Skip Navigation


Glycobiology Advance Access originally published online on September 14, 2006
Glycobiology 2006 16(12):158R-184R; doi:10.1093/glycob/cwl040
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplemental Data
Right arrow All Versions of this Article:
16/12/158R    most recent
cwl040v1
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 Ma, B.
Right arrow Articles by Taylor, D. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ma, B.
Right arrow Articles by Taylor, D. E.
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

REVIEW

Fucosylation in prokaryotes and eukaryotes

Bing Ma, Joanne L. Simala-Grant and Diane E. Taylor1

Department of Medical Microbiology and Immunology, University of Alberta, Edmonton, Alberta, Canada T6G 2H7


1 To whom correspondence should be addressed; e-mail: diane.taylor{at}ualberta.ca

Received on June 21, 2006; revised on August 25, 2006; accepted on August 25, 2006


    Abstract
 Top
 Abstract
 Introduction
 {alpha}1,2 Fucosylation
 {alpha}1,3/4 Fucosylation
 {alpha}1,6 Fucosylation
 Xyloglucan {alpha}1,2...
 O-Fucosylation
 Sequence homology and structural...
 Abnormal fucosylation in mammals
 Application of FucTs,...
 Concluding remarks
 Supplementary Data
 Acknowledgments
 Conflict of interest statement
 References
 
Fucosylated carbohydrate structures are involved in a variety of biological and pathological processes in eukaryotic organisms including tissue development, angiogenesis, fertilization, cell adhesion, inflammation, and tumor metastasis. In contrast, fucosylation appears less common in prokaryotic organisms and has been suggested to be involved in molecular mimicry, adhesion, colonization, and modulating the host immune response. Fucosyltransferases (FucTs), present in both eukaryotic and prokaryotic organisms, are the enzymes responsible for the catalysis of fucose transfer from donor guanosine-diphosphate fucose to various acceptor molecules including oligosaccharides, glycoproteins, and glycolipids. To date, several subfamilies of mammalian FucTs have been well characterized; these enzymes are therefore delineated and used as models. Non-mammalian FucTs that possess different domain construction or display distinctive acceptor substrate specificity are highlighted. It is noteworthy that the glycoconjugates from plants and schistosomes contain some unusual fucose linkages, suggesting the presence of novel FucT subfamilies as yet to be characterized. Despite the very low sequence homology, striking functional similarity is exhibited between mammalian and Helicobacter pylori {alpha}1,3/4 FucTs, implying that these enzymes likely share a conserved mechanistic and structural basis for fucose transfer; such conserved functional features might also exist when comparing other FucT subfamilies from different origins. Fucosyltranferases are promising tools used in synthesis of fucosylated oligosaccharides and glycoconjugates, which show great potential in the treatment of infectious and inflammatory diseases and tumor metastasis.

Key words: eukaryotes / fucosylation / FucTs / prokaryotes


    Introduction
 Top
 Abstract
 Introduction
 {alpha}1,2 Fucosylation
 {alpha}1,3/4 Fucosylation
 {alpha}1,6 Fucosylation
 Xyloglucan {alpha}1,2...
 O-Fucosylation
 Sequence homology and structural...
 Abnormal fucosylation in mammals
 Application of FucTs,...
 Concluding remarks
 Supplementary Data
 Acknowledgments
 Conflict of interest statement
 References
 
Fucosyltransferases (FucTs) are widely expressed in vertebrates, invertebrates, plants, and bacteria. They belong to the glycosyltransferase superfamily (EC 2.4.1.x.y), which is defined in the category of Carbohydrate-Active enZYmes (CAZY) (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html). FucTs catalyze the inverting reaction in which a fucose residue is transferred from the donor guanosine-diphosphate fucose (GDP-Fuc) to the acceptor molecules including oligosaccharides, glycoproteins, and glycolipids (Oriol et al., 1999Go). The fucosylated glycoconjugates are involved in a variety of biological and pathological processes.

FucT subfamilies
Based on the site of fucose addition, FucTs are classified into {alpha}1,2, {alpha}1,3/4, {alpha}1,6, and O-FucTs (Figures 1 and 2). The former three subfamilies of enzymes in eukaryotic organisms are type II transmembrane Golgi-anchored proteins containing an N-terminal cytoplasmic tail, a transmembrane domain, and an extended stem region followed by a large globular C-terminal catalytic domain facing the Golgi lumen (Nilsson et al., 1993Go, 1996Go) (Figure 3A). O-FucTs, however, are endoplasmic reticulum (ER)-localized soluble proteins and catalyze O-fucosylation in the ER (Luo and Haltiwanger, 2005Go; Okajima et al., 2005Go). An unusual non-Golgi {alpha}1,2 FucT in Dictyostelium is localized in both the cytoplasm and the nucleus and modifies Skp1 protein, a subunit of the SCF-E3 ubiquitin ligase (van Der Wel et al., 2001Go). This enzyme lacks any conserved {alpha}1,2 FucT motifs (see Sequence homology and structural prediction of FucT subfamilies) and is classified in CAZY family 74 (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html).


Figure 1
View larger version (26K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 1. Fucosylation sites of human {alpha}1,2, {alpha}1,3/4, and {alpha}1,6 FucTs. Both of the two main types of acceptor substrates, O-glycans (left) and N-glycans (right), contain poly-LacNAc chains (in brackets), but chitobiose (grey square) is only found in N-glycans. Fucosylation sites of human {alpha}1,2 and {alpha}1,6 FucTs are shown on the antennae of O- and N-glycans. For human {alpha}1,3/4 FucTs, only their preferred fucosylation sites are given, even though they are able to add fucose to either internal, middle, or distal sites of GlcNAc in poly-LacNAc chains. Of note, FUT7 cannot use Type II acceptor unless it is sialylated (in light gray shade). With the sialylated-Type II acceptor, FUT9 switches the preferred fucosylation site from distal to the inner most GlcNAc (in light gray shade). The black and the dark gray blocks at the bottom represent proteins. Asn, asparagine; Gal, galactose; GalNAc, N-acetylgalactosamine; GlcNAc, N-acetylglucosamine; Man, mannose; Neu5Ac, sialic acid; Ser, serine; Thr, threonine.

