Portal de Periódicos da CAPES

Variety of Nucleotide Sugar Transporters with Respect to the Interaction with Nucleoside Mono- and Diphosphates

2007; Elsevier BV; Volume: 282; Issue: 34 Linguagem: Inglês

10.1074/jbc.m611358200

1083-351X

Masatoshi Muraoka, Toshiaki Miki, Nobuhiro Ishida, Takahiko Hara, Masao Kawakita,

Genomics and Chromatin Dynamics

Nucleotide sugar transporters have long been assumed to be antiporters that exclusively use nucleoside monophosphates as antiport substrates. Here we present evidence indicating that two other types of nucleotide sugar transporters exist that differ in their antiport substrate specificity. Biochemical studies using microsomes derived from Saccharomyces cerevisiae cells expressing either human (h) UGTrel7 or the Drosophila (d) FRC (Fringe connection) transporter revealed that (i) efflux of preloaded UDP-glucuronic acid from the yeast microsomes expressing hUGTrel7 was strongly enhanced by UDP-GlcNAc added in the external medium, but not by UMP or UDP, suggesting that hUGTrel7 may be described as a UDP-sugar/UDP-sugar antiporter, and (ii) addition of UDP-sugars, UDP, or UMP in the external medium stimulated the efflux of preloaded UDP-GlcNAc from the yeast microsomes expressing dFRC to a comparable extent, suggesting that UDP, as well as UMP, may serve as an antiport substrate of dFRC. Antiport of UDP-sugars with these specific substrates was reproduced and definitively confirmed using proteoliposomes reconstituted from solubilized and purified transporters. Possible physiological implications of these observations are discussed. Nucleotide sugar transporters have long been assumed to be antiporters that exclusively use nucleoside monophosphates as antiport substrates. Here we present evidence indicating that two other types of nucleotide sugar transporters exist that differ in their antiport substrate specificity. Biochemical studies using microsomes derived from Saccharomyces cerevisiae cells expressing either human (h) UGTrel7 or the Drosophila (d) FRC (Fringe connection) transporter revealed that (i) efflux of preloaded UDP-glucuronic acid from the yeast microsomes expressing hUGTrel7 was strongly enhanced by UDP-GlcNAc added in the external medium, but not by UMP or UDP, suggesting that hUGTrel7 may be described as a UDP-sugar/UDP-sugar antiporter, and (ii) addition of UDP-sugars, UDP, or UMP in the external medium stimulated the efflux of preloaded UDP-GlcNAc from the yeast microsomes expressing dFRC to a comparable extent, suggesting that UDP, as well as UMP, may serve as an antiport substrate of dFRC. Antiport of UDP-sugars with these specific substrates was reproduced and definitively confirmed using proteoliposomes reconstituted from solubilized and purified transporters. Possible physiological implications of these observations are discussed. Cellular glycoconjugates, including glycoproteins, glycolipids, proteoglycans, and glucuronides, are glycosylated in appropriate ways in the endoplasmic reticulum (ER) 2The abbreviations used are: ERendoplasmic reticulumNSTsnucleotide sugar transportersGlcUAglucuronic acidhhumandDrosophilahUGThuman UDP-Gal transporter and the Golgi compartments. The catalytic centers of most glycosyltransferases involved in glycosylation processes face the lumen of these organelles. Therefore, a variety of nucleotide sugars have to be transported from the cytosol into the lumen of the ER and the Golgi apparatus by specific transporters to serve as the monosaccharide donors in oligo/polysaccharide chain elongation (1Hirschberg C.B. Robbins P.W. Abeijon C. Annu. Rev. Biochem. 1998; 67: 49-69Crossref PubMed Scopus (311) Google Scholar, 2Kawakita M. Ishida N. Miura N. Sun-Wada G.-H. Yoshioka S. J. Biochem. (Tokyo). 1998; 123: 777-785Crossref PubMed Scopus (55) Google Scholar). Several of such nucleotide sugar transporters (NSTs) that contribute to glycoconjugate biosynthesis have been cloned and characterized at the molecular level (3Gerardy-Schahn R. Oelmann S. Bakker H. Biochimie (Paris). 2001; 83: 775-782Crossref PubMed Scopus (87) Google Scholar). endoplasmic reticulum nucleotide sugar transporters glucuronic acid human Drosophila human UDP-Gal transporter It is likely that the concentrations of nucleotide sugars in the lumen of the ER and the Golgi apparatus, which are determined by the transport activity of NSTs, are among the factors that regulate glycosyltransferase activity. To gain a better understanding of how glycosylation is regulated, it may therefore be important to know how NST activity is controlled in the cell. NMP is so far the only known cellular substance that may affect NST activity under physiological conditions. The activities of several NSTs have been shown to be enhanced by an appropriate NMP that serves as an antiport substrate of the respective transporter (4Capasso J.M. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1984; 81: 7051-7055Crossref PubMed Scopus (110) Google Scholar, 5Waldman B.C. Rudnick G. Biochemistry. 1990; 29: 44-52Crossref PubMed Scopus (61) Google Scholar, 6Milla M.E. Clairmont C.A. Hirschberg C.B. J. Biol. Chem. 1992; 267: 103-107Abstract Full Text PDF PubMed Google Scholar, 7Hong K. Ma D. Beverley S.M. Turco S.J. Biochemistry. 