Skip Navigation


Glycobiology Advance Access originally published online on August 9, 2006
Glycobiology 2006 16(12):1194-1206; doi:10.1093/glycob/cwl035
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
16/12/1194    most recent
cwl035v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Sato, T.
Right arrow Articles by Narimatsu, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sato, T.
Right arrow Articles by Narimatsu, H.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org

Molecular cloning and characterization of a novel human ß1,3-glucosyltransferase, which is localized at the endoplasmic reticulum and glucosylates O-linked fucosylglycan on thrombospondin type 1 repeat domain

Takashi Sato1,3, Maiko Sato1,3, Katsue Kiyohara3, Maki Sogabe3, Toshihide Shikanai3,5, Norihiro Kikuchi3,5, Akira Togayachi3, Hiroyasu Ishida3, Hiromi Ito3, Akihiko Kameyama3, Masanori Gotoh3,4 and Hisashi Narimatsu2,3

3 Glycogene Function Team of Research Center for Glycoscience (RCG), National Institute of Advanced Industrial Science and Technology (AIST), and
4 GlycoGene, Inc., Open Space Laboratory Central-2, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan; and
5 Mitsui Knowledge Industry Co., Ltd., Honcho 1-Chome, Nakano-ku, Tokyo 164-8721, Japan


2 To whom correspondence should be addressed; e-mail: h.narimatsu{at}aist.go.jp

Received on May 11, 2006; revised on July 28, 2006; accepted on August 2, 2006


    Abstract
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
Protein O-linked fucosylation is an unusual glycosylation associated with many important biological functions such as Notch signaling. Two fucosylation pathways synthesizing O-fucosylglycans have been reported on cystein-knotted proteins, that is, on epidermal growth factor-like (EGF-like) domains and on thrombospondin Type 1 repeat (TSR) domains. We report here the molecular cloning and characterization of a novel ß1,3-glucosyltransferase (ß3Glc-T) that synthesizes a Glcß1,3Fuc{alpha}- structure on the TSR domain. We found a novel glycosyltransferase gene with ß1,3-glycosyltransferase (ß3GT) motifs in databases. The recombinant enzyme expressed in human embryonic kidney 293T (HEK293T) cells exhibited glucosyltransferase activity toward fucose-{alpha}-para-nitrophenyl (Fuc{alpha}-pNp). Thin-layer chromatography (TLC) analysis revealed that the product of the recombinant enzyme migrated to the same position as did the product of endogenous ß3Glc-T of Chinese hamster ovary (CHO) cells. The two products could be digested by ß-glucosidase from almond and by exo-1,3-ß-glucanase from Trichoderma sp. These results strongly suggested that the product has the structure of Glcß1-3Fuc. Therefore, we named this novel enzyme ß3Glc-T. Immunostaining revealed that FLAG-tagged ß3Glc-T is an enzyme residing in the endoplasmic reticulum (ER) via retention signal, "REEL," which is a KDEL-like sequence, at the C-terminus. The TSR domain expressed in Escherichia coli was first fucosylated by the recombinant protein O-fucosyltransferase 2 (POFUT2), after which it became an acceptor substrate for the recombinant ß3Glc-T, which could apparently transfer Glc to the fucosylated TSR domain. Our results suggest that a novel glycosyltransferase, ß3Glc-T, contributes to the elongation of O-fucosylglycan and that this occurs specifically on TSR domains.

Key words: glucosyltransferase / glycosyltransferase / O-fucose / thrombospondin / TSR


    Introduction
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
Protein O-fucosylation is known to be one of the most important carbohydrate modifications, and it regulates many biological functions, for example, the Notch signaling process (Bruckner et al., 2000Go; Moloney, Panin et al., 2000Go). So far, two guanine diphosphate-fucose (GDP-Fuc): protein O-fucosyltransferases (POFUTs) that catalyze protein O-fucosylation, named POFUT1 and POFUT2, have been identified (Wang et al., 2001Go; Luo, Nita-Lazar et al., 2006aGo). Each POFUT recognizes a specific domain structure of the protein that serves as a substrate. POFUT1 transfers a Fuc to Ser/Thr in the consensus sequence CXXGGS/TC of epidermal growth factor (EGF)-like repeat domains, which are found in several proteins (Harris and Spellman, 1993Go). O-Fucosylglycan structures found on the EGF domains of tissue-type plasminogen activator (Harris et al., 1991Go), factor VII (Bjoern et al., 1991Go), and factor XII (Harris et al., 1992Go) consist of monosaccharide, that is, Fuc is attached to the Ser/Thr residue in the consensus sequence, forming Fuc{alpha}1-Ser/Thr. By contrast, tetrasaccharide O-fucosylglycan, Sia{alpha}2,3/6Galß1,4GlcNAcß1,3Fuc{alpha}1-Ser/Thr, was found on the EGF domains of factor IX (Nishimura et al., 1992Go; Harris et al., 1993Go) and of Notch (Moloney, Shair et al., 2000aGo). Recent studies in Drosophila melanogaster have demonstrated the important roles of O-fucosylglycans in Notch signaling. Notch is a cell surface receptor containing 36 tandem EGF repeats in its extracellular domain, and it plays an essential role in numerous developmental events (Wharton et al., 1985Go; Mumm and Kopan, 2000Go). The addition of Fuc to Ser/Thr on the EGF domain of Notch is catalyzed by POFUT1 (Wang et al., 1996Go, 2001Go). Fringe was discovered to be the gene responsible for a mutant phenotype of D. melanogaster showing abnormal dorsal/ventral boundary formation in the wing. Biochemical investigations subsequently identified Fringe as a glycosyltransferase that transfers GlcNAc to the Fuc residue on the O-fucosylated EGF domain of Notch through a ß1-3 linkage. Thus, O-fucosylation of EGF domains by POFUT1 and the addition of GlcNAc to the O-Fuc by Fringe, which is a ß1,3-N-acetylglucosaminyltransferase (ß3GlcNAc-T), are essential for Notch activation through its ligands Delta and Serrate/Jagged (Irvine, 1999Go).

In addition to the GlcNAcß1,3Fuc{alpha}1-Ser/Thr structure on EGF domains, another unique disaccharide structure of O-fucosylglycan, Glcß1,3Fuc{alpha}1-Ser/Thr, was first detected in human urine (Hallgren et al., 1975Go; Klinger et al., 1981Go). Hofsteenge and coworkers first demonstrated that the thrombospondin 1 (TSP1) protein is modified with the disaccharide structure, Glcß1-3Fuc{alpha}-Ser/Thr. This modification also occurred in a consensus sequence of CSXS/TCG in three TSP Type 1 repeat (TSR) domains of human TSP1 (Hofsteenge et al., 2001Go). The human proteins propardin and F-spondin were also reported to contain the disaccharide structure Glcß1-3Fuc{alpha}-Ser/Thr on their TSR domains (Gonzalez de Peredo et al., 2002Go). The TSP1 literature has been extensively reviewed (Chen et al., 2000Go). TSP1 has been found in platelets, the extracellular matrix, and other tissues. It participates in cellular responses to growth factors, cytokines, and injury and is involved in multiple physiological and pathological events such as wound healing, inflammation, angiogenesis, and neoplasia. It consists of multiple domains, that is, N- and C-terminal globular domains, a procollagen homologous domain, and three types of repeated sequence motifs, Type 1, Type 2, and Type 3 (Lawler and Hynes, 1986Go; Bornstein, 2001Go). The Glcß1-3Fuc{alpha}-Ser/Thr structure was demonstrated to be in the CSXS/TCG consensus sequence within the Type 1 repeat (Hofsteenge et al., 2001Go; Luo, Nita-Lazar et al., 2006aGo). The TSR domain can bind to integrins, the integrin-associated protein (IAP/CD47), CD36, and proteoglycans (Chen et al., 2000Go). In particular, a 20 amino acid peptide containing the motif CSXS/TCG showed inhibitory activity against angiogenesis induced by VEGF and FGF-2 (Chen et al., 2000Go). However, there have been no reports of the contribution of O-fucosylglycan to their binding activities.

