Protein-protein interactions and membrane localization of the human organic solute transporter

An-Qiang Sun, Natarajan Balasubramaniyan, Ke Xu, Chuan Ju Liu, Vijaya M. Ponamgi, Hongguang Liu, Frederick J. Suchy

Abstract

Two proteins that mediate bile acid export from the ileal enterocyte, organic solute transporter (OST)-α and -β, have recently been identified. It is unclear whether these two proteins associate directly and how they interact to mediate transport function and membrane localization. In this study, the protein-protein interactions, transport functions, and membrane localization of human (h)OST-α and -β proteins were examined. The results demonstrated that coexpression of hOST-α and -β in transfected cells resulted in a three- to fivefold increase of the initial rate of taurocholate influx or efflux compared with cells expressing each protein individually and nontransfected cells. Confocal microscopy demonstrated plasma membrane colocalization of hOST-α and -β proteins in cells cotransfected with hOST-α and -β cDNAs. Protein-protein interactions between hOST-α and -β were demonstrated by mammalian two-hybrid and coimmunoprecipitation analyses. Truncation of the amino-terminal 50 amino acid extracellular residues of hOST-α abolished its interaction with hOST-β and led to an intracellular accumulation of the two proteins and to only background levels of taurocholate transport. In contrast, carboxyl-terminal 28 amino acid truncated hOST-α still interacted with hOST-β, and majority of this cytoplasmic tail-truncated protein was expressed on the basolateral membrane when it was stably cotransfected with hOST-β protein in Madin-Darby canine kidney cells. In summary, hOST-α and -β proteins are physically associated. The intracellular carboxyl-terminal domain of hOST-α is not essential for this interaction with hOST-β. The extracellular amino-terminal fragment of hOST-α may contain important information for the assembly of the heterodimer and trafficking to the plasma membrane.

  • bile acid transporter

bile acids are the end products of cholesterol catabolism and play a critical role in a multitude of biological processes including bile secretion and absorption of fat and fat-soluble vitamins. Bile acids are conserved by an efficient enterohepatic circulation that involves many transport proteins in the liver, intestine, and kidney. Organic solute transporter-α/β (Ost-α/β) is a novel heteromeric bile acid and sterol transporter expressed at the basolateral membranes of the epithelium in the ileum, kidney, and liver (2, 3, 5). Human (h)OST-α is composed of 340 amino acids and has a predicted structure by hydropathy analysis of seven transmembrane (TM) domains. hOST-β contains 128 amino acids and has one TM domain near the NH2-terminus. These two proteins function together to transport bile acids and other important steroid-derived molecules via a Na+-independent mechanism (3, 5). Transport by OST appears to occur by facilitated diffusion, as transport activity is not affected by depletion of intracellular ATP or by changes in transmembrane Na+, K+, H+, or Cl concentration gradients (3). Mouse Ost mRNA has only been detected in the kidney and intestine (5), whereas hOST-α and -β are expressed together in many tissues, including the liver, intestine, kidney, testis, heart, lung, and brain (2, 17). The mouse Ost-α and -β promoters contain functional farnesoid X receptor (FXR) and liver receptor homolog (LRH)-1 elements, which, respectively, mediate positive and negative feedback regulation by bile acids. The positive regulatory pathway appears to be dominant (6). Ost-α and -β are upregulated in response to cholestatic liver injury, and this response is dependent on the FXR bile acid receptor (4, 13, 21). Dawson et al. (5) have shown that coexpression of mouse Ost-α and -β was required to convert the Ost α-subunit to a mature glycosylated endoglycosidase H-resistant form, suggesting that coexpression facilitates the trafficking of mouse Ost-α through the Golgi apparatus. Immunolocalization studies have shown that coexpression was also necessary for plasma membrane expression of both mouse Ost-α and -β. These results indicate that mouse Ost-α and -β proteins are basolateral bile acid carriers and are largely responsible for bile acid efflux in the ileum and other Na+-dependent bile acid transporter (ASBT)-expressing tissues (5). However, so far, there is no experimental evidence to show that hOST-α and -β proteins associate directly and what domains of these proteins are involved in forming a heteromer. In this study, the protein-protein interactions, transport functions, and membrane localization of hOST-α and -β proteins were examined by a combination of molecular biological, biochemical, and confocal microscopic methods.

