°C. Alkaline phosphatase (AP) activity in the supernatant was measured using pNPP (Sigma). Quantitation of cell surface expression for Rim-tagged STRA6 proteins was performed similarly as described above except that anti-Rim tag monoclonal antibody (Rim3F4) was used.
Vitamin A uptake assay
3H-retinol/RBP production and cellular 3H-retinol uptake from 3H-retinol/RBP was performed similarly as described (26). All STRA6 variants were cotransfected with lecithin retinol acyltransferase (LRAT) for the vitamin A uptake assay. COS-1 cells express very low or undetectable levels of LRAT. Briefly, cells were washed with HBSS before incubation with 3H-retinol/RBP diluted in serum free medium (SFM) for 1 hour at 37°C. The reactions were stopped by removing the medium, washing the cells with HBSS, and solubilizing the cells in 1% Triton X-100 in PBS. Radioactivity was measured with a scintillation counter.
RBP binding assay
Quantitation of RBP binding to STRA6 was performed similarly as described using alkaline phosphatase-RBP fusion protein (AP-RBP) (26). We have shown previously that AP tagged at the N-terminus of RBP does not interfere with RBP’s interaction with either the RBP receptor in native cells or expressed STRA6 (26). Briefly, transfected cells grown on 12-well cell culture plates were washed once with HBSS and incubated with AP-RBP diluted in SFM at 37°C for 1 hour. The reaction was stopped by washing cells twice with HBSS. The cells were then lysed in 200 µl of cold PBS containing 1% Triton X-100 and protease inhibitors per well. The cell lysates were centrifuged at 3,000 g at 4°C for 5 minutes. The supernatants were removed and heated at 65°C for 1 hour. Twenty microliters of the heated lysate were mixed with 200 µl of pNPP (Sigma) and incubated at 37°C for 1 hour for the AP color reaction. The reactions were stopped by adding 50 µl 3 M NaOH and were transferred to a 96-well plate for reading in a microplate reader at 405 nm.
Cell surface biotinylation of STRA6
For cell surface biotinylation of STRA6, HEK293 cells transfected with STRA6 constructs for 24 hours were washed once with PBS before harvesting in PBS with 5 mM EDTA. The cells were further washed with PBS 3 times to remove serum protein after harvesting by gentle pelletting at 3000 rpm for 30 sec using Eppendorf 5417C centrifuge. Immediately before use, 2 mg NHS-PEO4-Biotin (Pierce) was dissolved in 170 µl of water. Then 30 µl of the biotin solution was added to 1 ml of cells suspended in PBS. After rotating at room temperature for 30 min, the biotinylation reaction was stopped by washing cells with PBS with 100 mM glycine. Since a large number of cell surface proteins can be biotinylated, biotinylated STRA6 needs to be purified to specifically detect its biotinylation. We inserted a 6XHis tag and a Rim tag (8) into a loop region of STRA6 (the M10 position in Figure 1), which is well exposed in native STRA6 (Figure 3). These two tags allow tandem-affinity purification of STRA6 from crude cellular membranes.
Figure 1.
Hydropathy plot of STRA6 and locations of Myc insertions. A. Kyle-Doolittle hydropathy plot for bovine STRA6. Positive scores indicate hydrophobicity and negative scores indicate hydrophilicity. B. Two computer programs for transmembrane domain prediction, SUSUI and TMPRED, predicted 11 transmembrane domains for bovine STRA6, although they cannot distinguish the two possible orientations. The locations of the Myc insertions are indicated in the first transmembrane topology model. The number in parenthesis indicates the residue after which the Myc tag was inserted.
Figure 3.
Quantitation of cell surface expression of STRA6-Myc series. To quantitate Myc epitope exposed to the cell surface for different STRA6-Myc proteins, anti-mouse alkaline phosphatase was used as the secondary antibody after anti-Myc antibody was bound to the cell surface on live cells. Activity of alkaline phosphatase bound to cells was quantitated. Activity of STRA6-M10, which has the highest activity, is defined as 100%. Statistical significance of cell surface expression levels was determined against that of M10 by Student’s t-test (***, P < 0.001).