 

Figure 2
View larger version (18K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 2. Fucosylation sites of plant {alpha}1,4 FucTs, core {alpha} 1,3 FucTs, core {alpha} 1,6 FucTs, xyloglucan {alpha} 1,2 FucTs, and O-FUT1 and 2 in the structures of plant biantennary N-glycan (A) (Vietor et al., 2003Go), xyloglucan (B) (Sarria et al., 2001Go), and proteins containing small cysteine-knot motifs with EGF repeat (left) or thrombospondin type repeat (right) consensus sequence (C), respectively. (Wang and Spellman, 1998Go; Wang et al., 2001Go; Luo, Koles, et al., 2006Go; Luo, Nita-Lazar, et al., 2006Go). FucTs that catalyze the fucose addition at specific linkages are shown by an arrow. Asn, asparagine; C, cysteine; Fuc, fucose; Gal, galactose; Glc, glucose; GlcNAc, N-acetylglucosamine; Man, mannose; S, serine; T, threonine; W, tryptophan; X, any amino acid; Xyl, xylose. The superscript on each of six conserved cysteine residue indicates their order in the small cysteine-knot motifs. The amino acids that link the conserved cysteine residues but are not part of the proposed consensus sequences are shown as a solid line.

 

Figure 3
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 3. Schematic structures of (A) mammalian {alpha}1,3/4 FucTs and (B) Helicobacter pylori {alpha}1,3/4 FucTs anchored at Golgi and H. pylori cell membranes, respectively. The two cylinders in (B) represent two putative amphipathic {alpha}-helices and the stem (B) corresponds to the heptad repeat region. +, positively charged amino acid residues; CAT, catalytic domain; COOH, C-terminus; H, hydrophobic amino acid residues; NH2, N-terminus; P, polar amino acid residue; TM, transmembrane domain. This figure was published previously (Ma et al., 2003Go).

 
GDP-fucose synthesis and transport
GDP-Fuc is synthesized in the cytoplasm (Becker and Lowe, 2003Go) through de novo synthesis and salvage pathways. De novo synthesis, accounting for 90% of the total GDP-Fuc production, is involved in converting GDP-mannose to GDP-Fuc via GDP-mannose 4,6-dehydrase (GMD) and GDP-keto-6-deoxymannaose 3,5-epimerase/4-reductase (also named FX) that contains dual activities (Tonetti et al., 1996Go) (Figure 4). The salvage pathway, accounting for approximately 10% of GDP-Fuc production, utilizes the free cytosolic fucose as substrate which is derived from an extracellular source or from lysosomal degradation (Becker and Lowe, 2003Go). Fucose is firstly phosphorylated by fucokinase to form fucose-1-phosphate, which is then converted to GDP-fucose by GDP-Fuc pyrophosphorylase (Niittymaki et al., 2004Go) (Figure 4). Subsequently, a GDP-Fuc transporter (Gft), anchored at the Golgi membrane (Figure 4), imports GDP-Fuc from the cytoplasm to the Golgi-lumen (Luhn et al., 2004Go), where GDP-Fuc becomes concentrated in vesicles and is recognized by Golgi-localized FucTs as a donor substrate (Hirschberg, 2001Go). Human Gft was reported to be a hydrophobic protein possessing ten transmembrane domains, with the amino and carboxy termini exposed to the cytosol (Hirschberg, 2001Go). Assembling into a homodimer, Gft couples GDP-Fuc import to the Golgi lumen with GMP export to the cytoplasm (Luhn et al., 2004Go; Helmus et al., 2006Go) (Figure 4). As O-FucTs are present and able to catalyze O-fucosylation in the ER (Luo and Haltiwanger, 2005Go; Okajima et al., 2005Go), one expects the presence of an ER-localized Gft, which has not yet been identified.


Figure 4
View larger version (55K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 4. Biosynthesis and transport of GDP-fucose and localization of FucTs in eukaryotic cells. GDP-Fuc is synthesized in the cytosol through the de novo synthesis and salvage pathways (Becker and Lowe, 2003Go; Niittymaki et al., 2004Go). GDP-Fuc is imported from the cytosol to the Golgi or the ER by GDP-Fuc transporters (Luhn et al., 2004Go). It is believed the ER-localized transporter is present, but it has not yet been identified. {alpha}1,2, {alpha}1,3/4 and {alpha}1,6 FucTs are Type II membrane proteins with a N-terminal transmembrane domain anchored at the Golgi-membrane and the C-terminal catalytic region exposed to the Golgi-lumen (Nilsson et al., 1993Go, 1996Go). {alpha}1,2 FucTs, {alpha}1,3/4 FucTs, {alpha}1,6 FucTs, and sialyltransferases, involved in terminal modifications of carbohydrate, reside in the late cisternae of the Golgi. The proper compartmentalization of {alpha}1,3/4 FucTs, {alpha}1,2FucTs, and sialyltransferases (circled in broken line) within the Golgi is sorted by the CF transmembrane conductance regulator (Rhim et al., 2001Go). Mutation of this protein is responsible for impaired glycosylation in CF patients. O-FUT1 and 2 are ER-localized soluble proteins (Luo and Haltiwanger, 2005Go; Okajima et al., 2005Go). Notch has been shown to be fucosylated first by O-FUT1 in the ER and then secreted to the cell surface aided by O-FUT1 as a chaperone (Okajima and Irvine, 2002Go; Okajima et al., 2003Go, 2005Go). Whether O-fucosylation catalyzed by O-FUT1 can take place other than ER remains to be determined. Proteins with TSR sequences are fucosylated by O-FUT2 (Luo, Koles, et al., 2006; Luo, Nita-Lazar, et al., 2006), but if it occurs in the ER awaits further examination.

 

    {alpha}1,2 Fucosylation
 Top
 Abstract
 Introduction
 {alpha}1,2 Fucosylation
 {alpha}1,3/4 Fucosylation
 {alpha}1,6 Fucosylation
 Xyloglucan {alpha}1,2...
 O-Fucosylation
 Sequence homology and structural...
 Abnormal fucosylation in mammals
 Application of FucTs,...
 Concluding remarks
 Supplementary Data
 Acknowledgments
 Conflict of interest statement
 References
 
{alpha}1,2 FucTs transfer fucose at an {alpha}1,2 linkage from GDP-Fuc to the galactose moiety of Galß1,4GlcNAc (Type II, also called LacNAc) or Galß1,3GlcNAc (Type I) structures, which are localized at the peripheral (antennae) position of acceptor molecules (Figures 1 and 5). {alpha}1,2 FucTs, belonging to CAZY family 11 (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html), were shown to adopt random bi bi mechanism for fucose transfer (Palcic et al., 1989Go).