2000; 39: 2013-2022Crossref PubMed Scopus (83) Google Scholar, 8Aoki K. Ishida N. Kawakita M. J. Biol. Chem. 2003; 278: 22887-22893Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar, 9Segawa H. Soares R.P. Kawakita M. Beverley S.M. Turco S.J. J. Biol. Chem. 2005; 280: 2028-2035Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar). The physiological significance of the antiport of NMP in glycoconjugate biosynthesis is underscored by the following observations: in a Golgi GDPase-defective mutant of Saccharomyces cerevisiae, the GMP concentration in the yeast-derived Golgi vesicles is decreased, which in turn results in reduced entry of GDP-mannose into the vesicles (10Berninsone P. Miret J.J. Hirschberg C.B. J. Biol. Chem. 1994; 269: 207-211Abstract Full Text PDF PubMed Google Scholar). The altered mannosylation of macromolecules in this mutant is likely due to the reduced availability of GDP-mannose in the Golgi apparatus (11Abeijon C. Yanagisawa K. Mandon E.C. Hausler A. Moremen K. Hirschberg C.B. Robbins P.W. J. Cell Biol. 1993; 122: 307-323Crossref PubMed Scopus (162) Google Scholar). Antiport substrates that could affect NST activities may not be limited to nucleoside monophosphates. Recently, we (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar, 15Goto S. Taniguchi M. Muraoka M. Toyoda H. Sado Y. Kawakita M. Hayashi S. Nat. Cell Biol. 2001; 3: 816-822Crossref PubMed Scopus (115) Google Scholar, 16Segawa H. Kawakita M. Ishida N. Eur. J. Biochem. 2002; 269: 128-138Crossref PubMed Scopus (77) Google Scholar) and others (7Hong K. Ma D. Beverley S.M. Turco S.J. Biochemistry. 2000; 39: 2013-2022Crossref PubMed Scopus (83) Google Scholar, 13Berninsone P. Hwang H.Y. Zemtseva I. Horvitz H.R. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3738-3743Crossref PubMed Scopus (116) Google Scholar, 14Selva E.M. Hong K. Baeg G.H. Beverley S.M. Turco S.J. Perrimon N. Hacker U. Nat. Cell Biol. 2001; 3: 809-815Crossref PubMed Scopus (109) Google Scholar, 17Suda T. Kamiyama S. Suzuki M. Kikuchi N. Nakayama K. Narimatsu H. Jigami Y. Aoki T. Nishihara S. J. Biol. Chem. 2004; 279: 26469-26474Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar) reported that some NSTs are able to transport more than one nucleotide sugar. This suggests the possibility that such NSTs may exchange different nucleotide sugars distributed on the two sides of membranes and that the antiport of one nucleotide sugar could be regulatory for the entry of another substrate nucleotide sugar. Although the physiological importance of a nucleotide sugar as a regulatory factor affecting the transport of another nucleotide sugar has not been directly demonstrated so far, the possible involvement of this type of regulation has been suggested for the cellular glucuronidation process. UDP-GlcNAc is known to stimulate UDP-glucuronosyltransferase in the ER (18Pogell B.M. Leloir L.F. J. Biol. Chem. 1961; 236: 293-298Abstract Full Text PDF PubMed Google Scholar). On the basis of biochemical analyses using rat liver ER, Bossuyt and Blanckaert (19Bossuyt X. Blanckaert N. Biochem. J. 1995; 305: 321-328Crossref PubMed Scopus (47) Google Scholar) proposed a model in which antiport of UDP-GlcNAc enhances translocation of UDP-glucuronic acid (GlcUA) into the ER lumen, which in turn stimulates glucuronidation. However, the transporter that should play a crucial role in this model has not been identified yet at the molecular level. We reported previously the molecular cloning and characterization of the human (h) UGTrel7 (UDP-Gal transporter-related isozyme 7), which is localized in the ER (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar). In this study, we further show that this transporter can transport UDP-GlcNAc and can act as a UDP-GlcUA/UDP-GlcNAc antiporter. We also investigated the antiport substrate specificity of the Drosophila (d) FRC (Fringe connection) transporter (14Selva E.M. Hong K. Baeg G.H. Beverley S.M. Turco S.J. Perrimon N. Hacker U. Nat. Cell Biol. 2001; 3: 809-815Crossref PubMed Scopus (109) Google Scholar, 15Goto S. Taniguchi M. Muraoka M. Toyoda H. Sado Y. Kawakita M. Hayashi S. Nat. Cell Biol. 2001; 3: 816-822Crossref PubMed Scopus (115) Google Scholar) and the human UDP-Gal transporter (hUGT) (16Segawa H. Kawakita M. Ishida N. Eur. J. Biochem. 2002; 269: 128-138Crossref PubMed Scopus (77) Google Scholar, 20Miura N. Ishida N. Hoshino M. Yamauchi M. Hara T. Ayusawa D. Kawakita M. J. Biochem. (Tokyo). 1996; 120: 236-241Crossref PubMed Scopus (121) Google Scholar) and found that various NSTs differed as to whether they can interact with nucleoside mono- and diphosphates. The implications of this variety in the control of nucleotide sugar transport are discussed. Materials—Radioactive compounds were purchased from American Radiolabeled Chemicals, Inc., except [acetyl-3H]UDP-GlcNAc and [5-3H]UMP were purchased from Moravek Biochemicals, Inc. The yeast strain CB001L was a kind gift from Dr. Y. Kikuchi (University of Tokyo). Soybean phospholipid (asolectin) was purchased from Nihon Pharmaceutical Co., Ltd. and was partially purified before use for reconstitution by the method of Sone et al. (21Sone N. Yoshida M. Hirata H. Kagawa Y. J. Biochem. (Tokyo). 