Tandem-affinity purification of STRA6
Tandem-affinity purification of 6xHis-Rim-tagged STRA6 was achieved by purification of 6XHis tag followed by purification of the Rim tag using a previously described protocol with modifications (8, 34). All procedures were performed on ice or at 4°C. Briefly, after cell surface biotinylation of cells expressing wild-type or mutant 6xHis-Rim-STRA6 (one 10 cm dish per construct), crude cellular membranes were prepared and suspended in 50 µl PBS containing protease inhibitors. Cell membranes were then solubilized with one volume of 1% dodecylmaltoside and 20% glycerol in PBS and quickly diluted by adding 8 volumes of dilution solution (PBS containing 0.75% CHAPS, 10% glycerol, 1mg/ml brain polar lipid, 2 mM β-mercaptoethanol and protease inhibitors). The mixture was incubated for 30 min on ice and insoluble material was removed by centrifugation at 16,000 g for 5 min. The soluble fraction was incubated in 50 µl of a slurry of TALON resin (Clontech) overnight in the presence of 5 mM imidazole. The resin was then washed three times with wash solution (PBS containing 0.75% CHAPS, 1mg/ml brain polar lipid and 10 mM imidazole). 6xHis-Rim-STRA6 was eluted in 0.75% CHAPS, 1mg/ml brain polar lipid and 100 mM imidazole in PBS. The eluted material was incubated for 3 hours with Rim3F4 antibody conjugated to sepharose beads. Sepharose beads were then washed quickly 3 times with 1% CHAPS, 1 mg/ml brain polar lipid in PBS. The beads were washed once more with 0.2% CHAPS, 1mg/ml brain polar lipid in PBS by incubating for 15 min. Finally, 6xHis-Rim-STRA6 was eluted in 80 µl of 0.2% CHAPS, 1 mg/ml brain polar lipid and 0.5 mg/ml Rim3F4 peptide in PBS (incubation time is 40 minutes). Biotinylated STRA6 in the eluted material was detected by Western blot using Streptavidin-HRP (Pierce) to detect biotin.
RESULTS
STRA6 is a largely hydrophobic protein, as indicated by its amino acid composition and its hydropathy profile (Figure 1). Two computer programs designed to predict transmembrane topology (SOSUI and TMPRED) both predict 11 transmembrane domains for STRA6 (Figure 1). However, computer-predicted topology is often different from real topology. Since the STRA6 protein is localized to the plasma membrane, its topology can be determined based on the accessibility of each region of STRA6 to the extracellular space. This is most commonly achieved by epitope tagging. We inserted the Myc tag into the 14 hydrophilic regions of bovine STRA6 (STRA6-Myc), including the N-terminal, the C-terminal, and all putative intracellular and extracellular loop regions (Figure 1). The least conserved regions were chosen for each insertion. These constructs are named STRA6-M1 to STRA6-M14.
Only when a Myc tag is exposed to the extracellular face of STRA6 can an anti-Myc antibody bind to the tag in live cells. Therefore, positive anti-Myc staining of live cells indicates that the Myc tag is localized to the extracellular region of STRA6. When a Myc tag is inserted into a hydrophilic region but cannot be stained by anti-Myc antibody in live cells, it is likely localized intracellularly. Alternatively, this Myc tag may prevent STRA6 from being properly expressed by the cell. For this reason, we need to perform the control experiments consisting of using anti-Myc antibody to stain fixed and permeabilized cells to make sure that STRA6 can still be detected in permeabilized cells. A third possibility for the lack of live cell surface staining by anti-Myc antibody is that the Myc insertion causes misfolding of STRA6 and prevents its proper expression on the cell surface. For this reason, we need to perform functional analysis of STRA6-Myc proteins including a live cell RBP binding assay and live cell vitamin A uptake assay from holo-RBP. These STRA6-dependent functional assays will test whether the STRA6-Myc proteins are functionally expressed on the cell surface.
In addition to determining the transmembrane topology of STRA6, these STRA6-Myc proteins may also help to reveal the role of each hydrophilic region in STRA6’s biochemical activity including RBP binding and vitamin A uptake. For example, if a STRA6-Myc protein is expressed on the cell surface, as assayed by live cell anti-Myc staining, but has reduced RBP binding, the region of STRA6 where the Myc tag is inserted is likely involved in RBP binding.
Live cell and permeabilized cell staining
The STRA6-Myc constructs were transfected into HEK293 cells. Anti-Myc immunostaining of live transfected cells revealed that cells transfected with STRA6-M1, STRA6-M3, STRA6-M5 and STRA6-M10 have the Myc epitope exposed on the cell surface of live cells, but cells transfected with the other 10 constructs do not (Figure 2). To confirm that the 10 negative STRA6-Myc constructs are expressing the protein, we performed anti-Myc immunostaining of fixed and permeabilized cells as a control for the proper expression of the tagged STRA6 (Figure 2). All STRA6-Myc constructs expressed the Myc-tagged STRA6 (Figure 2). This indicates that all Myc epitopes in the STRA6-Myc proteins were accessible and not hidden within the protein. Although 10 STRA6-Myc constructs are negative for live cell staining, their fixed and permeabilized staining signals are similar to those positive for live cell staining. We further quantitated the anti-Myc antibody binding using alkaline phosphatase-tagged secondary antibody (Figure 3). This quantitation confirmed the anti-Myc live cell staining data (M1, M3, M5 and M10 are on the extracellular side) and also identified STRA6-M10 as the construct with the strongest signal for Myc expression on the cell surface.