The addition of Fuc to Ser/Thr in the CSXS/TCG consensus sequence of the TSR domain is catalyzed by POFUT2, which is homologous to POFUT1. POFUT1 and POFUT2 seem to recognize the tertiary domain structure of specific sequences, and each has a rigid substrate specificity. POFUT1 transfers Fuc to the EGF domain but not to the TSR domain. By contrast, POFUT2 transfers it to the TSR domain but not to the EGF domain (Luo, Nita-Lazar et al., 2006Go). Regarding the elongation of a fucosylglycan on the TSR domain, Moloney and Haltiwanger (1999)Go identified the enzymatic activity transferring Glc to Fuc{alpha}-para-nitrophenyl (Fuc{alpha}-pNp) with a ß1,3 linkage in Chinese hamster ovary (CHO) cells. They detected the same activity in various tissues and cultured cells despite the limited distribution of Glcß1,3Fuc{alpha}-Ser/Thr structures. Very recently, Luo and others demonstrated that the ß1,3-glucosyltransferase (ß3Glc-T) activity detected in CHO cells can synthesize the disaccharide structure, Glcß1,3Fuc{alpha}-Ser/Thr, on the TSR domain but not on the EGF domain (Luo, Koles et al., 2006Go; Luo, Nita-Lazar et al., 2006Go). Although many features of this unique enzyme have been documented, a gene encoding this enzyme has not been identified yet in any species.

Glycosyltransferase genes may be classified into several families based on the conserved motif sequences that lead to their catalyzing linkages between donor and acceptor substrates, that is, ß1,3 and ß1,4 linkages. Recently, we published a paper describing the molecular cloning and characterization of two major families, that is, ß1,3-glycosyltransferase (ß3GT) and ß1,4-glycosyltransferase (ß4GT) family members, which transfer monosaccharides via ß1,3 and ß1,4 linkages, respectively. The members of the ß3GT family have three consensus amino acid stretches designated as ß3GT motifs in their catalytic domain (Iwai et al., 2002Go; Hiruma et al., 2004Go; Ishida et al., 2005Go), whereas the members of the ß4GT family have one consensus domain designated as the ß4GT motif (Sato et al., 2003Go; Gotoh et al., 2004Go). To date, 14 members of human ß3GT family, including five ß1,3-galactosyltransferases (ß3Gal-Ts) (Miyazaki et al., 1997Go; Amado et al., 1998Go; Kolbinger et al., 1998Go; Isshiki et al., 1999Go; Bai et al., 2001Go), seven ß3GlcNAc-Ts (Zhou et al., 1999Go; Shiraishi et al., 2001Go; Togayachi et al., 2001Go; Iwai et al., 2002Go; Seko and Yamashita, 2004Go; Ishida et al., 2005Go), and two ß1,3-N-acetylgalactosaminyltransferases (ß3GalNAc-T) (Okajima et al., 2000Go; Hiruma et al., 2004Go), have been identified. Another seven glycosyltransferase genes, including three mammalian Fringes (Lunatic Fringe [Lfng], Manic Fringe [Mfng], and Radical Fringe [Rfng]) (Rampal et al., 2005Go), the core 1-galactosyltransferase-1 (C1Gal-T1) (Ju et al., 2002Go), and three chondroitin synthases (ChSy/CSS1, CSS2, and CSS3) (Kitagawa et al., 2001Go; Yada, Gotoh et al., 2003Go; Yada, Sato et al., 2003Go), exhibit glycosyltransferase activities with a ß1,3 linkage and also have ß3GT motifs with weaker strictness. Thus, the ß3GT family currently consists of 21 reported members.

In this study, we have identified the 22nd member of the ß3GT family, an enzyme that exhibits a novel glycosyltransferase activity by transferring Glc to Fuc with a ß1,3 linkage. This is the first report of molecular cloning and characterization of a ß3Glc-T. The enzyme is localized in the endoplasmic reticulum (ER), and it extends the O-linked fucosylglycans on TSR domains.


    Results
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
Determination of the nucleotide sequence of ß3Glc-T cDNA and its putative amino acid sequence
We determined the full-length cDNA sequence by the 5' rapid amplification for complementary DNA (cDNA) ends (5'-RACE) method and registered it in the GenBank database with accession number AB101481 [GenBank] . The gene is located at 13q12.3 and is composed of at least 15 exons. The open-reading frame (ORF) consists of 1479 bp encoding a predicted 498 amino acid protein with a hydrophobic stretch at the N-terminus, and there are two possible N-glycosylation sites.

As shown in the phylogenetic tree of the ß3GT family (Figure 1), the novel enzyme ß3Glc-T was categorized into a minor cluster consisting of Rfng, C1Gal-T1, and ChSy/CSS1. On the contrary, many of the ß3Gal-Ts, including ß3GalNAc-Ts and ß3Gn-Ts, comprise a major cluster (Ishida et al., 2005Go). In the minor cluster, the amino acid sequence of ß3Glc-T had weak homology with that of Rfng, which is ß3GlcNAc-T synthesizing GlcNAcß1,3Fuc{alpha}-Ser/Thr (Rampal et al., 2005Go); C1Gal-T1, which is ß3Gal-T synthesizing Galß1,3GalNAc (Ju et al., 2002Go); and ChSy/CSS1, which is ß3GlcA-T synthesizing GlcAß1,3GalNAc (Kitagawa et al., 2001Go). These proteins had 27, 26, and 23% local identities, respectively, with ß3Glc-T. The amino acid sequence of the catalytic domain of ß3Glc-T was compared with the sequences of Rfng, C1Gal-T1, and ChSy/CSS1 (Figure 2). We have reported that members of the ß3GT family exhibiting glycosyltransferase activity and catalyzing a ß1,3 linkage possess three conserved motifs, termed ß3GT motifs (Ishida et al., 2005Go). ß3Glc-T contained three ß3GT motifs having a weak similarity to the others, and a DDD sequence within the second motif is thought to participate in divalent cation binding. During the preparation of this manuscript, Heinonen and others (2003)Go published same nucleotide sequence that was described as "ß3-glycosyltransferase-like"; however, they did not mention any glycosyltransferase activities. We determined that this enzyme exhibits a novel glycosyltransferase activity, transferring Glc to Fuc with a ß1,3 linkage, and we have designated it as ß3Glc-T.


Figure 1
View larger version (36K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 1. A phylogenetic tree of human ß3GTs and their substrate specificities. The phylogenetic tree of human ß3GTs was constructed with ClustalW based on the amino acid sequences (left). ß3Glc-T was shown by boldface type. The branch length indicates the evolutionary distance. Reaction products of each enzyme are shown on the right.

 

Figure 2
View larger version (40K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 2. Multiple alignment of amino acid sequences neighboring the ß3GT motifs of human Rfng, C1Gal-T1, ChSy/CSS1, and ß3Glc-T. Introduced gaps are shown with hyphens. The three ß3GT motifs are boxed. Identical and similar amino acids are shown with asterisks and dots, respectively.