Figure 5
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
 
Fig. 5. Structures of Lewis blood group antigens. Fuc, fucose; Gal, galactose; GalNAc, N-acetylgalatosamine; GlcNAc, N-acetylglucosamine.

 
{alpha}1,2 Fucosylation in mammals, Caenorhabditis elegans and Schistosoma mansoni
In the human genome, fut1, fut2, and sec1 (a pseudogene with a mutation causing a frameshift) encode the H, Se (secretor), and non-functional {alpha}1,2 FucTs, respectively. Products of fut1 is expressed mainly on erythrocyte membrane and vascular endothelium (Mollicone et al., 1995Go), whereas the fut2 is expressed on epithelial cells and in body fluids (i.e., saliva) (Avent, 1997Go). Human FUT1 and FUT2 are responsible for synthesis of H-antigen (Figure 5), and secretor status is determined by the secretor gene (fut2). Secretor-negative individuals, also called non-secretors, lack H-antigen and its derived structures in body exocrine tissues (Avent, 1997Go). Human blood group (H) and secretory (Se) {alpha}1,2 FucTs show different acceptor specificities; the former prefers Type I and Type II acceptors and is less efficient in fucose transfer to Galß1,3GalNAc (Type III) structures (Kyprianou et al., 1990Go; Sarnesto et al., 1990Go), whereas the latter is more active on Type I and Type III than on Type II acceptor (Sarnesto et al., 1990Go).

Among over two-dozen putative {alpha}1,2 FucTs from C. elegans, one of them, CE2FT-1, prefers the unusual sugars Galß1,4Xyl-R and Galß1,6GlcNAc-R as acceptors. CE2FT-1 is expressed in all developmental stages of C. elegans and shares only 5–10% sequence identity to {alpha}1,2 FucTs from human, rabbits, and mice (Zheng et al., 2002Go). It is unable to fucosylate Type III, Galß1,4Glcß-R, or lactose. Neither can it fucosylate acceptors that already contain an {alpha}1,3/4-linked fucose moiety (Zheng et al., 2002Go). Despite its unique features, CE2FT-1 is categorized into CAZY family 11 (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html).

Notably, glycoconjugates of S. mansoni contain Fuc{alpha}1,2 Fuc structures, indicating the existence of an {alpha}1,2 FucT that is capable of transferring Fuc to another Fuc residue (Marques et al., 2001Go). Such an {alpha}1,2 FucT from cercariae of the avian schistosome Trichobilharzia ocellata displayed a relatively low catalytic efficiency for the substrate Fuc{alpha}1,3GlcNAcß1,2Man because of its high Km (Hokke et al., 1998Go). It is not yet known whether this schistosomal enzyme contains {alpha}1,2 FucT motifs (see Sequence homology and structural prediction of FucT subfamilies), and if it belongs to an as yet unidentified subfamily.

{alpha}1,2 Fucosylation in bacteria
Several putative bacterial {alpha}1,2 FucTs have been identified to date, including proteins involved in colanic acid synthesis in Escherichia coli K12 and Salmonella enterica LT2 (Reeves et al., 2006Go), in O-antigen synthesis in Yersinia enterocolitica O8 (Reeves et al., 2006Go), the WbsJ in enteropathogenic E. coli O128 strain (Shao et al., 2003Go), and {alpha}1,2 FucTs from Helicobacter pylori (Wang, Boulton, et al., 1999Go; Wang, Rasko, et al., 1999Go). Of these enzymes, only H. pylori {alpha}1,2 FucTs have been functionally characterized (Wang, Boulton et al., 1999Go; Wang, Rasko, et al., 1999Go; Wang et al., 2000Go).

FucTs from H. pylori are responsible for the last steps in the synthesis of Lewis blood antigen structures on their lipopolysaccharide (LPS), which contains incompletely fucosylated repeating LacNAc chains (Chan et al., 1995Go; Sherburne and Taylor, 1995Go; Aspinall and Monteiro, 1996Go; Aspinall et al., 1996Go, 1997Go). More than 80% of H. pylori strains express Type II Lewis antigens (Lex and/or Ley), and half of them express both (Sherburne and Taylor, 1995Go; Aspinall and Monteiro, 1996Go; Wirth et al., 1996Go; Monteiro et al., 1998Go). A much smaller proportion of H. pylori strains express Type I Lewis blood group antigens (Lea and/or Leb) (Monteiro et al., 1998Go) and a very small number express sialyl-Lex (Wirth et al., 1996Go; Monteiro et al., 2000Go). Some H. pylori strains also possess O-antigens that show no reaction to antibodies against Lex, Ley, Lea, or Leb. These structures were designated non-typeable O-antigens (Rasko et al., 2001Go).

The gene encoding the {alpha}1,2 FucT in the H. pylori genome was named futC (Berg et al., 1997Go; Alm et al., 1999Go), which contains polyA–polyC slippery tracts at the 5' end. During DNA replication, the addition or deletion of one or more base pairs within these tracts occurs at a much higher rate than the normal mutation frequency (Appelmelk et al., 1998Go, 1999Go, 2000Go; Wang Boulton, et al., 1999Go; Wang, Rasko, et al., 1999Go; Wange et al., 2000). Moreover, futC contains imperfect TAA (or GAA or AAA) repeats immediately downstream of the poly C tract at the mid-region of the gene (Wang, Rasko, et al., 1999Go). The signature sequences, an internal Shine-Dalgarno-like context, a heptamer (AAAAAAG) and the downstream potential stem-loop structure, are also present, and they are hypermutable (Wang, Rasko, et al., 1999Go). During translation, the ribosome can slip to the –1 reading frame at a frequency of 50% to encode a full-length {alpha}1,2 FucT protein instead of two truncated products. Additionally, the futC gene in some H. pylori strains lacks a valid promoter region leading to lack of expression of an {alpha}1,2 FucT protein (Wang, Rasko, et al., 1999Go). Therefore, the on/off status of the futC gene can be controlled at both translational and transcriptional levels. The phenomenon of switching genes on and off by various genetic mechanisms is referred to as phase variation, which is a very common mechanism used by micro-organisms to switch on or off the expression of outer membrane proteins (Owen et al., 1996Go; Seifert, 1996Go; Bart et al., 1999Go; Neyrolles et al., 1999Go; Horino et al., 2003Go; Kyme et al., 2003Go; Zhang et al., 2004Go; Martin et al., 2005Go).