Figure 2.
Anti-Myc live cell staining and anti-Myc permeabilized staining of STRA6-Myc proteins. HEK293 cells transfected with STRA6-Myc constructs were either stained with anti-Myc antibody in live cells (left panel) or stained with anti-Myc antibody after the cells had been fixed and permeabilized. Anti-mouse Alexa Fluor 488 (green signal) was used as secondary antibody.
Vitamin A uptake and RBP binding activity of the STRA6-Myc proteins
To further characterize these proteins, we compared the vitamin A uptake activities of 14 Myc-tagged STRA6 constructs with the wild-type STRA6. Although 10 STRA6-Myc constructs do not express the Myc epitope on cell surface, 7 of them have significant vitamin A uptake activity from holo-RBP (Figure 4). Their vitamin A uptake activities from extracellular holo-RBP suggest that these proteins are functional and are expressed on the cell surface. Combined with their positive staining of permeabilized cells, these results suggest that the absence of Myc staining in live cells of these proteins is due to the intracellular localization of the Myc epitope. In contrast, STRA6-M7, STRA6-M8, and STRA6-M12 have little uptake activity (Figure 4). Since these 3 constructs do not have the Myc epitope expressed on the cell surface, the most likely explanation for both the absence of Myc epitope and vitamin A uptake activity is protein misfolding that results in non-functional STRA6. Therefore, M7, M8, and M12 are not informative in determining the intracellular and extracellular locations of these positions.
Figure 4.
Vitamin A uptake activities from holo-RBP of the STRA6-Myc series. All STRA6 variants were cotransfected with LRAT into COS-1 for the vitamin A uptake assay. For the vitamin A uptake assay, 3H-retinol/RBP was added to cells. After 1 hour of incubation, cells were washed with PBS, and total radioactivity from cell lysate was determined as uptake activity. Activity of wild-type STRA6 is defined as 100%. Statistical significance of vitamin A uptake activity was determined against the activity of wild-type STRA6 by Student’s t-test (*, P < 0.05, **, P < 0.01, ***, P < 0.001).
We also tested the RBP binding abilities of STRA6-M1 to STRA6-M14 proteins with wild-type STRA6 (Figure 5). Similar to the vitamin A uptake assay, STRA6-M2, STRA6-M4, STRA6-M6, STRA6-M9, STRA6-M11, STRA6-M13, and STRA6-M14 had significant RBP binding activity, indicating the proper expression of these proteins on the cell surface. Their absence of Myc epitope on cell surface is due to the localization of the epitope on intracellular domains. Similar to the vitamin A uptake assay, STRA6-M7, STRA6-M8, and STRA6-M12 have dramatically reduced RBP binding activity compared to the wild-type STRA6. In addition, STRA-M5 and STRA6-M10 also have significantly reduced RBP binding activity. Since STRA6-M5 and STRA6-M10 are expressed well on the cell surface (Figure 2 and 3), this result suggests that Myc insertion at these extracellular positions likely interferes with the binding of RBP to STRA6. The relative vitamin A uptake activity of each STRA6-Myc protein compared to the wild-type STRA6 is not always identical to the RBP binding activity. This difference may be explained by the distinct nature of these assays. The vitamin A uptake activity represents the cumulative activity of vitamin A uptake over a period of time (i.e., one hour), whereas the RBP binding activity represents steady state binding of RBP at a single time point. A related fact is that RBP’s binding to its receptor STRA6 is transient (data not shown). This transient nature is important for RBP’s delivery of vitamin A to cells because the uptake cannot continue if RBP fails to leave STRA6 after retinol uptake. Because retinol uptake from holo-RBP depends on both the binding of RBP and the turnover of RBP from STRA6, the steady state RBP binding activity of a STRA6-Myc mutant may not always predict its accumulated vitamin A uptake activity.
Figure 5.
RBP binding activities of the STRA6-Myc series. To measure RBP binding activity, AP-RBP was added to COS-1 cells transfected with the STRA6-Myc constructs for 1 hour. After washing off unbound AP-RBP with PBS, alkaline phosphatase activity associated with the cells was measured by color reaction using pNPP as substrate. Activity of wild-type STRA6 is defined as 100%. Statistical significance of RBP binding activity was determined against the activity of wild-type STRA6 by Student’s t-test (**, P < 0.01, ***, P < 0.001).