 
Additional genes possessing a sequence similar to human ß3Glc-T were found in the GenBank database for mouse (partial), Drosophila meganogaster (accession number AE003612), and Caenorhabditis elegans (accession number NP_504520 [GenBank] ). These genes are probably orthologous to human ß3Glc-T and contain three ß3GT motifs and a KDEL-like sequence at the C-terminus, which is thought to be a signal for retention in the ER. The mouse gene mß3Glc-T was also subcloned using a cDNA library derived from the brain of ICR mouse, and its full-length cDNA sequence was determined (Figure 3). It encoded a hypothetical 489 amino acid protein carrying a shorter hydrophobic domain of 15 amino acids at the N-terminus.


Figure 3
View larger version (74K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 3. Alignment of amino acid sequences of human and mouse ß3Glc-Ts by ClustalW. Introduced gaps are shown with hyphens. The putative signal sequences are underlined. The three ß3GT motifs are boxed. The ER retention signal-like sequences are written in boldface type. Identical and similar amino acids are shown with asterisks and dots, respectively.

 
Substrate specificity of ß3Glc-T
We determined substrate specificity of the truncated, soluble ß3Glc-T that was expressed in human embryonic kidney 293T (HEK293T) cells. The hydrophobic region at the N-terminus of ß3Glc-T was replaced by a FLAG tag, and the soluble form was collected from the conditioned medium. ß3Glc-T was detected as a single band corresponding to the predicted size by both Coomassie staining (Figure 4A, lane 2) and western blotting using an anti-FLAG M2 antibody (lane 4). Utilizing a variety of radioisotope-labeled uridine diphosphate (UDP) donors and monosaccharide acceptors with a pNp or benzyl (Bz) group, donor and acceptor substrates for ß3Glc-T were screened. As summarized in Table I, ß3Glc-T exhibited a glycosyltransferase activity only when UDP-14Glc was used as a donor substrate and Fuc{alpha}-pNp was used as an acceptor substrate. Fucosyloligosaccharides, that is, H-antigen Types 1 and 2, Galß1-3(Fuc{alpha}1-4)GlcNAc (Lea), and Galß1-4(Fuc{alpha}1-3)GlcNAc (Lex), were synthesized enzymatically and used as acceptor substrates. ß3Glc-T showed glucosyltransferase activity toward H-antigen Type 2 and Lea but not toward H-antigen Type 1 and Lex (Table I).


Figure 4
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 4. Determination of ß3Glc-T enzymatic activity and characterization of its reaction products by TLC. (A) The recombinant ß3Glc-T was purified with anti-FLAG M2 agarose from culture medium of HEK293T cells expressing ß3Glc-T and was applied to SDS–PAGE. The detection of proteins was performed with coomassie brilliant blue (CBB) staining (lane 2) and western blotting using an anti-FLAG M2 monoclonal antibody (lane 4). (B) Enzymatic activity was determined by [14C]Glc incorporation of UDP-Glc into Fuc{alpha}-pNP (lane 2). The reaction products were subjected to ß1,3-glucanase digestion and separated by TLC. The positions of reaction products are indicated by an arrow.

 

View this table:
[in this window]
[in a new window]

 
Table I. Substrate specificity of ß3Glc-T

 
Characterization of the ß3Glc-T reaction product
Because this novel enzyme, ß3Glc-T, is a member of the ß3GT family by virtue of having three ß3GT motifs in its amino acid sequence, the reaction product was assumed to be Glcß1,3Fuc. To prove this, we compared the reaction product of the recombinant ß3Glc-T with that of homogenates of CHO cells, which had been reported to contain the ß3Glc-T activity endogenously (Moloney and Haltiwanger, 1999Go). As shown in Figure 4B, the radioactive product synthesized by recombinant ß3Glc-T migrated to the same position on a thin-layer chromatography (TLC) plate as did the product resulting from endogenous activity of CHO cell lysates (compare lane 2 with lane 4). The CHO lysates from cells that overexpressed the ß3Glc-T gene synthesized significantly more product that migrated to the same position as that of the product of mock CHO cell lysates (lane 6). Furthermore, all signals disappeared after treatment with exo-ß1,3-glucanase from Trichoderma sp., which releases a terminal Glc bound with a ß1-3 linkage but not with ß1-4 or ß1-6 linkage (lanes 3, 5, and 7). ß-Glucosidase from almonds also digested these products (data not shown). These results strongly suggest that the reaction product of the recombinant human ß3Glc-T is Glcß1,3Fuc.

In vitro fucosylation and glucosylation of the EGF and TSR domains by POFUT1, POFUT2, and ß3Glc-T
The Glcß1,3Fuc{alpha}-Ser/Thr structure was identified on the TSR domain of platelet glycoprotein TSP1 (Hofsteenge et al., 2001Go) but not on the EGF domain, which also contains O-linked fucosylglycans. To determine whether ß3Glc-T recognizes this domain structure specifically, we performed in vitro fucosylation and glucosylation assays using the TSR and EGF domains, which were expressed in Escherichia coli for use as acceptor proteins. The recombinant POFUT1, POFUT2, and ß3Glc-T expressed in HEK293T cells (Figure 5A) showed preferences for particular proteins as acceptor substrates. As shown in Figure 5B, POFUT1 transferred radioactive Fuc to the EGF domain (lane 1) but not to the TSR domain (lane 3). By contrast, POFUT2 transferred Fuc to the TSR domain (lane 4) but not to the EGF domain (lane 2). Both enzymes therefore seem to recognize the domain structure of specific proteins. This is consistent with results from other laboratories that were reported previously (Luo, Nita-Lazar et al., 2006Go). The recombinant ß3Glc-T transferred Glc to the fucosylated TSR domain (lane 6) but not to the fucosylated EGF domain (lane 5). These results demonstrated that ß3Glc-T recognizes a fucosylated protein structure specifically, fucosylated TSR, and transfers a Glc to the fucosylated TSR domain with strict substrate specificity as well as POFUT2 adding the first Fuc to the TSR domain.


Figure 5
View larger version (25K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 5. In vitro fucosylation and glucosylation of the EGF and TSR domains by recombinant POFUT1, POFUT2, and ß3Glc-T. (A) The recombinant enzymes POFUT1, POFUT2, and ß3Glc-T were expressed in HEK293T cells and purified with anti-FLAG M2 agarose. Enzymes were separated by SDS–PAGE and probed by an anti-FLAG M2 monoclonal antibody. (B) In vitro fucosylation assays using the EGF and TSR domains and GDP-[14C]Fuc as substrates were performed for POFUT1 and POFUT2. In the in vitro glucosylation assay, fucosylation was first performed by POFUT1 for the EGF domain and POFUT2 for the TSR domain with nonradiolabeled GDP-Fuc, and then glucosylation was catalyzed by ß3Glc-T with UDP-[14C]Glc. Reaction mixtures were separated by SDS–PAGE, and gels were then subjected to autoradiography (above panels). Detection was by western blotting using an anti-6 x His antibody (below panels).

