α1,3-Fucosyltransferases (Fuc-Ts) convert N-acetyllactosamine (LN, Galβ1-4GlcNAc) to Galβ1-4(Fucα1-3)GlcNAc, the Lewis x (CD15, SSEA-1) epitope, which is involved in various recognition phenomena. We describe details of the acceptor specificity of α1,3-fucosyltransferase IX (Fuc-TIX). The unconjugated N- and O-glycan analogs LNβ1-2Man, LNβ1-6Manα1-OMe, LNβ1-2Manα1-3(LNβ1-2Manα1-6)Manβ1-4GlcNAc, and Galβ1-3(LNβ1-6)GalNAc reacted well in vitro with Fuc-TIX present in lysates of appropriately transfected Namalwa cells. Fuc-TIX reacted well with the reducing end LN of GlcNAcβ1-3′LN (underscored site reacted) and GlcNAcβ1-3′LNβ1-3′LN (both LNs reacted), but very poorly with the reducing end LN of LNβ1-3′LN. However, Fuc-TIX reacted significantly better with the non-reducing end LN as compared to the other LN units in the glycans LNβ1-3′LN and LNβ1-3′LNβ1-3′LNβ1-3′LN, confirming our previous data on LNβ1-3′LNβ1-OR. In contrast, the sialylated glycan Neu5Acα2-3′LNβ1-3′LNβ1-3′LNβ1-3′LN was fucosylated preferentially at the two most reducing end LN units. We conclude that Fuc-TIX is a versatile α1,3-Fuc-T, that (1) generates distal Lewis x epitopes from many different acceptors, (2) possesses inherent ability for the biosynthesis of internal Lewis x epitopes on growing polylactosamine backbones, and (3) fucosylates the remote internal LN units of α2,3-sialylated i-type polylactosamines.
Received on January 25, 2002; revised on February 21, 2002; accepted on February 21, 2002
α1,3-Fucosyltransferases (Fuc-Ts) transfer fucose from GDP-Fuc to N-acetyllactosamine (LN) forming the Lewis x (Lex, CD15, SSEA-1) epitope Galβ1-4(Fucα1-3)GlcNAc. Six human α1,3‐Fuc-Ts have been cloned: Fuc-TIII–VII and Fuc-TIX (Goelz et al., 1990; Kukowska-Latallo et al., 1990; Kumar et al., 1991; Lowe et al., 1991; Koszdin and Bowen, 1992; Weston et al., 1992a,b; Natsuka et al., 1994; Sasaki et al., 1994; Kaneko et al., 1999a). The α1,3-Fuc-T products, Lex and related carbohydrate structures, are involved in various recognition phenomena, such as recruitment of leukocytes to sites of infection and their homing to lymph nodes (Rosen, 1999; Vestweber and Blanks, 1999; Lowe, 2001), tumor metastasis (Kannagi, 1997), compaction of the embryo (Fenderson et al., 1984; Eggens et al., 1989), neural development (Marani and Mai, 1992; Götz et al., 1996; Streit et al., 1996; Ashwell and Mai, 1997), angiogenesis (Nguyen et al., 1993; Halloran et al., 2000), fertilization (Johnston et al., 1998; Oehninger et al., 1998), and bacterial adhesion (Ilver et al., 1998; Herron et al., 2000).
