Human Placental Na+-dependent Multivitamin Transporter

We have cloned the human Na+-dependent multivitamin transporter (SMVT), which transports the water-soluble vitamins pantothenate, biotin, and lipoate, from a placental choriocarcinoma cell line (JAR). The cDNA codes for a protein of 635 amino acids with 12 transmembrane domains and 4 putative sites for N-linked glycosylation. The human SMVT exhibits a high degree of homology (84% identity and 89% similarity) to the rat counterpart. When expressed in HRPE cells, the cDNA-induced transport process is obligatorily dependent on Na+ and accepts pantothenate, biotin, and lipoate as substrates. The relationship between the cDNA-specific uptake rate of pantothenate or biotin and Na+ concentration is sigmoidal with a Na+:vitamin stoichiometry of 2:1. The human SMVT, when expressed in Xenopus laevis oocytes, induces inward currents in the presence of pantothenate, biotin, and lipoate in a Na+-, concentration-, and potential-dependent manner. We also report here on the structural organization and chromosomal localization of the human SMVT gene. TheSMVT gene is ∼14 kilobase pairs in length and consists of 17 exons. The SMVT gene is located on chromosome 2p23 as evidenced by somatic cell hybrid analysis and fluorescence in situ hybridization.

Human placenta expresses a single transport system that transports the micronutrients pantothenate, biotin, and lipoate (1)(2)(3). Studies using human placental brush border membrane vesicles (1,2) and human placental choriocarcinoma cells (JAR) (3) have demonstrated that the stoichiometry of Na ϩ :substrate for this multivitamin transporter is 2:1, indicating that for every vitamin molecule entering the cell, two Na ϩ ions are cotransported into the cell. Since the substrates pantothenate, biotin, and lipoate carry one negative charge at physiological pH, the transport process is electrogenic. Thus the transport system is energized by both the Na ϩ gradient as well as the potential difference that exist across the cell membrane. Sodiumdependent pantothenate and biotin transport systems have also been individually characterized in brush border membrane vesicles isolated from the intestine and kidney of rat, rabbit, and human (4 -13). Although it is unequivocal that the transport system is Na ϩ -dependent, the electrogenicity of the transport system is highly controversial and much debated. It is also not known at this time whether a single transporter, like in placenta, is responsible for the transport of both biotin and pantothenate in the intestine and kidney. Said et al. (6 -8), characterizing biotin transport in intestinal brush border membrane vesicles of rat, rabbit, and humans, have shown that the transport process is electroneutral with a Na ϩ :substrate stoichiometry of 1:1. An electroneutral transport system for biotin has also been reported from rat and human renal brush border membrane vesicles (9,10). On the other hand, the transport systems for pantothenate and biotin in rabbit renal brush border membrane vesicles have been reported to be stimulated by an inside-negative membrane potential indicating that the transport systems are electrogenic (11)(12)(13).
Recently, a cDNA clone encoding the rat sodium-dependent multivitamin transporter (rSMVT) 1 has been isolated from a rat placental cDNA library in our laboratory (14). The cloned cDNA, when expressed in HeLa cells, induces Na ϩ -dependent pantothenate and biotin transport activities. The cDNA-specific uptake of both pantothenate and biotin is inhibitable by lipoate. The transporter has a Michaelis-Menten constant (K t ) of ϳ5 M for pantothenate and ϳ15 M for biotin. The nucleotide sequence of the cloned transporter predicts a protein of 68.6 kDa with 634 amino acids and 12 potential transmembrane domains. Comparison of the protein sequence of SMVT with other proteins entered in the Swiss-Prot data base shows significant sequence homology with known members of the Na ϩ -dependent glucose transporter family.