MATERIALS AND METHODS

Amplification of OST cDNA fragments from the human liver cDNA library.

hOST-α and -β cDNA fragments were amplified from the human liver Quick-Clone cDNA library (BD Biosciences Clontech) by PCR using specific primers according to the manufacturer's directions. The primers used for the PCR amplification were 5′-gattgctggagagaacgc-3′ (forward) and 5′-ttgtccaagccatccacc-3′ (reverse) for hOST-α and 5′-ctcgttgcacacgctacc-3′ (forward) and 5′-tgtgtctggcttaggatgg-3′ (reverse) for hOST-β. After subcloning into expression vectors, selected clones were isolated and purified, and the plasma DNA was sequenced. The DNA sequence alignment revealed that an amplified cDNA fragment was identical to the sequence of hOST-α reported by GenBank (Accession No. AY194243; data not shown). For the hOST-β protein, we obtained a cDNA fragment that has a 2-bp (nt no. G66 and G93) difference compared with the hOST-β sequence reported by GenBank (A66 and A93, Accession No. XM 058693; data not shown). However, these differences did not change the amino acid reading code, and the amino acid sequence of the amplified fragment was identical to hOST-β protein reported by GenBank (Accession No. XM 058693). Therefore, these two amplified cDNA constructs of hOST-α and -β were used in all of the experiments reported here.

Construction of epitope-tagged and mutant hOST.

A combination of restriction enzyme digestion and PCR were used to generate epitope-tagged and mutant transporters with hOST-α and -β cDNAs as templates. The PCR was done with oligonucleotide primers generated from DNA sequencing information. Wild-type and mutant hOST-α and -β cDNAs were subcloned into mammalian expression vector pcDNA3.1 (Invitrogen) and/or a green fluorescent protein (GFP) fusion protein vector (pEGFPN2; Clonetech, Palo Alto, CA) at the XhoI/SalI sites. For the mammalian two-hybrid (M2H) system (Promega), full-length wild-type hOST-α and -β cDNAs as well as fragments corresponding to amino acids 1–313 [carboxyl-end truncated (α-CT)] and amino acids 51–340 [NH2-end truncated (α-NT)] of hOST-α were amplified by PCR and cloned in-frame into the SalI/NotI sites of pBIND or pACT fusion vectors (Promega) as previously described (20). GFP fused wild-type and COOH/NH2-end truncated hOST-α and -β constructs were generated by insertion in-frame into the pEGFPN2 vector, as previously described (19). The pBudCE4.1 vector (Invitrogen) was designed for simultaneous expression of two genes in mammalian cell lines. The pBudCE4.1 vector contains a human cytomegalovirus (CMV) immediate-early promoter and a human elongation factor 1α-subunit (EF-1α) promoter for independent expression of two recombinant proteins. The fragments corresponding to wild-type and truncated hOST-α and -β were inserted in-frame into the PstI/XbaI sites of the CMV promoter of the pBudCE4.1 vector to generate 6× histidine (6xHis)-tagged constructs. GFP-fused wild-type and/or mutated hOST-α and -β were inserted in-frame into the NotI/XhoI sites of the EF-1α promoter of the pBudCE4.1 vector. All of the positive clones containing cDNA inserts were identified by restriction enzyme mapping and sequenced using the model 377 ABI automated DNA sequencer at the DNA Core Facility of Mount Sinai School of Medicine.

Establishment of transiently and stably transfected cell lines.

Cell culture and DNA plasmid transient transfection of COS-7 and HEK-293 cells were done as previously described (18). For the stably transfected cell model, the rat ileal (r)Asbt and wild-type or mutated hOST-α and -β (both constructed in the pBudCE4.1 vector) were stably expressed in Madin-Darby canine kidney (MDCK) cells as follows. On day 1, 60-mm plates were seeded with 5 × 104 rAsbt stably transfected MDCK cells (18). On day 2, cells were transfected with 3 μg of pBudCE4.1 with wild-type or mutated hOST-α and -β constructs using FuGENE 6 transfection reagent (Roche Applied Science). On day 3, cells were split into two 100-mm dishes in culture medium containing 0.9 mg/ml G418 (Invitrogen) and 250 μg/ml Zeocin (Invitrogen). After selection for ∼15 days, individual colonies were picked, expanded in 35-mm plates, and screened by Na+-dependent bile acid uptake activity for rAsbt and by confocal imaging and immunoblotting for GFP or 6xHis-tagged wild-type and mutated hOST-α and -β protein expression.

RT-PCR.

mRNA levels of transfected constructs were quantified by RT-PCR. For each target gene, the oligonucleotide primer sequences used in this study were as follows: 1) hOSTα, 5′-ctgcttctctcagcctccca-3′ (forward) and 5′-gggcaaagggtgttcttgta-3′ (reverse); 2) hOST-β, 5′-gctgctggaagagatgcttt-3′ (forward) and 5′-cctcatccaaatgcaggact-3′ (reverse); and 3) rAsbt, 5′-actcaggaacgattgtgatcc-3′ (forward) and 5′-ttgaccagctagtctagcca-3′ (reverse). RNA isolation and cDNA synthesis were carried out by the TRIzol and SuperScript Reverse Transcriptase methods (Invitrogen) as described by manufacturer. Briefly, total RNA was extracted from stably transfected MDCK cells using TRIzol reagent and precipitated by isopropanol. RNA was quantified by a sphectrophotometer, and equal amounts were used for the cDNA preparation from the different transfected cell lines; 2 μl of the total 20-μl volume were used. The PCR was carried out according to the GoldTaq Green Master Mix instructions from Promega.

