HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 21, 15868-15875, May 26, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/21/15868    most recent M001034200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ujita, M. Articles by Fukuda, M. Search for Related Content PubMed PubMed Citation Articles by Ujita, M. Articles by Fukuda, M. Social Bookmarking                      What's this?

Poly-N-acetyllactosamine Extension in N-Glycans and Core 2- and Core 4-branched O-Glycans Is Differentially Controlled by i-Extension Enzyme and Different Members of the 1,4-Galactosyltransferase Gene Family^*

Minoru Ujita, Anup K. Misra, Joseph McAuliffe, Ole Hindsgaul, and Minoru Fukuda^§

From the Glycobiology Program, Cancer Research Center, the Burnham Institute, La Jolla, California 92037

Received for publication, February 7, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Poly-N-acetyllactosamines are attached to N-glycans, O-glycans, and glycolipids and serve as underlying glycans that provide functional oligosaccharides such as sialyl Lewis^X. Poly-N-acetyllactosaminyl repeats are synthesized by the alternate addition of 1,3-linked GlcNAc and 1,4-linked Gal by i-extension enzyme (iGnT) and a member of the 1,4-galactosyltransferase (4Gal-T) gene family. In the present study, we first found that poly-N-acetyllactosamines in N-glycans are most efficiently synthesized by 4Gal-TI and iGnT. We also found that iGnT acts less efficiently on acceptors containing increasing numbers of N-acetyllactosamine repeats, in contrast to 4Gal-TI, which exhibits no significant change. In O-glycan biosynthesis, N-acetyllactosamine extension of core 4 branches was found to be synthesized most efficiently by iGnT and 4Gal-TI, in contrast to core 2 branch synthesis, which requires iGnT and 4Gal-TIV. Poly-N-acetyllactosamine extension of core 4 branches is, however, less efficient than that of N-glycans or core 2 branches. Such inefficiency is apparently due to competition between a donor substrate and acceptor in both galactosylation and N-acetylglucosaminylation, since a core 4-branched acceptor contains both Gal and GlcNAc terminals. These results, taken together, indicate that poly-N-acetyllactosamine synthesis in N-glycans and core 2- and core 4-branched O-glycans is achieved by iGnT and distinct members of the 4Gal-T gene family. The results also exemplify intricate interactions between acceptors and specific glycosyltransferases, which play important roles in how poly-N-acetyllactosamines are synthesized in different acceptor molecules.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Poly-N-acetyllactosamines are unique glycans having N-acetyllactosamine repeats (Gal14GlcNAc13)_n in one side chain (1). Poly-N-acetyllactosamines are attached to N-glycans (2-4), O-glycans (5-7), and glycolipids (8-10). Poly-N-acetyllactosamines are often modified to express differentiation antigens and functional oligosaccharides. One of those oligosaccharides is sialyl Le^X,¹ NeuNAc23Gal14(Fuc13)GlcNAcR discovered in human granulocytes and monocytes (11, 12). Sialyl Le^X and its sulfated forms, such as 6-sulfo sialyl Le^X, NeuNAc23Gal14[Fuc13(sulfo6)]GlcNAcR in mucin-type glycoproteins, have been shown to be ligands for E-, P-, and L-selectin (13-15).

Since these O-glycans are present as clusters in mucin-type glycoproteins, mucin-type glycoproteins can present multiple ligands to a selectin. In mucin-type glycoproteins of blood cells, sialyl Le^X can be found in core 2-branched oligosaccharides (5, 6, 16). Similarly, 6-sulfo sialyl Le^X in L-selectin ligands found in high endothelial venules are synthesized in core 2-branched oligosaccharides such as NeuNAc23Gal14[Fuc13(sulfo6)]GlcNAc16(Gal13)GalNAc1serine/threonine (17-19).

The enzyme responsible for core 2 branching is called core 2 1,6-N-acetylglucosaminyltransferase (C2GnT), and its cDNA has been cloned (20). When C2GnT was inactivated by gene targeting, leukocytes from those mutant mice exhibit much reduced binding to P-, L-, and E-selectin, although the effect on E-selectin binding was less severe than binding to P- and L-selectin (21).

In the gastrointestinal tract, oligosaccharides with core 3, GlcNAc13GalNAc, can be frequently found (22). In these tissues, core 3 is converted to core 4 by core 4 1,6-N-acetylglucosaminyltransferase (C4GnT), forming Gal14GlcNAc16(GlcNAc13)GalNAc (23, 24).