Lysine accessibility in live cells as a method to reveal cell surface exposed lysines
Of the three positions whose intra- or extracellular locations were not revealed by STRA6-Myc proteins (M7, M8 and M12), position M12 is likely intracellular based on the properties of its adjacent regions. Because positions M11, M13 and M14 are all intracellular and there are too many hydrophilic residues between M12 and M13 for this region to have a transmembrane domain, M12 is also likely intracellular. Another piece of indirect evidence for localizing M12 intracellularly is the presence of a non-conserved cysteine between M12 and M13. If M12 were extracellular, this cysteine located in the hydrophilic region immediately following M12 should also be extracellular. However, extracellular cysteines in proteins tend to be extremely conserved because they are oxidized to form disulfide bonds. Although unlikely, we cannot completely rule out the possibility that M12 is extracellular.
We designed an independent method to determine the locations of M7 and M8. We noticed two lysine residues (325 and 357) located immediately adjacent to the M7 and M8 insertion sites. Because only two other lysine residues (200 and 203, located near M5) are located in regions potentially exposed to the extracellular space, the accessibility of these lysine residues to membrane impermeable biotinylation reagent in live cells may help to determine their possible extracellular location. For this assay, we used the membrane impermeable biotinylation reagent NHS-PEO4-Biotin (Pierce), which reacts with lysine residues. Since a large number of cell surface membrane proteins can potentially be biotinylated, we used tandem-affinity purification to isolate STRA6 after cell surface biotinylation of live cells. We first created three STRA6 mutants (K200A/K203A/K325A/K357A, K200A/K203A, and K325A/K357A). Cell surface biotinylation followed by tandem-affinity purification of wild-type and the 3 STRA6 mutants showed that wild-type STRA6 and mutant K200A/K203A can be biotinylated in live cells, whereas mutants K200A/K203A/K325A/K357A and K325A/K357A showed no detectable biotinylation in live cells (Figure 6A). Western blots for the STRA6 protein detected successful purification of all STRA6 variants by tandem-affinity purification. This result narrows down the major exposed lysine residue(s) in live cell to residues 325 and 357, which are located between M7 and M8. We further created STRA6 mutants K325A and K357A and found that lysine 357 is the major lysine residue responsible for STRA6 biotinylation in live cells (Figure 6B). Because the K357A mutant is well expressed on the cell surface (Figure 6D), loss of biotinylation in the K357A mutant is due to the loss of the extracellularly accessible lysine. As residue 357 is immediately adjacent to position M8, this result suggests that M8 is extracellular.
Figure 6.
Cell surface biotinylation of STRA6 and identification of STRA6 mutants resistant to biotinylation. For tandem-affinity purification and cell-surface expression quantitation of STRA6, all STRA6 variants in this figure are tagged with a 6XHis tag and a Rim tag in an extracellular loop of bovine STRA6 (between residues 133 and 134). A. Western blots detecting biotin (upper picture) and STRA6 (lower picture) after tandem-affinity purification of cell surface biotinylated wild-type STRA6 and STRA6 mutants K200A/K203A/K325A/K357A, K200A/K203A, and K325A/K357A. B. Western blots detecting biotin (upper picture) and STRA6 (lower picture) after tandem-affinity purification of cell surface biotinylated wild-type STRA6 and STRA6 mutants K325A/K357A, K325A, and K357A. C. Live cell staining of STRA6 variants included in the biotinylation study. D. Quantitation of cell surface expression for wild-type STRA6 and STRA6 mutants K200A/K203A/K325A/K357A, K200A/K203A, K325A/K357A, K325A, and K357A using Rim3F4 monoclonal antibody. Expression level of wild-type STRA6 is defined as 100%.
There are other pieces of evidence suggesting the extracellular location of the region between M7 and M8. In an independent study of more than 900 random STRA6 mutants, we identified an essential RBP binding domain in the region of STRA6 located between M7 and M8 (unpublished result). This result suggests that the seventh “transmembrane domain” in the computer predicted model (Figure 1) is actually an extracellular domain. Consistently, this domain is the only computer-predicted “transmembrane domain” that has a significant number of hydrophilic amino acids. In bovine STRA6, 4 hydrophilic residues are located in this domain including aspartate, glutamate, and arginine residues. Interestingly, the equivalent region in mouse STRA6 is not predicted to be a transmembrane domain by the computer programs. The extracellular location of this domain (residues between M7 and M8) is also consistent with the intracellular location of M6 and M9, and the sixth and eighth transmembrane domains (Figure 1B) which are completely devoid of hydrophilic residues. The inaccessibility of the biotinylation reagent NHS-PEO4-Biotin to lysine 325 is likely due to its location in a protein domain that prevents efficient access by the biotinylation reagent. Interestingly, lysine 325 is located immediately adjacent to a transmembrane domain (it is the first hydrophilic residue after a transmembrane domain).
