Glycobiology Advance Access originally published online on June 13, 2007
Glycobiology 2007 17(8):857-867; doi:10.1093/glycob/cwm055
Molecular basis for acceptor substrate specificity of the human ß1,3-glucuronosyltransferases GlcAT-I and GlcAT-P involved in glycosaminoglycan and HNK-1 carbohydrate epitope biosynthesis, respectively
2 UMR 7561
3 UMR 7036 CNRS-Université Henri Poincaré Nancy 1, Faculté de Médecine, BP 184, 54505 Vandœuvre-lès-Nancy, France
4 INSERM U602, Université de Paris Sud XI, 94807 Villejuif, France
5 Department of Pathology and Neurosciences, University of Dundee, Dundee DD1 9SY, UK
1 To whom correspondence should be addressed; Tel: +33 383 68 39 72; Fax: +33 383 68 39 59; E-mail: sfg{at}medecine.uhp-nancy.fr
Received on March 8, 2007; revised on May 7, 2007; accepted on May 11, 2007
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
The human ß1,3-glucuronosyltransferases galactose-ß1,3-glucuronosyltransferase I (GlcAT-I) and galactose-ß1,3-glucuronosyltransferase P (GlcAT-P) are key enzymes involved in proteoglycan and HNK-1 carbohydrate epitope synthesis, respectively. Analysis of their acceptor specificity revealed that GlcAT-I was selective toward Galß1,3Gal (referred to as Gal2-Gal1), whereas GlcAT-P presented a broader profile. To understand the molecular basis of acceptor substrate recognition, we constructed mutants and chimeric enzymes based on multiple sequence alignment and structural information. The drastic effect of mutations of Glu227, Arg247, Asp252, and Glu281 on GlcAT-I activity indicated a key role for the hydrogen bond network formed by these four conserved residues in dictating Gal2 binding. Investigation of GlcAT-I determinants governing Gal1 recognition showed that Trp243 could not be replaced by its counterpart Phe in GlcAT-P. This result combined with molecular modeling provided evidence for the importance of stacking interactions with Trp at position 243 in the selectivity of GlcAT-I toward Galß1,3Gal. Mutation of Gln318 predicted to be hydrogen-bonded to 6-hydroxyl of Gal1 had little effect on GlcAT-I activity, reinforcing the role of Trp243 in Gal1 binding. Substitution of Phe245 in GlcAT-P by Ala selectively abolished Galß1,3Gal activity, also highlighting the importance of an aromatic residue at this position in defining the specificity of GlcAT-P. Finally, substituting Phe245, Val320, or Asn321 in GlcAT-P predicted to interact with N-acetylglucosamine (GlcNAc), by their counterpart in GlcAT-I, moderately affected the activity toward the reference substrate of GlcAT-P, N-acetyllactosamine, indicating that its active site tolerates amino acid substitutions, an observation that parallels its promiscuous substrate profile. Taken together, the data clearly define key residues governing the specificity of ß1,3-glucuronosyltransferases.
Key words: ß1,3-glucuronosyltransferases / acceptor substrate specificity / glycosaminoglycans / kinetics / site-directed mutagenesis
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
The human galactose-ß1,3-glucuronosyltransferase I (GlcAT-I) and galactose-ß1,3-glucuronosyltransferase P (GlcAT-P) are galactose-ß1,3-glucuronosyltransferases responsible for the formation of GlcAß1,3Gal linkages in different glycoconjugates. Phylogeny analysis indicated previously that these enzymes belong to a glycosyltransferase family including at least 119 members present in vertebrates, invertebrates, and plants (Fondeur-Gelinotte et al. 2006
).
GlcAT-I is an essential enzyme involved in heparin/heparan sulfate and chondroitin/dermatan sulfate biosynthesis (Bai et al. 1999). This inverting glycosyltransferase catalyzes the transfer of 
-D-glucuronic acid (GlcA) on Galß1, 3Galß1,4Xylß-O, completing the formation of the common growing linker region that is attached to a serine side-chain of the core protein of proteoglycans. There is increasing evidence for the vital role of proteoglycans in many biological processes such as cell proliferation, cell adhesion, blood clotting, inflammation, and wound healing (Raman et al. 2005). In most cases, the biological activities of proteoglycans are governed by interactions of their glycosaminoglycan (GAG) chains with growth factors, cytokines, morphogens, and a variety of protein ligands. Thus, GAGs as well as the glycosyltransferases involved in their synthetic pathway have recently received much attention as pharmacological targets (Sakai et al. 2006). The human GlcAT-I cDNA has been first isolated based on the sequence of GlcAT-P cloned from rat brain (Terayama et al. 1997; Kitagawa et al. 1998), thus allowing structural studies to be performed (Pedersen et al. 2000; Kakuda et al. 2004; Shiba et al. 2006). Since the enzyme operates at the branching point of the biosynthesis cascade of GAGs, GlcAT-I appeared to regulate the overall process of GAG synthesis. Its interest as a pharmacological target has been emphasized in osteoarthritis and cartilage repair (Venkatesan et al. 2004).
GlcAT-P catalyzes the addition of GlcA onto the terminal N-acetyllactosamine (Galß1,4GlcNAc) disaccharide of glycoproteins to form the HNK-1 (for human natural killer) epitope precursor GlcAß1,3Galß1,4GlcNAc (Oka et al. 1992). This trisaccharide sequence then serves as the acceptor substrate for 3-O-sulfotransferase, generating the HNK-1 epitope which is spatially and temporally regulated during the development of the nervous system. The HNK-1 carbohydrate is found on many different glycoproteins, glycolipids, and proteoglycans present in the extracellular matrix and on the cell surface. Some of these molecules are implicated in cell–cell and cell–substratum interactions, such as the neural cell adhesion molecule, myelin-associated glycoprotein, L1 protein, and PO glycoprotein (Bollensen and Schachner 1987). Thus, the GlcAT-I, GlcAT-P appears to play key roles in cell–cell and cell–extracellular matrix interactions.
