Molecular Cloning and Characterization of a Human Multisubstrate Specific Nucleotide-sugar Transporter Homologous to Drosophila fringe connection*

Nucleotide-sugar transporters are crucial components in the synthesis of glycoconjugates. We identified a novel human nucleotide-sugar transporter gene, hfrc1, which is homologous to Drosophila melanogaster fringe connection, Caenorhabditis elegans sqv-7, and human UGTrel7. HFRC1 was localized within the Golgi apparatus following its transient expression in HCT116 cells. In human tissues, hfrc1 and UGTrel7 exhibited similar tissue distributions, although hfrc1 transcripts showed a 10 times greater abundance than those of UGTrel7. The heterologous expression of HFRC1 in the yeast revealed the multisubstrate specific transport activity of HFRC1 (for UDP-N-acetylglucosamine (UDP-GlcNAc), UDP-glucose (UDP-Glc), and GDP-mannose (GDP-Man), with apparent Km values of 8.0, 2.1, and 0.14 μm, respectively). In the mammalian cells, HFRC1 transported UDP-GlcNAc and UDP-Glc, but not GDP-Man. Overexpression of the hfrc1 gene in HCT116 cells modulated the cell surface heparan sulfate expression status. These results suggest that HFRC1 takes part in the synthesis of heparan sulfate by regulating the level of UDP-GlcNAc, a donor substrate for the heparan sulfate synthases.

Nucleotide sugars, high-energy donor substrates for glycosyltransferases, are synthesized in the cytosol (or in the nucleus in the case of CMP-sialic acid). On the other hand, glycosylation reactions occur in the lumen of the endoplasmic reticulum (ER) 1 and Golgi apparatus (1,2). Translocation of nucleotide sugars from the cytosol into the lumen compartment is mediated by specific nucleotide-sugar transporters (NSTs) (1,2). In several studies, NSTs have been mentioned as possible crucial players in the synthesis of glycoconjugates (for reviews, see Refs. [3][4][5]. Recent studies have demonstrated that some NSTs are implicated in growth factor signaling during development through the regulation of proteoglycan synthesis. In Caenorhabditis elegans, sqv-7 mutants show a defect in vulva formation, with epithelial invaginations, during ontogeny (6). SQV-7 has been identified as a transporter of UDP-glucuronic acid (UDP-GlcUA), UDP-galacose (UDP-Gal), and UDP-N-acetylgalactosamine (UDP-GalNAc) (7). Biochemical analysis of sqv-7 mutants has demonstrated defects in both chondroitin and heparan sulfate biosynthesis in vivo (8). Furthermore, Selva et al. (9) and Goto et al. (10) reported that the Drosophila melanogaster gene of fringe connection (frc) is involved in embryonic Wingless/Hedgehog and fibroblast growth factor signaling (9,10). Frc encodes a multisubstrate-specific NST that transfers UDP-sugars into the Golgi apparatus (9,10). UDP-GlcUA and UDP-N-acetylglucosamine (UDP-GlcNAc) act as substrates for the synthesis of heparan sulfate. Embryos with a mutation of frc display severe segment polarity phenotypes (9).
Although sqv-7 and frc are considered to be orthologous to each other, it remains obscure which NSTs of humans have corresponding functions. In humans, the UGTrel7 protein has a similar multisubstrate specificity to those of FRC and SQV-7; namely, UDP-GlcUA/UDP-GalNAc (11). However, localization of UGTrel7 to the ER, not to the Golgi apparatus (11), leaves open the possibility that other transporters may be involved in proteoglycan synthesis in the Golgi apparatus.
The structural conservation present among NSTs enables the identification of many putative NST sequences from data bases. By a data base search, we identified a putative NST gene that is closely related to sqv-7, frc, and UGTrel7. HFRC1 had a multisubstrate specificity for UDP-GlcNAc/UDP-glucose (UDP-Glc)/GDP-mannose (GDP-Man) in the Golgi fraction of yeast cells expressing this gene. Furthermore, alteration in the expression of hfrc1 affected the heparan sulfate on the cell surface of mammalian cells. Thus, the present study raises the possibility that HFRC1 takes part in heparan sulfate synthesis by supplying UDP-GlcNAc. (Tokyo, Japan). Anti-ERp57 (MaP.ERp57) mAb and fluorescein isothiocyanate-conjugated anti-influenza hemagglutinin epitope (HA) mAb (HA-probe, F-7) were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Rhodamine-conjugated anti-mouse IgG mAb and horseradish peroxidase-conjugated anti-mouse IgG mAb were obtained from Bio-Rad. All other reagents were of the highest purity commercially available.

