Functional assignment of motifs conserved in b1,3-glycosyltransferases A mutagenesis study of murine UDP-galactose:b- N -acetylglucosamine b1,3-galactosyltransferase-I Martine Malissard, Andre  Dinter, Eric G. Berger and Thierry Hennet Institute of Physiology, University of Zu È rich, Switzerland The b1,3-glycosyltransferase enzymes identi®ed to date share several conserved r egions and c onserved c ysteine res- idues, all being located in t he putative catalytic d omain. To investigate the importance of these m otifs and cysteines for the e nzymatic act ivity, 1 4 m utants of the m urine b1,3- galactosyltransferase-I gene were constructed and expressed in Sf9 insect cells. Seven mutations abolished t he galacto- syltransferase act ivity. Kinetic analysis of the other seven active mutants revealed that t hree of them showed a three- fold to 21-fold high er apparent K m with regard to the donor substrate UDP-galactose relative to the wild-type enzyme, while two mutants had a sixfold t o 7.5-fold increase of the apparent K m value for the acceptor substrate N-acetylg lu- cosamine-b-p-nitrophenol. Taken together, our results indicate that the conserved residues W101 and W162 are involved in the b inding of the UDP-galactose donor, t he residue W315 in the binding of the N-acetylglucosamine-b-p- nitrophenol acceptor, and the domain including E264 appears t o participate in the binding of both substrates. Keywords: Gal transferase; GlcNAc transferase; mutagenesis; gene family. Glycosyltransferase enzymes account for the structural diversity of glycoconjugates found in all organisms. Based on amino-acid sequence similarity, glycosyltransferases can be classi®ed into at l east 27 different families [1]. I n contrast to the common c lassi®cation based on t he reaction catalyzed and the substrate s peci®city, the association by structural similarity allows one to putatively assign a glycosyltransferase function to proteins not previously suspected to catalyze such a reaction [2,3]. Also, the grouping by similarity may bring together glycosyl- transferases with a distinct s ubstrate speci®city, thereby providing some insight into the possible evolution from common ancestral genes. The latter i s true f or the family of b1,3-glycosyltransferases (b3GT), which comprises b1,3 -galactosylt ransferases (b3GalTs) [4±10], b1,3-N-acetyl- glucosaminyltransferases [11±13] and a b1,3-N-acetylgalac- tosaminyltransferase [14]. Within this g lycosyltransferase family, some e nzymes are c apable of using two donor substrates such as b3GalT-III, which accepts the donors UDP-Gal and UDP-GalNAc [5,14] or the Neisseria meningitidis lgtA enzyme that functions as both a b1,3-N- acetylglucosaminyltransferase and a b1,3-N-acetylgalactos- aminyltransferase [15]. Similarly, polyspeci®city towards acceptor substrates has been described for the b3GalT-V enzyme, which transfe rs G al to GlcNAc-based acceptors [8,9] and to th e GalNAc residue o f the g loboside Gb4 [16]. Structurally, other than the retention of a type-II trans- membrane topology, b3GTs do not show any similarity with the families of b1,4- [17] and a1,3-GalTs [18,19]. Within the b3GT family, the compar ison between the d ifferent proteins revealed several domains and amino-acid residues that are strongly conserved [11]. However, it is unclear whether this conservation re¯ects an evolutionary feature, where parts of the protein sequences are maintained primarily due to a late gene duplication event. Alternatively and more likely, the motifs may be conserved because they are involved in either the catalytic s ite and/or maintenance of conformation of the enzymes. To address the importance of these conserved motifs i n the enzymatic p roperties of b3GTs, a site-directed mutagenesis s tudy of the b3GT- speci®c motifs was performed using the full-length murine b3GalT-I gene as a model. To this end, a Staphylococcus aureus protein A (protA)-tagged f usion protein was gener- ated and expressed in insect cells taking advantage of the absence o f endogeneous b3GalT a ctivity in this host cell. We investigated the putative role of eight conserved motifs as well as the r ole of the six cysteines in the catalytic activity of murine b3GalT-I. Our results allowed a functional assign- ment of four domains an d s uggested that four cysteines m ay be