Recently, we and others have cloned cDNA encoding C4GnT (25, 26). This enzyme is unique in having a major C2GnT activity and a minor I branching activity in addition to C4GnT activity. The expression profile of this novel C2/C4/IGnT (termed C2GnT-mucin type) is consistent with the expression profile of core 4 oligosaccharides in various tissues (22, 25). In these tissues, C4GnT activity is always associated with C2GnT activity, suggesting that C2GnT-mucin type is probably the enzyme responsible for the formation of core 4-branched oligosaccharides in these tissues. Since core 4 oligosaccharides can be further modified to express sialyl Le^X in N-acetyllactosaminyl side chains, the synthesis of core 4 and its poly-N-acetyllactosaminyl extension provide a basis for the formation of functional oligosaccharides.

These results indicate that it is crucial to understand the synthesis of N-acetyllactosamine repeats in core 2- or core 4-branched O-glycans as well as in N-glycans. To this end, we have previously demonstrated (27) that poly-N-acetyllactosamines in core 2-branched O-glycans are synthesized with a newly discovered member of 1,4-galactosyltransferase, 4Gal-TIV (28) and a newly cloned i-extension enzyme (iGnT) (29). We have also found that 4Gal-TI, iGnT, and I-branching enzyme (IGnT) are involved in the synthesis of I-branched poly-N-acetyllactosamines (30). In the same study, we found that the addition of N-acetyllactosamine repeats to linear N-acetyllactosamine is preferred over the extension of N-acetyllactosamine to an I branch, mainly because galactosylation of a GlcNAc16 branch is inefficient (30). Third, we demonstrated that similar size and similar amounts of poly-N-acetyllactosamine are synthesized in both side chains of GlcNAc16(GlcNAc12)Man16Man1R, which was preformed by the action of N-acetylglucosaminyltransferase V. We found that this equal distribution of poly-N-acetyllactosamines is achieved because 4Gal-TI and iGnT have complementary branch specificity toward two side chains (31).

These results prompted us in the present study to determine how poly-N-acetyllactosamine synthesis is regulated in both N- and O-glycans. Our results demonstrate that 4Gal-TI is most efficient for poly-N-acetyllactosamine synthesis in N-glycan acceptors and that iGnT but not 4Gal-TI decreases its enzymatic activity as acceptors contain increasing numbers of N-acetyllactosamine repeats. We also demonstrated that 4Gal-TI predominantly galactosylates core 4-branched O-glycans. Moreover, we found that fewer N-acetyllactosamine extensions can be formed on core 4- than on core 2-branched oligosaccharides, mainly because iGnT is inefficient on core 4-branched acceptors.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation and Expression of cDNA Encoding iGnT-- The cDNA encoding iGnT was cloned into pcDNA3.1, resulting in pcDNA3.1-iGnT as described previously (29). pcDNAI-A, harboring cDNA encoding a signal sequence and an IgG binding domain of Staphylococcus aureus protein A, was constructed as described before (32). The catalytic domain of iGnT was cloned into this vector, resulting in pcDNAI-A·iGnT (29).

pcDNAI-A and pcDNAI-A·iGnT were separately transfected with LipofectAMINE Plus (Life Technologies) into COS-1 cells as described previously (29). The chimeric enzyme released into serum-free OPTI-MEM was used after adsorbing the protein A chimeric enzyme to IgG-Sepharose 6FF (Amersham Pharmacia Biotech) as described previously (33). Alternatively, the culture medium was concentrated 100- or 1,000-fold by a Centricon 10 concentrator (Amicon) and directly used as an enzyme source. In most of the studies, the concentrated culture medium was used for iGnT, since IgG-Sepharose-bound enzymes had a low activity as seen for other glycosyltransferases (34, 35). Typically, the activity of iGnT in the incubation mixture was 38.0 nmol/h/ml (for the addition of one GlcNAc) or 380.0 nmol/h/ml (for poly-N-acetyllactosamine synthesis), using 0.5 mM Gal14Glc1p-nitrophenol (Toronto Research Chemicals) as an acceptor. The medium from mock-transfected COS-1 cells contained less than one-fifth of iGnT activity compared with that derived from pcDNAI-A·iGnT-transfected COS-1 cells, as described (29).