Influx and efflux transport assays.

Na+-dependent and -independent taurocholate (TC) influx assays were performed in 12-well plates or using a Transwell filter culture system as previously described (15, 18). Briefly, for the Transwell filter system (Costar, Cambridge, MA), transfected and untransfected cells were grown to confluence for 4–5 days on 0.45-μm pore size Transwell filter inserts. [3H]TC uptake was performed at 37°C for 10 min. Confluent cell monolayers grown on Transwell filters were washed twice with warm sodium or choline uptake buffer [116 mM NaCl (or choline), 5.3 mM KCl, 1.1 mM KH2PO4, 0.8 mM MgSO4, 1.8 mM CaCl2, 11 mM d-glucose, and 10 mM HEPES; pH 7.4], and each well was incubated from the apical (0.2 ml) or basolateral (0.6 ml) side with uptake buffer containing 10 μM [3H]TC at the final concentration. After a 10-min incubation, uptake assays were terminated by aspirating the medium, and the filters were successively dipped into three beakers, each of which contained 100 ml of ice-cold uptake buffer. Filters were excised from cups, and attached cells were solubilized in 0.2 ml of 1% SDS and transferred into scintillation vials with 4 ml Optifluor (DuPont-New England Nuclear). Protein concentrations were determined with the Bio-Rad Protein Assay kit. Na+-independent [+H]estrone-3-sulfate (E3S) influx assays were performed as previously described (9). Stably transfected MDCK cells (see above) were used to examine hOST efflux transport activity using rAsbt as the bile acid loading pump. For Na+-independent bile acid efflux assays in the Transwell filter system, stably transfected MDCK cells were plated on 24-well plates with 6.4-mm Transwell filter inserts (Costar) and cultured in the culture medium containing 0.9 mg/ml G418 (Invitrogen) and 250 μg/ml Zeocin (Invitrogen). After ∼4–5 days, cells were treated with 10 mM sodium butyrate for 16 h at 37°C to induce the expression of the transfected genes. Cell monolayers were washed three times with Na+ transport buffer and incubated at 37°C for 10 min with Na+ transport buffer plus 10 μM [3H]TC added to the apical chambers. Cells were then washed three times in ice-cold choline buffer, choline buffer was added to both the upper and lower chambers, and cells were continually incubated at 37°C for 5 min. After the cell incubation, the choline buffer in the lower chamber and the cells were harvested to determine basolateral domain effluxes (accumulated radioactivity in the lower chamber buffer) and cell-associated radioactivity by scintillation counting. The percentage of Na+-independent TC efflux was calculated from the cell-associated radioactivity and the amounts in the basolateral and apical chambers at the end of the 5-min incubation with the choline buffer.

M2H assay for protein-protein interactions in vivo.

M2H experiments were performed as previously described (20). Briefly, when cell cultures reached ∼50% confluence in six-well plates, COS-7 cells were cotransfected with 1 μg each of the pBIND-hOST-α (pBIND-α) and pACT-hOST-β (pACT-β) plasmids using Lipofectamine (Life Technologies) according to the manufacturer's recommendations. Cultures were harvested 48 h after transfection and lysed. The firefly luciferase activity was determined according to the manufacturer's recommendations.

Indirect immunofluorescence microscopy.

Briefly, indirect immunofluorescence microscopy was performed on a confluent monolayer of transfected cells cultured on glass coverslips. Cells were fixed and permeabilized for 7 min in 100% of methanol at −20°C, followed by rehydration in PBS. Nonspecific sites were blocked with normal goat serum for 60 min at room temperature. The rabbit anti-6xHis antibody (1:1,000) was diluted in the blocking buffer (1% BSA, 10% goat serum, and 0.05% Tween 20 in PBS) at 4°C overnight. After being washed with PBS for 15–30 min, cells were incubated with goat anti-rabbit IgG antibodies conjugated to Texas red for 1 h. After being washed with PBS, cells on coverslips were inverted onto a drop of VectaShield. Fluorescence was examined with a Leica TCS-SP (UV) four-channel confocal laser scanning microscope in the Imaging Core Facility Microscopy Center of Mount Sinai School of Medicine. A semiquantitative image analysis of protein colocalization was performed using the MetaMorph program (Metamorph Offline, version 6.3 r3, Molecular Devices).

