Portal de Periódicos da CAPES

Structural Basis for the Inactivity of Human Blood Group O2 Glycosyltransferase

2004; Elsevier BV; Volume: 280; Issue: 1 Linguagem: Inglês

10.1074/jbc.m410245200

1083-351X

Hojun Lee, C.H. Barry, S.N. Borisova, Nina O.L. Seto, Ruixiang Blake Zheng, Antoine Blancher, Stephen V. Evans, Monica M. Palcic,

Glycosylation and Glycoproteins Research

The human ABO(H) blood group antigens are carbohydrate structures generated by glycosyltransferase enzymes. Glycosyltransferase A (GTA) uses UDP-GalNAc as a donor to transfer a monosaccharide residue to Fucα1-2Galβ-R (H)-terminating acceptors. Similarly, glycosyltransferase B (GTB) catalyzes the transfer of a monosaccharide residue from UDP-Gal to the same acceptors. These are highly homologous enzymes differing in only four of 354 amino acids, Arg/Gly-176, Gly/Ser-235, Leu/Met-266, and Gly/Ala-268. Blood group O usually stems from the expression of truncated inactive forms of GTA or GTB. Recently, an O2 enzyme was discovered that was a full-length form of GTA with three mutations, P74S, R176G, and G268R. We showed previously that the R176G mutation increased catalytic activity with minor effects on substrate binding. Enzyme kinetics and high resolution structural studies of mutant enzymes based on the O2 blood group transferase reveal that whereas the P74S mutation in the stem region of the protein does not appear to play a role in enzyme inactivation, the G268R mutation completely blocks the donor GalNAc-binding site leaving the acceptor binding site unaffected. The human ABO(H) blood group antigens are carbohydrate structures generated by glycosyltransferase enzymes. Glycosyltransferase A (GTA) uses UDP-GalNAc as a donor to transfer a monosaccharide residue to Fucα1-2Galβ-R (H)-terminating acceptors. Similarly, glycosyltransferase B (GTB) catalyzes the transfer of a monosaccharide residue from UDP-Gal to the same acceptors. These are highly homologous enzymes differing in only four of 354 amino acids, Arg/Gly-176, Gly/Ser-235, Leu/Met-266, and Gly/Ala-268. Blood group O usually stems from the expression of truncated inactive forms of GTA or GTB. Recently, an O2 enzyme was discovered that was a full-length form of GTA with three mutations, P74S, R176G, and G268R. We showed previously that the R176G mutation increased catalytic activity with minor effects on substrate binding. Enzyme kinetics and high resolution structural studies of mutant enzymes based on the O2 blood group transferase reveal that whereas the P74S mutation in the stem region of the protein does not appear to play a role in enzyme inactivation, the G268R mutation completely blocks the donor GalNAc-binding site leaving the acceptor binding site unaffected. The human A and B blood groups are produced by two closely related glycosyltransferase enzymes (1Yamamoto F. Clausen H. White T. Marken J. Hakomori S. Nature. 1990; 345: 229-233Crossref PubMed Scopus (869) Google Scholar, 2Watkins W.M. Adv. Hum. Genet. 1980; 10: 1-136PubMed Google Scholar, 3Palcic M.M. Seto N.O. Hindsgaul O. Transfus. Med. 2001; 11: 315-323Crossref PubMed Scopus (17) Google Scholar). Blood type A structures are synthesized by an α1-3 N-acetylgalactosaminyltransferase (GTA, 1The abbreviations used are: GTA, glycosyltransferase A; GTB, glycosyltransferase B; MOPS, 4-morpholinepropanesulfonic acid; aa, amino acids. 1The abbreviations used are: GTA, glycosyltransferase A; GTB, glycosyltransferase B; MOPS, 4-morpholinepropanesulfonic acid; aa, amino acids. EC 2.4.1.40) that transfers GalNAc from UDP-GalNAc to Fucα1-2Galβ-R (H)-terminating acceptors producing the A antigen GalNAcα1-3[Fucα1-2]Galβ-R. The B-synthesizing α1-3 galactosyltransferase (GTB, EC 2.4.1.37) transfers Gal from UDP-Gal to the same acceptors, producing the B antigen Galα1-3[Fucα1-2]Galβ-R. Individuals with blood type O do not express enzymes capable of modifying the H antigen (1Yamamoto F. Clausen H. White T. Marken J. Hakomori S. Nature. 