involved in the formation of two disul®de bonds. MATERIALS AND METHODS Generation of mutant b3GalT-I genes Site-directed mutagenesis of the murine b3GalT-I g ene was carried out by an overlapping PCR method described previously [20] using a pBluescript SKII + /b3GalT-I con- struct [5] as template. The mutant forms of b3GalT-I are Correspondence to T. Hennet, Institute of Physiology, Winter- thurerstrasse 190, 8057 Zu È rich, Switzerland. Fax: + 4 1 1635 6814, Tel.: + 41 1635 5080, E-mail: thennet@access.unizh.ch Abbreviations: b3GT, b1,3-glycosyltransferase; b3GalT, b1,3-galactosyltransferase; pNP, p-nitrophenol; protA, protein A. (Received 24 July 2001, revised 25 October 2001, accepted 30 October 2001) Eur. J. Biochem. 269, 233±239 (2002) Ó FEBS 2002 named according to the amino-acid residues substituted b y alanine and their respective position in the polypeptide sequence where the start methionine is number one. The mutagenic s ense oligonucleotides are listed in Table 1. A mutagenic primer was used in a PCR reaction with primer containing a SalIsiteatthe5¢ termini or primer c ontaining a s top codon ¯anked by a XbaIsiteatthe3¢ termini, whichever was appropriate. The two overlapping PCR Ôhalf- fragmentsÕ were puri®ed, combined, ®lled with the T 4 DNA polymerase and used as template for the second-round PCR with the two external restriction site-containing primers. PCR c onditions were 20 cycles of 30 s a t 9 4 °C, 30 s at 55± 58 °Cand90sat72°C. The restriction site-containing primers were a s follows 5¢-TAGTCGACGCTTCAAAG ATCTCCTGCCTCTA-3¢ and 5¢-ATATCTAGACTAA CATCTCAGATGCTTCTTGCTTGAC-3¢ introducing a SalIandXbaI site, respectively. The W315A and C326A mutants were generated through s ingle PCR reactions using the 5¢ restriction site-containing primer and an antisense oligonucleotide i ntroducing t he desired mutation (Table 1). After puri®cation, mutated full-length fragments were subcloned into pBluescript SKII + and veri®ed by DNA sequencing. Cloning of recombinant baculoviruses and expression in Sf9 cells Wild-type and mutant full-length b3GalT-I cDNAs in pBluescriptSKII + were released by SalIandXbaIand subcloned into a pFmel±protA vector [21] opened with SalI and XbaI. The recombinant baculoviruses were generated by transposon-mediated recombination [22] as des cribed previously [5]. Sf9 i nsect cells were infected at a multiplicity of 10 and f urther incubated at 27 °C before a ssaying for b3GalT-I a ctivity and Western blotting. b3GalT activity assays Baculovirus-infected Sf9 cells were washed with NaCl/P i and lysed in 2% Triton X-100 for 15 min on i ce. Nuclei were removed f rom the lysates by centrif ugation at 500 g. Galactosyltransferase activity w as assayed by incubating 10 lL of Sf9 cell lysate for 3 0 min at 37 °Cin50lL reactions of 50 m M cacodylate buffer, pH 6.6, 10 m M MnCl 2 ,0.5m M UDP-Gal, 10 m M GlcNAc-b-p-nitrophe- nol (pNP) and 1% Triton X-100. UDP-[ 14 C]Gal (10 5 c.p.m., Amersham Pharmacia B iotech) were a dded to standard assays, whereas 2.5 ´ 10 5 c.p.m. of UDP- [ 14 C]Gal (410 pmol) were a dded when kinetic parameters were determined. The reaction was stopped by adding 0.5 m L of ice-cold water. Samples were puri®ed on Sep- Pak C 18 cartridges (Waters) by washing with 15 mL of water and eluting with 5 mL of methanol. The amount of [ 14 C]Gal transferred to the a cceptor was measured in a b-scintillation counter (Rackbeta, Pharmacia). Apparent Michaelis constants (K m ) w ere determined b y nonlinear regression analysis ( GRAPHPAD PRISM ) of double-reciprocal plots of initial velocity vs. GlcNAcb-pNP concentration (0±20 m M ) a t a constant UDP-Gal concentration ( 0.5 m M ) or of the initial velocity vs UDP-Gal c oncentration (0±15 m M ) at a constant concentration of GlcNAcb-pNP (10 m M ). Western blot analysis Sf9 cell lysate was diluted 1 : 750 in Laemmli buffer [23], denatured 5 min at 95 °Cand15lL were analyzed by 10% SDS/PAGE. After blotting onto n itrocellulose membrane (Millipore) according to Towbin et al. [24], staining was performed with 1 : 3000-diluted biotinylated anti-ProtA Ig (Sigma) f ollowed b y streptavidin±horseradish peroxidase (diluted 1 : 5000; Fluka). The protA±b3GalT-I fusion protein was then detected by electrochemiluminescence (Amersham Pharmacia Biotech). RESULTS The multiple sequence alignment of