Expression of cDNAs Encoding 4Gal-TII, -TIII, -TIV, and -TV-- 4Gal-TII, -TIII, and -TIV were expressed in insect cells, and the supernatants from these transfected insect cells were used as an enzyme source as described previously (27, 28, 36). Human milk 4Gal-T preparation (Sigma) was directly used as 4Gal-TI (27). 4Gal-TV (37) was cloned and expressed in COS-1 cells as described previously (27). For comparing the enzymatic activities of different 4Gal-T samples, the final concentration of 4Gal-TI, -TII, -TIII, -TIV, and -TV was adjusted to 38.0 nmol/h/ml as measured using 0.5 mM GlcNAc1p-nitrophenol (Sigma) as an acceptor.

Synthesis of Oligosaccharides-- (Gal14GlcNAc13)_nGal14GlcNAc16Man16Man1O(CH₂)₇CH₃(octyl), where n = 0, 1, and 2, were synthesized, starting from the derivatives of Gal14GlcNAc and Man16Man1octyl, as described previously (27, 38). GlcNAc13Gal14GlcNAc16Man16Man1octyl and GlcNAc13Gal14GlcNAc13Gal14GlcNAc16Man16Man1octyl were prepared by Escherichia coli -galactosidase (5 units; Sigma) treatment of Gal14GlcNAc13Gal14GlcNAc16Man16Man1octyl and Gal14GlcNAc13Gal14 GlcNAc13Gal14GlcNAc16Man16Man1octyl, respectively, as described previously (27, 38). GlcNAc12Man16Man1octyl was synthesized essentially as described previously (39).

Gal14GlcNAc16(GlcNAc13)GalNAc1octyl (compound 1) was synthesized from octyl O-(3,4,6-tri-O-acetyl-2-deoxy-2-phthalimido--D-glucopyranosyl)-(13)-2-acetamido-4,6-O-benzylidene-2- deoxy--D-galactopyranoside (compound 2) and ethyl O-(2,3,4,6- tetra-O-acetyl--D-galactopyranosyl)-(14)-3,6-di-O-benzyl-2-deoxy-2-phthalimido-1-thio--D-glucopyranoside (compound 3). All glycosylation reactions were performed under nitrogen in the presence of 4-Å molecular sieves and monitored by thin layer chromatography. Briefly, compound 2 was allowed to couple with compound 3 in the presence of dimethyl(methylthio)sulfoniumtriflate to furnish octyl O-(2,3,4,6-tetra-O-acetyl--D-galactopyranosyl)-(14)-O-(3,6-di-O-benzyl-2-deoxy-2-phthalimido--D-glucopyranosyl)-(16)-[(3,4,6-tri-O-acetyl-2-deoxy-2-phthalimido--D-glucopyranosyl)-(13)]-2-acetamido-2-deoxy--D-galactopyranoside (compound 4). After conventional deprotection of the blocking groups of compound 4 (40), the reaction mixture was finally purified over silica gel followed by LH-20 Sephadex column chromatography, resulting in compound 1.

Compound 1 was treated with E. coli -galactosidase as described above and purified by a C₁₈-Sep-Pak column followed by LH-20 gel filtration as described previously (27, 38), resulting in GlcNAc16(GlcNAc13)GalNAc1octyl (compound 5).

Compound 2 was synthesized by conjugation of octyl 2-acetamido-4,6-O-benzylidene-2-deoxy--D-galactopyranoside and 3,4,6-tri-O-acetyl-2-deoxy-2-phthalimido--D-glucopyranosyl trichloroacetimidate using trimethylsilyl trifluoromethanesulfonate as catalyst. Compound 3 was prepared by glycosylation of 2,6-di-O-acetyl-3,4-di-O-chloroacetyl-1-chloro--D-galactopyranose with ethyl 3,6-di-O-benzyl-2-deoxy-2-phthalimido-1-thio--D-glucopyranoside in the presence of silver trifluoromethanesulfonate as promoter.

In a similar fashion, Gal14GlcNAc16(Gal13)GalNAc1octyl (compound 6) was synthesized by coupling of compound 3 with octyl O-(2,3,4,6-tetra-O-benzoyl--D-galactopyranosyl)-(13)-2-acetamido-2-deoxy--D-galactopyranoside in the presence of dimethyl(methylthio)sulfoniumtriflate as promoter followed by conventional deprotection of the blocking groups.