Coimmunoprecipitation.

MDCK cells stably cotransfected with epitope (GFP or 6xHis)-tagged wild-type or truncated hOST-α and hOST-β proteins were used for coimmunoprecipitation (co-IP). Cell monolayers in two 100-mm dishes were washed with PBS and lysed with 500 μl/dish of lysis buffer, which contained 0.5% Nonidet P-40, 50 mM Tris·HCl (pH 7.5), 1 mM EDTA, 150 mM NaCl, and protease inhibitor cocktails (10 μg/ml aprotinin, 20 μg/ml phosphoramidon, 40 μg/ml Pefabloc, 1 μg/ml leupeptin, and 1 μg/ml pepstatin), for 15 min at 4°C. The supernatant was collected following centrifugation; 200 μl of the cell extract from postnuclear supernatants were treated with protein A-Sepharose and preimmune serum and incubated with 6 μg rabbit polyclonal antibody against 6xHis tag (Abcam) or rabbit polyclonal GFP (FL) antibody (Santa Cruz Biotechnology) for 2 h at 4°C. Then, 100 μl of protein A-agarose were added, and samples were incubated overnight at 4°C in a rotating shaker. Immune complexes were recovered by centrifugation, and samples were washed three times with PBS and released from the beads by 100 μl of 1× Laemmli sample buffer. Total proteins (∼50 μg) from the cell extracts and immunoprecipitated proteins (∼1 μg) were analyzed by SDS-PAGE and Western blotted with corresponding antibodies to detect each protein.

RESULTS

Transport activity and cellular distribution of hOST-α and -β proteins.

To verify the transport function and cellular distribution of hOST-α and -β proteins amplified from the human liver cDNA library (see materials and methods), the transport activity of COS7 cells expressing OST-α and -β proteins together or individually was examined. The results showed that the initial rates of Na+-independent E3S and TC transport were increased about three- to fivefold in hOST-α and -β cDNA cotransfected cells compared with nontransfected cells, but there was no detectable increase of transport activity in cells individually transfected with hOST-α or -β cDNA (Fig. 1, A and B).

Fig. 1.

Transport activity of cells transfected with human organic solute transporter (hOST). Na+-independent [3H]estrone-3-sulfate (E3S) (A) and [3H]taurocholate (TC) (B) uptake were measured in confluent monolayers of hOST-α and -β transiently cotransfected or individually transfected COS7 cells grown on 12-well plates. Transfected cells were preincubated in 3 μM [3H]E3S (A) or 10 μM [3H]TC (B) at 37°C for 10 min, respectively. C and D: Na+-independent [3H]TC efflux (C) and Na+-dependent [3H]TC uptake (D) were measured in Madin-Darby canine kidney (MDCK) cells stably transfected with rat ileal Na+-dependent bile salt transporter (rAsbt), hOST-α and -β, or green fluorescent protein (GFP)-tagged hOST-α (α-GFP) and hOST-β (β-GFP) cDNA. Data represent means ± SE (bars) of triplicate determinations from at least 3 different cell culture preparations. *P < 0.05, significant difference from hOST-nontransfected COS7 cells. E: mRNA levels of rAsbt and wild-type and epitope-tagged hOST-α and -β constructs in stably transfected MDCK cells were obtained using the RT-PCR (see materials and methods), and the reaction products were visualized by staining with ethidium bromide. The results shown in D and E indicate that these stably transfected MDCK cells have similar gene expression levels for rAsbt and wild-type and epitope-tagged hOST-α and -β as well as equal TC uptake activity.

To demonstrate polarized basolateral efflux transport activity and the cellular distribution of these two proteins, GFP or 6xHis epitope-tagged OST-α and -β constructs were made. The Na+-independent efflux of TC was examined in MDCK cells stably transfected with rAsbt and hOST-α and -β cDNAs. In this model system, rAsbt acted as an apical bile acid loading pump in stably transfected MDCK cells. The bile acid efflux activity from the basolateral domain by hOST-α and -β proteins was measured by radioactive scintillation counting. Figure 1C shows that TC efflux was increased by approximately three- to fourfold in transfected MDCK cells. Cells cotransfected with GFP or 6xHis tag-fused OST-α and -β had similar efflux activity compared with wild-type OST-α and -β cotransfected cells (Fig. 1C). The Na+-dependent TC uptake activity from apical membrane by rAsbt of stably transfected MDCK cells is shown in Fig. 1D. Figure 1E shows the expression levels of rAsbt and wild-type and epitope-tagged hOST-α and -β in stably transfected MDCK cell lines by RT-PCR. The results shown in Fig. 1, D and E, indicate that these stably transfected cells have similar gene expression levels for rAsbt, hOST-α, and hOST-β and equivalent Na+-dependent TC uptake activity.