members of the b3GT protein family highlighted several conserved regions that were located in t he predicted luminal domain (Fig. 1). These motifs are spread across the polypeptide chain and are not clustered in d istinct regions as observed for example i n sialyltransferases [25]. It is of note t hat the Drosophila melanogaster protein B rainiac [26] shares the same con- Table 1. Sense strand oligonucleotides used for site-directed mutagenesis. Sequences are shown for the sense strand oligonucleotides of each complementary pair of primers used for site-directed mutagenesis. Underlined bases represent the mutations introduced. For the W 315A and C326A mutations, the an tisense primers were used in com bination with the 5¢ restriction s ite-containing primer. Mutation Oligonucleotide sequence (5¢)3¢) C73A AAATGAGCCCAACAAAG CCGAGAAAAACATT I97A-R98A AATTTGATGCTCGACAGGCT GCCGCGGAGACATGG W101A CAATCCGGGAGACA GCTGGTGATGAAAA F116A-L117A-L118A-G119A TAGCCACACTTG CAGCCGCGGCCAAAAATG W162A TTAATGGGGATGAGAGCGGTT GCCACTTTCT C167A AGATGGGTTGGCAACTTTC GCTTCAAAA D177A-D179A-F181A TGAAAACC GCCAGTGCTATTGCTGTGAACA P233A-P234A CCTGACAGCAACTACG CAGCGTTCTGTTCAG C236A AGCAACTATCCACCGTTC GCTTCAGGGACTG E264A TGCTTCATCTTG CTGACGTGTACGTGGGACT C271A ATGTGTACGTGGGACTGGCACTTCGAAAGC C295A AAAATGGCCTACAGTTTA GCTCGGTACC W315A CAGAATC GCCAATGACATGTCAAGGAAGAAGCATCTGAGATGTTAGTCTAGATAT C326A GTCAAGGAAGAAGCATCTGAGA GCCTAGTCTAGATAT 234 M. Malissard et al. (Eur. J. Biochem. 269) Ó FEBS 2002 served domains as b3GT proteins, suggesting that this protein may represent a member of this glycosylt ransferase family [2]. None of the conserved s tretches, except one motif, the so called DXD motif, found in b3GT enzymes were present in a1,3- and b1,4-GalTs [27]. To elucidate the functional relevance of several conserved residues f ound in b3GT proteins, we constructed 14 mutants of the murine b3GalT-I enzyme, where the amino acids of interest were changed to alanine. We ®rst chose to substitute the six cysteine residues of b3GalT-I a s four of them (C1, C2, C 5 and C6 in Fig. 1) were strictly conserved in all known b3GT proteins. S econdly, we modi®ed the boxes AIR (position 96), FLLG (position 116), DXD (position 177), PPX (position 233) and EDV (position 264). In addition, three tryptophan residues at positions 101, 162 and 315 were found in all b3GT proteins a s well as i n B rainiac. As tryptophan residues have previou sly been imp licated in the binding of UDP-Glc in a glycosyltransferase [28], we also mutated conserved tryptophans in our survey. The wild- type and mutant forms of the full-length murine b3GalT-I gene were fused with protA to enable the detection of the protein produced. The b3GalT-I constructs were expressed as recombinant b aculovirus in Sf9 insect cells. Western blot analysis con®rmed that all protA±mutant b3GalT-I were expressed at similar levels a nd exhibited the same molecular mass as the prot A-wild-type b3GalT-Iproteins(Fig.2). The wild-type an d mutant recombinant protA full-length b3GalT-I proteins remained localized intracellularly. There- fore, the galact osyltransferase activity was assayed in t he lysate of Sf9 cells harvested 7 2 h after infection. When assayed in presence of 10 m M GlcNAcb-pNP, t he wild-type protA±b3GalT-I c onstruct yielded an avera ge galactosyl- transferase a ctivity of 11.6 nmolámin )1 ámg protein )1 ,which is in the range of the activity measured w ith the untagged enzyme [5]. A ctivity assays performed w ith t he mutant forms of b3GalT-I revealed two groups. The mutations C73A, I97A-R98A, F116A-L117A-L118A-G119A, C167A, D177A-D179A-F181A, C295A and C326A abolished t he enzymatic activity of b3GalT-I. In contrast, residual activity was detected with the seven mutations W101A, W162A, P233A-P234A, C236A, E264A, C271A and W315A, which yielded 22, 15, 53, 98, 1 9, 20 and 12% of the wild-type protA±b3GalT-I activity, respectively (Fig. 3). No galacto- syltransferase activity was detected in the supernatant of Sf9 cells expressing the wildtype and mutant protA±b3GalT-I proteins indicating that the decrease or loss of activity found with some mutant forms was not caused by increased secretion (data not shown). T he mutant forms o f b3GalT-I, which retained a signi®cant galactosyltransferase activity, were u sed for comparative kinetic a nalysis. The K m values