In this study, we describe the molecular cloning and structural and functional characterization of the human sodium-dependent multivitamin transporter from a placental choriocarcinoma cell line (JAR). This transporter is homologous to the transporter cloned from the rat placenta. The cloned transporter has been characterized by expressing the transporter in human retinal pigmental epithelial (HRPE) cells and in Xenopus laevis oocytes and by analyzing its transport function by tracer uptake and electrophysiological methods. Northern blot analysis shows that the transcript of hSMVT of 3.2 kbp in size is widely expressed in human tissues. The exon-intron organization and chromosomal localization of the human SMVT gene are also presented. fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin as described before (15). The HRPE cell line number 165 originally provided by M. A. Del Monte (W. K. Kellogg Eye Center, Department of Ophthalmology, Ann Arbor, MI) was routinely maintained in Dulbecco's modified Eagle's medium/F-12 medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin as described before (16,17). Frogs (Xenopus sp.) were purchased from Nasco (Fort Atkinson, WI), and MEGAscript cRNA synthesis kit was obtained from Ambion (Austin, TX). SuperScript System for cDNA cloning, TRIzol reagent, oligo(dT) cellulose, and Lipofectin were purchased from Life Technologies, Inc. Restriction enzymes were purchased from New England Biolabs (Beverly, MA). Magna nylon transfer membranes were purchased from Micron Separations, Inc. (Westborough, MA). D-[ 14 C]Pantothenate (51.5 mCi/mmol) and [ 3 H]biotin (58.2 Ci/mmol) were procured from NEN Life Science Products. All of the unlabeled vitamins were purchased from Sigma.

Materials-The
Screening of the JAR Cell cDNA Library-The BlpI/StuI fragment of the rat placental SMVT cDNA was used as the probe. This fragment was 1.45 kbp long and included 75 bp 5Ј to the translational start site and most of the coding region. The probe was labeled with [␣-32 P]dCTP by random priming using the Ready-to-go oligolabeling beads. A JAR cell cDNA library was screened as described earlier (18,19). Positive clones were identified and the colonies purified by secondary screening.
DNA Sequencing-Taq DyeDeoxy terminator cycle sequencing was performed with an automated Perkin-Elmer Applied Biosystems 377 Prism DNA Sequencer to sequence both the sense and antisense strands of the cDNA. The sequence was analyzed using the GCG sequence analysis software package GCG version 7.B (Genetics Computer Group, Inc., Madison, WI).
Functional Expression in HRPE Cells-This was done using the vaccinia virus expression system (20) using HRPE cells as described before (14,21). Subconfluent HRPE cells grown on 24-well plates were first infected with a recombinant (VTF 7-3 ) vaccinia virus encoding T7 RNA polymerase and then transfected with the plasmid carrying the full-length cDNA. After 10 -12 h post-transfection, uptake measurements were made at room temperature with [ 14 C]pantothenate and [ 3 H]biotin. The uptake medium was 25 mM Hepes/Tris, pH 7.5, containing 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl 2 , 0.8 mM MgSO 4 , and 5 mM glucose. In most experiments, the time of incubation was 30 min. Endogenous transport was always determined in parallel using cells transfected with empty vector. This transport accounted for 15-25% of the transport measured in cells that were transfected with the vector carrying the cDNA insert. Therefore, the transport values measured in cells transfected with empty vector were always subtracted from the corresponding transport values measured in cells transfected with vector-cDNA to obtain the cDNA-specific uptake of the vitamins.
Functional Expression in X. laevis Oocytes-cRNA from the cloned hSMVT-cDNA was synthesized using the MEGAscript kit (Ambion) according to manufacturer's protocol. The cDNA was linearized using NotI, and the cDNA insert was transcribed in vitro using T7 RNA polymerase in the presence of an RNA cap analog. The resultant cRNA was purified by multiple extractions with phenol/chloroform and precipitated with ethanol.