Unlike eukaryotic enzymes, H. pylori {alpha}1,2 FucTs lack the N-terminal cytosolic tail and the transmembrane domain (Wang, Boulton, et al., 1999Go). In fact, the {alpha}1,2 FucT from H. pylori strain UA802 was demonstrated to be a soluble protein located in the cytoplasm (Wang, Boulton, et al., 1999Go). {alpha}1,2 FucTs from different H. pylori strains share very high sequence identity with one another (~95%), but very low identity (18–22%) with their mammalian counterparts (Oriol et al., 1999Go).


    {alpha}1,3/4 Fucosylation
 Top
 Abstract
 Introduction
 {alpha}1,2 Fucosylation
 {alpha}1,3/4 Fucosylation
 {alpha}1,6 Fucosylation
 Xyloglucan {alpha}1,2...
 O-Fucosylation
 Sequence homology and structural...
 Abnormal fucosylation in mammals
 Application of FucTs,...
 Concluding remarks
 Supplementary Data
 Acknowledgments
 Conflict of interest statement
 References
 
{alpha}1,3/4 FucTs, belonging to CAZY family 10 (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html), add fucose at {alpha}1,3 or {alpha}1,4 linkage to the GlcNAc moiety of Type II or Type I structures, which are localized at the peripheral (antennae) position of acceptor molecules (Oriol et al., 1999Go) (Figure 1). Kinetic studies suggested that human {alpha}1,3/4 FucTs employed an ordered sequential mechanism with donor GDP-Fuc binding first, followed by acceptor binding, and then product Lex being released followed by the GDP portion of the donor (Murray et al., 1996Go, 1997Go; Qiao et al., 1996Go).

{alpha}1,3/4 Fucosylation in mammals
Mammalian {alpha}1,2 and {alpha}1,3/4 FucTs are involved in the last steps of synthesis of A, B, and H Lewis blood antigens and Lewis-related carbohydrate antigens (i.e., Lex, Ley, Lea, Leb, sialyl-Lex, and sialyl-Lea) (Figure 5). Difucosylated Lewis antigens (Ley and Leb) can be synthesized via two pathways: terminal fucosylation (in an {alpha}1,2 linkage) followed by subterminal fucosylation (in an {alpha}1,3 or {alpha}1,4 linkage) or subterminal fucosylation followed by terminal fucosylation. Mammalian cells predominantly use the former pathway (Henry et al., 1995Go).

Genes fut3–7 and fut9 in the human genome encode six {alpha}1,3/4 FucTs, abbreviated FUT3–7 and FUT9 (or Fuc-TIII–VII and Fuc-TIX), all of which have {alpha}1,3 activity, but FUT3 and FUT5 also possess {alpha}1,4 activity. Human {alpha}1,3/4 FucTs are Type II membrane proteins (Figure 3A). Domain swapping studies demonstrated that the extended stem region, also called the hypervariable region, of FUT3 and FUT5 confers {alpha}1,4 activity (Legault et al., 1995Go; Xu et al., 1996Go; Nguyen et al., 1998Go). Specifically, an aromatic residue (Trp) within this region determines Type I acceptor recognition (Dupuy et al., 1999Go, 2004Go). In animals, Lea seems to be exclusively present in primates (Costache et al., 1997Go; Dupuy et al., 2002Go). Human fut10 and fut11 were described to encode two proteins that contain the characteristic {alpha}1,3 motifs (see Sequence homology and structural prediction of FucT subfamilies) and share homology with {alpha}1,3 FucT from Drosophila (Roos et al., 2002Go). The function and the acceptor specificity of FUT10 and FUT11 have not yet been defined.

Human FUT3–7 and FUT are able to transfer fucose to acceptors that contain a poly-LacNAc chain, but they have different preferable fucosylation sites (Supplementary Table 1 Figure 1), although FUT3, 5, and 6 share about 85% protein sequence identity. In addition, when these six enzymes transfer fucose to 3'-sialylated or 2'-fucosylated Type I or Type II structures, they also exhibited distinct preferences (Supplementary Table I). Likewise, human FUT3 and FUT5 behave differently in terms of {alpha}1,4 specificity. FUT5 fucosylates H type I only on core 2 [Galß1,3(GlcNAcß1,6)GalNAc-Ser/Thr] structures, whereas FUT3 fucosylates H type I much more efficiently on core 3 (GlcNAcß1,3GalNAc-Ser/Thr) than core 2 structures (Holgersson and Lofling, 2006Go).

It is noteworthy that FUT4 and FUT7 are the only human FucT enzymes expressed in leukocytes, where they are responsible for generation of the functional selectin ligands (Homeister et al., 2001Go; Lowe, 2003Go). Sialyl-Lex is the core recognition epitope for E-, P-, and L-selectins that mediate not only the lymphocyte homing but also the initial leukocyte–endothelial cell adhesion events in both acute and chronic inflammation (Kannagi, 2002Go; Lowe, 2003Go). Production of Lex and sialyl-Lex by FUT4 and FUT7 in leukocytes is therefore crucial for leukocyte trafficking. In addition, Lewis-related carbohydrates are also involved in embryogenesis, angiogenesis, microbe-host interactions, neural development, fertilization, host-microbe interactions, and tumor metastasis (Norman et al., 1998Go; Ley, 2003Go). For instance, Lex is widely expressed on early embryonic cells and primordial germ cells in vertebrates (Muramatsu and Muramatsu, 2004Go). Sialyl-Lex/a expressed on capillary endothelial cells (Nguyen et al., 1992Go) participates in capillary morphogenesis through interaction with E-selectin (Nguyen et al., 1993Go). Ehrlichia equi- and Ehrlichia phagocytophila-related bacteria that cause human granulocytic ehrlichiosis contain selectin-like structures on their cell surface. As a consequence, through binding to P-selectin glycoprotein ligand-1 (containing sialyl-Lex moiety) expressed on leukocytes, bacteria display tropistic binding to host leukocytes (Herron et al., 2000Go). In addition, sialyl-Lex/a are heavily expressed on malignant cancer cells and are believed to interact with endothelial E-selectin to promote tumor progression and hematogenous metastasis, while cancer cells are also able to induce endothelial E-selectin expression (Kannagi, 2004Go). As a result, sialyl-Lex/a is not only a cancer marker but also a marker for cancer diagnosis and prognosis, such that higher levels of sialyl-Lex/a is correlated with increased metastasis and poor prognosis (Kannagi, 2004Go; Magnani, 2004Go).