Distinct acceptor specificity patterns are emerging among the α1,3-Fuc-Ts. Fuc-TVII readily fucosylates α2,3-sialylated type 2 N-acetyllactosamine (LN) but reacts very weakly if at all with neutral LN (Natsuka et al., 1994; Sasaki et al., 1994). The opposite is true for Fuc-TIV (Kumar et al., 1991) and Fuc‐TIX (Kaneko et al., 1999a). The three very homologous enzymes, Fuc-TIII, Fuc-TV, and Fuc-TVI, react with both sialylated and nonsialylated LN (Koszdin and Bowen, 1992; Weston et al., 1992a,b). Furthermore, the most versatile α1,3‐Fuc-Ts, that is, Fuc-TIII and Fuc-TV, can react with both type II (Galβ1-4GlcNAc) and type I (Galβ1-3GlcNAc) lactosamines (Kukowska-Latallo et al., 1990; Weston et al., 1992a). In addition, the α1,3-Fuc-Ts show distinct preferences as to what position along a polylactosamine chain they act on. Fuc‐TVII preferentially reacts with α2,3-sialylated LN units at the nonreducing ends of polylactosamines (Niemelä et al., 1998), whereas Fuc-TIX preferentially fucosylates nonreducing end LN units of neutral polylactosamines (Nishihara et al., 1999). Both Fuc-TIV and Fuc-TV preferentially fucosylate the inner LN units in both sialylated and nonsialylated polylactosamines: Fuc-TV prefers the reducing end LN unit (Pykäri et al., 2000), whereas Fuc-TIV prefers middle LN units (Niemelä et al., 1998). Fuc-TIII and Fuc-TVI similarly fucosylate the inner LN units of short neutral polylactosamines (Nishihara et al., 1999), but their site specificity on longer and/or sialylated polylactosamines has not yet been studied.
The differences in the acceptor profiles could reflect the various biological roles of the different α1,3-Fuc-Ts. For example, the leukocyte Fuc-Ts, Fuc-TIV and Fuc-TVII, collaborate in the generation of functional selectin ligands (Homeister et al., 2001). They seem to be at least partly specialized in the way that Fuc-TVII directs the expression of P-selectin binding glycoforms of P-selectin glycoprotein ligand-1 and controls the rolling frequency of leukocytes, whereas Fuc-TIV directs the expression of E-selectin binding glycoforms of E-selectin ligand-1 and dictates the rolling velocity (Huang et al., 2000; Weninger et al., 2000). Fuc-TIX, which is also expressed in leukocytes (Kaneko et al., 1999a), has been shown to direct the synthesis of CD15 (nonsialylated Lex) in mature granulocytes (Nakayama et al., 2001), but its possible role in selectin ligand biosynthesis remains to be elucidated.
The Fuc-TIX (FUT9) gene sequence is highly conserved between human, mouse, rat, and hamster (Kudo et al., 1998; Kaneko et al., 1999a; Baboval et al., 2000; Patnaik et al., 2000), indicating that it has been under strong selective pressure during evolution, and thus suggesting that it has an essential role in the organisms. The human Fuc-TIX gene is localized to chromosome 6q16 (Kaneko et al., 1999b). It has a single exon, and it is transcribed as several transcripts of different size. The 2.5-kb transcript is ubiquitously expressed and particularly strongly in fetal tissues (Cailleau-Thomas et al., 2000). The expression of the 3.0-kb transcript is restricted to brain, stomach, spleen, and peripheral blood leukocytes (Kaneko et al., 1999a), whereas that of the 12-kb transcript is restricted to brain, kidney, placenta, and pancreas (Cailleau-Thomas et al., 2000). The most abundant expression level of this enzyme is found in both the developing and mature brain in human, rat and mouse (Kaneko et al., 1999a; Baboval et al., 2000; Cailleau-Thomas et al., 2000). Fuc-TIX expression has been shown to increase during neural differentiation of PC19 EC cells induced by retinoic acid (Osanai et al., 2001). Pax6, a transcription factor involved in brain patterning and neurogenesis, seems to control the spatially and temporally restricted expression of the Lex epitope in the developing brain by regulating Fuc-TIX gene expression (Shimoda et al., 2002). All this suggests that Fuc‐TIX expression is spatially and temporally regulated in different tissues and developmental stages and may have a specific role in the development of the brain.
The knowledge of the detailed acceptor specificity will help in clarifying the biological roles of Fuc-TIX. It has been shown that Fuc-TIX preferentially fucosylates the nonreducing end LN unit of LNβ1-3′LNβ1-OR (Nishihara et al., 1999). Fuc‐TIX does not react with type 1 lactosamines, α2,3-sialylated lactosamine, or lactose, but reacts with type 2 blood group H determinant, forming the Lewis y epitope, and with GalNAcβ1-4GlcNAc (LacdiNAc) (Cailleau-Thomas et al., 2000). In this study, further details of the acceptor specificity of Fuc-TIX were examined and the site-specificity of its action on polylactosamines was determined.