Structural studies suggested that GlcAT-I and GlcAT-P are organized as dimers, each subunit with a Rossman-like fold divided into two regions connected by a so-called DXD motif (Breton and Imberty 1999). The N-terminal region comprises the UDP-sugar-binding site and is terminated, in GlcAT-I, by a DDD sequence involved in the coordination of Mn2+ cations essential for enzyme activity (Gulberti et al. 2003). The C-terminal region includes the acceptor-binding site and is terminated by a C-terminal domain extending to the other monomer, which is thought to be important for substrate recognition both in GlcAT-I and in GlcAT-P (Kakuda et al. 2004; Gulberti et al. 2005). Although a lot of structural data have been gathered on these enzymes (Pedersen et al. 2000; Kakuda et al. 2004), their acceptor substrate specificity has been debated since the enzyme activity was first described in cartilage and brain extracts (Brandt et al. 1969; Helting and Roden 1969). It has been reported that GlcAT-I expressed in COS cells exhibited a very low activity toward lactosamine analogs in contrast to Galß1,3Gal derivatives, although overexpression of this enzyme in mammalian cells led to the presence of HNK-1 epitopes on the cell surface (Wei et al. 1999). Screening GlcAT-P activities toward disaccharide substrates linked to a hydrophobic aglycone suggested that this enzyme uses a broader range of acceptors and it was observed that overexpression of the recombinant protein can bypass GAG synthesis deficiency in mutant CHO cells (Wei et al. 1999).
To clarify these issues, we expressed the full-length human GlcAT-I and GlcAT-P cDNA in Pichia pastoris and analyzed the specificity of these two glucuronosyltransferases toward several related disaccharide substrates. We found that GlcAT-I catalyzed the transfer of GlcA only onto Galß1,3Gal, whereas GlcAT-P exhibited a more promiscuous activity. In an attempt to understand the molecular basis of the individual substrate specificity of these two glucuronosyltransferases, we conducted a comprehensive scanning mutagenesis of amino acid residues potentially involved in acceptor substrate interactions, based on structural information and multiple sequence alignment. In addition, we constructed chimeric enzymes to analyze the effect of the exchange of C-terminal ends between GlcAT-I and GlcAT-P that are predicted to play an important role in the specificity of these enzymes. These approaches, combined with computer-aided modeling, allowed us to delineate the contribution and the functional importance of several active-site residues in acceptor substrate binding and specificity.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Substrate specificity of the recombinant human GlcAT-I and GlcAT-P
To compare the acceptor substrate specificity of GlcAT-I and GlcAT-P, we analyzed the glucuronosyltransferase activity of these enzymes toward related disaccharide substrates analogous to the nonreducing end of the linkage region of GAGs, the HNK-1 epitope, and glycoproteins. As shown in Table I, GlcAT-I exhibited activity toward Galß1,3Gal (referred to as Gal2-Gal1) but was not able to use any other disaccharide tested as an acceptor substrate. In contrast, GlcAT-P catalyzed the glucuronosyltransferase reaction not only toward Galß1,3Gal but also toward Galß1,4GlcNAc and lactose (Galß1,4Glc). Low activity was detected toward lacto-N-biose (Galß1,3GlcNAc). These results provided evidence for a more promiscuous substrate profile of GlcAT-P compared with GlcAT-I. Nonetheless, further kinetic analysis performed in this study (Table III) indicated that the Km value of GlcAT-P toward Galß1,3Gal was about 7.5-fold for Galß1,4GlcNAc, suggesting that Galß1,4GlcNAc is the preferred substrate of the latter enzyme.
|
|
Identification of amino acid residues involved in the acceptor substrate recognition
To investigate the molecular basis underlying substrate recognition by GlcAT-I and GlcAT-P, we took advantage of the structural data available for these enzymes (Pedersen et al. 2000
; Kakuda et al. 2004). This information was used in combination with multiple sequence alignment, in an attempt to target potential amino acid residues involved in protein–substrate interactions. We showed in a previous phylogenetic analysis that GlcAT-I and GlcAT-P belong to a glycosyltransferase family containing 119 related members in vertebrates, invertebrates, and plants (Fondeur-Gelinotte et al. 2006). The multiple alignment of 40 selected sequences shown in Figure 1 identified eight peptide motifs conserved along the full-length of the sequences with conserved intermotif distances, suggesting that the genes coding for these enzymes derive from a common ancestor. In this study, we examined the sequence of motifs 5–7 (Figure 1), which contain the acceptor-binding site, as inferred from structural analysis (Pedersen et al. 2000; Kakuda et al. 2004). Our data showed that Glu227 in motif 5, Arg247 and Asp252 in motif 6, and Glu281 in motif 7 (in GlcAT-I) are highly conserved among animal species, suggesting that these residues are potentially responsible for the specificity of these enzymes toward the terminal galactose of the acceptor substrate (Gal2) on which GlcA is transferred. It is worth noting that these residues exhibit a lower degree of conservation among plant sequences, such as Arg247 and Asp252 in subfamily B and in subfamily A liliopsida, where they are replaced by methionine or leucine and glycine, respectively. This may indicate that these enzymes differ in their acceptor specificity, although experimental evidence is required to establish this point.
|
To address the functional importance of the residues potentially involved in Gal2 interactions, we conducted a systematic mutational analysis. For this purpose, we expressed wild-type and mutant GlcAT-I in P. pastoris and compared activities and kinetic parameters of the recombinant enzymes. Both conservative and nonconservative mutations of Glu227, Arg247, Asp252, and Glu281 (position of residues indicated in Figure 2) were carried out and the consequences on expression and activity were evaluated. Upon expression in P. pastoris, immunoblot analysis clearly indicated that the eight mutants were expressed to approximately the same level as the wild-type protein except for Arg247Lys, which was produced in a slightly higher amount (Figure 3A, inset). Interestingly, mutation of any of these residues produced a drastic effect on the enzyme activity. All alanine-substituted mutants were inactive toward Galß1,3Gal (Figure 3A). In addition, none of the conservative mutations performed was able to restore GlcAT-I enzyme activity (Figure 3A), thus precluding further kinetic studies. The structural analysis of GlcAT-I in complex with the acceptor substrate indicated that the OE1 atom of Glu227 is within hydrogen-bonding distance to O6 of Gal2 and the OD2 atom of Asp252 to O4. The OE2 atom of Glu281 is within 3-Å distance to O3 of Gal2, and NH1 of Arg247 is within hydrogen-bonding distance to O6 of Gal2 (Pedersen et al. 2000
). These residues form a close hydrogen bond network with the various hydroxyl groups of Gal2. The site-directed mutagenesis analysis carried out in this study indicated that any change in this network of interactions abolishes enzyme activity. This result is in agreement with the strong conservation of these amino acids among the members of the ß1,3-glucuronosyltransferase family in animals, as shown by our phylogenetic analysis (Fondeur-Gelinotte et al. 2006) and multiple sequence alignments (Figure 1).