 
Intracellular localization of ß3Glc-T
The mouse ortholog of ß3Glc-T (mß3Glc-T) exhibited the same enzyme activity, synthesizing Glcß1-3Fuc (data not shown). Human ß3Glc-T and mß3Glc-T have hydrophobic domains of 20 and 15 amino acids, respectively, at their N-termini (Figure 6A). This stretch of 15 amino acids is too short to be a transmembrane domain of a Golgi-retention glycosyltransferase with a Type II topology. In addition, both human and mouse ß3Glc-Ts had a KDEL-like sequence, REEL, at their C-termini. KDEL is a consensus sequence for ER retention. To determine whether ß3Glc-T was a secreted protein with a signal sequence, we inserted a FLAG tag just behind the predicted cleavage site for the signal sequence in human ß3Glc-T (pcDNA3.1 neo/ß3Glc-T-FLAG in Figure 6A). In addition, we constructed ß3Glc-T with a FLAG tag that did not contain the predicted ER retention signal, REEL, at the C-terminus (pcDNA3.1 neo/ß3Glc-T-FLAG {Delta} REEL in Figure 6A). These constructs were expressed in COS1 cells and were immunoprecipitated by an anti-FLAG antibody from both cell lysates and conditioned medium. As seen in Figure 6B, the recombinant ß3Glc-T having the REEL sequence was recovered equally from both the cell lysates and the conditioned medium. By contrast, the recombinant enzyme lacking the REEL sequence was almost totally recovered from conditioned medium and only slightly from cell lysates. These results strongly indicate that ß3Glc-T is an ER retention enzyme with a retention signal, REEL, and that it is secreted as a soluble protein.


Figure 6
View larger version (24K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 6. ß3Glc-T is a secreted enzyme and is retained in the ER via its ER retention signal, REEL. (A) Comparison of the N-terminal domains of human and mouse ß3Glc-T. A FLAG tag was introduced just behind the signal sequences predicted by SignalP 3.0 of human ß3Glc-T. The diagram shows the expression constructs of ß3Glc-T with or without the REEL sequence. (B) ß3Glc-T with or without the REEL sequence was transfected into COS1 cells and the protein was immunoprecipitated from cells and conditioned medium using anti-FLAG M2 agarose. Precipitants from the cells (left panel) and medium (right panel) were separated by SDS–PAGE and probed by an anti-FLAG M2 monoclonal antibody. (C) Immunostaining was performed for COS1 cells expressing ß3Glc-T with or without the REEL sequence. In the top and bottom panels, the anti-FLAG fluorescein isothiocyanate (FITC) conjugate (green) was used for costaining with anti-calreticulin/TXRD (red) as an ER marker. In the middle panels, anti-FLAG/TXRD (red) was used with anti-GM130/Alexa488 (green) as a Golgi marker. Scale bars mean 20 µm.

 
Immunostaining of COS1 transfectants revealed that ß3Glc-T colocalized at the ER with calreticulin, which is a typical ER marker, but not with GM130, a typical Golgi marker. Deletion of the REEL sequence resulted in almost complete disappearance of staining or in abnormal localization to regions besides the ER and Golgi (Figure 6C). The amounts of ß3Glc-T transcripts in these transfectants were quantitated by the real-time polymerase chain reaction (PCR) method, which indicated their levels were almost equal (data not shown).

Quantitative analysis of the ß3Glc-T transcripts in human tissues by real-time PCR
We determined the tissue distribution and expression levels of the human ß3Glc-T transcript by the real-time PCR method. The expression levels of ß3Glc-T in various tissues are shown relative to amounts of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcript (Figure 7). The ß3Glc-T transcript was ubiquitously expressed in many tissues, with testis and uterus showing relatively higher levels.


Figure 7
View larger version (22K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 7. Quantitative real-time PCR analysis of the ß3Glc-T transcript in human tissues. Standard curves for ß3Glc-T and GAPDH were generated by serial dilution of each plasmid DNA. The expression level of the ß3Glc-T transcript was normalized to that of the GAPDH transcript, which was measured in the same cDNAs from human tissues. Data were obtained from triplicate experiments and are indicated as the mean ± SD.

 

    Discussion
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
CHO cells have a disaccharide structure, Glcß1,3Fuc, and the ß3Glc-T activity responsible for the synthesis of this disaccharide structure has been detected in CHO cells (Moloney et al., 1997Go; Moloney and Haltiwanger, 1999Go). However, the gene encoding this enzyme has not been identified. In this study, we found a novel human glycosyltransferase having the ß3GT motif. It exhibited Glc-T activity toward Fuc and synthesized the Glcß1,3Fuc structure. This is the first report of the molecular cloning of ß3Glc-T and mß3Glc-T and of in vitro synthesis of Glcß1,3Fuc{alpha}Ser/Thr on a TSR domain using recombinant enzymes.

The comparison of ß3Glc-T with other members of the ß3GT family shows that Rfng, which has a ß3GlcNAc-T activity toward O-linked fucosylglycan on the EGF domain, is the closest family member by phylogenetic tree analysis (Figure 1); however, the similarity of the amino acid sequences of two human enzymes seems to be very low except for the ß3GT motifs (Figure 2). The ortholog analyses in the databases revealed that probable orthologous genes of ß3Glc-T are found in D. melanogaster and C. elegans. However, an ortholog of Rfng was found in D. melanogaster but not in C. elegans. O-Fucosylglycans may exist in C. elegans, because two genes that are probably orthologous to the human POFUT1 and POFUT2 genes, respectively, are found in C. elegans genome. On the basis of these findings, we presume that during the evolution of O-fucosylglycans, Rfng may have diverged from ß3Glc-T in D. melanogaster.

On screening with monosaccharide acceptors and UDP sugar donors, ß3Glc-T showed Glc-T activity toward Fuc{alpha}-pNp (Table I). In most fucosylated glycan structures, Fuc is positioned at the nonreducing terminus, except in the case of O-linked fucosylglycans. An in vitro Glc-T assay summarized in Table I indicated that ß3Glc-T showed significant Glc transfer activity toward H-antigen Type 2 and Lea; however, such glucosylated H-antigen Type 2 and Lea structures have never been reported. We consider that these glucosylated products produced by in vitro synthesis are not physiological, because ß3Glc-T is localized in the ER (Figure 6), whereas structures such as H-antigen Type 2 and Lea are synthesized in the Golgi apparatus. ß3Glc-T did not utilize H-antigen Type 1 and Lex as acceptor substrates (Table I). The distance between Fuc and the N-acetyl residue of GlcNAc in Lex has been chemically determined to be shorter than that in Lea. This may be one reason that ß3Glc-T cannot recognize Fuc of Lex as an acceptor.

ß3Glc-T has a hydrophobic domain at its N-terminus. At first, we thought this might be a transmembrane domain and that ß3Glc-T would therefore be a typical Type II membrane protein, like many other glycosyltransferases, residing in the Golgi apparatus. However, we found that the hydrophobic domain of the mouse ortholog consisted of only 15 amino acids, which is too short for a transmembrane domain to be retained in the Golgi membrane. In fact, mammalian glycosyltransferases retained in the Golgi apparatus possess a hydrophobic domain consisting of about 20 amino acids on average. Software to predict signal peptides, SignalP 3.0 (Bendtsen et al., 2004Go), predicted that the hydrophobic sequences of human and mouse ß3Glc-Ts are signal sequences not transmembrane domains for Golgi retention. Furthermore, ß3Glc-T also had a C-terminal sequence, REEL, that was similar to an ER retention signal. The deletion of this signal resulted in increased secretion of the recombinant protein into the medium while it almost disappeared from the cells (Figure 6). These results suggest that ß3Glc-T is a soluble enzyme secreted into the ER lumen that is then tethered with some ER components via its REEL sequence. Figure 6 shows that the wild-type ß3Glc-T, although mostly tethered in the cells, could also be secreted into the medium, although this might have occurred because of overexpression that exceeded the limited retention capacity of the ER.