for e ach f usion protein were determined for the donor substrate UDP-Gal and for the a cceptor substrate Glc- NAcb-pNP. The apparent K m values obtained for the protA±b3GalT-I c onstruct were s imilar to those d etermined Fig. 1. Alignment of b3GT protein sequences. The protein sequences of the m urine b1,3-GalTs b3GalT-I (GenBank a ccession AF029790), b3GalT-II (AF029791), b3GalT-III (AF029792), b3GalT-IV (AF082504), b3GalT-V (AF254738), murine b1,3-N-acetylglucosaminyltransferases b3GnT-I (AF092050), b3GnT-III (AY037785), b3GnT-IV (AY037786) and D. me lanogaster Brainiac (U41449) were aligned using the CLUSTALW algorithm [38]. Similar amino acids conserved in all proteins are shaded in black while the similarities found in at least seven proteins are shaded in gray. The positions of the am ino acids mutated in the present study are marked with white arrows. The positions of the six c ysteines of b3GalT-I (C1 to C6) are marked with b lack arrows. Ó FEBS 2002 Mutagenesis of b3GalT-I (Eur. J. Biochem. 269) 235 for the full-length b3GalT-I enzyme without the protA tag [5], indicating that the fusion with protA has no effect on the catalytic properties of b3GalT-I (Table 2). We found that the K m values for the donor substrate UDP-Gal were signi®cantly altered with the mutations W101A, W162A and E 264A, which increased t he K m value by about 3.7-, 8- and 21.6-fold, respectively. Similarly, t he mutations E26 4A and W315A caused a respective 7 .5- and 6-fold increase of the K m values for the acceptor s ubstrate G lcNAcb-pNP. Considering the dual substrate speci®city of some b3GT proteins [15,16], we also analyzed the donor and acceptor preference of the b3GalT-I m utants. T o exclude a switch i n substrate speci®city as a cause of the loss of galactosyl- transferase activity detected with some m utations, we tested Sf9 cell lysates in presence of the donor substrates UDP- GlcNAc and UDP-GalNAc, as well as the acceptor substrates GalNAcb-pNP and Galb1,4GlcNAcb-pNP. Do- nors and acceptors were assayed at concentrations of 0.5 m M and 10 m M , respectively. We failed to detect any novel substrate s peci®city due to the mutations introduced in the b3GalT-I enzyme (Table 3). DISCUSSION Comparison between glycosyltransferase enzymes with similar activities often bring to light several conserved residues. In assigning a f unctional signi®cance t o these amino a cids, site-directed mutagenesis represents the ®rst logical method of choice. In the present study, we have investigated in the murine b3GalT-I enzyme [5] the relevance of 14 positions, which are conserved among known b3GTproteinsaswellasintheDrosophila signaling protein Brainiac ( see Fig. 1). First, our study revealed that four of the six cysteine residues are essential for the galactosyltransferase activity. C ysteine is one of the most versatile amino acids in e nzymes as i t can be used for substrate binding, be part of the catalytic mechanism and be used for the mainten ance of proper c onformation. The loss of the galactosyltransferase activity in the C73A, C 167A, C295A and C326A mutants indicated that these cysteines may be implicated in the catalytic activity of b3GalT-I or in the formation of disul®de bridges. Similar results were obtained f or the b1,4-GalT-I enzyme, where it was shown that the rigidity of the protein core is maintained by two disul®de bridges [29]. Note that the two cysteines t hat show the l east conservation among b3GT proteins, i.e. C236 and C271 in b3GalT-I, are not absolutely required for enzymatic activity. The fact that the four essential cysteines are not surrounded by other conserved residues supports their involvement in disul®de b ridge formation rather than a direct participation in the catalytic activity. Mutations of the conserved motifs A IR, FLLG and DXD also abolished t he enzymatic a ctivity indicating that these stretches are impor- tant for the catalytic reaction. However, the exact role of these motifs in possible binding sites requires further investigations. The AIR a nd FLLG motifs are found in b1,3-GalTs and b1,3-N-acetylglucosaminyltransferases, thereby suggesting that they are more likely involved in mediating the b1,3-linkage speci®city r ather than in t he direct binding of the substrates. The loss of activity seen in the DXD mutant (D177A-D179A-F181A) did