The products were characterized by NMR spectroscopy and electron spray mass spectrometry. Compound 1; partial ¹H NMR:  4.78 (d, J_1,2 = 3.9 Hz, 1H, H-1), 4.55 (d, J_1,2 = 8.1 Hz, 1H, H-1), 4.52 (d, J_1",2" = 7.8 Hz, 1H, H-1"), 4.45 (d, J_1',2' = 7.8 Hz, 1H, H-1'), 1.99-2.03 (3 s, 9H, 3 NHAc). m/z (C₃₈H₆₇O₂₁N₃): [M + Na⁺] 924.9470; Calcd. 924.9474. Compound 5; partial ¹H NMR:  4.82 (d, J_{1, 2} = 3.6 Hz, 1H, H-1), 4.57 (d, J_1',2' = 8.4 Hz, 1H, H-1'), 4.52 (d, J_{1", 2"} = 8.4 Hz, 1H, H-1"), 4.20 (bs, 1H, H-4), 2.01-2.07 (3s, 9H, 3 NHAc). m/z (C₃₂H₅₇O₁₆N₃): [M + Na⁺] 739.8152; Calcd. 739.8157. Compound 6; partial ¹H NMR:  4.83 (d, J_{1, 2} = 3.6 Hz, 1H, H-1), 4.52 (d, J_{1", 2"} = 7.8 Hz, 1H, H-1"), 4.42 (2d, J_1',2' = 7.8 Hz and J_1,2 = 7.8 Hz, 2H, H-1' and H-1), 4.26 (dd, 1H, H-3), 4.17 (bs, 1H, H-4), 2.0 and 2.02 (2s, 6H, 2 NHAc). m/z (C₃₆H₆₄O₂₁N₂): [M + Na⁺] 883.8991, Calcd. 883.8996.

Detailed procedures of the synthesis will be published elsewhere.²

Poly-N-acetyllactosamine Synthesis by iGnT and 4Gal-Ts-- To assay poly-N-acetyllactosamine formation, 0.5 mM acceptor was incubated with different 4Gal-Ts (760.0 nmol/h/ml) and iGnT (380.0 nmol/h/ml) under the conditions described previously (27, 31). The incubation mixture was purified by a C₁₈-reverse phase Sep-Pak cartridge column (Waters), and the product was analyzed by HPLC using an NH₂-bonded silica column (Varian Micropak AX-5) as described previously (27, 30, 31). The radioactivity of aliquots in the effluent was determined.

Peaks containing one and two N-acetyllactosamine repeats were separately digested with Escherichia freundii endo--galactosidase for 18 h at 37 °C (41). The digest was purified by a Sep-Pak column and applied to a column of Bio-Gel P-2 (400 mesh) as described previously (27).

The Addition of Galactose by Different 4Gal-Ts-- To assay the transfer of galactose residues by 4Gal-T, the reaction mixture was exactly the same as described previously (27).

Similarly, the transfer of N-acetylglucosamine by iGnT was assayed exactly in the same way as described previously (27).

In all of the above incubations, the reaction mixture was incubated for 10 h to analyze the products or for 1 h to obtain kinetic parameters.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Synthesis of Poly-N-acetyllactosamines in N-Glycans by iGnT and Different Members of the 4Gal-T Family-- Previously, we have shown that iGnT and 4Gal-TI can form poly-N-acetyllactosamines in N-glycan acceptors (27). To determine if other members of the 4Gal-T family may be also involved in poly-N-acetyllactosamine formation in N-glycans, N-glycan acceptors were incubated with iGnT and different members of 4Gal-T.