Mature oocytes from X. laevis were isolated by treatment with collagenase A (1.6 mg/ml), manually defolliculated, and maintained at 18°C in modified Barth's medium supplemented with 10 mg/liter gentamycin (22). On the following day, oocytes were injected with 50 ng of cRNA. Oocytes injected with water served as control. The oocytes were used for electrophysiological studies 6 days after cRNA injection. Electrophysiological studies were done by the conventional two-microelectrode voltage clamp method (23)(24)(25). Oocytes were superfused with a NaCl-containing buffer (100 mM NaCl, 2 mM KCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 3 mM Hepes, 3 mM Mes, and 3 mM Tris, pH 7.5) followed by the same buffer containing varying amounts of either pantothenate, biotin, or lipoate. The membrane potential was held steady at Ϫ50 mV. For studies involving the current-voltage (I-V) relationship, step changes in membrane potential were applied, each for a duration of 100 ms in 20-mV increments. Kinetic parameters for the saturable transport of the three substrates were calculated using the Michaelis-Menten equation. Data were analyzed by nonlinear regression.
Northern Analysis-Tissue distribution of the hSMVT-specific transcripts was determined by Northern blot. A commercially available membrane blot containing size-fractionated poly(A) ϩ mRNA from several tissues of human origin (CLONTECH, Palo Alto, CA) was used for this purpose. The full-length hSMVT cDNA labeled with [␣-32 P]dCTP was used as the probe. Hybridization was performed at 42°C in 6ϫ SSPE, 50% formamide, 10ϫ Denhardt's solution, 2% SDS, and 100 g/ml salmon sperm DNA. Post-hybridization washing was done at high stringency conditions that involved a final wash at 60°C for 30 min in 0.5ϫ SSPE in 0.5% SDS.
Isolation and Characterization of Human SMVT Genomic Clone-This was done by screening a commercially available human placental genomic library (Lambda FIX II, Stratagene, La Jolla, CA) with a SalI/BglII fragment (ϳ0.65 kbp), comprising the 5Ј end of the human SMVT cDNA, as the probe. Positive clones were isolated and plaquepurified by secondary and tertiary screening. The size of the genomic insert was determined by digestion with SalI and size-fractionation of the released insert on agarose gel. Digestion of the genomic clone with XbaI released 3 fragments (1.1, 5.0, and ϳ9.0 kbp) of the genomic DNA insert. The 1.1-and 5-kbp fragments were subcloned into pSPORT. The complete nucleotide sequence of the cloned SMVT gene was obtained by sequencing the subcloned fragments and the phage DNA.
Chromosomal Localization-The chromosomal localization of the human SMVT gene was done by somatic cell hybrid analysis, and the regional assignment of the gene on the chromosome was accomplished by FISH. In somatic cell hybrid analysis, the mapping panel number 2 consisting of mouse-human and Chinese hamster-human somatic cell hybrids, obtained from the Human Genetic Cell Repository (NIGMS, National Institutes of Health, Camden, NJ) was used. Each hybrid cell line contained a single human chromosome. DNA was isolated from the hybrid cells and digested with PstI. The restriction fragments were size-fractionated, transferred to nylon filters, and probed with fulllength hSMVT cDNA as described previously (26). The cDNA was labeled with [ 32 P]dCTP by random priming.
For use in FISH, the entire phage DNA of the hSMVT genomic clone was labeled by nick translation with biotin-11-dUTP and used for hybridization to human metaphase chromosome spreads at a concentration of 50 ng/l. FISH was carried out essentially as detailed earlier (27). The chromosomes were stained by the chromycin A 3 /distamycin A/4,6-diamidino-2-phenylindole technique for fluorescent microscopic analysis.

RESULTS AND DISCUSSION
Isolation of Human Placental SMVT cDNA from a JAR Cell cDNA Library-Earlier studies using placental membrane vesicles have demonstrated the presence of SMVT in the brush border membrane of the human placenta (1, 2). We characterized the uptake of pantothenate and biotin in the choriocarcinoma cell line JAR, and we showed that these cells express a transporter identical to the transporter that is expressed in the placental brush border membrane (3). Unlike RNA isolated from the term placenta that is contaminated with RNA from other tissues like maternal decidua and non-trophoblast cells of the placenta, RNA highly specific to the trophoblast can be isolated from JAR cells in culture. With this in mind, we constructed and screened a JAR cell cDNA library to clone the human SMVT using the recently cloned rSMVT as the probe. Upon screening of approximately 3.6 ϫ 10 5 colonies, several positive colonies were obtained, one of which was arbitrarily chosen for further characterization. The cDNA was designated hSMVT for human sodium-dependent multivitamin transporter.