not com e as a surprise. In the past, site-directed mutagenesis o f aspartate residues in other DXD-containing enzymes, such as the yeast MNN1 mannosyltransferase [30] and the yeast chitin synthetase-2 [31], demonstrated that they are essential for the catalytic activity. In the large clostridial glucosyltrans- Fig. 2. Western blot analysis of wild-type protA±b3GalT-I and its m utants. The e xpres- sion of the p ro tA±b3GalT-I proteins was detected us ing a biotinylated antiprotA antibody. La ne 1, Sf9 cells; Lane 2, mock transfected Sf9; Lane 3 , protA±b3GalT-I; Lanes 4±17, mutants of protA±b3GalT-I with C73A, I 97A-R98A, W101A, F116A-L117A- L118A-G119A, W 162A, C167A, D 177A- D178A-F181A, P 233A-P234A, C236A, E264A, C2 71A, C295A, W315A and C326A, respectively. The muta nts of protA±b3GalT-I migrated si milarly to t he wild-type enzyme in the SDS/polyacrylamide gel. Fig. 3. b3GalT-I activity o f protA± b3GalT-I and its mutan ts. The activity detected for the mutant enzymes is i ndicated i n percentage o f the a ctivity measured w ith the wild-type b3GalT-I (11 665 pmoles of galactose tran sferred per min per mg protein). 236 M. Malissard et al. (Eur. J. Biochem. 269) Ó FEBS 2002 ferase toxin [32], using photoaf®nity labeling, the DXD motif h as been implicated in the binding of UDP-Glc a nd Mn 2+ . I t has been proposed that the DXD motif is involved in the folding of a small region of the protein required for catalysis or has a role i n the catalytic site. In this context, it is worth noting that DXD is found in inverting and nonin- verting transferases, which add different sugars to other sugars, phosphates and proteins. However, these DXD- containing glycosyltransferases all use nucleosides diphos- phate sugars a s donors and require divalent cations, usually manganese. In 1999, Gastinel et al. [29] resolved part of this issue b y providing structural informations for the DXD- containing bovine b1,4-GalT I. They showed that the phosphate groups of UDP-Gal a re close to the DVD motif, but they obtained no information concerning the divalent cation. More recently, the 3-D structure of t he rabbit N- acetylglucosaminyltransferase-I enzyme has been described previously [33]. In this protein, the DXD motif is present in the form of EDD and it was shown that the third position D213 makes the only direct interaction with the bound Mn 2+ ion. In addition, it makes a hydrogen bond with one of the metal coordinating water molecules, which itself is hydrogen bonded to the ®rst position of the motif E211. These residues are further constrained by the well de®ned octahedral geometry characteristic of Mn 2+ ion c oordina- tion [34]. As the phosphates of the nucleotide-sugar also coordinate the M n 2+ ion, the r elative orientation of the nucleotide-sugar and the conserved acidic r esidues is w ell de®ned. In the rabbit N-acetylglucosaminyltransferase-I, this arrangement t enders the GlcNAc moiety of the sugar donor for interaction with the ®rst position of t he motif. Owing to this geometry, this position would also be expected to play a Ôcarbohydrate b indingÕ role in other sugar-nucleoside diphosphate/Mn 2+ dependent glyco- syltransferases. Based on these results and knowing that murine b3GalT-I uses UDP-Gal as donor and Mn 2+ ion as cofactor [5], we may assume that the DXD motif of b3GalT-I is implicated in the binding of UDP-Gal and Mn 2+ . The mutants W101A, W162A, P233A-P234A, C236A, E264A, C271A and W3 15A displayed a r educed galactosyltransferase activity ( see F ig. 3). The decreased activity observed with the alanine mutants of P233A- P234A, C236 and C271 may be caused by minor alterations of the tertiary structure. I n f act, a d irect i nvolvement of these residues in the catalytic process or substrate binding seems unlikely as w e observed no m odi®cations o f K m values for the donor UDP-Gal and acceptor GlcNAcb- pNP. In contrast, kinetic an alysis provided evidence that W101, W162 and E264 are involved in the binding of UDP- Gal and that E264 and W315 are involved in the GlcNA cb- pNP binding site. The residue E264, which is part of the conserved EDV motif, is involved in both substrate b inding sites, a phenomenon already observed in sialyltransferases [35,36]. Mutagenesis analysis of the Galb1,4GlcNAc a2,6- sialyltransferase s howed that the residues