As shown in Fig. 1, A and F, 4Gal-TI together with iGnT efficiently formed poly-N-acetyllactosamines on both Gal14GlcNAc16Man16Man1octyl and Gal14GlcNAc12Man16Man1octyl. As many as four (possibly five) N-acetyllactosamine units were added to both acceptors. In contrast, together with iGnT, 4Gal-TII and -TIII added two N-acetyllactosamine units as a maximum, while 4Gal-TIV and -TV added only one N-acetyllactosamine unit (Fig. 1). Moreover, the total amount of N-acetyllactosamine incorporation by 4Gal-TII, -TIII, -TIV, or -TV and iGnT was less than half of that incorporated by 4Gal-TI and iGnT.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   HPLC analysis of the products after incubation of N-glycan acceptors with iGnT and different 4Gal-T. Gal14- GlcNAc16Man16Man1octyl (A-E) or Gal14GlcNAc12Man16Man1octyl (F-J) was incubated with 5 mM UDP-[3H]Gal, 5 mM UDP-[3H]GlcNAc, iGnT, and different members of 4Gal-T. 4Gal-TI (A and F), 4Gal-TII (B and G), 4Gal-TIII (C and H), 4Gal-TIV (D and I), and 4Gal-TV (E and J) were employed. Peaks 1-5 represent the products containing one (1) to five (5) N-acetyllactosaminyl repeats. The same amount of the enzyme, 760.0 nmol/h/ml (for 4Gal-Ts), determined using 0.5 mM GlcNAc1p-nitrophenol, or 380.0 nmol/h/ml (for iGnT), determined using 0.5 mM Gal14Glc1p-nitrophenol, was present in these experiments.

4Gal-TII, -TIII, -TIV and -TV Are Inefficient in Galactosylation of N-Glycan Acceptors-- The above results can be obtained if 4Gal-TII, -TIII, -TIV, and -TV are less efficient in galactosylation of the shortest acceptor containing only one N-acetylglucosamine (structure a in Fig. 2). Alternatively, these 4Gal-Ts are less efficient in galactosylation of acceptors containing more than one N-acetyllactosamine unit. As shown in Fig. 2A, 4Gal-TI is very efficient in galactosylation of the acceptors tested and barely decreased its galactosylation efficiency when acceptors contained increasing numbers of N-acetyllactosaminyl repeats. In contrast, 4Gal-TIV or -TV added much less galactose to N-glycan acceptors because their K_m is much higher than that for 4Gal-TI (Fig. 2, D and E, Table I). Moreover, 4Gal-TIV significantly decreases its V_max once the acceptor contains more than one N-acetyllactosamine unit (structure b in Fig. 2), while 4Gal-TV has much lower V_max than 4Gal-TI irrespective of the number of N-acetyllactosamine units present in the acceptors (Table I). 4Gal-TII has as low a V_max value as 4Gal-TV, but its K_m is lower than that of 4Gal-TI (Table I). Similarly, K_m for 4Gal-TIII is lower than that of 4Gal-TI, but 4Gal-TIII has lower V_max than 4Gal-TI (Table I). Moreover, 4Gal-TIII decreases its V_max significantly once the acceptor contains two N-acetyllactosamine units (structure c in Fig. 2, Table I). These are probably the reasons why 4Gal-TII or 4Gal-TIII, together with iGnT, managed to add two N-acetyllactosamine units, although the amount of the products was still less than that by 4Gal-TI and iGnT (Fig. 1). These results, as a whole, indicate that 4Gal-TI is most efficient in galactosylation of poly-N-acetyllactosaminyl N-glycan acceptors as well as a single N-acetyllactosaminyl N-glycan acceptor.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Dependence of 4GalTs on the concentration of N-glycan acceptors containing different numbers of N-acetyllactosamine repeats. GlcNAc16Man16Man1octyl (a); GlcNAc13Gal14GlcNAc16 Man16Man1octyl (b), or GlcNAc13Gal14GlcNAc13Gal14GlcNAc16Man16Man1octyl (c), depicted at the lower right, was incubated with different members of 4Gal-T and 5 mM UDP-[3H]Gal. The same amount of the enzyme, 38.0 nmol/h/ml, determined using 0.5 mM GlcNAc1p-nitrophenol, was used.

View this table: [in this window] [in a new window]   Table I Kinetic properties of iGnT and 4Gal-Ts

iGnT Becomes Less Efficient as Acceptors Become Longer by Having N-Acetyllactosamine Repeats-- The above results explain the reason for the difference in N-acetyllactosamine formation by different 4Gal-Ts. On the other hand, the amount of the products containing two or more N-acetyllactosamine units was significantly less than that containing one N-acetyllactosamine unit, regardless of which 4Gal-T was present.

To determine if the efficiency of iGnT changes when the acceptors contain increasing numbers of N-acetyllactosamine units, (Gal14GlcNAc13)_nGal14GlcNAc16Man16Man1octyl, where n = 0, 1, or 2, was incubated with iGnT.