The STRA6 topology model
The results of the epitope tagging experiments are summarized in Figure 7. The 14 Myc-tagged STRA6 proteins can be categorized into four groups. Group 1 (STRA6-M1 and STRA6-M3) has the Myc epitope expressed on the cell surface and has significant vitamin A uptake activity and RBP binding activity. For Group 1, the Myc epitope is inserted into the extracellular regions of STRA6, but the insertion does not substantially affect the interaction of RBP and STRA6 on the extracellular side. Group 2 (STRA6-M5 and STRA6-M10) also has the Myc epitope expressed on the cell surface but shows strong reduction in RBP binding activity. This is even more evident given the fact that they are better expressed on the cell surface than STRA6-M1 and STRA6-M3 (Figure 3). For Group 2, the Myc epitope is also inserted into the extracellular regions of STRA6, but these extracellular regions likely contribute to the interaction of STRA6 with RBP. Group 3 (STRA6-M2, STRA6-M4, STRA6-M6, STRA6-M9, STRA6-M11, STRA6-M13, and STRA6-M14) does not have the Myc epitope expressed extracellularly and has significant vitamin A uptake activity and RBP binding activity. For Group 3, the absence of anti-Myc antibody binding in live cells is due to the insertion of the epitope on the intracellular regions of STRA6, because these proteins are still expressed on the cell surface and anti-Myc antibody can stain the permeabilized cells. Group 4 (STRA6-M7, STRA6-M8, and STRA6-M12) also does not have the Myc epitope expressed extracellularly, but has no vitamin A uptake activity or RBP binding activity. For Group 4, the absence of extracellular localization of the Myc epitope is due to the absence of surface expression of these proteins. A transmembrane topology model for STRA6 (Figure 7) is proposed based on these experiments, the lysine accessibility experiment (Figure 6), and the logic that a transmembrane domain is flanked by an extracellular domain and an intracellular domain.
Figure 7.
Transmembrane topology model of STRA6. Upper panel, a summary of the properties of the STRA6-Myc proteins. In the second and third columns, a plus sign indicates positive staining and a minus sign indicates negative staining. Lower panel, a model for the transmembrane topology of STRA6. Locations of the Myc epitopes are indicated. The Myc epitopes that were on the extracellular side and substantially reduced STRA6’s RBP binding are marked by asterisks. The lysine residue in STRA6 shown by live cell biotinylation to be exposed on cell surface is also shown. The two extracellular cysteines (residue 44 and 195 of bovine STRA6) are also noted. Missense mutations in STRA6 associated with human birth defects (27) are indicated as black circles with white letters. These mutations are P90(91)L, P293(294)L,T321(322)P, T644(643)M and R655(654)C (numbers in parenthesis are amino acid residues in bovine STRA6).
Functional effects of the loss of extracellular cysteines in STRA6
There are two extracellular cysteines (residue 44 and 195 of bovine STRA6) in the STRA6 topology model (Figure 7). These cysteines are highly conserved. We tested the effect of mutating these two cysteines to alanines on STRA6’s functions (Figure 8). We found that the C44A and C195A mutants have significant losses of vitamin A uptake activities from holo-RBP (Figure 8A) and RBP binding activities (Figure 8B). However, their cell surface expression levels are similar to that of wild-type STRA6 (Figure 8C and 8D). These results suggest that the loss of RBP binding, instead of loss of cell-surface expression, is largely responsible for the loss of vitamin A uptake activity for these mutants. Since RBP is an extracellular ligand of STRA6, the effects of these mutations on RBP binding are consistent with the extracellular locations of these two cysteine residues.
Figure 8.
Comparison of wild-type STRA6 and C44A and C195A mutants. For live-cell staining and cell-surface expression quantitation of STRA6, all STRA6 variants in this figure are tagged with a Myc tag on an extracellular loop of bovine STRA6 (between residues 133 and 134). A. 3H-retinol uptake activity from 3H-retinol/RBP. B. AP-RBP binding activity. C. Quantitation of cell surface expression. D. Live cell staining using anti-Myc antibody. For A, B, and C, the activity of wild-type STRA6 is defined as 100%.