Parts of this data were presented as a poster at Glyco XVI, August 19–24, 2001, The Hague, The Netherlands (Toivonen et al., 2001a), and at Glycobiology 2001, November 14–17, 2001, San Francisco, CA (Toivonen et al., 2001b).
Fuc-TIX reactivities of disaccharide acceptors
The initial transfer rate of fucose to LN by lysates of Namalwa cells transfected with full-length human Fuc-TIX was on average 5 nmol fucose per h per mg lysate protein at 5 mM acceptor concentration. The type 1 acceptor (Galβ1-3GlcNAc) reacted only barely above the detection limit (4% compared to LN). LacdiNAc reacted well. GlcNAcβ1-4GlcNAc (chitobiose) reacted appreciably, whereas lactose was virtually nonreactive (Table I).
Fuc-TIX reactivities of LNs carrying acidic substituents
Charged groups had a major impact on the reactivity of the acceptor. Neu5Acα2-3′LN and Neu5Acα2-6′LN reacted at rates representing only 5% and 3%, respectively, of the reactivity of the LN control (Table I). 6-Sulfated LN was a much better acceptor for Fuc-TIX than the sialylated versions, as it reacted at a rate of 40% as compared to neutral LN (Table I).
Fuc-TIX reactivities of neutral trisaccharide acceptors containing LN
Neutral trisaccharides showed widely different Fuc-TIX reactivities. For example Fucα1-2′LN and GlcNAcβ1-3′LN showed decreased, yet strong reactivities, that is, 51% and 71%, respectively, of LN control. On the contrary, the Galβ1‐4′LN trisaccharide was virtually not an acceptor for Fuc-TIX at all. The addition of Gal to the reducing end of LN slightly hampered the acceptor activity, as could be seen with LNβ1-6Gal and LNβ1-3Galβ1-OMe (Table I).
Fuc-TIX reactivity of N- and O-glycan analogs
We also tested the Fuc-TIX reactivity of larger oligosaccharides representing N- and O-glycan structures. The N‐glycan analog LNβ1-6Manα-OMe showed the highest reactivity of all the single site acceptors tested. LNβ1‐2Manα1-3(LNβ1-2Manα1-6)Manβ1-4GlcNAc showed a reactivity level close to the control LN, whereas LNβ1-2Man reacted less well. The O-glycan Galβ1-3(LNβ1-6)GalNAc showed a high reactivity as well (Table I).
The site specificity of Fuc-TIX on linear polylactosamines
The site specificity of Fuc-TIX was analyzed by exoglycosidase digestions followed by paper chromatography, as shown in Figures 1–4. The site-specificity data derived this way is summarized in Figure 5. Surprisingly, the overall reactivity of most of the multisite polylactosamine acceptors was lower than that of the LN disaccharide (Table I).
Our earlier work has shown that whereas unconjugated LN is a good acceptor for Fuc-TIX, the internal LN of the polylactosamine LNβ1-3′LNβ1-OR is not (Nishihara et al., 1999). In the present studies we found that LNβ1-3′LN reacted with Fuc-TIX initially at half the rate of unconjugated LN (Table I). Degradation of the resulting early fucosylation product with mixed β-galactosidase and β-N-acetylhexosaminidase gave two [14C]-labeled products that were separated and identified by paper chromatography as Lexβ1-3′LN (peak 1 in Figure 1; 1327 cpm) and Lex (peak 2 in Figure 1; 109 cpm). The glycan Lexβ1-3′LN resisted the enzymatic degradation (Arakawa et al., 1974), whereas the isomeric product LNβ1-3′Lex was degraded to Lex trisaccharide. The data imply that the LN unit at the nonreducing end of the tetrasaccharide had accepted [14C]fucose at a rate which was 43% of the rate at which free LN reacted, whereas the GlcNAc at the reducing end of the tetrasaccharide had reacted at a rate that was only 3% of the rate of free LN (Figure 5).