Materials-GDP-[2-
Isolation of a Novel Human NST cDNA and Construction of Expression Plasmids-A novel putative NST gene was cloned using the same procedures as previously described (12). Briefly, a TBLASTN search was performed for the amino acid sequence of the open reading frame (ORF) of the GDP-fucose (GDP-Fuc) transporter gene (13). We identified a cDNA sequence encoding a full-length ORF (GenBank TM accession number XM_047286). To obtain this cDNA and create recombination sites for the GATEWAY TM cloning system (Invitrogen), we used two steps of attB adaptor PCR for the preparation of attB-flanked PCR products. For the first gene-specific amplification, a forward templatespecific primer with attB1 (5Ј-aaaaagcaggcttcccgcaggagatgacgg-3Ј) and a reverse template-specific primer with attB2 (5Ј-agaaagctgggtcgctcttcaaatccaaaca-3Ј) were used. PCR was performed using Platinum® Pfx DNA polymerase (Invitrogen) and a cDNA library derived from human colon tissue. The insertion of a complete attB adaptor and cloning into the pDONR TM 201 vector to create an entry clone for the subsequent subcloning were performed according to the instruction manual. The ORF of the UGTrel7 cDNA was obtained by the same procedures as mentioned above using the forward template-specific primer 5Ј-aaaaagcaggcttcgcagccatggcggaagt-3Ј and the reverse template-specific primer 5Ј-agaaagctgggtccactgctcctttcccctt-3Ј.
Expression vectors were inserted with three copies of the HA epitope tags (YPYDVPDYA) at the position corresponding to the C terminus of the expressing protein and converted to a Gateway destination vector with a conversion site. Each entry clone was subcloned into appropriate expression vectors using the Gateway Cloning System according to the instruction manual.
Transient Transfection and Immunofluorescence Microscopy-Transient transfection and immunofluorescence microscopy was performed as previously described (12). HCT116 cells were subcultured onto a 4-well Lab-Tek chamber slide (Nalge Nunc International, New York) in Dulbecco's modified Eagle's/Ham's medium containing 10% fetal bovine serum. After 24 h, cells were transfected with 0.25 g/well of pCXN2 (14) or pCXN2 inserted with HA-tagged hfrc1 using LipofectAMINE 2000 reagent (Invitrogen) according to the manufacturer's protocol. After 72 h, cells were fixed in phosphate-buffered saline (PBS) containing 4% paraformaldehyde for 30 min at 4°C, then permeabilized in permeabilizing buffer (PBS containing 5% bovine serum albumin and 0.1% Triton X-100) for 1 h at 4°C. Then, cells were stained with anti-␤1,4-galactosyltransferase 1 (␤4GalT1) mAb (15) or anti-ERp57 mAb for 16 h at 4°C, washed four times with permeabilizing buffer, and incubated with rhodamine-labeled anti-IgG mAb for 90 min at room temperature. After incubation, the cells were washed, then stained with fluorescein isothiocyanate-conjugated anti-HA mAb for 90 min at room temperature. Finally, cells were washed and mounted with PermaFluor (Thermo Shandon, Pittsburgh, PA). The fluorescence was observed under a fluorescence microscope.