DISCUSSION
STRA6 functions both as a membrane receptor for RBP and a transport protein that mediates cellular vitamin A uptake from holo-RBP. In this paper, we determined STRA6’s transmembrane topology, which is a fundamental aspect of the structural information for a multi-transmembrane domain protein. Compared with one orientation (extracellular N-terminus) of the computer predicted model (Figure 1), our experimentally-confirmed model is similar to the predictions in the N-terminal half of the protein. Our topology model suggests that STRA6 has 19 distinct domains including 5 extracellular domains, 9 transmembrane domains, and 5 intracellular domains (Figure 7). Many membrane transporters have 8–12 transmembrane domains (35). Although STRA6 represents a new membrane transport protein not homologous to any protein of known function, its number of transmembrane domains lies within this range. Most, if not all, proteins with more than 7 transmembrane domains function as membrane transporters or channels. A large number of transmembrane membranes domains potentially makes it more feasible to form a specific transmembrane pore, through which the ligand of the transporter or channel can pass through.
In addition to epitope tagging and lysine accessibility, the locations of cysteine residues also support this topology model. There are only two cysteine residues located extracellularly in the topology model (Figure 7). Consistent with the importance of extracellular cysteines in stabilizing protein structure through disulfide bond, this pair of cysteines is absolutely conserved from Xenopus STRA6 to human STRA6. We have shown that a mutation in either of these two cysteines leads to a substantial loss in RBP binding activity, supporting the extracellular locations of these residues (Figure 8). Due to the unique biochemical properties and functions of cysteines, non-conserved cysteines are seldom, if ever, found in extracellular domains of eukaryotic proteins. Consistently, all three non-conserved cysteine residues in bovine STRA6 are located intracellularly in the topology model (one near M6, one near M9, and one near M12).
This study identified two regions (M5 and M10) of STRA6 that are likely to participate in the interaction of STRA6 and RBP. Although STRA6-M5 and STRA6-M10 are expressed well on the cell surface, they have substantially reduced RBP binding activity. This model also reveals a large C-terminal region of STRA6. In addition to the evidence presented in this study, there is an additional piece of evidence supporting the intracellular orientation of the C-terminal domain. A polyclonal antibody recognizing an epitope on the C-terminal domain cannot recognize STRA6 on live cells, but can recognize STRA6 in permeabilized cells (data not shown). The C-terminal region of STRA6 is highly conserved between human, mouse, and bovine STRA6. Why does STRA6 need such a long intracellular C-terminal domain? It is likely involved in vitamin A uptake into cells and/or in the cellular targeting of STRA6. STRA6 is specifically targeted to the basolateral membrane of the RPE (26), a location consistent with its role in vitamin A uptake from the holo-RBP in choroidal blood. In addition, two pathogenic missense mutations in human STRA6 associated with severe birth defects are located on the C-terminus (Figure 7).
The elucidation of the transmembrane topology of STRA6 lays the groundwork for understanding the structure and function of this novel receptor and membrane transport protein. The transmembrane topology model demarcates STRA6 into extracellular, intracellular, and transmembrane domains. This general demarcation is critically important for future studies aimed at understanding the roles of these regions in STRA6 function and in elucidating the structural basis of its function both as a high-affinity receptor for RBP and in the transport of vitamin A into cells.
Abbreviations
- RBP
retinol binding protein
- Holo-RBP
retinol/RBP complex
- STRA6
stimulated by retinoic acid gene 6
- LRAT
lecithin retinol acyltransferase
- RPE
retinal pigment epithelium
Footnotes
This project was supported by National Institute of Health grant 1R01EY018144 (H.S.).
REFERENCES
- 1.Ross AC, Gardner EM. The function of vitamin A in cellular growth and differentiation, and its roles during pregnancy and lactation. Adv Exp Med Biol. 1994;352:187–200. doi: 10.1007/978-1-4899-2575-6_15. [DOI] [PubMed] [Google Scholar]