Structural Features of hSMVT cDNA-The hSMVT cDNA is 3162 bp long with an open reading frame of 1908 bp (including the termination codon), encoding a protein of 635 amino acids. The open reading frame is flanked by a 5Ј-noncoding sequence 391 bp long and a 863-bp long 3Ј-noncoding sequence (Fig. 1). The cDNA contains a poly(A) ϩ tail and a polyadenylation signal (AATAAA) in the 3Ј-noncoding region. The predicted molecular mass and pI of the protein encoded by the open reading frame are 68.7 and 8.6 kDa, respectively. Preceding the putative initiation codon is the Kozak consensus sequence (GAG GAT) (28). Hydrophobicity analysis of the predicted amino acid sequence using the algorithm of Kyte and Doolittle (29) with a window size of 20 amino acid residues indicated that the hSMVT protein contains 12 putative transmembrane domains. When modeled with both the amino terminus and the carboxyl terminus of the protein placed on the cytoplasmic side, there are four potential N-linked glycosylation sites (Asn 138 , Asn 489 , Asn 498 , and Asn 534 ), the first site between transmembrane domains 3 and 4 and the other three sites between transmem-brane domains 11 and 12. The deduced amino acid sequence also displays two sites (Ser 283 and Thr 286 ) with consensus sequence for protein kinase C-dependent phosphorylation (30,31) in the cytoplasmic loop between transmembrane domains 6 and 7. Amino acid sequence comparison of hSMVT against rSMVT shows a high degree of homology (84% identity and 89% similarity) between the two proteins indicating that the protein is highly conserved across the species. Blast search (32) against known sequences in the Swiss-Prot data base shows considerable homology with members of the sodium-dependent glucose transporter family confirming that the protein is a member of this family. Among the sodium-dependent glucose transporter family members, the closest relative to SMVT with respect to amino acid sequence similarity is the sodium-dependent iodide transporter (40% identity and 64% similarity) (33).
Functional Characterization of hSMVT following Expression in HRPE Cells-The cDNA library used in this study was constructed in pSPORT vector with cDNA inserts unidirectionally ligated in such a way that the sense transcription of the cDNA is under the control of the T7 promoter in the plasmid. This made it possible to employ the transient vaccinia virus expression system for functional characterization of the clone. Initial experiments involving the time course of radiolabeled pantothenate uptake into vector-and cDNA-transfected cells indicated that the uptake of pantothenate was higher in cDNA-transfected cells than in vector-transfected cells. The uptake was also found to be linear up to 60 min, and hence all subsequent experiments were done with a 30-min incubation (data not shown).
The ionic dependence of the cDNA-stimulated pantothenate uptake was investigated by measuring pantothenate transport in empty vector-and cDNA-transfected cells in the presence of various inorganic salts (Table I). Control uptake was measured in the presence of NaCl. Replacement of Na ϩ with other cations such as Li ϩ , K ϩ , choline, or N-methyl-D-glucamine almost completely abolished the pantothenate uptake in both vector-and cDNA-transfected cells indicating that Na ϩ is obligatory for the transport function. Such a marked inhibition was not seen when the Cl Ϫ in the buffer was replaced with gluconate suggesting the noninvolvement of Cl Ϫ in the transport process. Replacing Cl Ϫ with I Ϫ on the other hand reduced the uptake of the radiolabeled pantothenate by nearly 50%. It is not known at this time whether this inhibition is due to the change in the potential difference across the plasma membrane brought about by the prolonged incubation with NaI or if it is due to I Ϫ competing with pantothenate for the transport process. The latter hypothesis cannot be ruled out since, as mentioned earlier, hSMVT has very high sequence homology with the Na ϩ /I Ϫ cotransporter.