|
|
Investigation of determinants for GlcAT-I and GlcAT-P specificity
We next examined the role of the amino acid residues that may govern the selectivity of GlcAT-I toward Galß1,3Gal and the broader specificity of GlcAT-P. For this purpose, we analyzed the effect of the substitution of amino acids potentially involved in the recognition of the first carbohydrate of the disacccharide structure (Galß1,3Gal or Galß1,4GlcNAc), i.e. Gal1 and N-acetylglucosamine (GlcNAc) for GlcAT-I and GlcAT-P, respectively.
Structural studies suggested that the plane of the side-chain of Trp243 is parallel to Gal1 molecule and that Gln318 of the second GlcAT-I monomer is positioned to form a hydrogen bond with O6 of Gal1 (Pedersen et al. 2000). Furthermore, previous site-directed mutagenesis emphasized an important functional role for Lys317 adjacent to Gln318 (Gulberti et al. 2005). On the other hand, structural studies of GlcAT-P have suggested that Phe245 together with Val320 and Asn321 residues are key determinants of the specificity of this enzyme (Kakuda et al. 2004). Thus, it was interesting to determine the effect of substituting Trp243, Lys317, and Gln318 in GlcAT-I with their counterparts Phe, Val, and Asn, respectively, in GlcAT-P on GlcAT-I activity and to test whether these mutations may broaden the specificity of the enzyme. Mutants of Trp243, Gln318, and Lys317 in GlcAT-I were produced at about the same level compared with the wild type, except for Trp243Phe and Lys317Val, which were expressed in a slightly lower amount and for the triple mutant Trp243Phe/Lys317Val/Gln318Asn, which was highly expressed (Figure 3B, inset). Mutation of Trp243 to Phe reduced GlcAT-I enzyme activity by 5.5-fold (Figure 3B) and induced a strong increase in the Km value toward Galß1,3Gal (Table II). This mutation did not modify the substrate specificity of GlcAT-I, as we found that no activity toward the reference substrate of GlcAT-P, Galß1,4GlcNAc, could be detected (data not shown). We previously showed that mutation of Trp243 to Ala abolished the GlcAT-I activity (Gulberti et al. 2005). Thus, altogether our data highlight the importance of Trp at position 243 in GlcAT-I acceptor substrate binding and activity. On the other hand, mutation of Gln318 to Asn moderately decreased the enzyme activity of GlcAT-I toward Galß1,3Gal by about 45% (Figure 3B). Interestingly, the Gln318Asn mutant exhibited a similar Km value compared with the wild-type (Table II). In contrast, substitution of Lys317 by Val (its counterpart in GlcAT-P) abrogated the enzyme activity, similar to Ala substitution (Gulberti et al. 2005), underlining the critical importance of this residue in GlcAT-I function (Figure 3B). Interestingly, the conservative mutation Lys to Arg restored the enzyme activity toward Galß1,3Gal to about 80% that of the wild type. In addition, the Lys317Arg mutant exhibited a Km value in the same range as wild-type GlcAT-I, emphasizing the importance of the presence of a basic residue at position 317 (Table II).
|
Changing both Lys317 and Gln318 to Val and Asn, respectively, strongly reduced the enzyme activity (Figure 3B). The Vmax value of the double mutant was decreased 10-fold and its Km value exhibited a 2.4-fold increase (Table II). In addition, this double mutant was not able to use Galß1,4GlcNAc or any related disaccharide tested as acceptor substrate (data not shown). The triple mutation of Trp243, Lys317, and Gln318 to Phe, Val, and Asn, respectively, strongly reduced GlcAT-I enzyme activity, precluding kinetic analyses (Figure 3B). Finally, we constructed a chimeric protein by exchanging the C-terminal end of GlcAT-I (Glu312 to Val335) for that of GlcAT-P (Glu315 to Ile334), which contains the amino acids predicted to determine acceptor substrate specificity (construct represented in Figure 2). The chimeric enzyme, GlcAT-I–C-term-P, exhibited very low enzyme activity toward the GlcAT-I reference substrate, Galß1,3Gal (Figure 3B), and was not active toward other disaccharides tested (data not shown).
In a complementary approach, we cloned the human GlcAT-P cDNA and expressed the recombinant enzyme in P. pastoris. We thus examined the consequences of nonconservative mutations on Phe245, Val320, and Asn321 as well as the effect of substituting these residues by their counterparts Trp, Lys, and Gln, respectively, present in GlcAT-I. The recombinant proteins were produced to about the same level compared with the wild-type GlcAT-P, except for Phe245Trp and Asn321Gln mutants, which were expressed in slightly higher amounts (Figure 4, inset).
|
Interestingly, mutation of Phe245, Val320, and Asn321 either to Ala or to their counterpart residues in GlcAT-I (Trp, Lys, and Gln) moderately affected the activity of GlcAT-P toward Galß1,4GlcNAc (Figure 4). Determination of the kinetic parameters of the single or of the double mutant revealed that the Vmax values were at most 3-fold decreased and that the Km values were about 2-fold increased (Table III). However, it is noteworthy that the nonconservative substitution of Val320 by Ala dramatically increased the Km value of GlcAT-P toward Galß1,4GlcNAc, thus precluding the determination of the Vmax parameter of this mutant. This result emphasized the importance of the interactions between Galß1,4GlcNAc and Val320.