- 2.Blomhoff R. Overview of Vitamin A Metabolism and Function. In: Blomhoff R, editor. Vitamin A in Health and Disease. Marcel Dekker, Inc; 1994. pp. 1–35. [Google Scholar]
- 3.Mark M, Ghyselinck NB, Chambon P. Function of retinoid nuclear receptors: lessons from genetic and pharmacological dissections of the retinoic acid signaling pathway during mouse embryogenesis. Annu Rev Pharmacol Toxicol. 2006;46:451–480. doi: 10.1146/annurev.pharmtox.46.120604.141156. [DOI] [PubMed] [Google Scholar]
- 4.Dew SE, Ong DE. Specificity of the retinol transporter of the rat small intestine brush border. Biochemistry. 1994;33:12340–12345. doi: 10.1021/bi00206a042. [DOI] [PubMed] [Google Scholar]
- 5.Carlson A, Bok D. Promotion of the release of 11-cis-retinal from cultured retinal pigment epithelium by interphotoreceptor retinoid-binding protein. Biochemistry. 1992;31:9056–9062. doi: 10.1021/bi00152a049. [DOI] [PubMed] [Google Scholar]
- 6.Carlson A, Bok D. Polarity of 11-cis retinal release from cultured retinal pigment epithelium. Invest Ophthalmol Vis Sci. 1999;40:533–537. [PubMed] [Google Scholar]
- 7.Weng J, Mata NL, Azarian SM, Tzekov RT, Birch DG, Travis GH. Insights into the function of Rim protein in photoreceptors and etiology of Stargardt's disease from the phenotype in abcr knockout mice. Cell. 1999;98:13–23. doi: 10.1016/S0092-8674(00)80602-9. [DOI] [PubMed] [Google Scholar]
- 8.Sun H, Molday RS, Nathans J. Retinal stimulates ATP hydrolysis by purified and reconstituted ABCR, the photoreceptor-specific ATP-binding cassette transporter responsible for Stargardt disease. J Biol Chem. 1999;274:8269–8281. doi: 10.1074/jbc.274.12.8269. [DOI] [PubMed] [Google Scholar]
- 9.Ahn J, Wong JT, Molday RS. The effect of lipid environment and retinoids on the ATPase activity of ABCR, the photoreceptor ABC transporter responsible for Stargardt macular dystrophy. J Biol Chem. 2000;275:20399–20405. doi: 10.1074/jbc.M000555200. [DOI] [PubMed] [Google Scholar]
- 10.Newcomer ME, Ong DE. Plasma retinol binding protein: structure and function of the prototypic lipocalin. Biochim Biophys Acta. 2000;1482:57–64. doi: 10.1016/s0167-4838(00)00150-3. [DOI] [PubMed] [Google Scholar]
- 11.Zanotti G, Berni R. Plasma retinol-binding protein: structure and interactions with retinol, retinoids, and transthyretin. Vitam Horm. 2004;69:271–295. doi: 10.1016/S0083-6729(04)69010-8. [DOI] [PubMed] [Google Scholar]
- 12.Bok D, Heller J. Transport of retinol from the blood to the retina: an autoradiographic study of the pigment epithelial cell surface receptor for plasma retinol-binding protein. Exp Eye Res. 1976;22:395–402. doi: 10.1016/0014-4835(76)90177-9. [DOI] [PubMed] [Google Scholar]
- 13.Chen CC, Heller J. Uptake of retinol and retinoic acid from serum retinol-binding protein by retinal pigment epithelial cells. J Biol Chem. 1977;252:5216–5221. [PubMed] [Google Scholar]
- 14.Rask L, Peterson PA. In vitro uptake of vitamin A from the retinol-binding plasma protein to mucosal epithelial cells from the monkey's small intestine. J Biol Chem. 1976;251:6360–6366. [PubMed] [Google Scholar]
- 15.Heller J. Interactions of plasma retinol-binding protein with its receptor. Specific binding of bovine and human retinol-binding protein to pigment epithelium cells from bovine eyes. J Biol Chem. 1975;250:3613–3619. [PubMed] [Google Scholar]
- 16.Heller J, Bok D. Transport of retinol from the blood to the retina: involvement of high molecular weight lipoproteins as intracellular carriers. Exp Eye Res. 1976;22:403–410. doi: 10.1016/0014-4835(76)90178-0. [DOI] [PubMed] [Google Scholar]
- 17.Maraini G, Gozzoli F. Binding of retinol to isolated retinal pigment epithelium in the presence and absence of retinol-binding protein. Invest Ophthalmol. 1975;14:785–787. [PubMed] [Google Scholar]
- 18.Bhat MK, Cama HR. Gonadal cell surface receptor for plasma retinol-binding protein. A method for its radioassay and studies on its level during spermatogenesis. Biochim Biophys Acta. 1979;587:273–281. doi: 10.1016/0304-4165(79)90360-x. [DOI] [PubMed] [Google Scholar]
- 19.Torma H, Vahlquist A. Vitamin A uptake by human skin in vitro. Arch Dermatol Res. 1984;276:390–395. doi: 10.1007/BF00413360. [DOI] [PubMed] [Google Scholar]
- 20.Sivaprasadarao A, Findlay JB. The interaction of retinol-binding protein with its plasma-membrane receptor. Biochem J. 1988;255:561–569. [PMC free article] [PubMed] [Google Scholar]