GlcNAcβ1-3′LN was a better acceptor than LNβ1-3′LN (Table I). Enzymatic degradation of the fucosylation product of GlcNAcβ1-3′LN (Figure 2a) indicated that the acceptor had reacted only at the LN unit; the nonreducing end GlcNAc had not served as an acceptor site. Hence the data indicate that the reducing end LN of GlcNAcβ1-3LN was a surprisingly good acceptor for Fuc-TIX. The reducing end LN of LNβ1-3LN did not react essentially at all. GlcNAcβ1-3′LNβ1-3′LN was a better acceptor than LNβ1-3′LN as well (Table I). Enzymatic degradation of its fucosylation product (Figure 2b) showed that it had reacted quite well at both of its LN sites; the reactivity of the reducing end LN was increased ninefold as compared to LNβ1-3′LN.
In LNβ1-3′LN and LNβ1-3′LNβ1-3′LNβ1-3′LN, Fuc-TIX strongly preferred the terminal, nonreducing end site, the other sites reacting only very weakly (Figures 1, 3, and 5). Whereas sialylated LN reacted only marginally, the sialylated polylactosamine reacted at a rate of 51% of that of the LN control. Sialylation reversed the site specificity of Fuc-TIX on this long polylactosamine; the two most reducing end LN units in Neu5Acα2-3′LNβ1-3′LNβ1-3′LNβ1-3′LN showed good reactivity, as compared to the poor reactivity of the two LN units adjacent to Neu5Ac (Figures 3 and 5).
The site specificity of Fuc-TIX on a branched polylactosamine
Whereas the total reactivity of a branched polylactosamine acceptor LNβ1-3′(LNβ1-6′)LN was in the same range as that of the control LN, not all LN units in this glycan reacted equally. Fuc-TIX showed a preference toward the β1,3-linked arm, LNβ1-3′(LNβ1-6′)LN, over the β1,6-linked arm, LNβ1-3′(LNβ1-6′)LN. The LN unit at the branch point, LNβ1-3′(LNβ1-6′)LN, did not react at all (Figures 4–5).
The spatially and temporally restricted expression of Fuc-TIX transcripts, as well as its high level of sequence conservation between mammalian species, suggests that the enzyme participates in the biosynthesis of fucosylated lactosamines with important functions. The understanding of the detailed acceptor specificity of human Fuc-TIX, as well as its comparison to other Fuc‐Ts, will clarify its function. Our present results show that Fuc-TIX is a versatile enzyme capable of converting distal, nonreducing end LN units into distal Lex epitopes on many different kinds of unconjugated oligosaccharides. Fuc-TIX has been shown to synthesize the spatially and temporally regulated Lex epitope in the developing rat brain (Shimoda et al., 2002). The detailed structure that carries the epitope has not been solved, although N-linked polylactosamines on a proteoglycan have been implied (Shimoda et al., 2002).
The Km values of Fuc-TIX for the acceptors are not known, but the 5 mM acceptor concentrations used in the present expreriments probably represent saturating conditions for the enzyme. Therefore the reactivities presented here are likely to reflect saturation kinetics governed by Vmax values for the acceptors. The in vivo concentrations of the acceptors in the Golgi are difficult to determine and presently unknown. It is possible that the glycoconjugate acceptors are present in the Golgi in relatively high local concentrations. However, if the concentrations are considerably lower, the in vivo reactivities of the acceptors will be affected by their Km values as well.
Fuc-TIX was unable to transfer to distal GlcNAc residues at the nonreducing termini of GlcNAcβ1-3′LN and GlcNAcβ1-3′LNβ1-3′LN. Based on this, we suggest that LN is probably the smallest functional acceptor epitope for human Fuc-TIX. The importance of the galactose residue for the fucosyl transfer to the neighboring GlcNAc is further emphasized by the observation that even some individual free hydroxyl groups of the galactose residue are needed for the reactivity of the LN unit. Galβ1-4′LN, Neu5Acα2-6′LN, and the branch-bearing LN unit of LNβ1-3′(LNβ1-6′)LN were not functional acceptors for Fuc TIX. In contrast, Fucα1-2′LN and GlcNAcβ1-3′LN were functional acceptors. Together these data suggest that the hydroxyl groups at positions 4′ and 6′ of LN, but not those at positions 2′ and 3′, may be involved in the binding of Fuc-TIX to LN. The free 6′ hydroxyl group may be a general requirement for α1,3-Fuc-T action, as Fuc-Ts III–VI have been shown to be unable to react with 6′ deoxy LN (de Vries et al., 1995, 1997).