Next, we examined the subcellular distribution of hOST-α and -β proteins in transiently transfected HEK-293 cells and stably transfected MDCK cells. hOST-α (α-GFP) was expressed intracellularly in individually transfected HEK-293 and MDCK cells (Fig. 2). hOST-β (β-GFP) was located randomly in the plasma membrane and cytoplasm (Fig. 2). In contrast, in cells cotransfected with both hOST-α and -β cDNAs, these two proteins were colocalized to the plasma membrane of nonpolarized HEK-293 cells and in the basolateral membrane domain of polarized MDCK cells (Fig. 2). These results are consistent with the transport experiments and data from previous studies and indicate that the coexpression of hOST-α and -β proteins is required for transport function and specific polarized plasma membrane localization.

Fig. 2.

Confocal microscopy analysis of hOST-α and -β protein cellular distribution in transfected cells. HEK-293 cells (top) and MDCK cells (bottom) were individually tranfected or cotransfected with GFP-fused hOST-α (α-GFP) and hOST-β (β-GFP) constructs. The membrane expression of GFP-fused hOST-α and hOST-β in transfected HEK-293 and MDCK cells was detected by confocal fluorescence microscopy. Confocal photomicrographs of GFP-fused hOST were obtained on a confluent monolayer of transfected cells cultured on glass coverslips. The glass coverslip-grown cells were fixed in 100% methanol and mounted with Vectashield mounting medium. The fluorescence was examined with a Leica TCS-SP (UV) four-channel confocal laser scanning microscope.

Association of hOST-α and -β proteins.

There is no direct experimental evidence showing the direct physical association of hOST-α and -β proteins. To demonstrate the interaction of hOST-α and -β proteins in vivo, the M2H assay was used, in which the expression of the firefly luciferase gene is driven by five Gal4-specific enhancer elements. The complete coding sequence of hOST-α was inserted in-frame into the SalI/NotI sites of the pBIND fusion vector fused with the Gal4 DNA binding domain (pBIND-α). A cDNA fragment encoding hOST-β was inserted in-frame into the SalI/NotI sites of the pACT fusion vector fused with the VP16 transactivation domain (pACT-β). The two constructs were cotransfected into COS-7 cells. In the M2H system, the interaction between hOST-α and -β proteins brought together the Gal4 DNA binding domain and VP16 activation domains of the fusion proteins and activated the reporter gene in COS-7 cells. A greater than threefold increase of luciferase activity was observed in hOST-α/β (pBIND-α/pACT-β) cotransfected cells compared with pBIND-α/pACT vector or pBIND/pACT-β vector cotransfected cells (Fig. 3A). This result indicates that these two proteins associate physically in mammalian cells.

Fig. 3.

Interactions of hOST-α and -β proteins. A: interactions between hOST-α and hOST-β were examined by the mammalian two-hybrid (M2H) luciferase assay. Five copies of the Gal4 DNA recognition sequence in the reporter plasmid drive the expression of firefly luciferase. A fragment encoding the full open reading frame of hOST-α was linked to a sequence encoding the Gal4 DNA binding domain (GalDBD) in the pBIND vector (pBIND-α). A fragment encoding the full open reading frame of hOST-β was linked to a sequence encoding the VP16 activation domain (VP16AD) in the pACT vector (pACT-β). The plasmids, as indicated, were transfected into COS-7 cells. After cells had been incubated, firefly luciferase activity was determined in cell lysates. B: coimmunoprecipitation (co-IP) of 6xHis-tagged hOST-α (α-His) with GFP-fused hOST-β (β-GFP) proteins in stably cotransfected MDCK cells. MDCK cells stably cotransfected with 6xHis-tagged hOST-α and GFP-fused hOST-β (α-His + β-GFP) were used in this co-IP experiment. MDCK cells, nontransfected (MDCK) and stably cotransfected with 6xHis-tagged hOST-α and GFP (α-His + GFP), were used as negative controls. Total protein extracted from stably transfected MDCK cells was immunoprecipitated by a poly-His antibody (IP: his-Ab). Total protein (∼50 μg) extracted from stably transfected MDCK cells (cell extract) and coimmunoprecipitated proteins (∼1 μg) were isolated by SDS-PAGE and analyzed by Western blot (WB) using the GFP antibody to detect GFP-tagged hOST-β (WB: GFP Ab).