Two O-linked fucosylation pathways have been reported. The product of the first pathway is a tetrasaccharide structure on an EGF repeat, SA{alpha}2,3/6Galß1,4GlcNAcß1,3Fuc{alpha}-Ser/Thr, which participates in Notch signaling. This structure requires four enzymes, that is, POFUT1, Fringe, ß4Gal-T, and {alpha}2–3/6Sia-T, for its synthesis. POFUT1 is a unique enzyme that is secreted as a soluble protein and then is retained at the ER membrane, where it exhibits molecular chaperone activity by helping to fold EGF repeats, in addition to its FUT activity (Luo and Haltiwanger, 2005Go; Okajima et al., 2005Go). In Drosophila, Fringe (Dfng) is reported to be a Golgi retention enzyme having ß3GlcNAc-T activity toward the O-fucosylated EGF domain (Munro and Freeman, 2000Go). No ER retention signals resembling the KDEL sequence were found in Drosophila or mammalian Fringes (Dfng or Lfng, Mfng, and Rfng). The product of the second O-linked fucosylation pathway is a very unique disaccharide structure, Glcß1,3Fuc{alpha}-Ser/Thr, which has been found in human urine and in the TSR domain of the human platelet glycoprotein TSP1. Shao and Haltiwanger (2003)Go first reported that the specific protein O-fucosylation of the TSR domain is catalyzed by another POFUT, POFUT2. In this study, we demonstrated that ß3Glc-T, which synthesizes Glcß1,3Fuc{alpha}-Ser/Thr on the TSR domain, is localized in the ER. We therefore presume that POFUT2 also localizes to the ER, where it fucosylates the TSR domain in a similar manner to POFUT1 (Luo and Haltiwanger, 2005Go). Although POFUT2 has not confirmed to be an ER retention enzyme, it must be localized in the ER, because it functions before the activity of ß3Glc-T. In comparing the amino acid sequences of POFUT2 and POFUT1, it appears that POFUT2 also contains a signal peptide at its N-terminus. However, there is no apparent ER retention signal at the C-terminus of POFUT2. Very recently, Luo, Koles, and others (2006)Go reported that Drosophila POFUT2 is predominantly localized in the ER in Drosophila S2 cells. In the initial steps of the two pathways, the first fucosylation enzymes, that is, POFUT1 and POFUT2, may be localized in the ER. However, the second enzymes that catalyze extensions of the O-fucosylglycan, that is, Fringe and ß3Glc-T, differ in their localization, occurring at the Golgi and ER, respectively. In Notch signaling, GlcNAc modification by Fringe is critical for selection of the Notch ligands, Delta/Serrate/Lag2 proteins (Okajima et al., 2003Go; Xu et al., 2005Go; Yang et al., 2005Go). Thus, the expression of Fringe results in important regulation of Notch signaling. The different intracellular localizations of POFUT1 and Fringe may be important for independent regulation of these genes. In fact, structural analysis of fucosylglycans indicated that one structure is a fucosylated EGF domain that is not extended by Fringe, whereas the other is extended by the addition of GlcNAc, catalyzed by Fringe, and both are present in the cells. On the contrary, both POFUT2 and ß3Glc-T seem to be colocalized in the ER, where they synthesize the Glcß1,3Fuc{alpha}-Ser/Thr structure cooperatively, because the endogenous TSR domain in platelet TSP1 was fully modified by the addition of a disaccharide structure (Hofsteenge et al., 2001Go). All of the fucosylglycan on the TSR domains analyzed in mammalian cells to date possessed the disaccharide structure, Glcß1,3Fuc{alpha}-Ser/Thr (Hofsteenge et al., 2001Go; Gonzalez de Peredo et al., 2002Go; Luo, Nita-Lazar et al., 2006Go). This indicates that both enzymes are always regulated together in the cells. Taken together, these data indicate that the disaccharide structure, Glcß1,3Fuc{alpha}-Ser/Thr, on the TSR domain is essential to the function of TSP1. The gene knockdown of POFUT2 in C. elegans resulted in abnormal development (Menzel et al., 2004Go). The gene knockdown of ß3Glc-T may also show phenotypes similar to the POFUT2 mutant in development.

The TSR domain of TSP1 has been reported to be a multifunctional domain interacting with several ligands and to regulate several cellular responses (Chen et al., 2000Go; Bornstein, 2001Go). Some interactions of TSR domains with other proteins have been elucidated by inhibition assays using synthetic peptides incorporating the consensus sequence for O-fucosylation, CSXS/TCG (Hofsteenge et al., 2001Go). However, these assays to block the TSR domain function were performed with nonglycosylated peptides. Further experiments using glycopeptides having the Glcß1,3Fuc{alpha}-Ser/Thr structure may show more relevant inhibition results.

The consensus sequence preferred by POFUT2, CSXS/TCG, was found in more 70 human proteins other than TSP1 in protein databases. These included TSP2, the A disintegrin and metalloproteinase with TSP motifs (ADAMTS) family, and complement factors including C8, C9, properdin, F-spondin, semaphorin 5A, brain-specific angiogenesis inhibitor (BAI) 1, and so on. However, structural analyses to detect the presence of O-fucosylation have been performed on very few proteins (Hofsteenge et al., 2001Go; Gonzalez de Peredo et al., 2002Go). Future studies of protein O-fucosylation on TSR domains within these proteins and the detection of Glc elongation by ß3Glc-T will likely reveal novel biological functions of O-fucosylglycans.


    Materials and methods
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
Isolation of human ß3Glc-T cDNA
We performed a Blast search of the GenBank database using ß3GT motifs as query sequences and identified an expressed sequence tag sequence having the accession number XM_085042, which was predicted by GENSCAN software to contain a partial ORF. To obtain the complete ORF, we employed the 5'-RACE method using a Marathon-Ready cDNA Amplification Kit (Clontech, Palo Alto, CA). The sequences of the DNA fragments obtained by the 5'-RACE method were determined using a DYEnamic ET Terminator Cycle Sequencing Kit (Amersham Biosciences, Amersham, UK). Finally, a cDNA sequence encoding the full-length ORF was obtained by PCR using the Marathon-Ready cDNA of human brain (Clontech) as a template.

Construction and expression of the ß3Glc-T protein with FLAG peptide
The putative catalytic domain of human ß3Glc-T (amino acids 29–498) was expressed as a secreted protein fused with a FLAG peptide in HEK293T cells. An ~1.4 kb DNA fragment was amplified by PCR using the Marathon-Ready cDNA derived from human brain as a template and using two primers, 5'-GGAATTCTGAAGATACAAAGAAAGAGGT-3' and 5'-GCTCTAGAGTGATTTATAACTCCTCTCGAAAA-3'. The amplified fragment was digested with the restriction endonucleases EcoRI and XbaI and then inserted into the pFLAG-CMV1 vector (Sigma, St. Louis, MO) by a two-step ligation, because that fragment contained two EcoRI sites. The resulting plasmid, designated pFLAG-CMV1-ß3Glc-T, was transfected into HEK293T cells using LIPOFECTAMINE 2000 (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. A 50 mL volume of culture medium was mixed with anti-FLAG M2 agarose affinity gel (Sigma) and rotated slowly at 4°C overnight. The gel was then washed twice with 50 mM Tris-buffered saline (TBS; 50 mM Tris–HCl, pH 7.4, and 150 mM NaCl) and suspended in 100 µL of the assay buffer described below.