1990; 345: 229-233Crossref PubMed Scopus (869) Google Scholar). GTA and GTB exhibit characteristic mammalian glycosyltransferase topologies; they are type II integral membrane proteins with a short N-terminal cytoplasmic tail, a transmembrane domain, a proteolytically sensitive stem region, and a catalytic domain (4Paulson J.C. Colley K.J. J. Biol. Chem. 1989; 264: 17615-17618Abstract Full Text PDF PubMed Google Scholar). GTA and GTB are highly homologous enzymes differing in only 4 of 354 amino acids, Arg/Gly-176, Gly/Ser-235, Leu/Met-266, and Gly/Ala-268 (1Yamamoto F. Clausen H. White T. Marken J. Hakomori S. Nature. 1990; 345: 229-233Crossref PubMed Scopus (869) Google Scholar, 5Yamamoto F. Hakomori S. J. Biol. Chem. 1990; 265: 19257-19262Abstract Full Text PDF PubMed Google Scholar). Substitution of these four critical amino acids converts the donor specificity from that of GTA to that of GTB. Recently, single crystal x-ray diffraction studies of soluble forms of GTA and GTB in complex with H acceptor and UDP have provided a structural basis for substrate recognition (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar). The critical residue Leu/Met-266 dominates donor selection with Gly/Ala-268 having a significant but lesser effect. The origin of blood group O was initially shown to be a deletion or mutation in the GTA or GTB gene that gave inactive truncated enzyme (1Yamamoto F. Clausen H. White T. Marken J. Hakomori S. Nature. 1990; 345: 229-233Crossref PubMed Scopus (869) Google Scholar). More recently, an O2 enzyme (O03) was discovered that was a full-length form of GTA with three substitutions, P74S, R176G, and G268R (7Yamamoto F. McNeill P.D. Yamamoto M. Hakomori S. Bromilow I.M. Duguid J.K. Vox Sang. 1993; 64: 175-178Crossref PubMed Scopus (107) Google Scholar, 8Grunnet N. Steffensen R. Bennett E.P. Clausen H. Vox Sang. 1994; 67: 210-215Crossref PubMed Scopus (83) Google Scholar). The O2 glycosyltransferase showed no measurable transferase activity when expressed in Sf9 cells (9Amado M. Bennett E.P. Carneiro F. Clausen H. Vox Sang. 2000; 79: 219-226Crossref PubMed Scopus (19) Google Scholar). Since the P74S mutation is distant from the active site and the R176G mutation in GTA increases enzyme turnover (10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar), it is presumably the replacement of glycine with arginine at position 268 that adversely affects enzyme activity (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar, 11Yamamoto F. McNeill P.D. J. Biol. Chem. 1996; 271: 10515-10520Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). GTA transferase-like activity ranging from 0.006 to 0.25% of GTA levels have been detected in some concentrated blood group O serum samples (12Greenwell P. Glycoconj. J. 1997; 14: 159-173Crossref PubMed Scopus (129) Google Scholar). Here we prepare three mutant enzymes derived from truncated soluble R176G GTA: P74S, G268R, and P74S/G268R. For comparison, we also report the structure of the GTA R176G mutant enzyme, which was earlier shown to have an enzyme turnover rate greater than that of wild-type GTA (10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Each of these constructs is based on the same synthetic genes with codons optimized for facile mutagenesis and a high level of expression in Escherichia coli (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar, 10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Kinetics and single crystal x-ray diffraction were used to determine the effect of each mutation on enzyme activity. Materials and General Techniques—All molecular biology procedures were as described previously (10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar, 13Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Current Protocols in Molecular Biology. 