1977; 81: 519-528Crossref PubMed Scopus (81) Google Scholar). Yeast Strains, Transformation, and Culture—Saccharomyces cerevisiae strain YPH500 (MATα, ura3-52, lys2-801, ade2-101, trp1-Δ63, his3-Δ200, leu2-Δ1) or CB001L (MATα, leu2, trp1, ura3, prb–, pep4::LEU2) was used in the expression study. The DNA sequences encoding hUGTrel7, dFRC, and hUGT tagged with a hemagglutinin or His8 epitope at the C termini were introduced into the pYEX-BX vector. Transformation was performed using the lithium acetate method (22Adams A. Gottschling D.E. Kaiser C.A. Stearns T. Methods in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1997Google Scholar), and transformants were selected on selective medium containing 0.67% (w/v) Bacto-yeast nitrogen base without amino acids, 2% glucose, and auxotrophic supplements except uracil. For the preparation of membrane vesicles, the transformants were grown in liquid selective medium until they reached a density of A600 ∼ 0.8. Cupric sulfate was then added to the culture at a final concentration of 0.5 mm. The cells were cultured further for 2 h and harvested. Expression of the exogenous transporters was confirmed by Western blotting using antibodies against the epitope tags. Preparation of Yeast Microsomes—Membrane vesicles were prepared essentially as described previously (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar). Cells were washed with ice-cold 10 mm NaN3 and converted into spheroplasts in a spheroplast solution containing 1.4 m sorbitol, 50 mm potassium phosphate (pH 7.5), 10 mm NaN3, 0.25% (v/v) 2-mercaptoethanol, and 2 mg of Zymolyase 100T (Seikagaku Corp., Tokyo) per g of cells and incubated at 37 °C for 20 min. The spheroplasts were pelleted; resuspended in 5 volumes of lysis buffer containing 0.8 m sorbitol, 10 mm Hepes/Tris (pH 7.4), 1 mm EDTA, and a protease inhibitor mixture (Complete, EDTA-free, Roche Diagnostics); and homogenized using a Teflon homogenizer. The lysate was centrifuged at 1500 × g for 10 min to remove unlysed cells and debris. The supernatant was centrifuged at 10,000 × g for 10 min to yield a pellet of membrane vesicles (P10). The P10 fractions were resuspended in lysis buffer and used for transport assays and purification. Purification of hUGTrel7 and dFRC Transporters—The yeast P10 fraction from the His8-tagged hUGTrel7 or dFRC transformant was solubilized with deoxycholic acid at a final concentration of 1%, kept at 4 °C for 5 min, and centrifuged at 100,000 × g for 15 min. The supernatant was loaded onto a column of Talon resin (Clontech), which had been equilibrated with buffer A (25 mm Hepes/NaOH (pH 8.0), 1% deoxycholic acid, 20% glycerol, and EDTA-free Complete protease inhibitor mixture). The column was washed with buffer A, and the His8-tagged transporter was eluted with 25 mm Hepes/NaOH (pH 8.0) containing 300 mm imidazole, 1% deoxycholic acid, 20% glycerol, and EDTA-free Complete protease inhibitor mixture. Solubilization of the yeast P10 fractions with deoxycholic acid tended to lead to degradation of exogenous transporters, especially hUGTrel7. To minimize this degradation, strain CB001L, with vacuolar proteinases A and B disrupted, was used for purification and reconstitution studies. Reconstitution of Proteoliposomes from Purified Transporters—One part purified hUGTrel7 or dFRC transporter was mixed with 1 part asolectin (40 mg/ml) dispersed in buffer containing 25 mm Hepes/NaOH (pH 7.4), 1 mm EDTA, 200 mm NaCl, and 1% deoxycholic acid with or without substrates (2 mm) to be preloaded. The mixture was dialyzed overnight against a solution containing 25 mm Hepes/NaOH (pH 7.4), 1 mm EDTA, and 100 mm NaCl with or without substrates (1 mm) used for preloading with one change of the outer solution using Spectra/Por 6 (Mr cutoff = 25,000; Spectrum Laboratories, Inc.). The dialyzed mixture was then applied to Micro Bio-Spin P-30 columns (Bio-Rad) to remove substrates remaining outside of liposomes before use in the transport assay. Transport Assay—The transport assay was performed essentially as described previously (23Sun-Wada G.-H. Yoshioka S. Ishida N. Kawakita M. J. Biochem. (Tokyo). 1998; 123: 912-917Crossref PubMed Scopus (39) Google Scholar). Briefly, the transport reaction mixture (100 μl) used in assays contained yeast membrane vesicles (usually 50 μg of protein), 1 μm radiolabeled substrate, 0.8 m sorbitol, 10 mm Tris-HCl (pH 7.0), and 0.5 mm dimercaptopropanol. The reaction mixture also contained 1 mm EDTA unless stated otherwise (see Fig. 1, A, B, and D). The reaction mixture was incubated at 30 °C. The reaction was started by addition of membrane vesicles and terminated at the appropriate time by an 11-fold dilution with ice-cold stop buffer containing 0.8 m sorbitol, 10 mm Tris-HCl (pH 7.0), 1 μm nonradiolabeled substrate, and 1 mm EDTA. The entire reaction mixture was passed through a nitrocellulose filter (Millipore Corp.). The filter was washed three times with 1 ml of ice-cold stop buffer and dried, and the radioactivity trapped on the filter was determined. Analysis of Soluble Radioactive