Construction and expression of full-length ß3Glc-T and preparation of cell lysates
For the expression of full-length ß3Glc-T with a transmembrane domain, the PCR product amplified using the Marathon-Ready cDNA derived from human brain as a template for two primers, 5'-GGAATTCAGGATGCGGCCGCCCGCCTGCT–3' and 5'-GCTCTAGAGTGATTTATAACTCCTCTCGAAAA-3', was inserted into pcDNA3.1(+) (Invitrogen) as well as above after digestion with the endonucleases EcoRI and XbaI. The resulting plasmid, designated pcDNA3.1-ß3Glc-T, was transfected into CHO cells, and the transfected cells were selected in the presence of geneticin (0.6 mg/mL) (Invitrogen). After 3 weeks of exposure to geneticin, the cells stably expressing ß3Glc-T were subjected to enzymatic assays. The cells (~106) that received ß3Glc-T transfectant or mock transfectant were harvested by trypsin–ethylenediaminetetraacetic acid (EDTA) (Invitrogen) treatment and suspended in 500 µL of lysis buffer containing 1% Nonidet P-40 and a protease inhibitor cocktail (Sigma) in TBS (10 mM Tris–HCl, pH 7.5, and 0.15 M NaCl) after washing twice in TBS. After 15 min on ice, the cell suspensions were homogenized by sonication for 10 min on ice, and the supernatant was then centrifuged at 14,000 x g for 10 min at 4°C and used as an enzyme source.

Construction of ß3Glc-T without the predicted ER retention signal REEL at the C-terminus and of ß3Glc-T with a FLAG tag between the signal peptide and the catalytic domain
The version of ß3Glc-T without the four additional amino acids at the C-terminus was amplified by PCR from pcDNA3.1-ß3Glc-T as a template using two primers: 5'-AAGAGGTCCCAAGCTTCCGGGATGCGGCCTCCCGCGCT-3' and 5'-GCTCTAGAGTGATTTAAAAACCTTTCTGTGTCTCCT-3'. The PCR product was digested with BglII and XbaI and then replaced the BglII–XbaI fragment of pcDNA3.1-ß3Glc-T to construct pcDNA3.1-ß3Glc-T {Delta} REEL. The FLAG tag was inserted between the signal peptide predicted by SignalP 3.0 (Bendtsen et al., 2004Go) and the enzyme’s catalytic domain by PCR. Two PCR fragments containing the FLAG tag sequence were amplified by PCR from pcDNA3.1-ß3Glc-T as a template using two primers each: 5'-CGCAAATGGGCGGTAGGCGTG-3' and 5'-CTTGTCGTCATCGTCTTTGTAGTCAGCCAAACCAAAAGCCAGGGAGCA-3', containing the FLAG tag sequence at the 3' end, and 5'-GACTACAAAGACGATGACGACAAGTCTGAAGATACAAAGA-3' and 5'-CTCTTTAGTCTCTTGGTAAGCTT-3', containing the FLAG tag sequence at the 5' end. The two DNA fragments were then annealed using the FLAG tag sequences and were used as a template for PCR amplification. The amplified fragment was digested by HindIII, after which it replaced the HindIII fragment of pcDNA3.1-ß3Glc-T and pcDNA3.1-ß3Glc-T {Delta} REEL, yielding pcDNA3.1-ß3Glc-T-FLAG and pcDNA3.1-ß3Glc-T-FLAG {Delta} REEL, respectively.

Isolation of mouse ß3Glc-T cDNA
The mouse ortholog of ß3Glc-T was amplified by PCR from brain cDNA obtained from ICR mouse as a template using two primers for full-length expression: 5'-CCCAAGCTTCCGGGATGCGGCCTCCCGCGCT-3' and 5'-GCTCTAGATCTGTTCTATAATTCTTCTCTT-3'. For the soluble enzyme region, the primers were 5'-CCCAAGCTTTCTGAAGAGATAAAAGAAAAGGT-3' and 5'-GCTCTAGATCTGTTCTATAATTCTTCTCTT-3'. These fragments were digested with HindIII and XbaI and then inserted into the HindIII and XbaI sites of the expression vectors pcDNA3.1 (Invitrogen) and pFLAG-CMV1 (Sigma), respectively, to construct pcDNA3.1-mß3Glc-T and pFLAG-CMV1-mß3Glc-T.

Assay for glycosyltransferase activity
To determine the enzymatic activity, we utilized UDP-[14C]Glc, UDP-[14C]GlcNAc, UDP-[14C]Gal, UDP-[3H]GalNAc, UDP-[14C]glucuronic acid (GlcA), GDP-[14C]mannose (Man), and GDP-[14C]Fuc (American Radiolabeled Chemicals, St. Louis, MO) as donor substrates. For acceptor substrates, Fuc{alpha}-pNp, Fucß-pNp, Glc{alpha}-pNp, Glcß-pNp, GlcNAc{alpha}-Bz, GlcNAcß-Bz, Gal{alpha}-pNp, Galß-olto-Np (oNp), GalNAc{alpha}-Bz, GalNAcß-Bz, GlcAß-pNp, Man{alpha}-pNp, xylose-{alpha}-pNp (Xyl{alpha}-pNp), and Xylß-pNp were purchased from Calbiochem (La Jolla, CA) and Sigma. H-Antigen Types 1 and 2, Lea, and Lex structures for acceptor substrates were synthesized by sequential enzymatic reactions using GlcNAcß-Bz as a starting substrate, with recombinant human glycosyltransferases (ß3Gal-T5, ß4Gal-T1, FUT1, and FUT3) expressed in HEK293T cells. The products of each enzymatic reaction were separated by reverse phase high-performance liquid chromatography (HPLC) (Shimadzu, Kyoto, Japan) on an ODS-80Ts QA column (Tosoh, Tokyo, Japan) and analyzed by a matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) (Reflex IV, Bruker Daltonics, Billerica, MA). For the reaction in the Glc-T assay, 50 mM HEPES buffer (pH 7.0) containing 0.1 µCi UDP-[14C]Glc, 10 mM MnCl2, and 10 nmol of acceptor substrate was used. A 5 µL volume of enzyme source, which was affinity purified from conditioned medium, for 20 µL of each reaction mixture was added and incubated at 37°C for various periods. After incubation, the radioactive reaction products were separated from the free radioactive UDP-[14C]Glc using a Sep-Pak Plus C18 Cartridge (Waters, Milford, MA). For conditioning, a Sep-Pak Plus C18 Cartridge was washed with 1 mL of 100% methanol and then washed twice with 1 mL of water. The reaction mixtures were centrifuged, and the supernatants were loaded onto the equilibrated cartridge. The radioactive reaction products eluted by 1 mL of methanol after washing twice with 1 mL of water were divided into half volume. Half of the eluate was checked for radioactivity by liquid scintillation spectrophotometry, and the other half was dried using an evaporator and dissolved in an adequate volume of methanol. The dissolved residues were separated on a TLC plate using a solvent system of chloroform/methanol/CaCl2 (65:35:8, v/v/v), and the radioactive intensities of the bands were measured with a FLA-3000 Imaging Analyzer (Fuji Film, Tokyo, Japan).

ß-Glucosidase digestion of reaction products generated by ß3Glc-T
ß-Glucosidase from almonds (Sigma) and exo-1,3-ß-glucanase from Trichoderma sp. (Megazyme, County Wicklow, Ireland) were used to digest the reaction products generated by ß3Glc-T. The radioactive reaction products that were separated using a Sep-Pak Plus C18 Cartridge and then dried were next dissolved in the glucanase reaction buffer containing 50 mM sodium acetate (pH 5.0), 0.1 M NaCl, and 0.5 unit of ß-glucosidase or 0.2 unit of exo-1,3-ß-glucanase. After incubation at 37°C for 8 h, the digested products were subjected to TLC plate analysis after being separated using a Sep-Pak Plus C18 Cartridge.