When Galß1,3Gal, the specific substrate of GlcAT-I, was considered, the effect on the enzyme activity of the substitution of Phe245 and Asn321 in GlcAT-P by Ala appeared to be more severe. In particular, the nonconservative mutation of Phe245 completely abrogated the enzyme activity, emphasizing the importance of the change of Phe245 to Ala in the interaction with Galß1,3Gal, as shown for its counterpart in GlcAT-I. Mutation of Asn321 to Ala produced a large increase in the Km value, which precluded Vmax determination (Table III). Nonetheless, mutation of these residues to their counterpart in GlcAT-I (Trp, Gln, and Lys) did not change Km values and slightly decreased the Vmax value of GlcAT-P (at most 2-fold for the double mutant) (Table III). Taken together, the consequences produced by the mutations in GlcAT-P were less dramatic in terms of activity than their counterparts in GlcAT-I, indicating that the active site of GlcAT-P could, to some extent, tolerate amino acid substitutions. These results are consistent with the broader specificity of this enzyme compared with GlcAT-I. Finally, a chimeric protein in which the C-terminal end of GlcAT-P (Glu315 to Ile334) was replaced by that of GlcAT-I (Glu312 to Val335) (construct represented in Figure 2) exhibited no enzyme activity toward either the GlcAT-I reference substrate Galß1,3Gal or toward Galß1,4GlcNAc (Figure 3).
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
An early step in the assembly of the xylose–serine-linked proteoglycans is the transfer of GlcA to the C3 position of the terminal Gal residue in the carbohydrate–protein linkage region. A similar reaction occurs in the biosynthesis of HNK-1 antigens, and the possibility that these two processes were catalyzed by the same glycosyltransferases has been questioned. Early studies using embryonic chick brain and cartilage as enzyme sources suggested the existence of two separate enzymes catalyzing the transfer of GlcA onto a galactose residue (Curenton et al. 1991
). Cloning experiments indicated that the HNK-1 carbohydrate was completed by two glucuronosyltransferases (designated GlcAT-P and GlcAT-S) (Mitsumoto et al. 2000; Marcos et al. 2002), whereas GlcAT-I was responsible for the completion of the tetrasaccharide linkage region of proteoglycans (Kitagawa et al. 1998). However, Wei et al. (1999) provided clear evidence for the expression of HNK-1 epitopes after the heterologous expression of GlcAT-I in CHO cells. In addition, these authors showed that GlcAT-P could bypass a mutation in GlcAT-I abrogating GAG synthesis, suggesting an apparent overlap in the behavior of these glycosyltransferases.
We used several model disaccharides exhibiting the structure of the nonreducing end of GAGs and glycoproteins, which allow the specificity of the recombinant human GlcAT-I and GlcAT-P to be clearly distinguished. We found that GlcAT-I exhibited a strict substrate specificity toward Galß1,3Gal, as no activity could be detected toward any other disaccharide tested as potential acceptor substrate. In agreement with our data, Wei et al. (1999), using disaccharide substrates linked to a hydrophobic aglycone, found higher GlcAT-I activity toward digalactoside derivatives compared with Galß1,4Glc and Galß1,4GlcNAc analogs. These findings corroborate the idea that GlcAT-I may not be involved in the synthesis of the HNK-1 epitope. Our results clearly show that GlcAT-P, in contrast to GlcAT-I, displayed a more promiscuous substrate profile, reacting with the four disaccharide molecules tested. Nonetheless, we found that GlcAT-P exhibited a higher Km value for the GlcAT-I-specific substrate, Galß1,3Gal, suggesting that indeed Galß1,4GlcNAc, rather than Galß1,3Gal, represents the preferred substrate for GlcAT-P. Our results also show a low level of enzyme activity toward Galß1,3GlcNAc. On the other hand, Kakuda et al. (2005) reported that this enzyme does not transfer GlcA onto Galß1,3GlcNAc. However, it is noteworthy that the activity displayed by the recombinant enzyme expressed in our study was several times lower (about 20 times) than that toward other disaccharide acceptors. Interestingly, these authors showed that a related ß1,3-glucuronosyltransferase, GlcAT-S, was highly active toward Galß1,3GlcNAc (Kakuda et al. 2005; Shiba et al. 2006).
Despite previous structural and mutagenesis studies, the functional contribution of active-site amino acid residues proposed to be involved in substrate binding of ß1,3-glucuronosyltransferases is not completely elucidated yet. In an attempt to understand the molecular basis for substrate recognition in GlcAT-I and GlcAT-P, we first examined the role of four highly conserved amino acids in motifs 5–7 of the ß1,3-glucuronosyltransferases that are potentially involved in interactions with Gal2, the terminal galactose unit on which GlcA is transferred (Pedersen et al. 2000). Interestingly, our results on GlcAT-I show that amino acid exchange at any of these conserved positions is fundamentally deleterious for the enzyme activity. Examination of the structure of GlcAT-I, GlcAT-P, and EXTL2, involved in heparan sulfate synthesis (Pedersen et al. 2003), in complex with their substrate shows that the nonreducing end of the disaccharide or trisaccharide acceptor molecule is deeply buried in a cleft. Because the acceptor-binding site is largely solvent accessible, the hydroxyl groups of the pyranose ring of the saccharide unit interact through hydrogen bonds with amino acid residues such as Glu227, Arg247, Asp252, and Glu281 in GlcAT-I. Site-directed replacement of these residues clearly demonstrates that the integrity of a network of interactions between these amino acids and Gal2 is essential for the enzyme function. In addition, our phylogenetic analysis of the glycosyltransferases related to GlcAT-I showed that these residues are highly conserved among different animal species. It is thus reasonable to assume that an organization of the substrate active site similar to that of GlcAT-I governs binding and recognition of the acceptor galactose unit in ß1,3-glucuronosyltransferases, including GlcAT-P. It is worth noting that these residues show a lower degree of conservation in plants, which may be indicative of a different acceptor specificity. This is also true for residues involved in donor substrate binding. The absence of Arg156 in plant sequences has been suggested to be related to xylosyltransferase, rather than glucuronosyltransferase activity (Mitchell et al. 2007). Thus, further experimental studies are required to establish the function of the plant CAZY GT43 family members.
The aim of the second part of our study was to identify structural determinants that distinguish the specificity of GlcAT-I and GlcAT-P. The overall structure of GlcAT-P and GlcAT-I is very similar, except for few domains located in the C-terminal region extending in the neighboring monomer (Pedersen et al. 2000; Kakuda et al. 2004). In GlcAT-I, there is an insertion of four amino acid residues near the C-terminus and an additional C-terminal helix, which is not found in GlcAT-P. This led us to assess the consequences of exchanging the C-termini of these enzymes on their specificity. Our results provide evidence for a major deleterious effect of this modification on enzyme activity, confirming the critical functional role of the C-terminal region. However, the GlcAT-P-C-term-I and GlcAT-I-C-term-P chimeric proteins did not retain sufficient activity to detect a possible change in specificity.