Compounds Accumulated in Yeast Microsomes—The filters on which radioactivity accumulated in yeast microsomes was trapped were prepared as described above. To extract radioactive compounds, the filters were boiled for 2 min in 25 mm Tris-HCl (pH 7.5). The soluble radioactive compounds in the extract were separated by ionexchange chromatography on Vivapure Q mini H columns (Vivascience). Before application to the columns, an appropriate set of nonradioactive compounds (30 nmol of GlcUA and 30 nmol of UDP-GlcUA or 15 nmol of GlcNAc, 15 nmol of GlcNAc-1-P, and 15 nmol of UDP-GlcNAc) was added to the sample. Elution patterns of nonradioactive compounds were determined in separate experiments prior to the analyses of radioactive extracts. hUGTrel7 Utilizes UDP-GlcNAc as a Transport Substrate in Addition to UDP-GlcUA and UDP-GalNAc—In our previous study (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar), hUGTrel7 protein was expressed in S. cerevisiae, and the uptake of nucleotide sugars by microsomes expressing hUGTrel7 was examined. Although definite uptake of UDP-GlcNAc, as well as UDP-GlcUA and UDP-GalNAc, was observed, we were not certain that the uptake of UDP-GlcNAc represented a net uptake by hUGTrel7 because of the high background uptake into vector control vesicles (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar). To clarify whether hUGTrel7 is able to transport UDP-GlcNAc, the effects of UDP-GlcUA and UDP-GalNAc, both of which are authentic substrates of hUGTrel7, on UDP-GlcNAc uptake were examined (Fig. 1, A and B). The uptake of UDP-GlcNAc by hUGTrel7-expressing microsomes was inhibited by UDP-GlcUA and UDP-GalNAc, whereas addition of these nucleotide sugars had no effect on the uptake of UDP-GlcNAc by vector control microsomes. On the other hand, the uptake of UDP-Glc and GDP-Man was not affected by UDP-GlcUA or UDP-GalNAc in either hUGTrel7-expressing or control microsomes. These results strongly suggest that UDP-GlcNAc is in fact taken up by hUGTrel7. This was shown more directly by including 1 mm EDTA in the reaction mixture of the transport assay (Fig. 1C). Addition of EDTA dramatically reduced the background uptake of UDP-GlcNAc by vector control microsomes, probably as a result of the removal of divalent cations required by GlcNAc transferase, which may have largely contributed to the high background incorporation of this nucleotide sugar. In contrast, UDP-GlcNAc uptake by microsomes expressing hUGTrel7 was practically unchanged in the presence of EDTA and was much higher than that by control microsomes. The uptake of UDP-Glc and GDP-Man was greatly reduced in both vector control and hUGTrel7-expressing microsomes, and their uptake was not appreciably stimulated by expression of hUGTrel7. The uptake of UDP-GlcNAc was saturable with respect to its concentration (supplemental Fig. S1B). These results clearly indicate that hUGTrel7 recognizes UDP-GlcNAc as a transport substrate. It should be noted that UMP, which has been believed to serve as a substrate of a number of UDP-sugar transporters, was not taken up by hUGTrel7-expressing microsomes (Fig. 1C). The accumulation of UDP-GlcNAc in hUGTrel7-expressing microsomes was not observed when the microsomes were permeabilized by treatment with Staphylococcus aureus α-toxin (Fig. 1D). This indicates that the accumulation of UDP-GlcNAc is due to a net transport of UDP-GlcNAc across the microsomal membranes and that hUGTrel7 transports UDP-GlcNAc, as well as the previously identified substrates UDP-GlcUA and UDP-GalNAc. trans-Stimulation of UDP-GlcUA Transport Activity of hUGTrel7 by UDP-GlcNAc but Not by UMP or UDP—We reported previously that hemagglutinin-tagged hUGTrel7 protein transiently expressed in Chinese hamster ovary cells is localized in the ER (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar). This suggested the possibility that hUGTrel7 is the UDP-GlcUA transporter involved in supplying the substrate for glucuronidation in the ER (24Radominska-Pandya A. Czernik P.J. Little J.M. Battaglia E. Mackenzie P.I. Drug Metab. Rev. 1999; 31: 817-899Crossref PubMed Scopus (440) Google Scholar), the activity of which in the ER has long been recognized (25Nuwayhid N. Glaser J.H. Johnson J.C. Conrad H.E. Hauser S.C. Hirschberg C.B. J. Biol. Chem. 1986; 261: 12936-12941Abstract Full Text PDF PubMed Google Scholar, 26Bossuyt X. Blanckaert N. Biochem. J. 1994; 302: 261-269Crossref PubMed Scopus (41) Google Scholar). As it was reported or suggested that UDP-GlcNAc stimulates glucuronidation (18Pogell B.M. Leloir L.F. J. Biol. Chem. 1961; 236: 293-298Abstract Full Text PDF PubMed Google Scholar, 19Bossuyt X. Blanckaert N. Biochem. J. 1995; 305: 321-328Crossref PubMed Scopus (47) Google Scholar) by promoting the entry of UDP-GlcUA into the ER lumen in exchange for itself, we investigated whether UDP-GlcUA transport is enhanced by antiport of UDP-GlcNAc. Because the action of an antiporter should be bidirectional, we examined the release of UDP-GlcUA once incorporated into the microsomal membranes induced by several antiport substrate candidates in the experiments that follow. Yeast microsomes expressing hUGTrel7 were allowed to