The interaction of hOST-α and -β proteins was further established by co-IP (Fig. 3B) and colocalization by confocal microscopy (Fig. 4A). MDCK cells stably cotransfected with hOST-α-6xHis and hOST-β-GFP were used in these co-IP experiments. MDCK cells, nontransfected and stably cotransfected with hOST-α-6xHis and GFP, were used as controls. Cell extracts from stably transfected MDCK cells were immunoprecipitated by a poly-His antibody. Coprecipitated proteins were isolated by SDS-PAGE and analyzed by Western blot using the GFP antibody to detect GFP-tagged hOST-β protein. Figure 3B shows that co-IP of stably transfected MDCK cell lysates with the poly-His antibody for hOSTα-6xHis immunoprecipitated the GFP-fused hOST-β protein (lane α-His + β-GFP) but not the GFP protein (lane α-His + GFP). Total protein from stably transfected cells incubated with the poly-His antibody resulted in the precipitation of more than one band detected by the GFP antibody (Fig. 3B). The principal band corresponded to the hOST-β-GFP isoform based on the calculated molecular mass. Multiple forms of Ost-β were also seen in a previous study (3) of rat Ost-β. The observation of the other bands suggests the presence of other isoforms or variants different from hOST-β. It is notable that an additional band of GFP was observed in cells that expressed hOST-β-GFP (Fig. 3B, lane α-His + β-GFP). This finding has been seen in cells transfected with GFP-fused Asbt and Na+-bile acid cotransporter protein (Ntcp) (unpublished data) for reasons that are unclear. Immunofluorescence microscopy (Fig. 4A) demonstrated that the hOST-α and -β proteins were colocalized on the basolateral domain of stably transfected MDCK cells. The semiquantitative image analysis of protein colocalization suggested that >80% of hOST-α and -β proteins were colocalized on the basolateral plasma membrane of stably transfected MDCK cells.

Fig. 4.

Confocal images of MDCK cells cotransfected with wild-type or mutated hOST-α and -β cDNA. Stably transfected MDCK cells were grown on glass coverslips and fixed with cold 100% methanol. In stably transfected MDCK cells, GFP-fused wild-type hOST-α (α-GFP; A) and mutants [28 amino acids of COOH-end fragment-deleted hOST-α (α-CT; B) and 50 amino acids of NH2-end fragment-deleted hOST-α (α-NT; C)] were coexpressed with 6xHis-tagged hOST-β (β-His), which was detected by rabbit anti-6xHis antibodies and visualized by Texas red-labeled anti-rabbit IgG (red). Right, enlarged images from the merged (overlap) picture; arrows indicate the location of the enlarged image in the overlap picture. The yellow color indicates the colocalization of hOST-α and -β proteins.

Identification of protein interaction domains of hOST-α and -β.

By prediction analysis of protein structure, several potential protein interaction motifs for membrane targeting (such as di-leu and RRK membrane sorting signal motifs) were identified in the amino-terminus (NH2 end) and carboxyl-terminus (COOH end) tails of hOST-α protein. To further define the possible protein interaction domain(s) of hOST-α, the 28 amino acids of the COOH-end cytoplasmic tail (α-CT) or the 50 amino acids of the NH2-end extracellular fragment (α-NT) of hOST-α were deleted. Protein interactions of truncated hOST-α (α-NT or α-CT) with wild-type hOST-β were then tested in the M2H assay and by co-IP. The plasmids, as shown in Fig. 5, were cotransfected into COS-7 cells. After an incubation, the firefly luciferase activity was determined in cell lysates. The results showed that an approximately twofold increase of luciferase activity was observed in cells cotransfected by hOST-β (pACT-β) with pBIND-α-CT cDNAs compared with cells co-transfected with pBIND vector (Fig. 5A). The lower luciferase activity of the α-CT mutant compared with wild-type hOST-α and -β (a decrease of ∼35%) suggests that the protein binding affinity may be decreased by the COOH-terminal truncation of the hOST α-subunit. In contrast, luciferase activity in cells cotransfected by pBIND-α-NT with hOST-β (pACT-β) cDNA showed no significant difference compared with cells cotransfected with pBIND vector (Fig. 5A). For co-IP experiments, truncated constructs (α-NT or α-CT) were tagged with GFP, and GFP-tagged α-NT or α-CT were used to immunoprecipitate 6xHis-tagged wild-type hOST-β protein from stably cotransfected MDCK cells. The results demonstrated that the 6xHis-tagged hOST-β protein was only coimmunoprecipitated by α-CT but not by α-NT (Fig. 5B). These finding are consistent with the M2H study and confirm that the NH2-terminal extracellular tail deleted form of hOST-α does not interact with hOST-β protein. Na+-independent TC efflux in MDCK cells stably transfected with rAsbt, OST α-CT, and OST-β demonstrated that the transport activity was increased by about threefold in cells cotransfected with α-CT and hOST-β cDNA compared with nontransfected cells (Fig. 6A). In contrast, there was only background efflux transport activity in cells co-transfected with α-NT and hOST-β cDNA (Fig. 6A). The results shown in Fig. 6, B and C, demonstrate that these MDCK cells stably transfected with rAsbt, truncated hOST-α, and hOST-β cDNA had similar Na+-dependent TC uptake activity by rAsbt and similar gene expression levels of these stably transfected constructs. The interaction of truncated hOST-α with hOST-β protein was further demonstrated by confocal microscopy (Figs. 4, B and C, and 6D). The confocal images show that in cotransfected cells, the majority of COOH-end truncated hOST-α (α-CT-GFP) and hOST-β proteins were expressed on the plasma membrane of nonpolarized HEK-293 cells (Fig. 6D) and on the basolateral domain of stably transfected MDCK cells (Fig. 4B). However, the semiquantitative image colocalization analysis showed that the percentage of membrane colocalization of truncated hOST-α (α-CT-GFP) and hOST-β proteins was decreased (∼60% overlap) compared with wild-type hOST-α and -β proteins (∼80% overlap). In contrast, NH2-end truncated hOST-α (α-NT-GFP) and hOST-β proteins showed an intracellular distribution in both nonpolarized HEK-293 and polarized MDCK cells (Figs. 4C and 6D). In stably transfected MDCK cells, ∼30% of NH2-tail truncated hOST-α (α-NT-GFP) and hOST-β proteins overlapped in the endoplasmic reticulum and Golgi region. These results indicate that the COOH-end tail of hOST-α is not essential for interactions with the hOST-β protein for transport function and polarized cellular distribution. With regard to the NH2-end tail of hOST-α, the results are consistent with several possible interpretations. It is most likely that the NH2-end tail of hOST-α interacts directly with hOST-β and is an important determinant of cellular distribution and transport function. However, deletion of the NH2-end tail could affect hOST-α biogenesis and folding, thereby causing global effects on its structure and function.