Subcellular Fractionation of Yeast and Transport Assay-Yeast of the Saccharomyces cerevisiae strain W303-1a (MATa, ade2-1, ura3-1, his3-11,15, trp1-1, leu2-3,112, and can1-100) was transformed by the lithium acetate procedure (16) using a yeast expression vector, YEp352GAP-II (17). Transformed yeast cells were grown at 30°C in synthetic defined medium in which uracil was omitted for the selection of transformants. Subcellular fractionation and nucleotide-sugar transport assays were performed as described by Roy et al. (18). The cells were harvested, washed with ice-cold 10 mM NaN 3 , then converted into spheroplasts by incubation at 37°C for 30 min in spheroplast buffer (1.4 M sorbitol, 50 mM potassium phosphate, pH 7.5, 10 mM NaN 3 , 40 mM 2-mercaptoethanol, and 1 mg of Zymolyase 100T per g of cells). Spheroplasts were pelleted in a refrigerated centrifuge and washed twice with 1.0 M ice-cold sorbitol to remove Zymolyase. The cells were suspended in ice-cold lysis buffer (0.8 M sorbitol in 10 mM triethanolamine, pH 7.2, 5 g/ml pepstatin A, and 1 mM phenylmethylsulfonyl fluoride), then homogenized using a Dounce homogenizer. The lysate was centrifuged at 1,000 ϫ g for 10 min to remove unlysed cells and cell-wall debris. The supernatant was then centrifuged at 10,000 ϫ g for 15 min at 4°C (Beckman XL-90 ultracentrifuge with 90.1 Ti rotor), which yielded a pellet of the P10 membrane fraction. The supernatant was further centrifuged at 100,000 ϫ g yielding a pellet of the P100 membrane fraction and a supernatant of the S100 fraction. Then, 200 g of protein of each fraction were incubated in 100 l of reaction buffer (20 mM Tris-HCl, pH 7.5, 0.25 M sucrose, 5.0 mM MgCl 2 , 1.0 mM MnCl 2 , and 10 mM 2-mercaptoethanol) containing 1 M radiolabeled substrate at 30°C for 5 min. After incubation, the radioactivity incorporated in the microsomes was trapped with a 0.45-m nitrocellulose filter, and measured by liquid scintillation. The amount of incorporated substrate was calculated as the difference from the background value obtained from the time 0 assay for each sample.
Western Blot Analysis-To determine the expression status of the HA-tagged protein, subcellular fractions of yeast cells were analyzed by Western blot analysis. Fifty micrograms of protein of each sample were added to 3ϫ SDS sample buffer (New England Biolabs Inc.), then incubated at room temperature for 2 h. The samples were fractionated on a 5-20% gradient SDS-polyacrylamide gel (Daiichi Pure Chemicals Co., Ltd., Tokyo, Japan). The separated proteins were electrotransferred onto polyvinyldene difluoride membrane. HA-tagged proteins were immunostained with anti-HA mouse mAb and horseradish peroxidase-conjugated anti-mouse IgG mAb. Bound horseradish peroxidase was detected using ECLϩplus (Amersham Biosciences) according to the manufacturer's directions.
Quantitative Analysis of hfrc1 and UGTrel7 Transcripts in Human Tissues by Real Time PCR-The amount of hfrc1 and UGTrel7 transcripts in human tissues was determined by real time PCR. Total RNA was extracted from human tissues by the method of Chomczynski and Sacchi (19). First-strand cDNA was synthesized using a SuperScript II first-strand synthesis kit (Invitrogen) according to the manufacturer's instructions. Real time PCR was performed using a qPCR Mastermix QuickGoldStar (EUROGENTEC) and ABI PRISM 7700 Sequence Detection System (Applied Biosystems). Multiplex real time PCR was used to compare the amounts of hfrc1 and UGTrel7 transcripts in normal human tissues. The sequences of the PCR primer pairs and TaqMan probes used for each gene were as follows. For the quantitation of hfrc1, the forward primer 5Ј-cctgacgacagcagtggtt-3Ј, reverse primer 5Ј-ctccaccgattaatatcccaat-3Ј, and probe 5Ј-agccatcaagaatgtatccgttgcct-3Ј were used. The probe was labeled at the 5Ј-end with the reporter dye FAM, and at the 3Ј-end with the quencher dye TAMRA. For the detection of UGTrel7, the forward primer 5Ј-cattgcgtatttcacaggagatg-3Ј, reverse primer 5Ј-ctgcagaagaaagagggtgtca-3Ј, and probe 5Ј-cccagccttcaaactccacagccttt-3Ј were used. The probe was labeled at the 5Ј-end with Yakima Yellow and at the 3Ј-end with Darquencher. The relative amounts of hfrc1 and UGTrel7 transcripts were normalized with respect to the human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcripts in the same cDNA.
Stable Transfection-HCT116 cells were subcultured onto 6-cm dishes in Dulbecco's modified Eagle's/Ham's (1:1) medium containing 10% fetal bovine serum. After 24 h, the cells were transfected with 8 g of pCXN2 vector or pCXN2 inserted with HA-tagged hfrc1 using Lipo-fectAMINE 2000 reagent according to the manufacturer's protocol. After 48 h, transfectants were selected by the addition of 600 g/ml Geneticin (Invitrogen) to the medium, and cultured for 1 month.