incorporate UDP-[3H]GlcUA for 3 min until a plateau was reached (supplemental Fig. S2), and then an excess of various compounds was added to test their trans-stimulating effect on UDP-GlcUA efflux from the microsomes, which was assessed by determining the amount of UDP-[3H]GlcUA remaining after incubation for another 30 s (Table 1). The ion-exchange chromatography analysis suggested that the majority of soluble radioactive compounds accumulated in the microsomes during incubation for 3 min remained as UDP-[3H]GlcUA. The elution profile of radioactive compounds extracted from the microsomes was consistent with that expected for UDP-GlcUA (Table 2). Addition of UDP-GlcNAc, UDP-GalNAc, and UDP-GlcUA, the substrates of hUGTrel7, significantly reduced the amount of UDP-[3H]GlcUA remaining (Table 1). The reduction was not due to a simple displacement of UDP-[3H]GlcUA that was bound to the external surface of microsomes by these nucleotide sugars because the amount of UDP-GlcUA associated with the microsomes was greatly diminished when they were mildly permeabilized with α-toxin. Under this condition, little (if any) UDP-[3H]GlcUA could be chased with UDP-GlcNAc. Metabolites of UDP-GlcUA and UDP-GlcNAc (viz. GlcUA, GlcNAc-1-P, and GlcNAc) that might be formed from the nucleotide sugars by the action of microsomal enzymes did not induce the efflux of UDP-[3H]GlcUA (Table 1). In another experiment, UDP-[3H]GlcUA was not significantly associated with the microsomes if they were obtained from mock-transformed cells, and UDP-[3H]GlcUA was not released at all from the microsomes in which hUGTrel7 was not expressed (Table 1). These results indicate that the accumulation and release of UDP-GlcUA observed are due to net transport of nucleotide sugars mediated by hUGTrel7 and that these substrates may be transported by hUGTrel7 in either direction in an antiport mode. This was further substantiated by the observation that addition of UDP-GlcUA in the external medium stimulated the efflux of preloaded UDP-[3H]GlcNAc from the microsomes and that this efflux definitely depended on the expression of hUGTrel7 in the microsomal membranes (Tables 3 and 4). Glucuronides, including p-nitrophenyl glucuronide and phenolphthalein glucuronide, did not induce the efflux of UDP-GlcUA (Table 1). Similarly, UMP did not affect the level of UDP-GlcUA remaining in the microsomal vesicles, which is consistent with the poor uptake of UMP by hUGTrel7-expressing microsomes. When UDP was added at a relatively high concentration, a small stimulation of UDP-GlcUA efflux was noted.TABLE 1UDP-[3H]GlcUA efflux from microsomes expressing hUGTrel7 is induced by UDP-GlcNAc but not by UMPTransport countersubstrateUDP-[3H]GlcUA remaining in microsomesEffluxpmol/mg%Experiment 1Control3.82 ± 0.090UMP4.01 ± 0.19-5.2 ± 4.9UDP3.22 ± 0.0415.7 ± 1.0UDP-GlcNAc2.34 ± 0.1738.7 ± 4.5UDP-GalNAc2.52 ± 0.1534.1 ± 4.0UDP-GlcUA2.77 ± 0.1227.4 ± 3.3UDP-Gal3.42 ± 0.2110.5 ± 5.4CMP-Sia4.30 ± 0.34-12.6 ± 8.8GlcNAc3.81 ± 0.070.3 ± 1.9GlcNAc-1-P3.81 ± 0.140.3 ± 3.6GlcUA3.54 ± 0.127.4 ± 3.2Nitrophenyl-GlcUA3.86 ± 0.25-1.0 ± 6.7Phenolphthalein-GlcUA3.95 ± 0.14-3.4 ± 3.6Experiment 2Control (α-toxin-treatedaMicrosomes treated with α-toxin as described for Fig. 1D.)1.01 ± 0.06UDP-GlcNAc (α-toxin-treatedaMicrosomes treated with α-toxin as described for Fig. 1D.)0.73 ± 0.01Experiment 3Control (mockbMicrosomes in which hUGTrel7 was not expressed.)1.10 ± 0.04UDP-GlcNAc (mockbMicrosomes in which hUGTrel7 was not expressed.)1.28 ± 0.19UDP-GalNAc (mockbMicrosomes in which hUGTrel7 was not expressed.)1.16 ± 0.04UDP-GlcUA (mockbMicrosomes in which hUGTrel7 was not expressed.)1.27 ± 0.29a Microsomes treated with α-toxin as described for Fig. 1D.b Microsomes in which hUGTrel7 was not expressed. Open table in a new tab TABLE 2Analysis of UDP-[3H]GlcUA-related compounds recovered from yeast microsomes after the nucleotide sugar transport reactionCompoundRecoveryUnboundEluate 1Eluate 2%GlcUA10000UDP-GlcUA0057 ± 1Microsomal extract0051 ± 3 Open table in a new tab TABLE 3Induced UDP-[3H]GlcNAc efflux from microsomes expressing hUGTrel7 by UDP-GlcUATransport countersubstrateUDP-[3H]GlcNAc remaining in microsomesEffluxpmol/mg%Experiment 1Control5.44 ± 0.070UDP-GlcUA3.73 ± 0.0231.4 ± 0.01Experiment 2Control (α-toxin-treated)0.74 ± 0.02UDP-GlcUA (α-toxin-treated)0.62 ± 0.05Experiment 3Control (mock)1.41 ± 0.20UDP-GlcUA (mock)1.30 ± 0.02 Open table in a new tab TABLE 4Analysis of UDP-[3H]GlcNAc-related compounds recovered from yeast microsomes after the nucleotide sugar transport reactionCompoundRecoveryUnboundEluate 1Eluate 2%GlcNAc10000GlcNAc-1-P083 ± 46 ± 2UDP-GlcNAc010 ± 280 ± 1Extract 11 ± 17 ± 377 ± 2Extract 21 ± 17 ± 179 ± 4 Open table in a new tab Similar experiments were carried out using proteoliposomes reconstituted from solubilized and purified hUGTrel7 to eliminate possible contribution by any endogenous microsomal component and to exclude the possibility that the substrates of hUGTrel7 might be degraded by enzymes in yeast microsomes. hUGTrel7 tagged with His8 was solubilized from yeast microsomes expressing the transporter and purified by metal affinity column chromatography (Fig. 2A). As shown in Fig. 2B, proteoliposomes reconstituted from the purified hUGTrel7 transporter were active in UDP-[3H]GlcUA transport, and the uptake was strongly dependent on preloading UDP-GlcNAc in the liposomes. Neither UMP nor UDP preloaded in liposomes significantly stimulated the transport of UDP-GlcUA across the membranes. To analyze the interaction of UMP, UDP, and UDP-sugars with hUGTrel7 more quantitatively, we measured the UDP-GlcNAc transport activity of the transporter in the presence of UMP, UDP, and UDP-GlcUA using yeast microsomes expressing hUGTrel7 (Fig. 3). As with other experiments in this study, EDTA was added to the transport reaction mixture to minimize the endogenous uptake of UDP-GlcNAc (supplemental Fig. S3A). UDP-GlcUA inhibited the UDP-GlcNAc transport activity of hUGTrel7, and Dixon plot analysis gave a Km value of 2.5 μm for UDP-GlcNAc and a Ki value of 4 μm for UDP-GlcUA. UDP slightly inhibited UDP-GlcNAc transport activity, but its Ki value was too high to be determined accurately. UMP did not affect the uptake of UDP-GlcNAc at all. These results clearly indicate that UDP-GlcNAc, but not UMP or UDP, serves as a countersubstrate for the UDP-GlcUA transport system. This is consistent with the results shown in Fig. 2B. The dFRC transporter, a homolog of hUGTrel7 that has been shown to transport various UDP-sugars, including UDP-GlcUA and UMP (15Goto S. Taniguchi M. Muraoka M. Toyoda H. Sado Y. Kawakita M. Hayashi S. Nat. Cell Biol. 2001; 3: 816-822Crossref PubMed Scopus (115) Google Scholar) (supplemental Fig. S1), was also reconstituted into liposomes following solubilization and purification from yeast microsomal membranes expressing dFRC (Fig. 2C). With the reconstituted dFRC transporter, UDP-GlcUA transport was absolutely dependent on the preloading of UDP-GlcNAc, UDP, or UMP (Fig. 2D). Distinct Interactions of UMP and UDP with dFRC and hUGT—The results shown above strongly suggest that UMP and UDP may serve as specific substrates of the dFRC transporter. This is in contrast to the currently accepted understanding that NSTs generally utilize nucleoside monophosphates but not nucleoside diphosphates, which might underscore the importance of NDPase (4Capasso J.M. Hirschberg C.B. Proc. Natl. Acad. Sci. U. S. A. 1984; 81: 7051-7055Crossref PubMed Scopus (110) Google Scholar, 7Hong K. Ma D. Beverley S.M. Turco S.J. Biochemistry. 2000; 39: 2013-2022Crossref PubMed Scopus (83) Google Scholar, 9Segawa H. Soares R.P. Kawakita M. Beverley S.M. Turco S.J. J. Biol. Chem. 2005; 280: 2028-2035Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar, 10Berninsone P. Miret J.J. Hirschberg C.B. J. Biol. Chem. 1994; 269: 207-211Abstract Full Text PDF PubMed Google Scholar, 11Abeijon C. Yanagisawa K. Mandon E.C. Hausler A. Moremen K. Hirschberg C.B. Robbins P.W. J. Cell Biol. 1993; 122: 307-323Crossref PubMed Scopus (162) Google Scholar). We therefore analyzed the interaction of UMP and UDP with dFRC and hUGT in more detail. We first examined the effects of various nucleotides and nucleotide sugars on the efflux of UDP-[3H]GlcNAc incorporated in microsomes from dFRC-expressing yeast (Tables 4 and 5). Both UMP and UDP were almost as effective as UDP-GlcNAc and UDP-GlcUA in stimulating the efflux of UDP-[3H]GlcNAc accumulated in the microsomal vesicles prior to their addition (Table 5). On the other hand, UMP and UDP-Gal induced the efflux of UDP-[3H]Gal from microsomes from hUGT-expressing yeast, but UDP was much less effective in stimulating the efflux (Table 6). It should be noted that even 100 μm UDP was much less effective than 10 μm UMP in stimulating UDP-[3H]Gal efflux from microsomes expressing hUGT. This suggested that much less than 10% of UDP (if any) was metabolized into UMP in this assay system. In contrast, UDP-[3H]GlcNAc efflux from microsomes expressing dFRC was stimulated much more effectively by 100 μm UDP than by 10 and 50 μm UMP, confirming that the UDP-[3H]GlcNAc efflux was not due to the UMP derived from UDP. These data indicate that UMP and UDP are able to stimulate the movement of appropriate nucleotide sugars across the microsomal membrane mediated by dFRC in both directions, whereas only UMP is able to act on hUGT in this way.TABLE 5Induced UDP-[3H]GlcNAc efflux from microsomes expressing dFRCTransport countersubstrateUDP-[3H]GlcNAc remaining in microsomesEffluxpmol/mg%Experiment 1Control17.4 ± 0.40UMP (10 μm)14.6 ± 0.616.1 ± 3.3UDP (10 μm)14.9 ± 0.614.5 ± 3.2UDP-GlcNAc (10 μm)14.4 ± 0.817.3 ± 4.4UMP (50 μm)13.4 ± 0.422.9 ± 2.5UMP (100 μm)13.0 ± 0.425.2 ± 2.3UDP (100 μm)11.6 ± 0.533.4 ± 3.1UDP-GlcNAc (100 μm)11.9 ± 0.331.4 ± 1.5UDP-GlcUA (100 μm)12.1 ± 0.330.6 ± 1.2GlcNAc (100 μm)17.1 ± 0.51.5 ± 2.7GlcNAc-1-P (100 μm)16.7 ± 0.64.2 ± 3.5GMP (100 μm)17.0 ± 0.22.5 ± 1.1Experiment 2Control (α-toxin-treated)3.3 ± 0.6UMP (α-toxin-treated; 100 μm)2.1 ± 0.2UDP (α-toxin-treated; 100 μm)1.9 ± 0.1UDP-GlcUA (α-toxin-treated; 100 μm)2.2 ± 0.4Experiment 3Control (mock)1.4 ± 0.2UMP (mock; 100 μm)1.2 ± 0.1UDP (mock; 100 μm)1.5 ± 0.2UDP-GlcUA (mock; 100 μm)1.3 ± 0.1 Open table in a new tab TABLE 6Induced UDP-[3H]Gal efflux from microsomes expressing hUGTTransport