1. John Wiley & Sons, Inc., New York1997: 8.5.7-8.5.10Google Scholar, 14Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, NY1989Google Scholar, 15Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S. Eur. J. Biochem. 1995; 234: 323-328Crossref PubMed Scopus (48) Google Scholar, 16Seto N.O.L. Compston C.A. Evans S.V. Bundle D.R. Narang S.A. Palcic M.M. Eur. J. Biochem. 1999; 259: 770-775Crossref PubMed Scopus (88) Google Scholar). The original GTA and the GTA R176G gene sequences (aa 54-354) have been described (10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar, 15Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S. Eur. J. Biochem. 1995; 234: 323-328Crossref PubMed Scopus (48) Google Scholar, 16Seto N.O.L. Compston C.A. Evans S.V. Bundle D.R. Narang S.A. Palcic M.M. Eur. J. Biochem. 1999; 259: 770-775Crossref PubMed Scopus (88) Google Scholar). All chimeric GTA/GTB enzymes are referred to by a series of four letters corresponding to the origin of each of the four residues where AAAA is GTA and BBBB is GTB. We reported earlier the interesting kinetics and donor specificity of the chimeric enzyme BAAA (aa 54-354, Gly-176, Gly-235, Leu-266, Gly-268) (10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Truncation of a further 10 amino acids from the N terminus gave higher expression levels and solubility; therefore, all mutants in this study are derived from wild-type GTA and GTB enzymes (both corresponding to aa 63-354). Cloning of GTA Arg-176 —GTA R176, also denoted as BAAA (aa 63-354), was made by PCR amplification using the BAAA (aa 54-354) gene as template together with the forward primer MIN-2 (5′-A TAT GAATTC ATG GTT TCC CTG CCG CGT ATG GTT TAC CCG CAG CCG AA), which introduced an EcoRI site (underlined) in the 5′ end, and the reverse primer PCR-3B (5′-ATA ATT AAGCTT CTA TCA CGG GTT ACG AAC AGC CTG GTG GTT TTT), which introduced a HindIII site (underlined) in the 3′ end. The following PCR profile was used for the construction of all the clones: 94 °C, 3 min (94 °C, 30 s, 55 °C, 30 s, 72 °C, 1 min) for 30 cycles. After gel purification, the PCR products were digested with EcoRI and HindIII for 2 h at 37 °C and were ligated into pCWΔlac vector (17Gegner J.A. Dahlquist F.W. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 750-754Crossref PubMed Scopus (163) Google Scholar) opened with EcoRI/HindIII for 1 h at room temperature. Each ligation was transformed into BL21 competent cells (Novagen). All insert and plasmid purifications were made by Qiagen Plasmid Purification System (Qiagen Inc., Chatsworth, CA). All ligations were made by the use of T4 DNA ligase (Invitrogen) at room temperature overnight. The clones were characterized by triplicate DNA sequence analysis of the entire coding region. Cloning of BAAA P74S—The BAAA P74S mutant was constructed by PCR using BAAA (aa 63-354) described above as a template and the forward primer HJL01 (5′-A TAT GAA TTC ATG GTT TCC CTG CCG CGT ATG GTT TAC CCG CAG TCC AAA GTT CTG ACC CCA TGC CG-3′), which was designed with a single codon substitution (CCG → TCC) at codon 74 and an EcoRI site at the 5′ end and the reverse primer PCR-3B. The amplified genes were digested by restriction enzymes (EcoRI and HindIII) and cloned as described above. Cloning of BAAA G268R—The BAAA G268R mutant was constructed by recombinant PCR using the BAAA (aa 63-354) clone as a template. The first PCR