The substrate specificity of the transporter was evaluated by assessing the ability of various unlabeled vitamins and vitamin analogs to inhibit the transport of labeled pantothenate in pSPORT-transfected and pSPORT-cDNA-transfected cells (Table II). The cDNA-specific uptake was determined in these experiments, by subtracting the transport measured in the corresponding empty vector-transfected cells. Pantothenate, biotin, and oxidized lipoate were able to inhibit the uptake of radiolabeled pantothenate almost completely. Desthiobiotin, a biotin analog, also inhibited the uptake significantly indicating that the tetrahydrothiophene ring of biotin is not recognized by the transporter. Iminobiotin and diaminobiotin, two other biotin analogs, both of which lack the keto group at the second position of the imidazole ring of biotin, did not inhibit pantothenate uptake indicating that this group is involved in the interaction of biotin with the transporter. The two other watersoluble vitamins tested, namely niacinamide and thiamine, also had no influence on cDNA-induced pantothenate uptake. These results are similar to that obtained in pantothenate uptake studies using placental brush border membrane vesicles and JAR choriocarcinoma cells (2,3).
The kinetics of pantothenate uptake was determined by measuring the rate of pantothenate uptake at varying concentrations of pantothenate (0.25-10 M) in HRPE cells transfected with the plasmid vector carrying the cDNA insert. Simultaneous measurements at identical concentrations of pantothenate were also done in empty vector-transfected cells. The values obtained were subtracted from the corresponding values of transport measured in cDNA-transfected cells to obtain cDNA-specific uptake. The data obtained were analyzed first by nonlinear regression analysis ( Fig. 2A) and confirmed by linear regression (Fig. 2A, inset). The kinetics of biotin uptake by the cloned transporter was also determined similarly by measuring the rate of biotin uptake at varying concentrations of biotin (0.5-20 M) (Fig. 2B). As can be seen from the figures, the cDNA-induced transport of both biotin and pantothenate is saturable. The experimental data were found to fit best to a model describing the uptake by a single carrier. The linear regression analysis (Eadie-Hofstee transformation) of the data for the cDNA-specific uptake yielded a linear plot for pantothenate and biotin. The kinetic parameters for the carrier-mediated uptake, K t (Michaelis-Menten constant) and V max (maximal velocity), were 1.5 Ϯ 0.2 M and 196.0 Ϯ 12.9 pmol/ 10 6 cells/30 min, respectively, for pantothenate and 3.2 Ϯ 0.7 M and 157.6 Ϯ 18.2 pmol/10 6 cells/30 min, respectively, for biotin. The K t values of the cloned transporter for the transport of pantothenate and biotin are comparable to the values reported using placental brush border membranes (1, 2) and JAR cells (3).
The three physiological substrates of SMVT are pantothenate, biotin, and lipoate. All three substrates possess a long side chain containing a carboxylate group. In addition, biotin has an imidazole-and thiazole-fused ring structure. Threedimensional structural analysis of the pantothenate molecule indicates that it too exists in a pseudo-ring conformation with a hydrogen bond between the amino group of the ␤-alanine moiety and the hydroxyl group in the pantoic acid moiety. Lipoate can exist in two states. In its oxidized state, it forms a ring structure by forming a disulfide linkage. In the reduced state with two sulfhydryl groups, lipoate does not have a ring structure. We investigated whether hSMVT can differentiate between these two forms of lipoate. We measured the uptake rate of labeled pantothenate in vector-and cDNA-transfected cells in the presence of varying concentrations of oxidized and reduced lipoate. When reduced lipoate was used, dithiothreitol at a final concentration of 1 mM was used in the buffer to prevent the lipoate from getting oxidized. Dithiothreitol by itself at a concentration of 1 mM had no effect on the uptake of the vitamin (data not shown). The cDNA-specific transport was determined, and the data were analyzed by nonlinear regression analysis (Fig. 3). Both oxidized and reduced lipoate inhibited the uptake of pantothenate; however, the latter was significantly less potent than the former. The IC 50 values (i.e. the concentration necessary to cause 50% inhibition) for oxidized lipoate was 2.7 Ϯ 0.2 M and that for reduced lipoate 7.3 Ϯ 0.8 M. Since the concentration of radiolabeled pantothenate used in the assay was 1 M and the K t of hSMVT for pantothenate is 1.5 M, the K I value for oxidized lipoate (calculated using the equation K I ϭ IC 50 /( 1 ϩ [L]/K t ) where L is the concentration of pantothenate in the assay) is 1.6 Ϯ 0.2 M and that for reduced lipoate is 4.4 Ϯ 0.5 M. Thus, the affinity of hSMVT for oxidized lipoate which has the ring conformation is approximately 3-fold higher than that for the reduced lipoate.