S320, G321, V335 and E339, all belonging to the S-sialyl motif [25], participate in the binding o f the donor substrate CMP-sialic acid as well as in the binding of the accep tor substrate a sialo a 1 -acid glycoprotein. The involvement of aromatic residues, such as Table 3. Donor and acceptor speci®city of the protA±b3GalT-I mutant constructs. UDP-Gal, UDP-GlcNAc and UDP-GalNAc w ere used a t 0.5 m m. GlcNAcb-pNP, GalNAcb-pNP and Gal(b1,4)Glcb-pNP (l actoseb-pNP) were used at 10 m M . Results re present averages of three mea- surements. All values are glycosyltransferase activity in pmolámin )1 ámg protein )1 . Donors are UDP-G al, UDP-GlcNAc a nd UDP-G alNAc. Acceptors are GlcNAcb-pNP and GalNAcb-pNP. UDP-Gal UDP-GlcNAc UDP-GalNAc GlcNAcb-pNP GalNAcb-pNP Lactoseb-pNP GalNAcb-pNP Lactoseb-pNP Sf9mock 335 132 75 12 135 b3GalT-I 11 665 134 160 14 148 C73A 276 140 80 23 129 I97A-R98A 989 166 265 13 64 W101 2571 94 76 40 54 F116A-L117A-L118A-G119A 420 158 133 73 76 W162A 2073 108 166 15 125 C167A 612 151 88 26 129 D177A-D179A 541 94 76 40 60 P233A-P234A 7135 102 115 33 66 C236A 11 579 153 101 88 146 E264A 2589 108 93 22 67 C271A 3286 128 99 65 141 C295A 412 156 82 93 134 W315A 1430 166 112 12 66 C326A 208 131 77 72 129 Table 2. Kinetic p arameters for t he protA±b3GalT-I constructs. Apparent K m value (m M ) UDP-Gal GlcNAcb-pNP b3GalT-I 1.5 8.3 W101A 5.4 9.3 W162A 11.8 12.8 P233A-P234A 1.5 8.3 C236A 1.9 10.7 C271A 2.5 12.0 E264A 31.6 62.0 W315A 1.7 50.0 Ó FEBS 2002 Mutagenesis of b3GalT-I (Eur. J. Biochem. 269) 237 W101 and W162 of the b3GalT-I e nzyme, in the binding of UDP-sugar donor substrates has been previou sly docu- mented. In fact, tryptophan residues located NH 2 -proxi- mally to the DXD motif of clostridial glucosyltransferases have also been implicated in the binding of UDP-Glc [32]. In addition, the crystal structures of SpsA f rom Bacillus subtilis [37] and of the bovine b1,4-GalT-I [29], two DXD- containing glycosyltransferases, s howed an aromatic resi- due, which is involved in the stacking of t he uracil ring of the cosubstrate in the catalytic fold. As found for the trypto- phan analog in the bacterial glucosyltransferase [28] and for W101 and W162 in murine b3GalT-I, this conserved aromatic residue is located NH 2 -proximally to t he DXD motif, and there is no strictly de®ned d istance between this residue and t he latter m otif. Therefore, it is tempting to speculate that W101 and W 162 represent analogous residues to the aromatic residues of SpsA and bovine b1,4-GalT-I. Taken together, our results provided evidence that several of t he conserved m otifs common t o b3GT enzymes are required for proper enzymatic activity. Our results also suggested the formation of possible disul®de bonds, which hold the enzyme in a c onformation that is required f or its catalytic activity. The conservation between the different members of the b3GT family suggests a common evolutionary origin; structural requirements for the catalysis of a b1,3-glycosidic linkage probably main- tained the motifs in t he evolving polypeptides. ACKNOWLEDGEMENTS We thank Bea Berger and Claudia Ruedin for their technical assistance. This work was supported by the Swiss National Science Foundation Grants 31-58577.99 to T. H., 5000-57797 to EGB. M. M. was supported by a scholarship of the M arie Heim-Vo È gtlin Foundation. REFERENCES 1. Campbell, J.A., Davies, G.J., Bulone, V. & Henrissat, B. (1997) A classi®cation of nucleotide-diphospho-sugar glycosyltransfe- rases based on amino acid sequence similarities. Biochem. J. 326, 929±939. 2. Yuan, Y.P., Schultz, J., Mlodzik, M. & Bork, P. ( 1997) Secreted Fringe-like signaling molecules may be glycosyltransferases. Cell 88, 9±11. 3. Moloney,D.J.,Panin,V.M.,Johnston,S.H.,Chen,J.H.,Shao,L., Wilson, R., Wang, Y ., Stanley, P., Irvine, K.D., H altiwanger, R.S. & V ogt, T.F. (2000) Fringe is a glycosyltransferase that modi®es Notch. Na tu re 406, 369±375. 4. Miyazaki, H., Fukumoto , S., Okada, M ., Hasegawa, T. & Furu- kawa, K. (1997) Expression cloning of rat cDNA encoding UDP- galactose: G D2 b1, 3-galactosyltransferase that determin es th e expression of G D1b /G M1 /G A1 . J. Biol. Chem. 272, 24794±24799. 5. Hennet, T., Dinter, A., Kuhnert, P., Mattu, T.S., Rudd, P .M. & Berger, E .G. (1998) G enomic cloning a nd expression of three murine UDP-galactose:b-N-acetylgluco samine b1,3-galactosyl- transferase genes. J. Bio l . Chem. 273, 5 8±65. 