was performed using the outside forward primer MIN2 together with the internal reverse primer HJL02 (5′-ACC GAA GAA ACG ACC CAG GTA GTA GAA GTC ACC-3′) that contains a single codon substitution (ACC → AGC) at codon 268. A second PCR was performed using the internal forward primer HJL03 (5′-C CTG GGT CGT TTC TTC GGT GGT TCC GTT CAG-3′) that contains a single codon substitution (GGT → CGT) at codon 268 together with the outside reverse primer PCR-3B. The outside forward and reverse primers included the EcoRI and HindIII restriction sites. The two overlapping PCR products were annealed together and amplified by PCR with the outside primers MIN2 and PCR-3B, digested with EcoRI and HindIII, and cloned as described above. Cloning of P74S/G268R—The BAAA P74S/G268R mutant was constructed by PCR using the BAAA G268R clone as a template together with the forward primer HJL01 and the reverse primer PCR-3B. The amplified gene was digested with restriction enzymes EcoRI and HindIII and cloned as described above. Protein Purification—Mutant enzymes were purified from E. coli as described previously (18Seto N.O.L. Compston C.A. Szpacenko A. Palcic M.M. Carbohydr. Res. 2000; 324: 161-169Crossref PubMed Scopus (45) Google Scholar, 19Marcus S.L. Polakowski R. Seto N.O.L. Leinala E. Borisova S. Blancher A. Roubinet F. Evans S.V. Palcic M.M. J. Biol. Chem. 2003; 278: 12403-12405Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). Expression levels for all of the mutants were good, and the yields of final purified proteins ranged from 26 mg/liter for the G268R mutant to 108 mg/liter for the P74S/G268R double mutant. Protein concentration was determined with a Bio-Rad protein assay kit based on the Bradford method (20Bradford M.M. Anal. Biochem. 1976; 72: 248-254Crossref PubMed Scopus (211983) Google Scholar) using bovine γ-globulin as a standard. Kinetic Characterization—Kinetic characterizations were carried out on all mutant enzymes using a Sep-Pak radiochemical assay with the hydrophobic acceptor Fucα1-2Galβ-O(CH2)7CH3 (21Palcic M.M. Heerze L.D. Pierce M. Hindsgaul O. Glycoconj. J. 1988; 5: 49-63Crossref Scopus (277) Google Scholar). Assays were carried out at 37 °C in a total volume of 15 μl containing substrates and enzyme in 50 mm MOPS buffer, pH 7.0, with 20 mm MnCl2 and 1 mg/ml bovine serum albumin. Seven different concentrations of the donor or acceptor were used, at a high concentration of the alternate substrate. The amount of substrate consumed was less than 15% to ensure linear initial reaction rates. The kinetic parameters Vmax and Km were derived from the best fit of the data to the Michaelis-Menten equation by using nonlinear regression with the GraphPad PRISM 3.0 program. Crystallography—All mutants related to O2 enzyme were crystallized using conditions similar to the native GTA and GTB enzymes (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar). Data were collected on a Rigaku R-AXIS4++ area detector at distances of 72 and 100 mm and exposure times between 4.0 and 5.0 min for 0.5° oscillations. X-rays were produced by an MM-002 generator (Rigaku/MSC) coupled to Osmic "Blue" confocal x-ray mirrors with power levels of 30 watts (Osmic). The crystals were frozen and maintained under cryogenic conditions at a temperature of -160 °C using a CryoStream 700 crystal cooler (Oxford). All structures were solved by using molecular replacement techniques with wild-type GTA or GTB (Protein Data Bank accession codes 1LZ0 and 1LZ7, respectively) as a starting model and were subsequently refined using the program CNS (22Brünger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16919) Google Scholar). Molecular images were generated using the program SETOR (23Evans S.V. J. Mol. Graphics. 