Flow Cytometric Analysis-After 16 h of subculture, cells were harvested with PBS containing 1 mM EDTA, and washed with PBS containing 0.1% bovine serum albumin. Then, 100 l of cell suspensions (1 ϫ 10 6 cells) were incubated with anti-heparan sulfate mAb or antichondroitin 4-sulfate mAb for 30 min on ice, and washed three times with 1.5 ml of PBS containing 1% bovine serum albumin. Then, cells were resuspended in 100 l of fluorescein isothiocyanate-conjugated goat anti-mouse IgM and incubated for 30 min on ice. Flow cytometric analysis was performed using EPICS ELITE flow cytometer (Beckman-Coulter). As the negative control, cells were treated with 10 milliunits/ml of heparitinase I or 100 milliunits/ml of chondroitinase ABC for 90 min at 37°C before immunoblotting. Data were analyzed using WinMDI 2.8 software (The Scripps Research Institute Cytometry software page).

RESULTS
Cloning of Human Hfrc1 cDNA-Using the TBLASTN algorithm, we identified a cDNA sequence (GenBank TM accession number XM_047286) homologous to the GDP-Fuc transporter gene. We cloned the full-length ORF as described under "Experimental Procedures," and named it hfrc1. The phylogenetic tree of the transporter genes in human, Drosophila, and C. elegans is shown in Fig. 1A. Hfrc1 is more closely related to UGTrel7, frc, and sqv-7 than to the GDP-Fuc transporter gene. The phylogenetic tree also demonstrates that both hfrc1 and UGTrel7 are orthologs of Drosophila frc (Fig. 1A). An alignment of the amino acid sequences of HFRC1, SQV-7, FRC, and UGTrel7 is shown in Fig. 1B. HFRC1 consists of 337 amino acids with a calculated mass of 36.7 kDa. Hydropathy analysis and predictions of the transmembrane helices of the amino acid sequence revealed that HFRC1 are Type III transmembrane proteins with seven transmembrane domains. HFRC1 showed 54.3, 46.6, and 46.6% identities to UGTrel7, FRC, and SQV-7, respectively. There is one potential N-glycosylation site in the HFRC1 sequence (double underlined). The gene structures of hfrc1 and UGTrel7 are shown in Fig. 1C. The hfrc1 gene is mapped on human chromosome 9q22.33, and the mRNA is composed of 12 exons. On the other hand, the UGTrel7 gene is mapped on human chromosome 1p32-p31, and the mRNA is composed of 12 exons.
Subcellular Localization of HFRC1 in Mammalian Cells-Hfrc1 has a significant similarity to human UGTrel7 and Drosophila frc. FRC is primarily localized to the Golgi apparatus (10), whereas UGTrel7 is localized to the ER (11). To investigate the subcellular localization of HFRC1, HA-tagged HFRC1 protein was transiently expressed in a mammalian cell line, and immunostained. A mammalian expression vector, pCXN2, was inserted with the ORF of hfrc1 or UGTrel7 with three copies of the HA epitope tag at the C terminus, and transfected transiently into HCT116 cells. The cells were double-immunostained with anti-HA mAb and anti-␤4GalT1 mAb or anti-ERp57 mAb, and the immunofluorescence was observed microscopically. As shown in Fig. 2, HFRC1-HA was co-localized with ␤4GalT1, which is a typical protein of trans-Golgi localization (15). In contrast, UGTrel7-HA was co-localized with ERp57, which is a typical protein of the ER (Fig. 2). These results indicate that HFRC1 is localized to the Golgi apparatus, unlike UGTrel7.
Tissue Distribution of Hfrc1 and UGTrel7 Transcripts in Human Tissues-Although hfrc1 and UGTrel7 are closely related genes, the subcellular localization of HFRC1 is different from that of UGTrel7. To investigate whether hfrc1 and UGTrel7 have similar tissue distributions, the expression levels of hfrc1 and UGTrel7 transcripts in human tissues were analyzed using real time PCR.