Furthermore, the structural analysis and sequence alignment indicated several amino acid residues that differ between GlcAT-I and GlcAT-P, which were proposed to be important for Gal1 and GlcNAc recognition, respectively (Pedersen et al. 2000; Kakuda et al. 2004). We thus examined the hypothesis that these residues may determine the distinct specificity of the two ß1,3-glucuronosyltransferases with regard to their acceptor substrate. To pursue this hypothesis, we generated mutants in which these residues were converted to their counterpart in GlcAT-I and GlcAT-P. A major finding of this study is the critical role of Trp243 and its counterpart Phe245 in the substrate recognition and specificity of GlcAT-I and GlcAT-P, respectively. Substantiating the functional importance of Trp243, we show a dramatic decrease in GlcAT-I activity when this residue was changed to Phe, together with a large increase in Km. In line with these results, we previously observed a marked reduction and a loss of activity upon mutation of this residue to Ala and Phe, respectively, when using a digalactoside linked to a hydrophobic aglycone as substrate (Gulberti et al. 2005). Examination of the structure of GlcAT-I in complex with Galß1,3Gal, as illustrated in Figure 5A, shows that the stacking interaction between the aromatic ring of Trp243 and the pyranose ring of Gal1 is important for binding the saccharide unit. Although examination of this structure also indicates potential hydrogen bond formation between Gln318 and O6 of Gal1, we found that mutation of this residue to Asn moderately affected the enzyme activity and did not change the Km value of GlcAT-I toward Galß1,3Gal. This result indicates that this interaction does not play a prominent functional role, as previously suggested (Gulberti et al. 2005). In contrast, our results emphasize the importance of Lys317. This residue could be effectively replaced by Arg, when using Galß1,3Gal as substrate, but not by Ala (Gulberti et al. 2005), nor by its Val counterpart in GlcAT-P, highlighting the importance of a positively charged residue at this position. In a previous study, we showed that the replacement of Lys317 by Arg restores to a lesser extent the loss of enzyme activity caused by Ala substitution, when using Galß1,3Galß-1-O-methoxyphenyl as substrate. It is reasonable to suggest that the presence of a hydrophobic aglycone linked to Gal2 in the vicinity of Arg at position 317 may affect GlcAT-I activity. However, further structural studies are required to verify this hypothesis. Since computer-assisted modeling of the acceptor-active site does not provide evidence for direct interaction between Gal2 and Lys317, our data are suggestive of a structural role for this residue in the organization of the acceptor active site. Altogether, our data support the idea that appropriate positioning of Galß1,3Gal with regard to Trp243 is probably a major determinant of GlcAT-I's preference for this substrate.
|
To elucidate further the basis for the difference in substrate profile of GlcAT-I and GlcAT-P, the model structures of the enzymes with acceptor sugars were built using energy minimization and molecular dynamic calculations. No satisfactory docking of Galß1,4GlcNAc in the GlcAT-I active site could be achieved, but a loss of stacking interaction with Trp243 was observed. This analysis is consistent with the finding that GlcAT-I does not use Galß1,4GlcNAc as acceptor substrate. In contrast, the model calculation of GlcAT-P in complex with Galß1,3Gal reveals that the aromatic ring of Phe245 can be stacked with Gal1, as with GlcNAc in the crystal structure of GlcAT-P in complex with Galß1,4GlcNAc (Figure 5B). This observation corroborates the site-directed mutagenesis effects observed on GlcAT-P with regard to its activity toward Galß1,3Gal. Remarkably, we observed that the mutation of Phe245 to Ala abrogated Galß1,3Gal activity, whereas the enzyme retained most Galß1,4GlcNAc activity, suggesting that, like its counterpart in GlcAT-I, Phe245 plays an important role in GlcAT-P's capacity to use Galß1,3Gal. It is interesting to note that, in a similar way, Shiba et al. (2006)
generated a double GlcAT-S mutant that exhibited a marked decrease in Galß1,3GlcNAc but not in Galß1,4GlcNAc activity, suggesting that Trp234 and/or Ala309 (equivalent to Phe245/Asn321 in GlcAT-P) are important for recognizing the acceptor substrate, especially the ß1,3-linkage.
Furthermore, we show that mutations of GlcAT-P active site residues potentially involved in GlcNAc interactions produced slight effects on Galß1,4GlcNAc activity, except for substituting Val320 by Ala or Lys, which induced a marked increase in Km values. This result was not unexpected, as it has been suggested that Val320 interacts with the N-acetyl group of the disaccharide substrate (Kakuda et al. 2000). It can be hypothesized that the hydrophobic interaction predicted between C
of Val320 and the N-acetyl group of GlcNAc (C8) is not possible when the size of the side-chain at position 320 is reduced by Ala substitution but can be restored by the longer side chain of Val.
In conclusion, this work has demonstrated that (i) a network of interactions with Gal2 is essential for acceptor substrate binding, (ii) the stacking interaction between Trp243 and Phe245 and Gal1 or GlcNAc is a critical determinant with regard to the specificity of GlcAT-I and GlcAT-P, respectively, and (iii) interactions with Gal1 are more stringent in the case of GlcAT-I than with GlcNAc, in the case of GlcAT-P, consistent with the broader specificity of the latter enzyme. This study brings us a few steps closer to an understanding of ß1,3-glucuronosyltransferase specificity and may provide a useful platform for the design of specific inhibitors and acceptors for the analysis of glucuronosyltransferases and as potential therapeutic agents.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Materials
UDP-GlcA, Galß1,4GlcNAc, Galß1,4Glc, Galß1,3GlcNAc, and antirabbit alkaline phosphatase-conjugated immunoglobulins were from Sigma (L'Isle D'Abeau, St Quentin Fallavier, France). Bacterial and yeast culture media were obtained from Difco (Becton Dickinson, Pont de Claix, France). The P. pastoris expression system and competent Escherichia coli cells were purchased from Invitrogen (Groningen, The Netherlands). Restriction enzymes and T4 DNA ligase were provided by New England Biolabs (Hitchin, UK). The DNA gel extraction kit and mini and midi plasmid kits were purchased from Qiagen (Hilden, Germany). The QuickChange site-directed mutagenesis kit was from Stratagene (La Jolla, CA). Galß1,3Gal was synthesized as previously described (Lattard et al. 2006
).