{alpha}1,4 Fucosylation in plants
Plant N-glycans are in the forms of oligomannosidic (Man>5GlcNAc2), paucimannosidic and complex types (Lerouge et al., 1998Go). The Lea moiety, localized at the antennae of N-glycans (Figure 2A), has been detected not only in Monocotyledons and Dicotyledons (Fitchette-Laine et al., 1997Go; Lerouge et al., 1998Go; Bakker et al., 2001Go; Wilson, 2001Go; Leonard et al., 2002Go, 2005Go; Castilho et al., 2005Go), but also in Physcomitrella patens (a monoecious moss) and various bryophyte species (Vietor et al., 2003Go). Indeed, the bryophyte species seem to contain similar types of N-glycans as those found in higher plants (Vietor et al., 2003Go).

Although Lea is expressed in all plant tissues (flowers, leaves, roots, and seedlings), the {alpha}1,4 FucT activity is predominant in young tissues (leaves and roots) (Lerouge et al., 1998Go). It appears that young growing tissues require de novo synthesis of Lea, which might play a role in cell elongation and/or differentiation. Moreover, peak {alpha}1,4 FucT activity is also detected in tobacco seeds, male flowers, and during tobacco male gametophyte development (pollen maturation, pollen germination, and tube elongation) (Joly et al., 2002Go), suggesting that Lea likely plays an important role in plant reproductive development. Additionally, plant {alpha}1,4 fucosylation is also suggested to be involved in cell-to-cell communication and/or recognition (Fitchette-Laine et al., 1997Go).

The FucT responsible for Lea production was isolated from Arabidopsis thaliana (Leonard et al., 2002Go) and sycamore cells (Fitchette-Laine et al., 1997Go) and was examined to have exclusive {alpha}1,4 activity. Based on the sequence alignment, these plant {alpha}1,4 FucTs contain the Trp residue (Bakker et al., 2001Go; Leonard et al., 2002Go) that confers the specificity of fucose transfer at {alpha}1,4 linkages for human FUT3 and FUT5 (Dupuy et al., 1999Go, 2004Go). Whether the Trp residue of plant {alpha}1,4 FucTs also confers Type I acceptor specificity awaits experimental examination.

{alpha}1,3/4 Fucosylation in S. mansoni
One major feature of schistosomal glycoconjugates is that they are rich in fucose and GalNAc moieties but lack sialic acid. Human schistosomes (blood flukes) are digenic trematodes causing schistosomiasis, and they have several developmental stages. Schistosomal eggs firstly are discharged from the human host and enter the intermediate host (snails) where cercariae develop. Cercariae can penetrate human skin and enter the human bloodstream becoming the schistosomula, which develop into sexually mature adults after encountering a partner of the opposite sex (Cummings and Nyame, 1996Go). Fucosylated glycoconjugates are observed in all developmental stages; however, much higher levels of FucT activities were found in egg extracts in comparison with cercarial or adult worm extracts (Marques et al., 2001Go), suggesting that FucT activities are differentially regulated during development.

Schistosomal {alpha}1,3/4 FucTs are able to use both unsubstituted, {alpha}1,2-fucosylated and {alpha} 2,3-sialylated Type II-structure, no matter whether these sugar moieties are in the form of oligosaccharides, glycoproteins, or glycolipids. Nevertheless, they use Type I series acceptor molecules much less efficiently (Hokke et al., 1998Go; Marques et al., 2001Go). Oligomeric Lex and polymeric Lex (n > 25) are detected not only on the O-linked glycans of gut-associated circulating cathodic antigen (van Dam et al., 1994Go; Cummings and Nyame, 1996Go) but also on N-glycans of schistosome eggs (Cummings and Nyame, 1996Go). Moreover, schistosomal glycoconjugates contain the fucose moiety linked to the GalNAc group within the LacdiNAc(GalNAcß1,4GlcNAc)-repeat structure forming Fuc{alpha}1,3GalNAcß1,4GlcNAc (Wuhrer et al., 2006Go). Additionally, a pseudo Ley [Fuc{alpha}1,3Galß1,4(Fuc{alpha}1,3)GlcNAcß1-R] structure was found in cercarial extracts containing a Fuc{alpha}1,3Gal linkage (Wuhrer et al., 2000Go; Hokke and Deelder, 2001Go). These data suggest that either schistosomal {alpha}1,3/4 FucTs have fairly loose acceptor specificity or some as yet unidentified FucT subfamilies with unique acceptor recognition are present and await discovery.

Notably, the first cloned schistosomal FucT gene was discovered to share identical nucleotide sequences with mouse fut7 gene except its 5' untranslated sequence and the first 39 translated base pairs (Smith et al., 1996Go; Marques et al., 1998Go). It has been proposed that a possible horizontal transfer of mouse DNA to the parasite might have occurred (Oriol et al., 1999Go), but such transmission remains to be examined closely.

In addition to being involved in many aspects of the parasite’s life cycle, schistosomal fucosylated glycoconjugates play a significant role in modifying the host’s immune responses (Cummings and Nyame, 1996Go, 1999Go). S. mansoni egg antigens and the glycolipids derived from cercariae and their excretory/secretory products have the ability to bind to dendritic cell-specific ICAM-3-grabbing non-integrin (DC-SIGN) with high affinity (Appelmelk et al., 2003Go), and the binding ligands are Lex and pseudo-Ley [Fuc{alpha}1,3Galß1,4(Fuc{alpha}1,3)GlcNAcß1-R] determinants (van Die et al., 2003Go; Van Liempt et al., 2004Go; Meyer et al., 2005Go). Upon egg deposition in murine schistosomiasis, the Th1 response is down-regulated and the Th2-type response becomes predominant (Velupillai and Harn, 1994Go). Moreover, the oligosaccharide that contains trisaccharide Lex on both schistosome eggs and schistosomula caused proliferation of the spleen B cells and induced the production of interleukin-10 and prostaglandin E2, both of which are down-regulators of the Th1 cell response. The nature of the oligosaccharide interaction with B cells is currently unknown (Velupillai and Harn, 1994Go).