The results shown in Fig. 3 indicate that the efficiency of iGnT significantly decreases when the acceptor contains two N-acetyllactosamine units (structure B in Fig. 3; see also Table I). In contrast, the efficiency of 4Gal-TI was barely affected by the increasing size of the acceptors (Fig. 2A, Table I). Since 4Gal-TI is abundantly present in various tissues, this enzyme is judged to be sufficiently present to form poly-N-acetyllactosamines in N-glycans in various tissues. These results, combined together, indicate that the action of iGnT is most likely a rate-limiting step in poly-N-acetyllactosamine formation in N-glycans.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Dependence of iGnT on the concentration of N-glycan acceptors containing different numbers of N-acetyllactosamine repeats. Gal14GlcNAc16Man16Man1octyl (A), Gal14GlcNAc13Gal14GlcNAc16Man16Man1octyl (B), or (Gal14GlcNAc13)2Gal14GlcNAc16Man16Man1octyl (C), depicted at the bottom, was incubated with iGnT and 5 mM UDP-[3H]GlcNAc. The same symbols as in Fig. 2 are used. The same amount of the enzyme, 38.0 nmol/h/ml, determined using 0.5 mM Gal14Glc1p-nitrophenol, was used.

Poly-N-acetyllactosamine Synthesis of Core 4-Branched O-Glycans-- Recently, we cloned the cDNA encoding a novel C2GnT, C2GnT-mucin type (25). This enzyme is unique in having core 4 branching activity in addition to core 2 branching activity. Since core 4 branch, GlcNAc16(GlcNAc13)GalNAc, is galactosylated in nature to form Gal14GlcNAc16(GlcNAc13)GalNAc, core 4 branches also can be further modified to form functional oligosaccharides such as sialyl Lewis^X in addition to core 2 branches. It is thus important to determine if N-acetyllactosamine formation in core 4 branches is more efficient than that in core 2 branches.

As shown in Fig. 4A, 4Gal-TI is most efficient in galactosylation of core 4 branches. We then incubated a galactosylated core 4 acceptor, Gal14GlcNAc16(GlcNAc13)GalNAc1octyl with iGnT and 4Gal-TI. As shown in Fig. 5A, core 4 was efficiently converted to those containing one or two N-acetyllactosamine units. To determine the structure of the product, peak 1 eluted at fractions 31-34 was digested with endo--galactosidase. The digested product eluted at the position corresponding to Gal14GlcNAc13Gal (Fig. 5D), indicating that one N-acetyllactosamine extension took place either at the Gal14GlcNAc16 branch or at the GlcNAc13GalNAc side chain. When the peak 2 eluted at fractions 38-40 was digested with endo--galactosidase, Gal14GlcNAc13Gal and GlcNAc13Gal were released (data not shown). These results indicate that peak 2 represents (*Gal14*GlcNAc13)_mGal14GlcNAc16[(*Gal14*GlcNAc13)_n*Gal14GlcNAc13]GalNAc1octyl, where m + n = 2 and radioactive sugars are labeled by asterisks.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Dependence of different members of 4Gal-T on the concentration of linear or core 3- or core 4-branched oligosaccharides. GlcNAc16(GlcNAc13)GalNAc1octyl (A), GlcNAc13GalNAc1p-nitrophenol (B), or Gal14GlcNAc16(GlcNAc13)GalNAc1octyl (C) was incubated with 5 mM UDP-[3H]Gal and 4Gal-TI (), -II (), -TIII (), -TIV (), or -TV (). The same amount of the enzyme was used as in Fig. 2.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Analysis of the products after incubation of core 4- and core 2-branched oligosaccharide acceptors. A and C, HPLC analysis of the products derived from core 4, Gal14GlcNAc16(GlcNAc13)GalNAc1octyl (A) and core 2, Gal14GlcNAc16(Gal13)GalNAc1p-nitrophenol (C) after incubation with iGnT and 4Gal-TI. B, HPLC analysis of the products from core 2, Gal14GlcNAc16(Gal13)GalNAc1octyl after incubation with iGnT and 4Gal-TIV. Peaks 1 and 2 in A and B represent the products containing one (1) and two (2) N-acetyllactosamine repeats. For A-C, the incubation conditions are the same as for Fig. 1. D-F, Bio-Gel P-2 gel filtration of the products obtained after endo--galactosidase digestion of peak 1 in A (D), B (E), and C (F). Peaks  and  denote the elution positions of GlcNAc13Gal and Gal14 GlcNAc13Gal, respectively.