The gene expression profiles of these genes in various human tissues are shown in Fig. 3. All transcript levels are shown relative to that of GAPDH. Hfrc1 and UGTrel7 displayed similar tissue distributions, each showing high levels of expression in the colon, stomach, lung, and leukocyte. However, the expression level of hfrc1 transcripts was 10 times that of UGTrel7 (note the different scales in the two panels in Fig. 3).
HFRC1 Is a Multisubstrate Specific Protein That Transports UDP-GlcNAc, UDP-Glc, and GDP-Man-The yeast S. cerevisiae is an organism that has been widely used for the analysis of NSTs by heterologous expression. To express HFRC1 protein in S. cerevisiae, the yeast expression vector YEp352GAP-II was inserted with the ORF of hfrc1, and introduced into W303-1a yeast. The expression status of HA-tagged HFRC1 protein in the yeast was analyzed by Western blotting using antibody against the HA epitope tag. HA-tagged HFRC1 protein was successfully expressed in the Golgi-enriched P100 fraction as a 42-kDa protein (Fig. 4A). The transport activity of HFRC1 for nucleotide sugars into P100 fractions was examined using radiolabeled substrates. As shown in Fig. 4B, the incorporation into the P100 fraction prepared from yeast cells (those expressing HFRC1 over that shown for the mock) was 6.2 times for UDP-GlcNAc, 4.2 times for UDP-Glc, and 1.9 times for GDP-Man. No significant difference was observed between hfrc1 and the mock in the transport of other nucleotide sugars. The substrate concentration dependences of UDP-GlcNAc, UDP-Glc, and GDP-Man transport by HFRC1 are shown in Fig. 4, C-E. HFRC1 showed saturable transport activity with increases in substrate concentration, and the apparent K m values for UDP-GlcNAc, UDP-Glc, and GDP-Man were estimated to be 8.0, 2.1, and 0.14 M, respectively.
Because the level of endogenous transport of GDP-Man is extremely high in S. cerevisiae, we also tested the transport activity of HFRC1 in a mammalian cell line. HCT116 cells stably expressing either HA-tagged hfrc1 or mock vector (vector alone) were analyzed for the uptake of these substrates into the P100 fraction. The hfrc transfectant showed 5.4 times the hfrc1 transcript level of the mock transfectant (6.6 versus 1.2 ϫ 10 Ϫ3 /GAPDH transcript). The P100 fraction of the hfrc1 transfectant showed 1.7 times the UDP-GlcNAc transport activity and 1.8 times the UDP-Glc transport activity of the mock transfectant (380.6 versus 228.8 pmol/5 min/mg of protein, and 910.7 versus 493.2 pmol/5 min/mg of protein, respectively). However, in contrast to the yeast expression data, a difference in GDP-Man transport activity was hardly detectable between the hfrc1 transfectant and mock transfectant (168.9 versus 175.9 pmol/5 min/mg of protein). These results indicate that

HFRC1 transports UDP-GlcNAc and UDP-Glc in the mammalian cells, but transports GDP-Man alone in the yeast.
Hfrc1 Has a Role in Heparan Sulfate Synthesis-Heparan sulfate molecules are found on the cell surface as glucosaminoglycan chains covalently attached to a core protein. Heparan sulfate has a repeat of GlcUA and GlcNAc, whereas chondroitin sulfate has a repeat of GlcUA and GalNAc in the glucosaminoglycan chains. Because HFRC1 can transport UDP-GlcNAc, the possibility needed to be considered that HFRC1 might contribute to the elongation of heparan sulfate by supplying the donor substrate. To investigate whether HFRC1 is involved in the synthesis of heparan sulfate, HCT116 cells stably expressing HA-tagged hfrc1 or mock vector were immunostained with an antibody against heparan sulfate, and analyzed by flow cytometry. As shown in Fig. 5A, the cell surface expression levels of heparan sulfate on hfrc1 transfectant cells were increased compared with that on mock transfectant cells. On the other hand, the cell surface expression levels of chondroitin sulfate were not different between the hfrc1 and mock transfectants. These results suggest that hfrc1 is involved in the synthesis of heparan sulfate by supplying UDP-GlcNAc, a donor substrate for the heparan sulfate synthases. DISCUSSION In the present study, we reported the molecular cloning and functional characterization of a novel human multisubstratespecific NST, which is homologous to Drosophila frc and C. elegans sqv-7. In humans, hfrc1 also exhibited a significant similarity to another NST, UGTrel7. The phylogenetic tree indicates that both hfrc1 and UGTrel7 are orthologs of Drosophila frc (Fig. 1A). In Drosophila, frc is the only gene corresponding to these two human NSTs. We could not find any putative NST sequence homologous to hfrc1 or UGTrel7 from Drosophila data bases, except frc. This raises the possibility that UGTrel7 and hfrc1 may share a common ancestral origin. In fact, these two genes have a similar exon-intron organization, both being composed of 12 exons (Fig. 1C). Furthermore, hfrc1 and UGTrel7 displayed the same expression profiles in human tissues (Fig. 3), suggesting that these two genes are under similar regulation. Hfrc1 and UGTrel7 were mapped on different chromosomes, 9q22.33 and 1p32-p31, respectively. These locations are near the regions of 9q33-34 and 1q21-25, respectively, which are paralogous to the major histocompatibility complex. Kasahara et al. (20) suggested in their report that the genes in the major histocompatibility complex region of jawed vertebrates have risen as a result of ancient chromosomal duplications. These data suggest that hfrc1 and UGTrel7 have evolved from a common ancestral gene through either segmental chromosome duplication or gene duplication.