It has been shown that Lex repeats evoked high titres of specific immunoglobulin M (IgM) antibodies in S. mansoni-infected animals and humans (Ko et al., 1990Go; Srivatsan et al., 1992Go; van Dam et al., 1994Go). These antibodies are believed to cross-react with the repeating Lex structures expressed on the surface of granulocytes, leading to the lysis of granulocytes in the presence of complement (Nyame et al., 1996Go; van Dam et al., 1996Go). Furthermore, the LacdiNAc and fucosylated LacdiNAc repeats are also immunogenic and are involved in host–parasite interactions (Nyame et al., 1999Go, 2000Go).

{alpha}1,3/4 Fucosylation in bacteria
Putative bacterial {alpha}1,3/4 FucTs have been identified in E. coli (Daniels et al., 1992Go), Vibrio cholerae (Stroeher et al., 1997Go), Rickettsia conorii (Ogata et al., 2000Go), Salmonella enterica serovar Typhi (McClelland et al., 2001Go; Parkhill, Dougan, et al., 2001Go), Yersinia pestis (Parkhill, Wren, et al., 2001Go), Yersinia pseudotuberculosis (Chain et al., 2004Go), and Mesorhizobium loti (Sullivan et al., 2002Go), but only {alpha}1,3/4 FucTs from H. pylori have been extensively characterized.

The H. pylori genome contains two paralogous genes futA and futB that encode two {alpha}1,3/4 FucTs, FutA and FutB, respectively (Berg et al., 1997Go; Alm et al., 1999Go). Similar to the futC gene, futA and futB also contain the polyA-polyC tract near the 5' end, and these poly-nucleotide tracts have a higher frequency of addition or deletion of one or more base pairs during DNA replication, leading to on/off status of the gene (Wang et al., 2000Go).

Unlike mammalian cells, difucosylated Lewis antigens (Ley and Leb) on H. pylori LPS are synthesized primarily through subterminal fucosylation (in {alpha}1,3 or {alpha}1,4 linkage) followed by terminal fucosylation (in {alpha}1,2 linkage) (Wang et al., 2000Go). Additionally, H. pylori {alpha}1,3/4 FucTs lack the N-terminal cytosolic tail and transmembrane domain that are present in mammalian counterparts; instead, they have a heptad-repeat region at the C-terminus, which is absent in mammalian enzymes (Ge et al., 1997Go; Martin et al., 1997Go; Rasko, Wang, Palcic, et al., 2000Go). {alpha}1,3/4 FucTs from different strains share >70% sequence identity, containing a highly conserved internal catalytic domain but a divergent N- and C- termini (Rasko, Wang, Palcic, et al., 2000Go). Most H. pylori {alpha}1,3/4 FucTs (i.e., from H. pylori strains NCTC11639 and NCTC11637) have exclusive {alpha}1,3 activity (Ge et al., 1997Go; Martin et al., 1997Go), but some (i.e., H. pylori strains UA948 and UA1111) possess both {alpha}1,3 and {alpha}1,4 activities (Rasko, Wang, Monteiro, et al., 2000;Go Rasko, Wang, Palcic, et al., 2000Go). {alpha}1,3/4 FucT from strain DMS6709 was recently reported to contain primarily {alpha}1,4 activity and only little {alpha}1,3 activity (Rabbani et al., 2005Go). Domain swapping and site-directed mutagenesis studies showed that the C-terminal hypervariable region (immediately upstream of the heptad-repeat region) (Ma et al., 2003Go), an aromatic residue (Tyr) in particular (Ma et al., 2005Go), of H. pylori {alpha}1,3/4FucT from strain UA948 is responsible for Type I acceptor recognition.

The functions of fucosylated carbohydrates on H. pylori LPS are not completely understood despite extensive investigation. As gastric mucosa secretes highly glycosylated molecules that contain Lewis blood group structures: Lea and Leb predominantly from the gastric surface epithelia and Lex and Ley primarily from the gastric deep glands (Cordon-Cardo et al., 1986Go; Kobayashi et al., 1993Go; Taylor et al., 1998Go; Lee, Choe, et al., 2006Go), molecular mimicry was therefore proposed to be a possible mechanism used by H. pylori to evade the immune response and thus maintain a long-term infection (Appelmelk et al., 1997Go). The same function has also been suggested for the sialyl-Lex expressed on the cell surface of some oral bacteria that are associated with infected endocarditis, such as Streptoccocus pyogenes, Porphyromonas gingivalis, Actinobacillus actinomycetemcomitans and Eikenella corrodens (Hirota et al., 1995Go). The sialyl-Lex structures not only may camouflage these bacteria and aid them in traveling away from their normal habitats but also are involved in binding to endothelial selectins to initiate inflammation (Hirota et al., 1995Go).

It was proposed that H. pylori Lewis antigens may also induce production of auto-antibodies causing antigastric auto-reactivity leading to tissue damage (Wirth et al., 1997Go; Heneghan et al., 2001Go). Further evidence indicated that Lewis antigens were not the antigenic component (Faller et al., 1998Go; Yokota et al., 1998Go) but that a structure in the polysaccharide chain of LPS served that function (Yokota et al., 1998Go). The same conclusion had been made for Helicobacter mustelae that express the blood group antigen A structure on the LPS core region (Monteiro et al., 1997Go). Serum from H. mustelae-infected ferrets displayed no reaction with blood group antigen A or B, and the antibodies produced were not absorbed by red blood cells expressing blood group A or by H. mustelae whole cells (Monteiro et al., 1997Go). This suggests that Lewis blood group antigen structures on bacterial LPS are not the antigenic elements causing autoimmune responses.

Among fucosylated carbohydrate structures, Lex on H. pylori LPS was discovered to function as an adhesin involved in the tropistic binding of H. pylori to the apical surface of gastric mucosal epithelial cells and to cells of the gastric pits (Edwards et al., 2000Go). However, accumulating evidence supports the idea that Leb binding adhesin (BabA) in H. pylori plays the most significant role in the binding process (Ilver et al., 1998Go), whereas Lewis blood antigens on bacterial cell surface are not the prerequisite for H. pylori to colonize or to adhere. Lex-mediated binding, in fact, plays a minor role in both the processes and functions only in some but not all strains (Guruge et al., 1998Go; Suresh et al., 2000Go; Takata et al., 2002Go; Altman et al., 2003Go). A recent study confirmed that Lex-mediated adhesion is only significant when BabA–Leb binding is absent (Sheu et al., 2006Go). Nevertheless, the presence of fucosylated oligosaccharides on H. pylori LPS has been found to be correlated with the occurrence of more severe gastric diseases (Monteiro et al., 2001Go; Rasko et al., 2001Go; Eaton et al., 2004Go).