The effect of Na ϩ on the kinetics of pantothenate and biotin uptake was investigated by measuring the uptake of radiolabeled pantothenate and biotin in HRPE cells transfected with hSMVT cDNA in the presence of varying concentrations of extracellular Na ϩ . The concentration of NaCl in the extracellular medium was varied from 0 to 140 mM. The cDNA-specific uptake was determined by subtracting the uptake measured  simultaneously in cells transfected with the pSPORT vector. The relationship between the uptake rate and the Na ϩ concentration was sigmoidal (Fig. 4), suggesting the involvement of more than one Na ϩ per pantothenate molecule transported. The data were fit to the Hill equation, and the Hill coefficient, which is the number of Na ϩ ions interacting with the carrier, was calculated. The value was 1.9 Ϯ 0.3 for the uptake of pantothenate (Fig. 4A) and 1.8 Ϯ 0.1 for the uptake of biotin (Fig. 4B). These values were confirmed from the slope of the Hill plots (Fig. 4, insets). This indicates that, for every pantothenate/biotin molecule transported, 2 Na ϩ ions are cotransported. Since both pantothenate and biotin are monovalent anions at physiological pH, the transport process is electrogenic. Thus, both the Na ϩ gradient as well as the difference in the membrane potential across the cell membrane energize the transport process. This greatly increases the concentrative capacity of the transport process. Functional Characterization of hSMVT following Expression in X. laevis Oocytes-A more direct way of characterizing a rheogenic transport system is by electrophysiological means using the two-microelectrode, voltage clamp technique. Since the transport of a substrate molecule across the cell membrane mediated by SMVT is associated with a net transfer of positive charge into cell, the SMVT function should be associated with inward currents under voltage clamp conditions in this experimental approach. The results of the electrophysiological experiments done with SMVT-expressing oocytes are given in Fig. 5. Perifusion of oocytes with buffer containing 10 M either pantothenate, biotin, or lipoate in the presence of NaCl induced inward currents (ϳ25-35 nA) (Fig. 5A). The inward currents were absent when NaCl in the buffer was replaced with Nmethyl-D-glucamine chloride and also in control oocytes injected with water, showing that the substrate-induced currents are dependent on the presence of Na ϩ in the uptake medium and SMVT expression (data not shown). The substrate-induced inward currents were dependent on the testing membrane potential (Fig. 5B). The currents increased in magnitude as the testing membrane potential became more hyperpolarized, demonstrating that the transport activity of SMVT, measured by the magnitude of the substrate-induced inward currents, is stimulated when the membrane potential is made more negative. These data demonstrate unequivocally that SMVT is a potential-sensitive transporter. In addition, since only transportable substrates can induce the inward currents in SMVTexpressing oocytes, the above data also demonstrate for the first time that lipoate is indeed a transportable substrate. Fig. 6 details the kinetics of hSMVT, characterized electrophysiologically, using pantothenate as the substrate. The SMVT-induced currents were dependent on the testing membrane potential as well as on the concentration of pantothenate (Fig. 6A). The current increased in magnitude with increasing concentrations of pantothenate and increasing hyperpolarization of the testing membrane potential. The substrate-induced currents showed saturation kinetics with respect to pantothenate concentration at different testing membrane potentials (Fig. 6B). The kinetic constants K t and I max for the transport of pantothenate at different testing membrane