Fuc-TIX efficiently fucosylated LNβ1-6Manα1-OMe and LNβ1-2Man. In addition to being N-glycan branches, LNβ1‐2Man and LNβ1-6(LNβ1-2)Man structures occur in the brain O-linked to serine or threonine (Finne et al., 1979; Chiba et al., 1997; Smalheiser et al., 1998; Chai et al., 1999). α2,3‐Sialylated LNβ1-2Manα1-Ser/Thr on dystroglycan is required for its binding to laminin (Chiba et al., 1997). Disturbances in the synthesis of this structure lead to muscular dystrophy and neuronal migration disorder (Yoshida et al., 2001). Because sialylation and fucosylation of LNβ1-2Manα1-Ser/Thr seem to be mutually exclusive (Chiba et al., 1997; Smalheiser et al., 1998; Chai et al., 1999), it is possible that fucosylation regulates the binding of dystroglycan to laminin and therefore cell adhesion.
The Fuc-TIX reactivities of polylactosamines were strongly influenced by small changes in the acceptor structure. Even changes at locations relatively far away from the reacting GlcNAc residue had a strong effect in some cases. For example, the LN unit of GlcNAcβ1-3′LN reacted well, whereas the analogous reducing end LN unit of LNβ1-3′LN reacted extremely slowly. Futhermore, both LN units of GlcNAcβ1-3′LNβ1-3′LN reacted equally well, unlike those of LNβ1-3′LN. We have previously described a similar long distance effect, caused by an α1,3-linked fucose residue in a polylactosamine chain, on the acceptor reactivity of an adjacent LN with a β1,6-GlcNAc transferase (Leppänen et al., 1997a). These observations demonstrate that modifications situated relatively far away from the actual acceptor monosaccharide may affect the binding of the acceptor oligosaccharide to the active site of the enzyme. Alternatively, two substrate molecules may be involved in the regulation of the reactions, the second one acting as an allosteric effector.
The high reactivities of the LN units of GlcNAcβ1-3′LN and GlcNAcβ1-3′LNβ1-3′LN imply that Fuc-TIX may participate in biosynthesis of internally fucosylated polylactosamines, if the resulting products of the GlcNAcβ1-3′Lex type serve as acceptors for subsequent β1,4-galactosylation. We have previously shown that the β4Gal T(s) in bovine milk catalyze these reactions readily in vitro (Räbinä et al., 1998). Cells transfected with Fuc-TIX show (Neu5Acα2-3Galβ1-4GlcNAcβ1-3Galβ1-4(Fucα1-3)GlcNAc (VIM-2) reactivity (Nakayama et al., 2001), which implies that they express Neu5Acα2-3′LNβ1-3′Lex epitopes. In the light of our data the VIM-2 epitope on Fuc-TIX transfected cells may be synthesized on growing polylactosamines, where α1,3-fucosylation of the penultimate GlcNAc preceeds the distal β1,4-galactosylation and sialylation reactions.
Our experiments confirmed the observations that Fuc-TIX works poorly with Neu5Acα2-3′LN (Kaneko et al., 1999b; Nishihara et al., 1999; Cailleau-Thomas et al., 2000). Similarly, the LN unit adjacent to Neu5Ac in Neu5Acα2-3′LNβ1-3′LNβ1-3′LNβ1-3′LN reacted only very weakly. However, the two LN units located furthest away from the Neu5Ac reacted threefold better. The data imply that even sialylated polylactosamines may be fucosylated quite effectively by Fuc-TIX if the i-type chains become long enough before terminal sialylation. In contrast to this, the neutral glycan LNβ1-3′LNβ1-3′LNβ1-3′LN reacted almost exclusively at the nonreducing end LN unit (underscored). The difference represents yet another example of a long-distance effect of a distant structural element (Neu5Ac) on the catalysis rate. The capability of Fuc-TIX to fucosylate the inner LN units of a polylactosamine chain, either during elongation or in preexisting long sialylated polylactosamines, suggests that it may have a role in the generation of the sialylated, multiply fucosylated polylactosamine selectin ligands (Stroud et al., 1996; Wilkins et al., 1996; Handa et al., 1997). The polylactosamine site-specificity profile of Fuc-TIX differs from those of the other α1,3-Fuc-Ts studied so far (Figure 6). The strong preference for the distal LN unit of neutral polylactosamines seems to be unique for Fuc-TIX (Nishihara et al., 1999). However, α2,3-sialylation of the polylactosamine acceptor shifts the site specificity of Fuc-TIX toward those of Fuc-TIV and Fuc-TV. In fine detail the site specificity of Fuc‐TIX on sialylated polylactosamines seems to be intermediate between Fuc-TIV and Fuc-TV. Fuc-TIV prefers middle LN units (Niemelä et al., 1998), and Fuc-TV shows a strong preference toward the most reducing end LN unit (Pykäri et al., 2000), whereas Fuc-TIX reacts equally well with the two most reducing end LN units.