1993; 11: 134-138Crossref PubMed Scopus (1249) Google Scholar). In this study, a series of mutant blood group glycosyltransferases were produced and characterized to determine the effect of each modification on enzyme structure and function. These mutants are based on the gene sequence for an O2 enzyme discovered in blood banking laboratories that yields a triple mutant of GTA with arginine 176 replaced with glycine, proline 74 replaced with serine, and glycine 268 replaced with arginine. These enzymes were all cloned as soluble truncated proteins (aa 64-354) of the catalytic domain and expressed in E. coli. The purification by successive ion-exchange and affinity chromatography on a donor-based UDP-hexanolamine resin was straightforward for all constructs, suggesting no impairment of UDP binding. Kinetic constants were determined for each purified enzyme at a high concentration of the alternate substrate (Table I). The kcat for BAAA enzyme (R176G) was larger than that of wild-type GTA, consistent with our report for an enzyme with the same mutation but with 10 additional amino acids on the N terminus (10Seto N.O.L. Palcic M.M. Hindsgaul O. Bundle D.R. Narang S.A. J. Biol. Chem. 1997; 272: 14133-14138Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Except for the P74S mutant, which was comparable with that of the wild-type GTA, the purified enzymes with mutations corresponding to the human glycosyltransferase O2 showed very low activity. The kcat for the G268R mutant was 4 × 104 times less than that of GTA and 3.3 × 104 times less than that of the P74S mutant. The P74S/G268R mutant showed the lowest activity of all, with a kcat of only 3.4 × 10-5 s-1.Table IKinetic constants for mutant glycosyltransferasesEnzymeKAaKA is the Michaelis-Menten constant Km for the acceptor (Fuca1-2Galβ-O-(CH2)7CH3) determined at 1 mm donor.KBbKB is the Michaelis-Menten constant Km for the donor UDP-GalNAc determined at high concentrations of acceptor 0.4-0.8 mm.kcatkcat/KAkcat/KBμmμms-1mm-1s-1mm-1s-1AAAA (GTA)cData are from Ref. 19.9.98.717.517702010BAAA (R176G)dThe name BAAA indicates that the first critical residue on GTA (Arg-176) has been mutated to the corresponding residue in GTB (Gly).48354810001370BAAA P74SeAll mutations were based on the BAAA clone as a template.7511014.6195133BAAA G268R1577000.000440.00280.00063BAAA P74S/G268R4223000.0000340.000810.000015a KA is the Michaelis-Menten constant Km for the acceptor (Fuca1-2Galβ-O-(CH2)7CH3) determined at 1 mm donor.b KB is the Michaelis-Menten constant Km for the donor UDP-GalNAc determined at high concentrations of acceptor 0.4-0.8 mm.c Data are from Ref. 19Marcus S.L. Polakowski R. Seto N.O.L. Leinala E. Borisova S. Blancher A. Roubinet F. Evans S.V. Palcic M.M. J. Biol. Chem. 2003; 278: 12403-12405Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar.d The name BAAA indicates that the first critical residue on GTA (Arg-176) has been mutated to the corresponding residue in GTB (Gly).e All mutations were based on the BAAA clone as a template. Open table in a new tab The O2 mutant enzymes also showed an increase in the Km values for both acceptor and donor (listed in Table I as KA and KB, respectively) over the wild-type GTA and the BAAA enzymes. For the P74S mutant, the Km for donor was 7-fold higher than GTA, whereas the effect on acceptor was less. The P74S/G268R mutant showed a large increase in Km for donor and a minor effect on acceptor Km (Table I). The presence of even these low turnover numbers suggests that the observed activity arises from either the target mutant enzyme or revertants (24Ly H.D. Withers S.G. Annu. Rev. Biochem. 