The efficient addition of one or two N-acetyllactosamine units was observed when the core 2-branched oligosaccharide was incubated with 4Gal-TIV and iGnT (Fig. 5B). The product containing one N-acetyllactosamine unit produced Gal14GlcNAc13Gal after endo--galactosidase treatment (Fig. 5E). On the other hand, the core 2 oligosaccharide incubated with iGnT and 4Gal-TI produced GlcNAc13Gal14GlcNAc16(Gal13)GalNAc1p-nitrophenol and Gal14GlcNAc13Gal14GlcNAc16(Gal13)GalNAc1p-nitrophenol (peaks 1 and 2 in Fig. 5C), indicating that only one N-acetyllactosamine unit was added as a maximum. The structure of peak 1 (Fig. 5F) and peak 2 was confirmed by endo--galactosidase treatment. These results confirmed the previous finding that 4Gal-TIV is involved in poly-N-acetyllactosamine synthesis of core 2-branched oligosaccharides (27).

Both iGnT and 4Gal-TI Are Less Efficient Toward Core 4-branched Acceptors than Core 2-branched Acceptors-- The above results also demonstrate that the amount of the products from the core 4-branched acceptor incubated with 4Gal-TI and iGnT is less than half of that derived from the core 2-branched acceptor incubated with 4Gal-TIV and iGnT (Fig. 5, compare A and B).

To determine why the core 4 oligosaccharide is a less favorable acceptor than the core 2 oligosaccharide, the iGnT and 4Gal-TI activities were tested on various acceptors. First, among 4Gal-Ts, 4Gal-TI worked most efficiently on GlcNAc13GalNAc1R (Fig. 4B). In contrast, the same enzyme worked less efficiently on Gal14GlcNAc16(GlcNAc13)GalNAc1R (Fig. 4C, Table II). This is probably due to a competition between UDP-Gal and the galactose residue in Gal14GlcNAc16(GlcNAc13)GalNAc1R. It is likely that galactosylation of the core 4 branch, GlcNAc16(GlcNAc13)GalNAc, is also reduced once GlcNAc16(Gal14 GlcNAc13)GalNAc is formed. We then tested if the iGnT is less efficient when one N-acetyllactosamine unit is present in a core 4-branched acceptor. Fig. 6 illustrates that iGnT acts less efficiently on the core 4 acceptor, Gal14GlcNAc16(GlcNAc13)GalNAc1R than on the core 2 acceptor, Gal14GlcNAc16(Gal13)GalNAc1R (see also Table II). This decrease toward the core 4-branched acceptor may also be due to competition between the N-acetylglucosamine residue in the acceptor and UDP-GlcNAc.

View this table: [in this window] [in a new window]   Table II Kinetics properties of iGnT and rGal-Ts Arrowheads and arrows indicate where GlcNAc and Gal are added, respectively.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Dependence of iGnT activity on the concentration of core 2- or core 4-branched oligosaccharides. Core 2 Gal14GlcNAc16(Gal13)GalNAc1octyl () or core 4 Gal14GlcNAc16(GlcNAc13)GalNAc1octyl () was incubated with iGnT and 5 mM UDP-[3H]GlcNAc. The same amount of the enzyme was used as in Fig. 3.

These results indicate that poly-N-acetyllactosamine formation on core 4-branched oligosaccharides is not efficient, most likely because acceptor structures compete with donor substrates.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The present study demonstrates that poly-N-acetyllactosaminyl extension in N-glycans is achieved mainly by 4Gal-TI and iGnT. Although 4Gal-TII and -TIII together with iGnT can also add N-acetyllactosamine repeats in N-glycans, their efficiency was less than half of that by 4Gal-TI and iGnT. Moreover, the products by 4Gal-TII or -TIII did not contain more than two N-acetyllactosamine repeats, while the products by 4Gal-TI included those containing four and possibly five N-acetyllactosaminyl repeats (Fig. 1). Considering that 4Gal-TI is widely distributed in various tissues (36, 42), these results, taken together, indicate that 4Gal-TI is mainly responsible for poly-N-acetyllactosamine synthesis in N-glycans.

The present study also demonstrated that 4Gal-TIII and -TIV become less efficient in poly-N-acetyllactosamine synthesis when the acceptor contains more than one N-acetyllactosamine unit (Fig. 2, Table I). Similarly, iGnT acts much less efficiently on Gal14GlcNAc13Gal14GlcNAc16Man1R than Gal14GlcNAc16 Man1R (Fig. 3, Table I).