Plasmid construction and site-directed mutagenesis
Cloning of GlcAT-I cDNA and subcloning into pPICZB (Invitrogen) for the heterologous expression of the full-length protein in the methyltrophic yeast P. pastoris have been previously described (Ouzzine et al. 2000).
Human GlcAT-P sequence was amplified by polymerase chain reaction from a human brain cDNA library (Clontech, Palo Alto, CA) using a sense primer (5'-AAAGAATTCAATATGCCGAAGAGACGGGACATC-3') containing an EcoRI site (boldface), together with an antisense primer (5'-AAACTCGAGTCAGATCTCCACCGAGGGGT-3') containing an XhoI site (boldface). The amplified fragment was digested with EcoRI and XhoI, and then subcloned into the yeast expression vector pPICZB (Invitrogen), which had been digested with the same enzymes to produce pPICZ-GlcAT-P.
Construction of amino acid-substituted mutants of GlcAT-I and GlcAT-P was carried out using pPICZ-GlcAT-I and pPICZ-GlcAT-P as template, respectively, with the QuickChange site-directed mutagenesis kit according to the manufacturer's recommendations. Primers used for the mutations of each vector are available as Supplementary data. Mutants were systematically checked by sequencing, and the various mutants were individually expressed in P. pastoris, as described subsequently.
For the construction of GlcAT-I and GlcAT-P chimeras, an AflII site was created by site-directed mutagenesis, changing Glu312 and Glu315, respectively, to a leucine residue. Following strand exchange of the 3' end of the corresponding sequences to create GlcAT-I-C-term-P and GlcAT-P-C-term-I (Figure 2), the leucine was substituted by the initial glutamate residue.
Heterologous expression in the yeast P. pastoris
Each pPICZ-derived vector containing native or mutant sequence for GlcAT-I and GlcAT-P was transformed into the P. pastoris SMD1168 yeast strain using the P. pastoris Easy Comp Transformation kit (Invitrogen). Transformants were selected on YPD plates containing yeast extract, peptone, dextrose, and 100 µg/mL of Zeocin (Invitrogen). The cells were grown in buffered glycerol/complex medium prior to induction by methanol [2% (v/v)] for 48 h at 30°C in a rotary shaker (215 rpm). Yeast cells were harvested by centrifugation at 3000g for 10 min, broken with glass beads, and further submitted to differential centrifugations, as previously described (Ouzzine et al. 1999). The 100 000g pellet corresponding to the membrane fraction was resuspended by Dounce homogenization in 5 mM HEPES (pH 7.4) buffer and 0.25 M sucrose.
Protein analysis and electrophoresis
Protein concentration was evaluated by the method of Bradford (1976). Protein samples were separated on a 12% (w/v) sodium dodecyl sulphate (SDS)–polyacrylamide gel under reducing conditions (Laemmli 1970). The proteins were electrotransferred to ImmobilonPTM membranes (Millipore, Saint-Quentin-en-Yveline, France). The blots were developed using polyclonal anti-GlcAT-I primary antibodies and alkaline phosphatase-conjugated anti-rabbit immunoglobulins as secondary antibodies, as previously described (Ouzzine et al. 2002). The amount of GlcAT-I and GlcAT-P mutants expressed in yeast cells was evaluated by scanning densitometry relative to wild-type protein taken as reference, and activities were normalized accordingly.
Glucuronosyltransferase assay and kinetics
The activity of recombinant GlcAT-I, GlcAT-P, and mutants was evaluated using Galß1,3Gal, Galß1,4GlcNAc, Galß1,4Gal, and Galß1,3GlcNAc as acceptor substrates. Standard reactions were performed in 100 mM acetate buffer (pH 6.5) containing 10 mM MnCl2, 50–100 µg of proteins, 10 mM acceptor substrate, and 5 mM UDP-GlcA in a total volume of 50 µL at 37°C for 60 min. The reaction product was analyzed by high-performance liquid chromatography (HPLC) after chromophore labeling using aniline (Wang et al. 1984), as previously described (Fondeur-Gelinotte et al. 2006). Activities are expressed per milligram of microsomal protein and are normalized to the amount of wild-type protein evaluated by scanning densitometry. Data are expressed as the mean ± SE of three determinations. Data are considered as significantly different from activity obtained with the wild-type enzyme when P < 0.05 (Student t test).
For the evaluation of kinetic parameters from initial velocity data, varying concentrations of acceptor substrates (0–40 mM for Galß1,3Gal and 0–20 mM for Galß1,4GlcNAc) were used at a constant concentration of UDP-GlcA (5 mM). Km and Vmax values were determined using nonlinear least squares analysis of the data fitted to the Michaelis–Menten rate equation (v = Vmax x [S]/Km + [S]) using the curvefitter program of Sigmaplot 9.0 (Systat Software Inc., San Jose, CA) (Segel 1975). Parameters were calculated from two independent determinations, with assays performed in duplicate.
Multiple sequence alignment
Phylogeny analysis of the ß1,3-glucuronosyltransferase family was carried out with Phylowin® (http://pbil.univ-lyon1.fr/software/phylowin.html) software using neighbor joining, observed distances, and 500 bootstrap replicates, as previously described (Fondeur-Gelinotte et al. 2006). Protein and DNA alignments were performed by ClustalW (Thompson et al. 1994) and saved in Pir format. The Pir alignment was used for the selection of informative positions by G-Blocks (Castresana 2000).