In vitro fucosylation and glucosylation on the human TSR domain and on the human factor VII EGF-like domain
The third TSR domain in human TSP1 was amplified by PCR using quick-clone cDNA (Clontech) derived from bone marrow as a template and using the following two primers: 5'-GGGGTACCTGACGCCTGCCCCATCAATGGA-3' and 5'-CCGCTCGAGAGGCATCCATCAATTGGACAGT-3'. The human factor VII EGF-like domain was also amplified by PCR from quick-clone cDNA (Clontech) derived from liver as a template using two primers: 5'-GGGGGTACCTAGTGATGGGGACCAGTGTGCCT-3' and 5'-GCTCTAGATCATCCTTGTGCGTCTCACA-3'. Each PCR fragment was subcloned into the KpnI and XbaI sites of pMT/BiP/V5-His (Invitrogen). For bacterial expression, SmaI–BamHI fragments from each subcloned plasmid were ligated into EcoRV and BamHI sites of pET20b(+) to construct pET20b-TSR and pET20b-EGF. These were then transformed into the E. coli BL21 (DE3) strain. Purifications of the TSR domain and of the EGF domain expressed as a periplasmic protein were carried out using BD TALON metal affinity resin (Clontech), according to manufacturer’s instructions. The proteins were eluted from the resin with 250 mM imidazole, 50 mM Tris (pH 8.0), and 0.15 M NaCl and dialyzed against 50 mM Tris (pH 8.0) and 0.15 M NaCl. The amounts of the TSR and EGF domains were estimated by western blotting with a penta-His horseradish peroxidase (HRP) conjugate (Qiagen, Valencia, CA), and the intensities of each band were compared with those of a 6 x His Protein Ladder (Qiagen). Assays were then carried out using the proteins as acceptor substrates for the POFUTs. The cDNAs of POFUT1 and POFUT2 were amplified using two primers—POFUT1, 5'-CACCTCCTGGGACCCGGCCGGTTA-3' and 5'-CTCCGGCCAGAATCAGAACTC-3', and POFUT2, 5'-CACCCAGGAGTTCTGGCCCGGACAA-3' and 5'-TCCTCAGTAGGTGATCTTCCAGT-3'—and were then subcloned into pENTR/D-TOPO (Invitrogen) constructing entry clones for GATEWAY Cloning Technology (Invitrogen). The POFUT1 and POFUT2 genes were transferred from pENTR/D-TOPO to pFLAG-CMV-3-DEST, which was constructed using Gateway Vector Conversion System (Invitrogen) and pFLAG-CMV-3 (Sigma), according to the instruction manuals, and were transfected into HEK293T cells for enzyme expression. Presence of the recombinant POFUT1 and POFUT2 enzymes was confirmed by western blotting with an anti-FLAG monoclonal antibody peroxidase conjugate (Sigma). POFUT activity was assayed in 20 µL of reaction mixture containing 50 mM HEPES (pH 7.5), 10 mM MnCl2, 0.1 µCi of GDP-[14C]Fuc, and 3.0 µg of the TSR or EGF domain. For ß3Glc-T reactions using the TSR or EGF domain as substrate, ß3Glc-T was added to the reaction mixtures with 0.1 µCi UDP-[14C]Glc after POFUT reactions with 50 µM GDP-Fuc. Reaction mixtures were separated by 15–25% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE), and radioactive intensities of the bands were measured with a FLA-3000 Imaging Analyzer.

Intracellular localization of ß3Glc-T in COS1 cells
The COS cells were transfected with pcDNA3.1-ß3Glc-T-FLAG or pcDNA3.1-ß3Glc-T-FLAG {Delta} REEL, and the transfected cells were selected in the presence of geneticin (0.6 mg/mL) (Invitrogen). After 3 weeks of exposure to geneticin, the cells stably expressing ß3Glc-T-FLAG or ß3Glc-T-FLAG {Delta} REEL were subjected to immunostaining using an anti-FLAG fluorescein isothiocyanate (FITC) conjugate (Sigma). Anti-calreticulin (Stressgen, British Columbia, Canada) was used as a marker for the ER and was detected by anti-rabbit IgG (H+L)-TXRD (Southern Biotechnology Associates, Birmingham, AL). The rabbit anti-FLAG antibody (Sigma) was detected by anti-rabbit IgG (H+L)-TXRD and was used for coimmunostaining with anti-GM130 (Clontech), a marker for the Golgi apparatus, which was detected by anti-mouse IgG1-Alexa488 (Molecular Proves, Eugene, OR). The cells (~106) transfected with pcDNA3.1-mock, pcDNA3.1-ß3Glc-T-FLAG, or pcDNA3.1-ß3Glc-T-FLAG {Delta} REEL were homogenized by sonication for 10 min on ice in lysis buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, and 1% Triton X-100), and the supernatants were centrifuged at 14,000 x g for 10 min at 4°C and then subjected to immunoprecipitation. Supernatants and conditioned media from each transfectant were mixed with anti-FLAG M2 agarose and rotated at 4°C overnight. The proteins were then separated by 12.5% SDS–PAGE and blotted with an anti-FLAG M2 monoclonal antibody peroxidase conjugate (Sigma).

Quantitative analysis of ß3Glc-T in human tissues by real-time PCR
For the quantification of human ß3Glc-T transcripts, we employed the real-time PCR method, as described in detail previously (Iwai et al., 2002Go). Total RNA of various human tissues was purchased from Clontech. cDNA templates were synthesized from the total RNA with a SuperScript II first-strand synthesis system (Invitrogen). Standard curves for the endogenous control, GAPDH cDNA, were generated by serial dilution of pCR2.1 (Invitrogen) DNA containing the GAPDH gene. The primer set and probe for the ß3Glc-T gene were as follows: ß3Glc-T, the forward and reverse primers were 5'-GACACACAGCCCTCTCTTCCA-3' and 5'-TGGGAACTTGATGAGAAAGGTAGTC-3', respectively, and the probe, 5'-AGGCTCGGCCGGTGGATTACCCTA-3', contained a minor groove binder. PCR products were continuously measured with an ABI PRISM 7700 Sequence Detection System (Applied Biosystems, Foster City, CA). The relative amounts of the transcripts were normalized to the amount of GAPDH transcript in the same cDNA samples.


    Acknowledgments
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
This work was performed as part of the R&D Project of the Industrial Science and Technology Frontier Program (R&D for Establishment and Utilization of a Technical Infrastructure for Japanese Industry) supported by the New Energy and Industrial Technology Development Organization (NEDO).


    Conflict of interest statement
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
None declared.


    Footnotes
 
The nucleotide sequence reported in this article has been registered in the GenBank/EBI Data Bank with accession number AB101481 for human and AB253762 for mouse.