6. Kolbinger, F., Strei, M.B. & Katopodis, A.G. (1998) Cloning of a human UD P-galactose:2-acetamido-2-deoxy- D -glucose 3b-galac- tosyltransferase catalyzing the form ation of type 1 chains. J. Biol. Chem. 273, 433±440. 7. Amado, M., Almeida, R., Carneiro, F., Levery, S.B., Holmes, E.H., Nomoto, M., Hollingsworth, M.A., Hassan, H., Schwien- tek, T., Nielsen, P.A., Bennett, E.P. & Clausen, H. (1998) A family of h uman b-3-galactosyltransferases ± characterization of four members o f a UDP-galactose-b-N-acetylglucosami ne/b-N-acetyl- galactosamine b-1,3-galacto syltransferase family. J. Bio l. C he m. 273, 12770±12778. 8. Isshiki, S., Togayachi, A., Kudo, T., Nishihara, S., Watanabe, M., Kubota, T., Kitajima, M., Shiraishi, N., Sasaki, K., Andoh, T. & Narimatsu, H. (1999) Cloning, expression, and characterization of a novel UDP-galactose: b-N-acetylglucosamine b1,3-galactosyl- transferase (b3Gal-T5) responsible for synthesis of type 1 chain i n colorectal and pancreatic e pithelia and t umor cells derived there- from. J. Biol. Chem. 274, 1 2499±12507. 9. Zhou, D., Berger, E.G. & Hennet, T. (1999) Molecular clonin g of a human UDP-galactose: GlcNAcb1,3G alNAc b1,3 galactosyl- transferase gene encoding an O-linked core3-elongation enzyme. Eur. J. Biochem. 263, 571±576. 10. Bai, X., Zhou, D., Brown, J.R., H ennet, T. & Esko, J .D. (2001 ) Biosynthesis of the linkage region of glycosaminoglycans: c loning and activity of galactosyltransferase II, the sixth member of the b1,3galactosyltransferase f amily (b3GalT6). J. Biol. Chem. DOI M107339200. 1 11. Shiraishi, N., Natsume, A., T ogayachi, A ., E ndo, T., Akashima, T.,Yamada,Y.,Imai,N.,Nakagawa,S.,Koizumi,S.,Sekine,S., Narimatsu, H. & Sasaki, K. (2000) I de nti®cation and character- ization of three novel b1,3-N-acetylglucosaminyltransferases structurally related to the b1,3-galactosyltransferase family. J. Biol. Chem. 276, 3498±3507. 12. Togayachi, A ., Akashima, T., Ookubo, R ., Kudo, T., Nishihara, S., I wasaki, H., Natsume, A., Mio, H., Inokuchi, J ., Irimura, T ., Sasaki, K. & Narimatsu, H. (2001) Molecular c loning and c har- acterization of UDP-GlcNAc: lactosylceramide b1,3-N-acetyl- glucosaminlytransferase (b3Gn-T5), an essential enzyme for the expression of th e HNK-1 an d Lewis X epitopes on g lycolip ids. J. Biol. Chem. 276, 22032±22040. 13. Henion, T.R., Zhou, D., W olfer, D.P., Jungalwala, F.B. & Hennet, T. (2001) Cloning of a m ouse b1,3 N-acetylglucosami- nyltransferase GlcNAc (b1,3) Gal (b1,4) Glc±ceramide synthase gene encoding the key regulator of lacto±series glycolipid bio- synthesis. J. Biol. Chem. 276, 3 0261±30269. 14. Okajima, T., Nakamura, Y ., Uchikawa, M., Haslam, D.B., Numata, S.I., Furukawa, K., U rano, T. & Furukawa, K. (2000) Expression cloning of h uman globoside synthase c DNAs. Identi- ®cation o f b3Gal-T3 as UDP-N-acetylgalactosamine:globotriao- sylceramide b1,3-N-acetylgalactosaminyltransferase. J. Biol. Chem. 275, 40498±40503. 15. Blixt, O., Van D ie, I., Norberg, T. & van den Eijnden, D.H. (1999) High-level ex pression o f the Neisseria meningitidis lgtA gene in Escherichia coli and c haracterization of the e ncoded N-acetylglu- cosaminyltransferase as a useful catalyst i n the synthesis of GlcNAc b1±3Gal and GalNAc b1±3Gal linkages. Glycobiology 9, 1061±1071. 16. Zhou,D.,Henion,T.,Jungalwala,F.B.,Berger,E.G.&Hennet,T. (2000) The b1,3-galactosyltransferase b3GalT-V is a stage-speci®c embryonic a ntigen-3 (SSEA-3) synthase. J. Biol. Chem. 275 , 22631±22634. 17. Lo, N.W., Shaper, J.H., Pevsner, J. & Shaper, N.L. (1998) The expanding b4-galactosyltransferase gene family: m essages from the databanks. Glycobiology 8, 517±526. 18. Joziasse, D.H., Shaper, J .H., van d en Eijnden, D.H., Van Tunen, A.J. & Shaper, N.L. (1989) Bovine a1±3-galactosyltransferase: isolation and characte rization of a cDNA clone. i de nti®cation of homologous sequences in human genomic DNA. J. Biol. Che m. 264, 14290±14297. 19. Yamamoto, F., Clausen, H., White, T., Marken, J. & Hakomori, S. (1990) Molecular genetic basis of t he histo-blood g ro up ABO system. Nature 345, 229±233. 20. Higuchi, R., Krummel, B. & Saiki, R.K. (1988) A general method of in vitro preparation and speci®c mutagenesis of DNA frag- ments: study o f protein and DNA interactio ns. Nucleic Acids Res. 16, 7351±7367. 