In conclusion, interesting new features of human Fuc-TIX emerge from our experiments. First, Fuc-TIX is a versatile enzyme, capable of reacting with many types of acceptor oligosaccharides. Second, the interaction between the enzyme and its i-type polylactosamine acceptors is influenced by long-distance effects of acceptor modification. Our data show that Fuc-TIX has its own unique acceptor and site-specificity profile that is distict from the other α1,3-Fuc-Ts. The clarification of the detailed specificities of the individual members of glycosyltransferase families will eventually help us understand the different biological roles of the various isoenzymes.
Materials and methods
Transfected cells and cell lysates
The transfection of Namalwa cells stably expressing Fuc-TIX has been described previously (Kaneko et al., 1999a). For the enzyme assays the cells were sonicated in 20 mM HEPES buffer, pH 7.2, containing 1% Triton X-100.
LN, Galβ1-3GlcNAc, chitobiose, and Galβ1-4′LN were from Sigma. Fucα1-2′LN and Neu5Acα2-6′LN were from Dextra (Reading, UK). Neu5Acα2-3′LN was from Oxford Glycosystems (Abingdon, UK). Lactose was from BDH Chemicals (Poole, UK). 2′-Fucosyllactose was from Biocarb (Lund, Sweden). LNβ1-2Man, LNβ1-6Manα1-OMe, GalNAcβ1-4GlcNAcβ1-OMe, LNβ1-3Galβ1-OMe, LNβ1-6Gal, and LNβ1-3′(LNβ1-6′)LN were synthesized enzymatically as described elsewhere (Renkonen et al., 1990; Pykäri et al., 2000). Galβ1-3(LNβ1-6)GalNAc was synthesized as described in Maaheimo et al. (1995); GlcNAcβ1-3′LN and LNβ1-3′LN as described in Renkonen et al. (1991) and GlcNAcβ1-3′LNβ1-3′LN as described in Leppänen et al. (1997b). LNβ1-3′LNβ1-3′LNβ1-3′LN was synthesized as described elswhere (Alais and Veyrières, 1990). Neu5Acα2-3′LNβ1-3′LNβ1-3′LNβ1-3′LN was synthesized by sialylating LNβ1-3′LNβ1-3′LNβ1-3′LN with rat recombinant α2,3-N-sialyltransferase (Calbiochem, La Jolla, CA) as described Pykäri et al. (2000). LNβ1-2Manα1-3(LNβ1-2Manα1-6)Manβ1-4GlcNAc was a kind gift of G. Strecker (Université des Sciences et Technologies de Lille, Villeneuve D’Ascq, France). 6-SO3-LN was obtained by galactosylating 6-SO3-GN (Sigma) with bovine milk β1,4-galactosyltransferase (Sigma) essentially as described in Brew et al. (1968), purified on a MonoQ 10/10 column as described in Maaheimo et al. (1995), and analyzed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (Nyman et al., 1998), where it gave a single peak at m/z 461.96, which was assigned to [M‐H]– of SO3Hex1HexNAc1 (calculated m/z 462.20). All the acceptors were quantitated by comparing their UV 214 absorbance to external GlcNAc and Neu5Ac (Sigma; >99% and >98%, respectively), which were weighed, dissolved in water, and run in gel filtration high-performance liquid chromatography on a Superdex Peptide 10/30 column (Amersham Pharmacia Biotech) with UV 214 detection.