Recently, the Lewis blood antigens of H. pylori were discovered to interact with DC-SIGN expressed on dendritic cells to produce increased amount of interleukin-10 (Appelmelk et al., 2003Go; Bergman et al., 2004Go). Being a Th2 cytokine, interleukin-10 blocks the Th1 response and promotes Th2 activation, thus modulating the host immune response (Bergman et al., 2004Go). Surfactant protein D, a pathogen-associated molecular pattern recognition receptor, is also expressed on human gastric mucosa and plays a role in innate immunity (Madsen et al., 2000Go). The surfactant protein D has a low affinity for fucose but a high affinity for glucose and galactose (Madsen et al., 2000Go). Because H. pylori Lewis antigen expression is regulated by phase variation at the rate of ~0.2–0.5% (Appelmelk et al., 1998Go), the levels of fucosylation, glucosylation, or galactosylation on H. pylori LPS fluctuate; so, binding affinity of H. pylori to surfactant D may also be modified leading to altered host immune response (Khamri et al., 2005Go). In summary, fucosylated structures on bacterial LPS, through molecular mimicry, enhancing adhesion and modulating host immune response, are believed to aid bacteria in adapting effectively to their niche to maintain a persistent infection.

Core {alpha}1,3 fucosylation
Some {alpha}1,3/4 FucTs can transfer fucose from GDP-Fuc to the innermost GlcNAc moiety of the chitobiose unit of the core Asn-linked glycans at an {alpha}1,3 linkage (Figure 2A). These enzymes, present in insects, plants, parasites, C. elegans, and Drosophila melanogaster but absent in mammals and bacteria (Staudacher et al., 1995Go; Fabini et al., 2001Go; Paschinger et al., 2005Go) are named core {alpha}1,3 FucTs, which also belong to CAZY family 10 (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html).

One major feature of plant N-glycans is the presence of ß1,2-xylose and {alpha}1,3-fucose on the trimannosyl core region (Wilson and Altmann, 1998Go; van Die et al., 1999Go; Wilson et al., 2001Go; Leonard et al., 2004Go; Castilho et al., 2005Go). The core {alpha}1,3 FucT isolated from mung beans is capable of transferring fucose to the GlcNAc moiety at the reducing end of N-glycopeptides or related structures carrying GlcNAc2-Man3-GlcNAc2 (Staudacher et al., 1995Go). Similar to plants, schistosomal N-glycan also contains core {alpha}1,3-fucose and core ß1,2 xylose moieties, and these structures on egg-derived glycoproteins have been reported to induce strong Th2 cytokine responses and elicit production of IgG1 (a Th2-associated isotype) but not IgG2b (a Th1-associated isotype) antibodies in S. mansoni-infected C57BL/6 mice (Faveeuw et al., 2003Go).

Comparison of mammalian and H. pylori {alpha}1,3/4 FucTs
As both mammalian and H. pylori {alpha}1,3/4 FucTs have been well characterized, they are the best FucT representatives for inter-kingdom comparison. Alignment of the H. pylori {alpha}1,3/4 FucTs with their mammalian counterparts demonstrated that the significant homology is observed only in a very short region within the catalytic domain where the two {alpha}1,3/4 FucT motifs are localized (Ge et al., 1997Go; Martin et al., 1997Go); yet strikingly, these two enzyme families seem to share remarkable functional similarities.

First, H. pylori {alpha}1,3/4 FucTs lack both an N-terminal tail and a transmembrane domain. Instead, they contain a heptad-repeat region followed by two putative predicted amphipathic helices at the C-terminus (Ge et al., 1997Go; Martin et al., 1997Go; Rasko, Wang, Palcic, et al., 2000Go; Ma et al., 2003Go). The leucine-zipper-like motif in the heptad-repeat region was suggested to mediate dimer formation (Ge et al., 1997Go; Martin et al., 1997Go). Recent thermal denaturation studies of the carboxyl terminal truncated H. pylori FucTs confirmed that the heptad repeats facilitate protein folding thus help to maintain a stable protein structure, which is aligned with the dimer formation model (Lin et al., 2006Go). The amphipathic helices may act as membrane anchors with the hydrophobic face embedded in the membrane and the positive charges interacting with phospholipid headgroups (Ma et al., 2003Go) (Figure 3B). Accordingly, H. pylori and mammalian FucTs share similar domain architecture but with opposite topology (Figure 3). As a consequence, the Type I acceptor recognition site for both mammalian and H. pylori {alpha}1,3/4 FucTs has been localized in their hypervariable stem regions (Legault et al., 1995Go; Xu et al., 1996Go; Nguyen et al., 1998Go; Ma et al., 2003Go) and was determined by a single aromatic residue (Dupuy et al., 1999Go, 2004Go; Ma et al., 2005Go). In addition, the membrane anchor region (N-terminus of mammalian FucTs versus the C-terminal amphipathic helices of H. pylori enzymes) can be truncated without significantly diminishing the enzyme activity, whereas removal of one amino acid at the C-terminus of human FucT V or ten amino acids from N-terminus of H. pylori FucTs (from strains NCTC11639 and UA948) almost completely abolished enzyme activity (Xu et al., 1996Go; Ge et al., 1997Go; Lin et al., 2006Go; Ma et al., 2006Go).

Second, similar to mammalian {alpha}1,3/4 FucTs, the hydroxyl group at C-6 of galactose in Type I and Type II acceptors is essential for recognition by H. pylori {alpha}1,3/4 FucTs (de Vries et al., 1995Go; Du and Hindsgaul, 1996Go; Gosselin and Palcic, 1996Go; De Vries et al., 1997Go; Ma et al., 2006Go). Like human enzymes, H. pylori {alpha}1,3/4 FucTs were able to use both the sialylated Type I and sialylated Type II acceptors (de Vries et al., 1995Go; De Vries et al., 1997Go; Sherwood et al., 2002Go; Rabbani et al., 2005Go; Ma et al., 2006Go). Kinetic studies showed that H. pylori {alpha}1,3/4 FucTs possess kinetic parameters comparable to their mammalian counterparts (de Vries et al., 1995Go; De Vries et al., 1997Go; Nguyen et al., 1998Go; Rabbani et al., 2005Go; Ma et al., 2006Go). In addition, both human and H. pylori {alpha}1,3/4 FucTs seem to catalyze fucose transfer following a sequential mechanism, with binding of donor molecule first followed by the acceptor (Qiao et al., 1996Go; Ma et al., 2006Go). All these functional similarities indicate that mammalian and bacterial {alpha}1,3/4 FucTs very likely share a conserved mechanistic and structural basis for fucose transfer. Such striking inter-kingdom functional similarity might also be present in other FucT subfamilies but remains to be identified.