Molecular modeling
The model structure of GlcAT-I in complex with Galß1,4GlcNAc (Figure 5A) and of GlcAT-P in complex with Galß1,3Gal (Figure 5B) was built by energy minimization and molecular dynamics calculations based on the initial structure available from the Protein Data Bank code 1FGG
[PDB]
(Pedersen et al. 2000) and Protein Data Bank code 1V84 (Kakuda et al. 2004), respectively. The best docking result for each ligand was minimized using the AMBER 8.0 program (Case et al. 2004). To avoid the missing residue regions in both Protein Data Bank files, all the minimizations were performed in a sphere of a 12-Å radius from the saccharide molecule, whereas all of the residues outside this sphere were kept fixed. Charges of both ligands were calculated using the GAUSSIAN98 package (Frisch et al. 1998) and the HF/6-31 G* basis set. Atom-centered charges were fitted with Antechamber of the AMBER 8.0 software package. Docking of Galß1,4GlcNAc in GlcAT-I and of Galß1,3Gal in GlcAT-P was performed using UCSF DOCK 6.0 (Moustakas et al. 2006).
|
Supplementary data |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Supplementary data are available at Glycobiology online (http://glycob.oxfordjournals.org/).
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
None declared.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Dr J-C Jacquinet (UMR 6005 CNRS-Université d'Orléans) is most gratefully acknowledged for providing Galß1,3Gal substrate. We also thank M-H Piet for her technical assistance. This work was supported by grants from Fonds National pour la Science (ACI no. 0693, Biologie Cellulaire, Moléculaire et Structurale, BMCS152 2004), Agence Nationale pour la Recherche (GT-GAG NT05-3_42251/2005), IT2B CNRS-INSERM, PRO-A INSERM, and PHRC regional programs. This work was also funded partially by CNRS (GDR 2590), by Association pour la Recherche sur le Cancer (ARC, 3611), and by "Contrat d'Interface International" between INSERM (S.F.-G.) and the University of Dundee, UK (M.W.H.C.).
|
Abbreviations |
|---|
GAG, glycosaminoglycan; Galß1,4Glc, lactose; Galß1,3GlcNAc, lacto-N-biose; Galß1,4GlcNAc, N-acetyllactosamine; GlcA,
-D-glucuronic acid; GlcAT-I, galactose-ß1, 3-glucuronosyltransferase I; GlcAT-P, galactose-ß1, 3-glucuronosyltransferase P; GlcNAc, N-acetylglucosamine; HPLC, high-performance liquid chromatography; P. pastoris, Pichia pastoris; SDS, sodium dodecyl sulphate. |
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataConflict of interest statementAcknowledgmentsReferences |
|---|
Bai X, Wei G, Sinha A, Esko JD. Chinese hamster ovary cell mutants defective in glycosaminoglycan assembly and glucuronosyltransferase I. J Biol Chem (1999) 274:13017–13024.
Bollensen E, Schachner M. The peripheral myelin glycoprotein P0 expresses the L2/HNK-1 and L3 carbohydrate structures shared by neural adhesion molecules. Neurosci Lett (1987) 82:77–82.[CrossRef][Web of Science][Medline]
Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal Biochem (1976) 72:248–254.[CrossRef][Web of Science][Medline]
Brandt AE, Distler J, Jourdian GW. Biosynthesis of the chondroitin sulfate–protein linkage region: purification and properties of a glucuronosyltransferase from embryonic chick brain. Proc Natl Acad Sci USA (1969) 64:374–380.
Breton C, Imberty A. Structure/function studies of glycosyltransferases. Curr Opin Struct Biol (1999) 9:563–571.[CrossRef][Web of Science][Medline]
Case DA, Darden TA, Cheatham TE III, Simmerling CL, Wang J, Duke RE, Luo R, Merz KM, Wang B, Pearlman DA, et al. AMBER 8. (2004) San Francisco, CA: University of California.
Castresana J. Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol Biol Evol (2000) 17:540–552.
Curenton T, Ekborg G, Roden L. Glucuronosyl transfer to galactose residues in the biosynthesis of HNK-1 antigens and xylose-containing glycosaminoglycans: one or two transferases? Biochem Biophys Res Commun (1991) 179:416–422.[CrossRef][Web of Science][Medline]
Fondeur-Gelinotte M, Lattard V, Oriol R, Mollicone R, Jacquinet J-C, Mulliert G, Gulberti S, Netter P, Magdalou J, Ouzzine M, et al. Phylogenetic and mutational analyses reveal key residues for UDP-glucuronic acid binding and activity of ß1,3-glucuronosyltransferase I (GlcAT-I). Protein Sci (2006) 15:1667–1678.[CrossRef][Web of Science][Medline]
Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Zakrzewski VG, Montgomery JA, Stratmann RE, Burant JC, et al. Gaussian (Revision A.1). (1998) Pittsburgh (PA): Gaussian Inc.
Gulberti S, Fournel-Gigleux S, Mulliert G, Aubry A, Netter P, Magdalou J, Ouzzine M. The functional glycosyltransferase signature sequence of the human ß1,3-glucuronosyltransferase is a XDD motif. J Biol Chem (2003) 278:32219–32226.
Gulberti S, Lattard V, Fondeur M, Jacquinet J-C, Mulliert G, Netter P, Magdalou J, Ouzzine M, Fournel-Gigleux S. Phosphorylation and sulfation of oligosaccharide substrates critically influence the activity of human ß1,4-galactosyltransferase 7 (GalT-I) and ß1,3-glucuronosyltransferase I (GlcAT-I) involved in the biosynthesis of the glycosaminoglycan– protein linkage region of proteoglycans. J Biol Chem (2005) 280:1417–1425.
Helting T, Roden L. Biosynthesis of chondroitin sulfate. I. Galactosyl transfer in the formation of the carbohydrate–protein linkage region. J Biol Chem (1969) 244:2790–2798.
Kakuda S, Sato Y, Tonoyama Y, Oka S, Kawasaki T. Different acceptor specificities of two glucuronyltransferases involved in the biosynthesis of HNK-1 carbohydrate. Glycobiology (2005) 15:203–210.
Kakuda S, Shiba T, Ishiguro M, Morita I, Oka S, Kawasaki T, Wakatsuki S, Kato R. Structural basis for acceptor substrate recognition of a human glucuronyltransferase, GlcAT-P, an enzyme critical in the biosynthesis of the carbohydrate epitope HNK-1. J Biol Chem (2004) 279:22693–22703.