As shown in Fig. 5, an alteration in the expression of hfrc1 affected heparan sulfate on the cell surface of mammalian cells. UDP-GlcNAc, a transport substrate for HFRC1, can be utilized by the enzymes that catalyze the elongation of heparan sulfate. Because the elongation of proteoglycan takes place within the Golgi apparatus (2,(21)(22)(23), the localization of HFRC1 to the Golgi apparatus would allow it to regulate the donor-substrate level for the enzymes. Heparan sulfate is known to be involved in a variety of biological phenomena, such as morphogenesis, development, angiogenesis, blood coagulation, cell adhesion, and lipid metabolism (23,24). It is also involved in a variety of signaling pathways, in particular those of fibroblast growth factor (25), Wnt/Wingless (26 -28), Decapentaplegic (27), and Hedgehog (29). Selva et al. (9) reported that frc mutant embryos display defects in Wingless, Hedgehog, and fibroblast growth factor signaling (9). Hfrc1 is expected to be involved in growth factor signaling through the regulation of heparan sulfate synthesis in a way similar to that observed for frc.
The substrate specificity of HFRC1 also suggests the possibility that HFRC1 may be able to modulate Notch activity in humans by supplying UDP-GlcNAc. The interaction of the Notch receptor with its ligands, Serrate and Delta, is modulated by the addition of GlcNAc to an O-linked Fuc on the Notch by FRINGE enzyme (30 -33). In Drosophila, mutation of frc leads to a defect in Notch maturation (10). Here, HFRC1 exhibited an apparent K m value for UDP-GlcNAc comparable with that of FRC (8 versus 7.8 M, respectively). It would be interesting to investigate whether hfrc1 can rescue the defects in glycosylation and signaling caused by frc mutations.
We also observed that HFRC1 has a UDP-Glc transport activity. HFRC1 is the first UDP-Glc transporter to be reported in a mammal, although the physiological significance of UDP-Glc transport in the Golgi apparatus is obscure. Possibly, it is involved in some glycosylation process in the Golgi apparatus, such as the addition of O-linked glucose (34) or the glucosyltransferase reaction by ␤4GalT1 in the presence of ␣-lactalbumin (35). On the other hand, we found that HFRC1 transported GDP-Man alone in the yeast, not in the mammalian cells. This difference is perhaps because of the abundance of antiport molecules for GDP-Man within the yeast Golgi apparatus. It has been suggested that GDP-Man is transported into the lumen of the Golgi apparatus of yeast and Leishmania, but not into that of mammals (3). Whether or not HFRC1 can act as a real GDP-Man transporter in mammalian cells needs to be evaluated in future investigations.
In humans, another UDP-GlcNAc transporter exists within the Golgi apparatus (36). We did not investigate whether HFRC1 and this UDP-GlcNAc transporter operate in a complementary fashion or in different ways within the Golgi apparatus. Identification of the glycoconjugates that are affected by these genes may hold the key to a clarification of their functions. Analysis using RNA interference of each gene may also help to elucidate the significance of these transporters in glycosylation.