countersubstrateUDP-[3H]Gal remaining in microsomesEffluxpmol/mg%Experiment 1Control12.4 ± 0.40UMP (10 μm)7.2 ± 0.342.3 ± 2.4UDP (10 μm)11.5 ± 0.37.6 ± 2.0UDP-Gal (10 μm)8.0 ± 0.235.8 ± 1.7UMP (100 μm)4.6 ± 0.163.0 ± 0.8UDP (100 μm)10.1 ± 0.218.5 ± 1.5UDP-Gal (100 μm)4.7 ± 0.161.9 ± 0.5Experiment 2Control (α-toxin-treated)2.6 ± 0.1UMP (α-toxin-treated; 100 μm)2.4 ± 0.2UDP-Gal (α-toxin-treated; 100 μm)2.4 ± 0.1Experiment 3Control (mock)3.3 ± 0.2UMP (mock; 100 μm)3.3 ± 0.3UDP (mock; 100 μm)3.0 ± 0.2UDP-Gal (mock; 100 μm)3.1 ± 0.1 Open table in a new tab We then measured the UDP-GlcNAc transport activity of dFRC in the presence of UMP and UDP, as we did with hUGTrel7, to analyze the interaction of UMP, UDP, and UDP-sugars more quantitatively. These two nucleotides inhibited the UDP-GlcNAc transport activity of dFRC to very similar extents, and Dixon plot analysis gave a Km value of 3 μm for UDP-GlcNAc and Ki values of 7.5 and 5.0 μm for UMP and UDP, respectively (Fig. 4A and supplemental Fig. S3B). Similar analysis of the results with hUGT gave a Km value of 3 μm for UDP-Gal and revealed that UMP inhibited the transport activity with a Ki value of 7 μm. On the other hand, UDP did not affect the uptake of UDP-Gal at all, indicating that it did not interact with the transporter (Fig. 4B and supplemental Fig. S3C). In a previous study (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar), we measured the uptake of nucleotide sugars by yeast microsomes expressing hUGTrel7 and showed that UDP-GlcUA and UDP-GalNAc are good substrates for this transporter. However, the results with UDP-GlcNAc were inconclusive because of considerable background uptake of this nucleotide sugar even by vector control microsomes. In this work, we were able to show definitely that UDP-GlcNAc is a substrate of hUGTrel7. This was made possible by our more recent finding that depletion of divalent cations in the reaction mixture is effective in reducing the background uptake of nucleotide sugars, including GDP-Man, UDP-GlcNAc, and UDP-Glc, by yeast microsomes without affecting the NST-mediated uptake of nucleotide sugars (15Goto S. Taniguchi M. Muraoka M. Toyoda H. Sado Y. Kawakita M. Hayashi S. Nat. Cell Biol. 2001; 3: 816-822Crossref PubMed Scopus (115) Google Scholar). NSTs do not seem to require divalent cations such as Mg2+ for their activity. This is consistent with a recent report showing that reconstituted Leishmania LPG2 transporter also transports GDP-Man in the absence of divalent cations (9Segawa H. Soares R.P. Kawakita M. Beverley S.M. Turco S.J. J. Biol. Chem. 2005; 280: 2028-2035Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar). The high background incorporation that is abolished by depletion of Mg2+ may be due mainly to endogenous glycosylation activities in the yeast microsomes because GDP-Man uptake, for instance, in the presence of Mg2+ was unaffected by permeabilization of microsomal membranes by α-toxin treatment (Fig. 1D). Glucuronidation of thousands of endobiotic and xenobiotic compounds is carried out by UDP-glucuronosyltransferases in the ER lumen (24Radominska-Pandya A. Czernik P.J. Little J.M. Battaglia E. Mackenzie P.I. Drug Metab. Rev. 1999; 31: 817-899Crossref PubMed Scopus (440) Google Scholar), but the transporter that supplies UDP-GlcUA for the transferases has not been identified. The glucuronidation reaction is activated by UDP-GlcNAc, and this activation was proposed to be accounted for by the presence of a putative UDP-GlcUA transporter that is activated by UDP-GlcNAc (18Pogell B.M. Leloir L.F. J. Biol. Chem. 1961; 236: 293-298Abstract Full Text PDF PubMed Google Scholar, 27Berry C. Hallinan T. Biochem. Soc. Trans. 1976; 4: 650-652Crossref PubMed Scopus (37) Google Scholar). The ability of ER-derived microsomes to transport UDP-GlcUA, stimulated by luminal UDP-GlcNAc but not by luminal UMP, was shown later by Bossuyt and Blanckaert (19Bossuyt X. Blanckaert N. Biochem. J. 1995; 305: 321-328Crossref PubMed Scopus (47) Google Scholar), and they proposed the existence of a UDP-GlcUA/UDP-GlcNAc antiporter in the ER membrane. We demonstrated previously that hUGTrel7 is localized in the ER when hUGTrel7 cDNA is expressed in Chinese hamster ovary cells (12Muraoka M. Kawakita M. Ishida N. FEBS Lett. 2001; 495: 87-93Crossref PubMed Scopus (108) Google Scholar). The substrate specificity of this transporter was examined in detail in this study, and we found that it is able to work as a UDP-GlcUA/UDP-GlcNAc antiporter. It should also be pointed out that hUGTrel7 is not able to use UMP as a countersubstrate for the UDP-GlcUA transport system. Taken together, these facts indicate that it is highly likely that hUGTrel7 represents the ER UDP-GlcUA/UDP-GlcNAc antiporter proposed by Bossuyt and Blanckaert (19Bossuyt X. Blanckaert N. Biochem. J. 1995; 305: 321-328Crossref PubMed Scopus (47) Google Scholar). UDP-GlcUA/phenol glucuronide antiport activity was observed previously in rat liver microsomes (28Banhegyi G. Braun L. Marcolongo P. Csala M. Fulceri R. Mandl J. Benedetti A. Biochem. J. 1996; 315: 171-176Crossref PubMed Scopus (34) Google Scholar), but the efflux of UDP-GlcUA was not enhanced by phenol