Kitagawa H, Tone Y, Tamura J, Neumann KW, Ogawa T, Oka S, Kawasaki T, Sugahara K. Molecular cloning of glucuronosyltransferase I involved in the biosynthesis of the glycosaminoglycan–protein linkage region of proteoglycans. J Biol Chem (1998) 273:6615–6618.
Laemmli UK. Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature (1970) 227:680–685.[CrossRef][Medline]
Lattard V, Fondeur-Gelinotte M, Gulberti S, Jacquinet J-C, Boudrant J, Netter P, Magdalou J, Ouzzine M, Fournel-Gigleux S. Purification and characterization of a soluble form of the recombinant human galactose-ß1,3-glucuronosyltransferase I expressed in the yeast Pichia pastoris. Protein Expr Purif (2006) 47:137–143.[CrossRef][Web of Science][Medline]
Marcos I, Galan JJ, Borrego S, Antinolo G. Cloning, characterization, and chromosome mapping of the human GlcAT-S gene. J Hum Genet (2002) 47:677–680.[CrossRef][Web of Science][Medline]
Mitchell RAC, Dupree P, Shewry PR. A novel bioinformatics approach identifies candidate genes for the synthesis and feruloylation of arabinoxylan. Plant Physiol (2007) 144:43–53.
Mitsumoto Y, Oka S, Sakuma H, Inazawa J, Kawasaki T. Cloning and chromosomal mapping of human glucuronyltransferase involved in biosynthesis of the HNK-1 carbohydrate epitope. Genomics (2000) 65:166–173.[CrossRef][Web of Science][Medline]
Moustakas DT, Lang PT, Pegg S, Pettersen E, Kuntz ID, Brooijmans N, Rizzo RC. Development and validation of a modular, extensible docking program: DOCK 5. J Comput Aided Mol Des (2006) 20:601–619.[CrossRef][Web of Science][Medline]
Oka S, Terayama K, Kawashima C, Kawasaki T. A novel glucuronyltransferase in nervous system presumably associated with the biosynthesis of HNK-1 carbohydrate epitope on glycoproteins. J Biol Chem (1992) 267:22711–22714.
Ouzzine M, Gulberti S, Levoin N, Netter P, Magdalou J, Fournel-Gigleux S. The donor substrate specificity of the human ß1,3-glucuronosyltransferase I (GlcAT-I) towards UDP-–glucuronic acid is determined by two crucial histidine and arginine residues. J Biol Chem (2002) 277:25349–25445.
Ouzzine M, Gulberti S, Netter P, Magdalou J, Fournel-Gigleux S. Structure/function of the human Galß1,3-glucuronosyltransferase. Dimerization and functional activity are mediated by two crucial cysteine residues. J Biol Chem (2000) 275:28254–28260.
Ouzzine M, Magdalou J, Burchell B, Fournel-Gigleux S. An internal signal sequence mediates the targeting and retention of the human UDP-glucuronosyltransferase 1A6 to the endoplasmic reticulum. J Biol Chem (1999) 274:31401–31409.
Pedersen LC, Dong J, Taniguchi F, Kitagawa H, Krahn JM, Pedersen LG, Sugahara K, Negishi M. Crystal structure of an 1,4-N-acetylhexosaminyltransferase (EXTL2), a member of the exostosin gene family involved in heparan sulfate biosynthesis. J Biol Chem (2003) 278:14420–14428.
Pedersen LC, Tsuchida K, Kitagawa H, Sugahara K, Darden TA, Negishi M. Heparan/chondroitin sulfate biosynthesis. Structure and mechanism of human glucuronyltransferase I. J Biol Chem (2000) 275:34580–34585.
Raman R, Sasisekharan V, Sasisekharan R. Structural insights into biological roles of protein–glycosaminoglycan interactions. Chem Biol (2005) 12:267–277.[CrossRef][Web of Science][Medline]
Sakai S, Akiyama H, Sato Y, Yoshioka Y, Linhardt RJ, Goda Y, Maitani T, Toida T. Chondroitin sulfate intake inhibits the IgE-mediated allergic response by down-regulating Th2 responses in mice. J Biol Chem (2006) 281:19872–19880.
Segel IH. Enzyme kinetics. (1975) New York: John Wiley & Sons.
Shiba T, Kakuda S, Ishiguro M, Morita I, Oka S, Kawasaki T, Wakatsuki S, Kato R. Crystal structure of GlcAT-S, a human glucuronyltransferase, involved in the biosynthesis of the HNK-1 carbohydrate epitope. Proteins (2006) 65:499–508.[CrossRef][Web of Science][Medline]
Terayama K, Oka S, Seiki T, Miki Y, Nakamura A, Kozutsumi Y, Takio K, Kawasaki T. Cloning and functional expression of a novel glucuronosyltransferase involved in the biosynthesis of the carbohydrate epitope HNK1. Proc Natl Acad Sci USA (1997) 94:6093–6098.
Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res (1994) 22:4673–4680.
Venkatesan N, Barre L, Benani A, Netter P, Magdalou J, Fournel-Gigleux S, Ouzzine M. Stimulation of proteoglycan synthesis by glucuronosyltransferase-I gene delivery: a strategy to promote cartilage repair. Proc Natl Acad Sci USA (2004) 101:18087–18092.
Wang WT, Le Donne NC, Ackerman B, Sweeley CC. Structural characterisation of oligosaccharides by high-performance liquid chromatography, fast-atom bombardment-mass and exoglycosidase digestion. Anal Biochem (1984) 141:366–381.[CrossRef][Web of Science][Medline]
Wei G, Bai X, Sarkar AK, Esko JD. Formation of HNK-1 determinants and the glycosaminoglycan tetrasaccharide linkage region by UDP-GlcUA:galactoseß1,3-glucuronosyltransferase. J Biol Chem (1999) 274:7857–7864.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Tone, L. C. Pedersen, T. Yamamoto, T. Izumikawa, H. Kitagawa, J. Nishihara, J.-i. Tamura, M. Negishi, and K. Sugahara 2-O-Phosphorylation of Xylose and 6-O-Sulfation of Galactose in the Protein Linkage Region of Glycosaminoglycans Influence the Glucuronyltransferase-I Activity Involved in the Linkage Region Synthesis J. Biol. Chem., June 13, 2008; 283(24): 16801 - 16807. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||