- 21.Shingleton JL, Skinner MK, Ong DE. Characteristics of retinol accumulation from serum retinol-binding protein by cultured Sertoli cells. Biochemistry. 1989;28:9641–9647. doi: 10.1021/bi00451a015. [DOI] [PubMed] [Google Scholar]
- 22.MacDonald PN, Bok D, Ong DE. Localization of cellular retinol-binding protein and retinol-binding protein in cells comprising the blood-brain barrier of rat and human. Proc Natl Acad Sci U S A. 1990;87:4265–4269. doi: 10.1073/pnas.87.11.4265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sivaprasadarao A, Findlay JB. Structure-function studies on human retinol-binding protein using site-directed mutagenesis. Biochem J. 1994;300(Pt 2):437–442. doi: 10.1042/bj3000437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Smeland S, Bjerknes T, Malaba L, Eskild W, Norum KR, Blomhoff R. Tissue distribution of the receptor for plasma retinol-binding protein. Biochem J. 1995;305(Pt 2):419–424. doi: 10.1042/bj3050419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Liden M, Eriksson U. Development of a versatile reporter assay for studies of retinol uptake and metabolism in vivo. Exp Cell Res. 2005;310:401–408. doi: 10.1016/j.yexcr.2005.08.002. [DOI] [PubMed] [Google Scholar]
- 26.Kawaguchi R, Yu J, Honda J, Hu J, Whitelegge J, Ping P, Wiita P, Bok D, Sun H. A membrane receptor for retinol binding protein mediates cellular uptake of vitamin A. Science. 2007;315:820–825. doi: 10.1126/science.1136244. [DOI] [PubMed] [Google Scholar]
- 27.Pasutto F, Sticht H, Hammersen G, Gillessen-Kaesbach G, Fitzpatrick DR, Nurnberg G, Brasch F, Schirmer-Zimmermann H, Tolmie JL, Chitayat D, Houge G, Fernandez-Martinez L, Keating S, Mortier G, Hennekam RC, von der Wense A, Slavotinek A, Meinecke P, Bitoun P, Becker C, Nurnberg P, Reis A, Rauch A. Mutations in STRA6 Cause a Broad Spectrum of Malformations Including Anophthalmia, Congenital Heart Defects, Diaphragmatic Hernia, Alveolar Capillary Dysplasia, Lung Hypoplasia, and Mental Retardation. Am J Hum Genet. 2007;80:550–560. doi: 10.1086/512203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Golzio C, Martinovic-Bouriel J, Thomas S, Mougou-Zrelli S, Grattagliano-Bessieres B, Bonniere M, Delahaye S, Munnich A, Encha-Razavi F, Lyonnet S, Vekemans M, Attie-Bitach T, Etchevers HC. Matthew-Wood Syndrome Is Caused by Truncating Mutations in the Retinol-Binding Protein Receptor Gene STRA6. Am J Hum Genet. 2007;80:1179–1187. doi: 10.1086/518177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Borjigin J, Nathans J. Insertional mutagenesis as a probe of rhodopsin's topography, stability, and activity. J Biol Chem. 1994;269:14715–14722. [PubMed] [Google Scholar]
- 30.Canfield VA, Levenson R. Transmembrane organization of the Na,K-ATPase determined by epitope addition. Biochemistry. 1993;32:13782–13786. doi: 10.1021/bi00213a005. [DOI] [PubMed] [Google Scholar]
- 31.Guan L, Kaback HR. Site-directed alkylation of cysteine to test solvent accessibility of membrane proteins. Nat Protoc. 2007;2:2012–2017. doi: 10.1038/nprot.2007.275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Bouillet P, Sapin V, Chazaud C, Messaddeq N, Decimo D, Dolle P, Chambon P. Developmental expression pattern of Stra6, a retinoic acid-responsive gene encoding a new type of membrane protein. Mech Dev. 1997;63:173–186. doi: 10.1016/s0925-4773(97)00039-7. [DOI] [PubMed] [Google Scholar]
- 33.Szeto W, Jiang W, Tice DA, Rubinfeld B, Hollingshead PG, Fong SE, Dugger DL, Pham T, Yansura DG, Wong TA, Grimaldi JC, Corpuz RT, Singh JS, Frantz GD, Devaux B, Crowley CW, Schwall RH, Eberhard DA, Rastelli L, Polakis P, Pennica D. Overexpression of the retinoic acid-responsive gene Stra6 in human cancers and its synergistic induction by Wnt-1 and retinoic acid. Cancer Res. 2001;61:4197–4205. [PubMed] [Google Scholar]
- 34.Sun H, Smallwood PM, Nathans J. Biochemical defects in ABCR protein variants associated with human retinopathies. Nat Genet. 2000;26:242–246. doi: 10.1038/79994. [DOI] [PubMed] [Google Scholar]
- 35.Hediger MA, Romero MF, Peng JB, Rolfs A, Takanaga H, Bruford EA. The ABCs of solute carriers: physiological, pathological and therapeutic implications of human membrane transport proteins Introduction. Pflugers Arch. 2004;447:465–468. doi: 10.1007/s00424-003-1192-y. [DOI] [PubMed] [Google Scholar]