potentials were calculated by nonlinear regression analysis. The kinetic constant K t , a parameter indicative of the affinity of SMVT for the substrate, decreased as the testing membrane potential became more hyperpolarized (Fig. 6C). At a testing membrane potential of Ϫ50 mV, the K t was 1.9 Ϯ 0.4 M, whereas the K t decreased to 1.0 Ϯ 0.2 M when the testing membrane potential was changed to Ϫ150 mV. This suggests that the affinity of SMVT for pantothenate increases at hyperpolarizing membrane potentials. The kinetic constant I max , a parameter indicative of the maximal velocity of the SMVT-mediated transport activity, increased as the testing membrane potential became more hyperpolarized (Fig. 6D). The value for I max increased from 38.7 Ϯ 2.8 nA at Ϫ50 mV to 290.4 Ϯ 14.8 nA at Ϫ150 mV. This shows that SMVT is activated by an inside-negative membrane potential.
Identical experiments were also carried out using biotin and oxidized lipoate as the substrates, and similar results were obtained (data not shown). The affinity constant K t , calculated for biotin and oxidized lipoate, when the transport measure-ments were made in SMVT-expressing oocytes with membrane potential clamped at Ϫ50 mV, was 19.5 Ϯ 4.6 and 7.3 Ϯ 1.0 M, respectively. The affinity constants measured using the electrophysiological technique following expression in Xenopus oocytes are significantly higher compared with values obtained using expression in mammalian cells. Although the reason for this is not clear, it may most probably be due to the differences in post-translational processing (e.g. N-linked glycosylation) of the transporter in the two expression systems.
Northern Blot Analysis-Tissue distribution of SMVT mRNA in human tissues was studied using a commercially available membrane blot containing size-fractionated poly(A) ϩ mRNA from eight tissues: heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas (Fig. 7). A 3.2-kbp hybridizing signal was detected in all these tissues. Among the tissues tested, the intensity of the hybridizing signal was the highest in the placenta. Moderate signals were seen in the kidney, liver, and pancreas. The detection of the SMVT-specific transcripts in the kidney in the present study lends credence to the expression of Na ϩ -dependent, electrogenic transport system for pantothenate and biotin in the renal tissue. The fact that the transcripts are detected in all the tissues tested hints at the possible involvement of SMVT in the uptake of pantothenate, biotin, and lipoate in all the tissues.
Isolation of hSMVT Genomic Clone-A genomic clone containing the human SMVT gene was isolated by screening 6 ϫ 10 5 phage colonies from an amplified human genomic library. SalI digestion of the DNA isolated from the positive clone released a single insert of ϳ15 kbp. Digestion with XbaI released three fragments of 1.1, 5.0, and ϳ9.0 kbp. Sequential Southern hybridization analysis of the SalI and XbaI restriction-digested fragments of the genomic clone, first with the SalI/BglII fragment (ϳ0.65 kbp) of the cDNA specific to the 5Ј end of hSMVT and then with the HincII/XbaI fragment (ϳ0.8 kbp) of the cDNA, specific to the 3Ј end of hSMVT, was performed. Both probes recognized the ϳ15-kbp SalI fragment, indicating that the isolated genomic clone contained the complete SMVT gene (data not shown). Of the three XbaI fragments, the probe specific to the 5Ј end hybridized to the 5.0and ϳ9.0-kbp fragments, and the probe specific to 3Ј end hybridized only to the ϳ9.0-kbp fragment (data not shown). The hybridization analysis and subsequent sequence analysis of the genomic clone led to the conclusion that the 5.0-kbp and the adjacent ϳ9.0-kbp fragment of the genomic clone contained the complete sequence of the hSMVT cDNA. The 1.1-kbp XbaI fragment was 3Ј to the gene.