1 These authors contributed equally to the work and both should be considered as first authors. Back


    Abbreviations
 
Bz, benzyl; cDNA, complementary DNA; CHO, Chinese hamster ovary; ChSy/CSS, chondroitin synthase; core 1, Galß1,3GalNAc{alpha}1-serine/threonine; EGF, epidermal growth factor; ER, endoplasmic reticulum; Fuc, fucose; FUT, fucosyltransferase; Gal, galactose; GalNAc, N-acetylgalactosamine; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GDP, guanine diphosphate; Glc, glucose; GlcA, glucuronic acid; GlcNAc, N-acetylglucosamine; HEK, human embryonic kidney; Lea, Galß1-3(Fuc{alpha}1-4)GlcNAc; Lex, Galß1-4(Fuc{alpha}1-3)GlcNAc; Lfng, Lunatic Fringe; Man, mannose; Mfng, Manic Fringe; ORF, open-reading frame; PCR, polymerase chain reaction; pNp, para-nitrophenyl; POFUT, protein O-fucosyltransferase; RACE, rapid amplification for cDNA ends; Rfng, Radical Fringe; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; TBS, Tris-buffered saline; TLC, thin-layer chromatography; TSP, thrombospondin; TSR, TSP Type 1 repeat; UDP, uridine diphosphate; Xyl, xylose; ß3Gal-T, ß1,3-galactosyltransferase; ß3GlcNAc-T, ß1,3-N-acetylglucosaminyltransferase; ß3Glc-T, ß1,3-glucosyltransferase; ß3GT, ß1,3-glycosyltransferase; ß4GT, ß1,4-glycosyltransferase


    References
 Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Acknowledgments
 Conflict of interest statement
 References
 
Amado M., Almeida R., Carneiro F., Levery S.B., Holmes E.H., Nomoto M., Hollingsworth M.A., Hassan H., Schwientek T., Nielsen P.A., others P.A. (1998) A family of human beta3-galactosyltransferases. Characterization of four members of a UDP-galactose:beta-N-acetyl-glucosamine/beta-nacetyl-galactosamine beta-1,3-galactosyltransferase family. J. Biol. Chem. 273:12770–12778.[Abstract/Free Full Text]

Bai X., Zhou D., Brown J.R., Crawford B.E., Hennet T., Esko J.D. (2001) Biosynthesis of the linkage region of glycosaminoglycans: cloning and activity of galactosyltransferase II, the sixth member of the beta 1,3-galactosyltransferase family (beta 3GalT6). J. Biol. Chem. 276:48189–48195.[Abstract/Free Full Text]

Bendtsen J.D., Nielsen H., von Heijne G., Brunak S. (2004) Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340:783–795.[CrossRef][ISI][Medline]

Bjoern S., Foster D.C., Thim L., Wiberg F.C., Christensen M., Komiyama Y., Pedersen A.H., Kisiel W. (1991) Human plasma and recombinant factor VII. Characterization of O-glycosylations at serine residues 52 and 60 and effects of site-directed mutagenesis of serine 52 to alanine. J. Biol. Chem. 266:11051–11057.[Abstract/Free Full Text]

Bornstein P. (2001) Thrombospondins as matricellular modulators of cell function. J. Clin. Invest. 107:929–934.[CrossRef][ISI][Medline]

Bruckner K., Perez L., Clausen H., Cohen S. (2000) Glycosyltransferase activity of Fringe modulates Notch-Delta interactions. Nature 406:411–415.[CrossRef][Medline]

Chen H., Herndon M.E., Lawler J. (2000) The cell biology of thrombospondin-1. Matrix Biol. 19:597–614.[CrossRef][ISI][Medline]

Gonzalez de Peredo A., Klein D., Macek B., Hess D., Peter-Katalinic J., Hofsteenge J. (2002) C-mannosylation and o-fucosylation of thrombospondin type 1 repeats. Mol. Cell Proteomics 1:11–18.[Abstract/Free Full Text]

Gotoh M., Sato T., Kiyohara K., Kameyama A., Kikuchi N., Kwon Y.D., Ishizuka Y., Iwai T., Nakanishi H., Narimatsu H. (2004) Molecular cloning and characterization of beta1,4-N-acetylgalactosaminyltransferases IV synthesizing N,N'-diacetyllactosediamine. FEBS Lett. 562:134–140.[CrossRef][ISI][Medline]

Hallgren P., Lundblad A., Svensson S. (1975) A new type of carbohydrate-protein linkage in a glycopeptide from normal human urine. J. Biol. Chem. 250:5312–5314.[Abstract/Free Full Text]

Harris R.J., Leonard C.K., Guzzetta A.W., Spellman M.W. (1991) Tissue plasminogen activator has an O-linked fucose attached to threonine-61 in the epidermal growth factor domain. Biochemistry 30:2311–2314.[CrossRef][Medline]

Harris R.J., Ling V.T., Spellman M.W. (1992) O-linked fucose is present in the first epidermal growth factor domain of factor XII but not protein C. J. Biol. Chem. 267:5102–5107.[Abstract/Free Full Text]

Harris R.J. and Spellman M.W. (1993) O-linked fucose and other post-translational modifications unique to EGF modules. Glycobiology 3:219–224.[Abstract/Free Full Text]

Harris R.J., van Halbeek H., Glushka J., Basa L.J., Ling V.T., Smith K.J., Spellman M.W. (1993) Identification and structural analysis of the tetrasaccharide NeuAc alpha (2–>6) Gal beta (1–>4) GlcNAc beta (1–>3) Fuc alpha 1–>O-linked to serine 61 of human factor IX. Biochemistry 32:6539–6547.[CrossRef][Medline]

Heinonen T.Y., Pasternack L., Lindfors K., Breton C., Gastinel L.N., Maki M., Kainulainen H. (2003) A novel human glycosyltransferase: primary structure and characterization of the gene and transcripts. Biochem. Biophys. Res. Commun. 309:166–174.[CrossRef][ISI][Medline]

Hiruma T., Togayachi A., Okamura K., Sato T., Kikuchi N., Kwon Y.D., Nakamura A., Fujimura K., Gotoh M., Tachibana K., others K. (2004) A novel human beta1,3-N-acetylgalactosaminyltransferase that synthesizes a unique carbohydrate structure, GalNAcbeta1-3GlcNAc. J. Biol. Chem. 279:14087–14095.[Abstract/Free Full Text]

Hofsteenge J., Huwiler K.G., Macek B., Hess D., Lawler J., Mosher D.F., Peter-Katalinic J. (2001) C-mannosylation and O-fucosylation of the thrombospondin type 1 module. J. Biol. Chem. 276:6485–6498.[Abstract/Free Full Text]

Irvine K.D. (1999) Fringe, Notch, and making developmental boundaries. Curr. Opin. Genet Dev. 9:434–441.[CrossRef][ISI][Medline]

Ishida H., Togayachi A., Sakai T., Iwai T., Hiruma T., Sato T., Okubo R., Inaba N., Kudo T., Gotoh M., others M. (2005) A novel beta1,3-N-acetylglucosaminyltransferase (beta3Gn-T8), which synthesizes poly-N-acetyllactosamine, is dramatically upregulated in colon cancer. FEBS Lett. 579:71–78.[CrossRef][ISI][Medline]

Isshiki S., Togayachi A., Kudo T., Nishihara S., Watanabe M., Kubota T., Kitajima M., Shiraishi N., Sasaki K., Andoh T., Narimatsu H. (1999) Cloning, expression, and characterization of a novel UDP-galactose: beta-N-acetylglucosamine beta1,3-galactosyltransferase (beta3Gal-T5) responsible for synthesis of type 1 chain in colorectal and pancreatic epithelia and tumor cells derived therefrom. J. Biol. Chem. 274:12499–12507.[Abstract/Free Full Text]

Iwai T., Inaba N., Naundorf A., Zhang Y., Gotoh M., Iwasaki H., Kudo T., Togayachi A., Ishizuka Y., Nakanishi H., Narimatsu H. (2002) Molecular cloning and characterization of a novel UDP-GlcNAc: GalNAc-peptide beta1,3-N-acetylglucosaminyltransferase (beta 3Gn-T6), an enzyme synthesizing the core 3 structure of O-glycans. J. Biol. Chem. 277:12802–12809.[Abstract/Free Full Text]

Ju T., Brewer K., D’Souza A., Cummings R.D., Canfield W.M. (2002) Cloning and expression of human core 1 beta1,3-galactosyltransferase. J. Biol. Chem. 277:178–186.[Abstrac