glucuronide in yeast microsomes expressing hUGTrel7. Milla et al. (6Milla M.E. Clairmont C.A. Hirschberg C.B. J. Biol. Chem. 1992; 267: 103-107Abstract Full Text PDF PubMed Google Scholar) observed that transport of UDP-GlcUA into proteoliposomes reconstituted from Golgi proteins was stimulated by preloading UMP, which was ineffective in stimulating hUGTrel7-mediated efflux of UDP-GlcUA. These observations suggest that there may be other UDP-GlcUA transporters in mammalian cells. This study has demonstrated that there are at least three NST types with respect to antiport substrate specificity. hUGT is a representative of the first type, nucleotide sugar/nucleoside monophosphate antiporters. For this type of transporter to work, nucleoside diphosphate formed concomitantly with glycosylation has to be hydrolyzed by nucleoside diphosphatase in the ER or Golgi lumen to nucleoside monophosphate, which in turn serves as the countersubstrate for nucleotide sugar transport (29Abeijon C. Mandon E.C. Hirschberg C.B. Trends Biochem. Sci. 1997; 22: 203-207Abstract Full Text PDF PubMed Scopus (101) Google Scholar). It should also be pointed out that nucleoside diphosphate is inhibitory to glycosyltransferases (30Khatra B.S. Herries D.G. Brew K. Eur. J. Biochem. 1974; 44: 537-560Crossref PubMed Scopus (150) Google Scholar). The hydrolysis of nucleoside diphosphate is thus physiologically important for cellular glycosylation to proceed continuously. In fact, disruption of the GDA1 gene, which codes for guanosine diphosphatase in the Golgi apparatus of S. cerevisiae, results in reduced entry of GDP-Man into the Golgi apparatus and defective mannosylation of lipids and proteins in vivo (11Abeijon C. Yanagisawa K. Mandon E.C. Hausler A. Moremen K. Hirschberg C.B. Robbins P.W. J. Cell Biol. 1993; 122: 307-323Crossref PubMed Scopus (162) Google Scholar). This strict requirement for nucleoside diphosphatase may not be laid if the second type of nucleotide sugar transporter, such as dFRC, which exchanges nucleoside diphosphate with nucleotide sugar, is involved in supplying the substrate for glycosyltransferases. The possibility that there is a compartment in which glycosylation proceeds without nucleoside diphosphatase was suggested in the case of Schizosaccharomyces pombe (31D'Alessio C. Trombetta E.S. Parodi A.J. J. Biol. Chem. 2003; 278: 22379-22387Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar). In such a compartment and with this type of NST, the ratio between cytoplasmic and luminal nucleoside diphosphate concentrations would be an important regulatory factor of glycosylation. The intracellular localizations of FRC, SFL (Sulfateless), and RHO (Rhomboid) proteins do not overlap exactly in Drosophila cells (32Yano H. Yamamoto-Hino M. Abe M. Kuwahara R. Haraguchi S. Kusaka I. Awano W. Kinoshita-Toyoda A. Toyoda H. Goto S. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 13467-13472Crossref PubMed Scopus (93) Google Scholar), suggesting that dFRC may be confined in a specialized compartment. It is intriguing that the synthesis of Fringe oligosaccharide might be controlled by the cellular UDP level. The third type of NST, nucleotide sugar/nucleotide sugar antiporters such as hUGTrel7, uses neither nucleoside monophosphate nor nucleoside diphosphate as a countersubstrate. In this type of NST, the transport activity of one nucleotide sugar substrate is affected by the luminal concentration of other nucleotide sugar substrates. For instance, UDP-GlcUA cannot enter the ER lumen via hUGTrel7 in the absence of luminal UDP-GlcNAc or UDP-GalNAc because of the lack of antiport substrates. In other words, functioning of this type of NST must be coupled with the activity of the first or second type of NST that shares nucleotide sugar substrates with it, as noted by Bossuyt and Blanckaert (19Bossuyt X. Blanckaert N. Biochem. J. 1995; 305: 321-328Crossref PubMed Scopus (47) Google Scholar). Consequently, the activity of the third type of NST would be regulated through the cellular level of the nucleotide sugars shared by these different types of NSTs. This suggests the possibility that the activity of cellular processes affecting the cellular level of "the regulatory (shared) nucleotide sugar" may affect reactions, in addition to those using the regulatory nucleotide sugar directly, that take place in the ER or Golgi lumen using distinct nucleotide sugars that are transported by the third type of NST. The possibility that nucleotide sugar concentrations in the ER and Golgi compartments may be differently regulated by nucleotides and nucleotide sugars depending on the antiport properties of nucleotide sugar transporters and that glycoconjugate biosynthesis may be affected at least partly by this mechanism has not been seriously considered so far. It should be noted that the formulation of this new concept awaited the recognition that NSTs are variable with respect to their antiport substrate specificity, which was demonstrated definitively for the first time in this study. We thank Dr. Hideki Yashiroda (Tokyo Metropolitan Institute of Medical Science) for helpful advice. Download .pdf (.54 MB) Help with pdf files