Structural Features of hSMVT Gene-The SMVT gene is  (Table III).
Chromosomal Localization-We have mapped the chromosomal location of the gene encoding the cloned placental SMVT. Southern blot hybridization of restriction digests of DNA from Chinese hamster-human and mouse-human hybrid cells with the hSMVT cDNA probe showed six human-specific fragments of 4.2, 2.5, 1.5, 1.35, 1.3, and 1.25 kbp (data not shown). Three fragments of 5.5, 2, and 1.6 kbp were detected in DNA of Chinese hamster and four bands of 4, 2.5, 2.2, and 1.6 kbp were present in mouse DNA. All six human-specific bands showed perfect segregation with human chromosome 2. Regional localization of the SMVT gene was carried out by fluorescent in situ hybridization (FISH). The human SMVT genomic clone was used as the probe. In situ hybridization of biotin-labeled genomic clone to human metaphase chromosome spreads corroborated the initial findings of Southern blot analysis on the localization of the SMVT gene to human chromosome 2 (Fig. 9). Twenty metaphase spreads were scored, and each showed hybridization signal on the short arm of chromosome 2, band p23. Although several inherited metabolic disorders have been mapped to this region of the chromosome, none of them appears to be due to a defect in SMVT function. On the other hand, several cases of biotin-responsive multiple carboxylase deficiency have been reported where a defect in the transport of biotin is suspected (34 -37). Unlike the more prevalent forms of multiple carboxylase deficiency where the activity of either holocarboxylase synthetase or biotinidase enzyme is invariably low, the activities of both the enzymes are normal in these patients. This disorder is characterized with lactic acidosis and organic aciduria, treatable with pharmacological doses of biotin. Neither the identity of the gene responsible for the biotinresponsive multiple carboxylase deficiency nor its chromosomal localization is known. SMVT may most likely be the defective FIG. 6. Characteristics of pantothenate-induced inward currents in hSMVT cRNA injected oocytes. The composition of the uptake buffer was 100 mM NaCl, 2 mM KCl, 1 mM MgCl 2 , 1 mM CaCl 2 , and 3 mM Mes/Hepes/Tris, pH 7.5. A, influence of testing membrane potential (V test ) on pantothenate-induced currents (I) at various concentrations of pantothenate. B, influence of pantothenate concentration on pantothenate-induced currents (I) at various testing membrane potentials. C, influence of testing membrane potential on K t for pantothenate. D, influence of testing membrane potential on I max for pantothenate-induced currents.
FIG. 7. Northern blot analysis of SMVT mRNA transcripts in human tissues. A commercially available blot containing 2 g of poly(A) ϩ RNA from human heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas was used. The poly(A) ϩ RNA in the blot has been size-fractionated on a denaturing formaldehyde-agarose gel and transferred onto nylon membrane. The hybridization and posthybridization washes were done at high stringency conditions. The sizes of the hybridizing RNA were determined using RNA size standards run in parallel in an adjacent lane. gene in these patients. Future studies should be able to establish a role for SMVT in biotin-responsive multiple carboxylase deficiency.
In conclusion, we have isolated a cDNA (hSMVT) from a human placental trophoblast cell line cDNA library that, when expressed in HRPE cells, induces Na ϩ -dependent transport of pantothenate, biotin, and lipoate. The functional characteristics of the induced activity are similar to those of the pantothenate/biotin transport activity described in human placental brush border membrane and in JAR cells. The cDNA has also been expressed in X. laevis oocytes and characterized by the electrophysiological technique. The data presented here not only constitute the first direct demonstration of the electrogenic nature of the transport process but also for the first time provide evidence to show that lipoate is a transportable substrate of this transporter. We have also isolated the gene that codes for hSMVT, sequenced it, and characterized the exonintron organization of the gene. The gene maps to human chromosome 2p23.