Fig. 5.

Identification of the hOST-α and -β protein interacting domain. A: examination of the interaction between truncated hOST-α and wild-type hOST-β by the M2H luciferase assay. A fragment encoding truncated hOST-α (α-CT or α-NT) was linked to a sequence encoding Gal4DBD in the vector pBIND. A fragment encoding the full open reading frame of hOST-β was linked to a sequence encoding VP16AD in the pACT vector. The plasmids, as indicated, were transfected into COS-7 cells. After cells had been incubation, firefly luciferase activity was determined in cell lysates. All experiments were performed at least twice with triplicate samples and are depicted as means ± SE. B: co-IP of truncated hOST-α and wild-type hOST-β proteins in stably cotransfected MDCK cells. GFP-tagged COOH-terminal truncated hOST-α (α-CT-GFP) or NH2-terminal truncated hOST-α (α-NT-GFP) was cotransfected with 6xHis-tagged wild-type hOST-β protein (β-6His) into MDCK cells, respectively. Total protein extracted from stably transfected MDCK cells was immunoprecipitated by the polyclonal GFP (FL) antibody. Total protein (∼50 μg) extracted from stably transfected MDCK cells (cell) and co-IP proteins (∼1 μg) were isolated by SDS-PAGE and analyzed by WB using a poly-His antibody to detect 6xHis-tagged wild-type hOST-β protein. The results demonstrated that the 6xHis-tagged hOST-β protein was only coimmunoprecipitated by COOH-terminal truncated hOST-α-GFP (α-CT) but not by NH2-terminal truncated hOST-α-GFP (α-NT).

Fig. 6.

Effects of truncation of hOST-α protein on TC efflux and cellular distribution. A and B: Na+-independent [3H]TC efflux (A) and Na+-dependent [3H]TC uptake (B) were measured in stably transfected MDCK cells that had been transfected with rAsbt, truncated hOST-α (α-CT or α-NT), and wild-type hOST-β cDNA. Data represent means ± SE (bars) of triplicate determinations from at least 3 different cell culture preparations. C: RT-PCR demonstrated similar gene expression levels of rAsbt, truncated hOST-α, and wild-type hOST-β constructs in stably transfected MDCK cells. D: HEK-293 cells were cotransfected with GFP-fused truncated hOST-α (α-CT-GFP or α-NT-GFP) and wild-type hOST-β cDNA. The membrane expression of GFP-fused truncated α-CT or α-NT was detected by confocal fluorescence microscopy, and photomicrographs of GFP-fused hOST mutants were obtained on a confluent monolayer of transfected HEK-293 cells cultured on glass coverslips.