These results suggest that the di-N-acetyllactosaminyl structure may have a different conformation from mono-N-acetyllactosaminyl structure and such a different conformation is not favorable for the action of iGnT, 4Gal-TIII, and 4Gal-TIV. These results are consistent with the fact that two N-acetyllactosamine units can be detected more widely than three N-acetyllactosamine units in a variety of glycoproteins (43-47). In particular, there are only a few cases where soluble glycoproteins contain three or more N-acetyllactosamine units (see, for example, Ref. 48). This is exemplified by human recombinant erythropoietin produced in Chinese hamster ovary cells. In that case, the majority of the recombinant erythropoietin was found to contain two N-acetyllactosamine units, and only less than 5% contained three N-acetyllactosamine units in a side chain (44-46).

The present study demonstrated that iGnT decreases its efficiency as an acceptor contains more N-acetyllactosamine repeats (Fig. 3). In contrast, 4Gal-TI only marginally decreases its efficiency when acceptors contain increasing numbers of N-acetyllactosamine repeats (Fig. 2A). These results indicate that the action of iGnT is probably a rate-limiting step for poly-N-acetyllactosamine synthesis in N-glycans, although there are other factors that control poly-N-acetyllactosamine synthesis.

First, it has been shown recently that a branched GlcNAc16(GlcNAc12)Man16Man1R structure serves as a much better acceptor than an unbranched GlcNAc16Man16Man1R or GlcNAc12Man16Man1R (31). This is probably due to the acquisition of a favorable conformation for iGnT and 4Gal-TI by having the branched structure (49). Second, it has been demonstrated that membrane glycoproteins contain more and longer poly-N-acetyllactosamines than soluble, secretory glycoproteins (50). This may be due to the fact that iGnT has a shorter transmembrane domain than many other glycosyltransferases (29). Third, it is necessary for glycoproteins to move slowly through the Golgi apparatus for poly-N-acetyllactosamine synthesis (51). Otherwise, the formation of N-acetyllactosamine is immediately followed by sialylation or other modifications that prevent further addition of N-acetyllactosamine. In addition, it appears that only selected numbers of glycoproteins are enriched with poly-N-acetyllactosamines. For example, Lamp-1 and Lamp-2 are always the major carriers for poly-N-acetyllactosamines in various cells (52-54). It is possible that Lamp-1 and Lamp-2 satisfy all of the criteria mentioned above.

It has been demonstrated that core 2-branched O-glycans contained shorter and fewer N-acetyllactosamine repeats than N-glycans when O-glycans and N-glycans were analyzed in the same Lamp molecules (52-55) or the same Chinese hamster ovary cells (43, 56). This is probably due to the intrinsic nature of 4Gal-TIV, which acts less efficiently on longer poly-N-acetyllactosamine units, as shown in synthetic substrates utilized in the present study and lacto-series glycolipids (28). In the present study, we demonstrated that core 4-branched O-glycans are galactosylated by 4Gal-TI but are even less efficient acceptors for poly-N-acetyllactosamine formation than core 2-branched O-glycans (Figs. 5 and 7).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Proposed biosynthetic pathways of core 2- and core 4-branched O-glycans. -N-Acetylgalactosamines are transferred to serine or threonine residues in a polypeptide by -N-acetylgalactosaminyltransferase (GalNAc-T). This is followed by the action of core 1 1,3-galactosyltransferase (3Gal-T), forming Gal13GalNAc1R (core 1). Core 1 is then converted to GlcNAc16(Gal13)GalNAc1R (core 2) by core 2 1,6-N-acetylglucosaminyltransferase (C2GnT) and then Gal14GlcNAc16(Gal13)GalNAc1R by 4Gal-TIV. Poly-N-acetyllactosamines will be added on galactosylated core 2 by alternate actions of iGnT and 4Gal-TIV. Alternatively, GalNAc1R can be extended by core 3 1,3-N-acetylglucosaminyltransferase (3GlcNAc-T), forming GlcNAc13GalNAc1R (core 3). This will be followed by core 4 1,6-N-acetylglucosaminyltransferase (C4GnT) to form GlcNAc16(GlcNAc13)GalNAc1R (core 4). Core 4 is galactosylated by 4Gal-TI, resulting in Gal14GlcNAc16(Gal14Gl