238 M. Malissard et al. (Eur. J. Biochem. 269) Ó FEBS 2002 21.Zhou,D.,Malissard,M.,Berger,E.G.&Hennet,T.(2000) Secretion and pu ri®cation of recombinant b1±4 ga lactosyltrans- ferase from insect cells using pFmel-protA, a novel tran sposition - based baculovirus transfer vector. Arch. Biochem. Biophys. 374, 3±7. 22. Luckow, V.A., Lee, S.C., Barry, G.F. & Olins, P.O. (1993) E- cient generation of infectious recombinant baculoviruses by site- speci®c t ransposon- mediat ed inse rtio n o f foreign genes i nto a baculovirus genom e propagated in Escherichia coli. J. Virol. 67, 4566±4579. 23. Laemmli, U.K. (1970) Cleavage of structural proteins d uring the assembly of the h ead of bacteriophage T4. Nature 227, 680±685. 24. Towbin, H., Staehelin, T. & Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl Acad. Sci. USA 76, 4350±4354. 25. Livingston, B.D. & Paulson, J.C. (1993) Polymerase chain reac- tion cloning of a developmentally regulated member of the sialyltransferase g ene family. J . Biol. C hem. 268, 11504±11507. 26. Goode, S., Wright, D. & Mahowald, A.P. (19 92) The neurogenic locus brainiac cooperates with the Dro sophila EGF r ec eptor t o establish t he ovarian follicle and to determine its dorsal-ventral polarity. Development 116 , 177±192. 27.Saxena,I.M.,Brown,R.M.Jr.,Fevre,M.,Geremia,R.A.& Henrissat, B. (1995) Multidomain architecture of b-glycosyl transferases: implications for mechanism of action. J. Bacteriol. 177, 1419±1424. 28. Busch, C., Hofmann, F., Gerhard, R. & Aktories, K. (2000) Involvement of a conserved tryptophan residue in the UDP- glucose binding of large clostridial cytotoxin glycosyltransferases. J. Biol. Chem. 275, 13228±13234. 29. Gastinel, L .N., Cambillau, C. & Bourne, Y. ( 1999) Crystal structures of the b ovine b4galactosyltransferase catalytic domain and its complex with uridine diphosphogalac tose. EMBO J. 18, 3546±3557. 30. Wiggins, C.A. & Munro, S. (1998) Activity of the yeast MNN1 a-1,3-mannosyltransfe rase requires a motif conserved in m any other f amilies of glycosyltransferases. Pro c. Natl Acad. Sc i. USA 95, 7945±7950. 31. Nagahashi, S., Sudoh, M., Ono, N., Sawada, R., Yamaguchi, E., Uchida, Y., Mio, T., Takagi, M., Arisawa, M. & Yamada-Okabe, H. (1995) Characterization of chitin synthase 2 o f Saccha romyce s cerevisiae. Implication of two highly conserved domains as pos- sible c atalytic sites. J. Biol. Chem. 270, 1 3961±13967. 32. Busch, C., Hofmann, F., Selzer, J., Munro, S., Jeckel, D. & Aktories, K . (1998) A c ommo n motif of eu karyotic glycosyl- transferases is essent ial for the enzyme activity o f large clostridial cytotoxins. J. Biol. Chem. 273, 19566±19572. 33. Unligil, U.M., Zhou, S., Yuwaraj, S ., Sarkar, M., Schachte r, H. & Rini, J.M. (2000) X-ray cry stal structure of rabbit N-acetylglu- cosaminyltransferase I: catalytic mechanism and a new protein superfam ily. EMBO J. 19, 5 269±5280. 34. Karlin, S., Zhu, Z.Y. & Karlin, K.D. (1997) The extended envi- ronment of mononuclear metal centers in protein structures. Proc. Natl Acad. Sci. USA 94, 14225±14230. 35. Datta, A.K. & Paulson, J.C. (1995) The sialyltransferase Ôsialyl- motifÕ participates in binding the donor substrate CMP-NeuAc. J. Biol. Chem. 270, 1497±1500. 36. Datta, A.K., Sinha, A. & Paulson, J.C. (1 998) Mutation of the sialyltransferase S-sialylmotif alters the kinetics of t he donor and acceptor substrates. J. Biol . Chem. 27 3, 9608±9614. 37. Charnock, S.J. & Davies, G.J. (1999) Structure of the nucleotide- diphospho-sugar transferase, SpsA from Bacillus subtilis,innative and nucleotide-complexed f orms. Biochemistry 38 , 6380±6385. 38. Thompson, J.D., Higgins, D.G. & Gibson, T.J. (1994) Clustal W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-speci®c gap penalties and weight matrix choice . Nucleic Acids Res. 22, 4673± 4680. Ó FEBS 2002 Mutagenesis of b3GalT-I (Eur. J. Biochem. 269) 239 . TGCTTCATCTTG CTGACGTGTACGTGGGACT C27 1A ATGTGTACGTGGGACTGGCACTTCGAAAGC C29 5A AAAATGGCCTACAGTTTA GCTCGGTACC W31 5A CAGAATC GCCAATGACATGTCAAGGAAGAAGCATCTGAGATGTTAGTCTAGATAT C32 6A. CAATCCGGGAGACA GCTGGTGATGAAAA F11 6A- L11 7A- L11 8A- G11 9A TAGCCACACTTG CAGCCGCGGCCAAAAATG W16 2A TTAATGGGGATGAGAGCGGTT GCCACTTTCT C16 7A AGATGGGTTGGCAACTTTC GCTTCAAAA D17 7A- D17 9A- F181A