DISCUSSION

Recent studies (2, 8) have indicated that OST-α/β is a major bile acid and steroid transporter located on the basolateral membrane of enterocytes, renal tubular cells, and cholangiocytes. OST-mediated transport requires the coexpression of two distinct gene products: a 340-amino acid, 7-potential TM domain protein (OST-α) and a 128-amino acid, single TM domain polypeptide (OST-β). OST transports TC and other sterols, including E3S and dehydroepiandrosterone-3-sulfate as well as digoxin and prostaglandin E2.

In oocytes, OST-α and -β proteins can target to the plasma membrane independently and together (17). However, in transfected MDCK cells, coexpression of mouse Ost-α and -β proteins was required for plasma membrane trafficking and transport function (5). Colocalization of Ost-α and -β proteins on the basolateral membrane has been convincingly demonstrated in previous studies, but whether these two proteins interact directly and the mechanism(s) of this interaction are unknown. Two dileucine motifs in the NH2-terminus and one RRK/RKR motif in the COOH-terminus of the OST-α protein are conserved in the human, rat, and mouse but not in the skate (2, 17). The dileucine motif may be important for assembly and trafficking of channel proteins to the plasma membrane (11). The RRK/RKR motif may also be involved in protein trafficking (2, 7). However, the molecular mechanisms through which OST-α and -β proteins interact to ensure basolateral localization and transport activity remain unclear.

Our results demonstrate, for the first time, that hOST-α and -β proteins are physically associated and that the predicted cytosolic COOH-terminal domain of hOST-α is not essential for this interaction with hOST-β. After deletion of the extracellular NH2-terminus of hOST-α, mutant hOST-α failed to interact with hOST-β, traffic to the plasma membrane, or exhibit transport activity in transfected mammalian cells. Thus, this protein interaction was required for hOST-α and hOST-β polarized basolateral membrane localization and transport activity in transiently transfected HEK-293 and COS-7 cells and stably transfected MDCK cells. Moreover, the truncation of the COOH-end 28-amino acid tail of hOST-α also resulted in a significant decrease in transport activity and membrane colocalization with hOST-β protein compared with wild-type hOST-α. These data suggest that the cytoplasmic tail of hOST-α may functionally serve as a support domain to optimize the association of the complex. Based on these data, it is more likely the hOST-α/β protein complex interacts first intracellularly and then moves to the plasma membrane. If this protein-protein interaction is interrupted, then hOST, especially the hOST-α protein, will not be properly targeted to the plasma membrane.

Posttranslational regulation of hOST may occur through interactions with multiple accessory proteins necessary for membrane trafficking, localization, and functional expression. Little is known about the proteins that bind and regulate the membrane trafficking of hOST. Our data demonstrated that removal of the NH2 end of hOST-α protein abolished its interaction with hOST-β. However, confocal microscopy demonstrated that the NH2-end truncated hOST-α protein was still largely colocalized with the hOST-β protein intracellularly, but may do so through a mechanism other than protein-protein interaction, such as residence in a common vesicular compartment. The hOST complex may be similar to a Na+ channel in that the seven-TM hOST-α may function as a core transport unit and the one-TM hOST-β protein may serve as a supporting unit to modulate OST transport function and polarized basolateral plasma membrane localization (1, 10). Whatever the case, stable protein-protein interaction, and not just colocalization within the cell, appears to be essential to these processes. It is notable that in the yeast two-hybrid system, a protein interaction of hOST-α and hOST-β was not observed (data not shown). That may be due to incorrect protein folding in yeast or the lack of posttranslational modifications or external cellular factor(s) that are not present in yeast (21). Whether other regions of OST-α and OST-β and/or cellular factors are also involved in subunit modulation is unclear. Further studies are required to clarify these questions.

In summary, in this study, we proved direct experimental evidence to demonstrate the association of hOST-α and hOST-β proteins. The NH2-end extracellular domains of hOST-α may play a critical role in complex assembly and trafficking to the plasma membrane. The COOH-terminal predicted cytosolic domain of hOST-α is not essential for complex assembly and trafficking to the plasma membrane. The COOH-terminus cytoplasmic tail of hOST-α may serve as a supporting domain to optimize the functional transporter complex formation.

GRANTS

This work was supported in part by the National Institutes of Health (NIH) Grant 5-R37-HD-020632-21 (to F. J. Suchy). Confocal laser scanning microscopy was performed at the Mount Sinai School of Medicine-Confocal Laser Scanning Microscopy core facility, which was supported with funding from NIH Shared Resources Grant 5-R24-CA-095823-04, National Science Foundation Major Research Instrumentation Grant DBI-9724504, and NIH Shared Instrumentation Grant 1-S10-RR-09145-01.

Acknowledgments

We thank Mohammad Shahid for assistance with the initiation of this work.

Present address of H. Liu: Dept. of Medicine, Saint Barnabas Medical Center, Livingston, NJ 07039.

Footnotes

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

REFERENCES

View Abstract