GDP-[14C]fucose [1 nmol, 100,000 cpm/nmol, prepared by mixing GDP-fucose (Sigma) and GDP-[14C]fucose (Amersham Pharmacia Biotech)] and the individual acceptors (50 nmol) were incubated for 1 h at 37°C in 10 µl 50 mM 3-(N-morpholino)propanesulfonic acid, pH 7.2, containing 10 mM HEPES, pH 7.2, 10 mM fucose, 5 mM ATP, and 0.5% Triton X-100 and lysates of Namalwa cells transfected with Fuc-TIX (28 µg protein, assayed by the BCA kit of Pierce). The reactions were terminated by adding 10 µl ethanol followed by 100 µl ice-cold water.
The acceptors and reaction products were desalted and separated from unreacted donor in a mixed bed of ion exchange resins (Dowex AG-50W [H+ form] and Dowex AG‐1 [acetate form] from Bio-Rad). The neutral oligosaccharides were eluted with water, and the singly charged sialylated oligosaccharides were eluted with 0.5 mM acetic acid as described in Koszdin and Bowen (1992). The reaction products were quantitated by subjecting aliquots of the mixtures of unreacted acceptor and labeled product to liquid scintillation counting. A background value (amounting to 3% of the total radioactivity associated with the reference acceptor LN), obtained by incubating the lysate with GDP-[14C]fucose without an exogenous acceptor, was subtracted from all readings.
The site specificity of Fuc-TIX action on polylactosamine acceptors was analyzed by treating the radiolabeled reaction products of sialylated acceptors with Arthrobacter ureafaciens sialidase (EC 188.8.131.52 Roche Molecular Biochemicals, Basel, Switzerland), and the resulting desialylated glycans as well as the neutral reaction products with a mixture of jack bean β‐galactosidase (EC 184.108.40.206 Seikagaku, Tokyo, or Glyko, Novato, CA) and jack bean β-N-acetylhexosaminidase (EC 220.127.116.11, Sigma) essentially as described in Niemelä et al. (1998). Descending paper chromatography of the digests was carried out as described elsewhere (Renkonen et al., 1989; Niemelä et al., 1998). The digestion products were identified by comparing their migration rates with those of the synthetic oligosaccharides Lexβ1-3′LNβ1-3′LNβ1-3′LN, Lexβ1-3′LNβ1-3′LN, Lexβ1-3′LN and Lex, described elsewhere (Niemelä et al., 1995, 1998, 1999; Pykäri et al., 2000).
This work was supported by the Academy of Finland (grants 38042, 40901, and 44318); the Technology Development Center, Helsinki (grant 40896); the Emil Aaltonen Foundation; and the Finnish Graduate School of Bioorganic Chemistry.
Fuc-T, fucosyltransferase; LacdiNAc, GalNAcβ1-4GlcNAc; Lex, Lewis x; LN, N-acetyllactosamine; VIM-2 (Neu5Acα2-3Galβ1-4GlcNAcβ1-3Galβ1-4(Fucα1-3)GlcNAc.
To whom correspondence should be addressed; E-mail: email@example.com
|Acceptor||Relative Fuc-TIX reactivitya|
|Acceptor||Relative Fuc-TIX reactivitya|
aAll of the assays were carried out twice, the error from the mean being below 8%.
bA dash indicates reactivity below 0.03.
cError from the mean 20%.
dError from the mean 14%.
2Institute of Biotechnology and Department of Biosciences, University of Helsinki, P.O. Box 56, 00014 Helsinki, Finland; 3Division of Cell Biology, Institute of Life Science, Soka University, 1-236 Tangi-cho, Hachioji, Tokyo 192-8577, Japan; 4Laboratory of Gene Function Analysis, Institute of Molecular and Cell Biology (IMCB), National Institutes of Advanced Industrial Science and Technology (AIST), Central-2, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan; and 5Biomedicum and Haartman Institute, Department of Bacteriology and Immunology, University of Helsinki, P.O. Box 63, 00014 Helsinki, Finland