Recently, it was reported that the heptad-repeat number of H. pylori FutA and FutB protein is directly correlated with the size of O-antigen polymer being fucosylated (Nilsson et al., 2006Go). The authors proposed a model in which FutA and FutB formed heterodimers or homodimers in which the number of heptad repeats controls the distance of the active fucosylation site from a fixed point and thus determines the size of the O-antigen polymer that is fucosylated. Although this fixed point was not explicitly articulated in the study, it should reside in the bacterial inner membrane where both FucT enzymes and O-antigen polymer substrates are anchored. Our previous dynamic light scattering experiments verified that the FutA protein from H. pylori strain ACTC11639 forms a dimer (Ma et al., 2006Go), however two issues remain to be addressed. First, it is not yet known whether dimerization is essential for enzyme activity. If it were not, then the idea that the heptad-repeat number controls the size of fucosylated O-antigen polymer would be independent of dimerization. Second, there is still no evidence that FutA and FutB can form a heterodimer, and no data to indicate whether heterodimer formation is preferable to homodimer formation as suggested by Nilsson (2006). When the heptad-repeat number of FutA and FutB varies significantly, one would expect the heptad-repeat-mediated interaction between homodimers to be much stronger than that between heterodimers. Moreover, some H. pylori strains (i.e., NCTC11639 and UA948) only express one full-length functional FucT protein (either FutA or FutB) (Ge et al., 1997Go; Rasko, Wang, Palcic, et al., 2000Go) where merely homodimers can be formed.


    {alpha}1,6 Fucosylation
 Top
 Abstract
 Introduction
 {alpha}1,2 Fucosylation
 {alpha}1,3/4 Fucosylation
 {alpha}1,6 Fucosylation
 Xyloglucan {alpha}1,2...
 O-Fucosylation
 Sequence homology and structural...
 Abnormal fucosylation in mammals
 Application of FucTs,...
 Concluding remarks
 Supplementary Data
 Acknowledgments
 Conflict of interest statement
 References
 
FucTs that add fucose to the innermost GlcNAc moiety of the chitobiose unit of the core Asn-linked glycans at an {alpha}1,6 linkage are designated {alpha}1,6 FucTs (Figures 1 and 2A) (Miyoshi et al., 1999Go; Yamaguchi et al., 1999Go, 2000Go). {alpha}1,6 FucTs show the highest activity in the presence of Mn2+, Mg2+, and Ca2+ but remain active when the metal cations are absent (Chazalet et al., 2001Go; Ihara et al., 2006Go). Recent kinetic studies supported that human {alpha}1,6 FucT adopts a rapid equilibrium random mechanism for fucose transfer (Ihara et al., 2006Go), in contrast to previous prediction of the ping-pong bi bi mechanism (Takahashi, Ikeda, Tateishi, et al., 2000Go). {alpha}1,6 FucTs are grouped into CAZY family 23 (http://afmb.cnrs-mrs.fr/CAZY/fam/acc_GT.html).

{alpha}1,6 Fucosylation in mammals
The synthesis of the oligosaccharide precursor (Glc3Man9GlcNAc2) of N-glycans begins in the cytoplasm and is trimmed and assembled at the ER membrane on a dolichyl-pyrophosphate carrier by a series of ER-localized glycosidases and glycosyltransferases. {alpha}1,6 FucT, acting at late Golgi cisternae (Kornfeld and Kornfeld, 1985Go), requires an unsubstituted ß1,2-linked GlcNAc on the {alpha}1,3-mannose arm of the core N-glycans (Longmore and Schachter, 1982Go; Voynow et al., 1991Go; Uozumi et al., 1996Go).

Human {alpha}1,6 FucT, encoded by fut8, is widely expressed in mammalian tissues (Miyoshi et al., 1999Go; Yamaguchi et al., 1999Go, 2000Go). Two Arg residues, in motif I (see Sequence homology and structural prediction of FucT subfamilies) common to {alpha}1,2, {alpha}1,6 and O-FucTs, were shown to be involved in GDP-Fuc binding (Takahashi, Ikeda, Tateishi, et al., 2000Go). A recent study showed that the sugar moiety of GDP-Fuc did not contribute significantly to the binding and recognition by FUT8. Instead, the guanine nucleotide and diphosphate portions of GDP-Fuc are more important (Ihara et al., 2006Go).

{alpha}1,6 Fucosylation in S. mansoni, insects, and C. elegans
The schistosomal egg glycoproteins contain non-fucosylated, {alpha}1,6-monofucosylated, core {alpha}1,3/{alpha}1,6-difucosylated, and xylosylated/{alpha}1,6-fucosylated forms of N-glycans, whereas the latter two forms of glycan are not characteristic for cercariae and adult worms (Khoo et al., 1997bGo, 2001Go; Faveeuw et al., 2003Go). The core {alpha}1,6 fucosylated diantennary N-glycan in schistosomal adult worms is composed of Hex3–4HexNAc6–12Fuc1–6 carrying dimers in the form of not only LacNAc (Type II) but also LacdiNAc (GalNAcß1,4GlcNAc) with or without fucose {alpha}1,3 linked to GlcNAc residues in the antennae (Wuhrer et al., 2006Go). Indeed, the presence of a significant amount of LacdiNAc structure is a unique feature of schistosomal glycoconjugates.

N-glycan core structure with double fucosylation at {alpha}1,6- and core {alpha}1,3-linkages, seemingly absent in vertebrates, is present not only in schistosomes (Khoo et al., 1997aGo,b, 2001) but also in C. elegans (Paschinger et al., 2005Go) and insects (i.e., Mamestra brassicae, honey bee and Drosophila) (Staudacher et al., 1991Go; Staudacher and Marz, 1998Go). The core {alpha}1,3 FucT of Drosophila prefers core {alpha}1,6-fucosylated glycans over the non-fucosylated forms (Fabini et al., 2001Go), and honeybee-venom-gland extracts exhibit the core {alpha}1,3 activity with monofucosylated (at 1,6 linkage) N-glycan acceptor (Staudacher et al., 1991