1999; 68: 487-522Crossref PubMed Scopus (266) Google Scholar); the observation of an elevated donor Km suggests that the activity that is observed is not due to wild-type enzyme. Details of the data collection and refinement results for the enzymes in the unliganded and liganded forms are shown in Tables II and III, respectively. Diffraction data were collected to a maximum resolution of 1.59-1.49 Å, with final Rwork ranging from 0.203 to 0.210 and Rfree ranging from 0.232 to 0.253. All structures showed excellent electron density along the entire length of the polypeptide chain with the exception of the disordered loop (aa 176-195) and the final 10 amino acid residues at the C terminus, both of which were also absent in the native GTA/GTB structures (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar). The electron density surrounding the active site of the P74S/R176G/G268R triple mutant is shown in Fig. 1a.Table IIData collection and refinement results for O2-related mutants grown in the absence of substratesP74SG268RP74S and G268RResolution (Å)20-1.5920-1.5920-1.49Space groupC2221C2221C2221a (Å)52.752.752.9b (Å)150.1150.1150.3c (Å)79.779.779.8Rmerge (%)aValues in parentheses represent high resolution shell.,bRmerge=∑|Iobs−Iave|/∑Iave.4.9 (39.0)3.4 (30.2)3.7 (29.3)Completeness (%)aValues in parentheses represent high resolution shell.99.3 (98.6)93.5 (78.7)91.6 (54.5)Unique reflections42,62540,45847,938RefinementResolution20-1.820-1.820-1.8Rwork (%)cRwork=∑||Fo|−|Fc||/∑|Fo|.20.920.920.7Rfree (%)d10% of reflections were omitted in Rfree calculations.24.424.323.8No. water215246268r.m.s.er.m.s., root mean square. bond (Å)0.0060.0050.005r.m.s. angle (°)1.301.301.32a Values in parentheses represent high resolution shell.b Rmerge=∑|Iobs−Iave|/∑Iave.c Rwork=∑||Fo|−|Fc||/∑|Fo|.d 10% of reflections were omitted in Rfree calculations.e r.m.s., root mean square. Open table in a new tab Table IIIData collection and refinement results for O2-related mutant crystals grown in the presence of donor substrates and H acceptorsP74S + DIaDI, Deoxy inhibitor. + UDP-GalNAcG268R + HA + UDP-GalNAcP74S and G268R + HAbHA, H-antigen. + UDP-GalNAcResolution (Å)20-1.4920-1.5520-1.55Space groupC2221C2221C2221a (Å)52.652.852.7b (Å)149.1150.1150.3c (Å)79.179.879.5Rmerge (%)cValues in parentheses represent high resolution shell.,dRmerge=∑|Iobs−Iave|∑Iave.6.2 (27.2)5.3 (36.5)3.2 (25.9)Completeness (%)cValues in parentheses represent high resolution shell.95.0 (61.0)95.0 (72.1)96.6 (74.1)Unique reflections48,63044,09644,636RefinementResolution20-1.920-1.820-1.8Rwork (%)eRwork=∑||Fo|−|Fc||/∑|Fo|.21.020.720.3Rfree (%)f10% of reflections were omitted in Rfree calculations.25.323.423.2No. water243226227r.m.s.gr.m.s., root mean square. bond (Å)0.0060.0050.005r.m.s. angle (°)1.321.331.31a DI, Deoxy inhibitor.b HA, H-antigen.c Values in parentheses represent high resolution shell.d Rmerge=∑|Iobs−Iave|∑Iave.e Rwork=∑||Fo|−|Fc||/∑|Fo|.f 10% of reflections were omitted in Rfree calculations.g r.m.s., root mean square. Open table in a new tab In this report, the catalytic domain corresponding to the human blood group O2 glycosyltransferase was produced and characterized. The O2 enzyme is a triple mutant of GTA with arginine 176 replaced with glycine, proline 74 replaced with serine, and glycine 268 replaced with arginine. Except for the P74S mutant, the purified mutants for human glycosyltransferase O2 from E. coli showed very low enzymatic activity. To determine the effect of each mutation on the transferase activity, three mutants (P74S, G268R, and P74S/G268R) derived from BAAA were prepared, purified, and analyzed by enzyme kinetics and single crystal x-ray diffraction. All enzymes were mutated from and compared with the BAAA "super-A" enzyme, which has higher A activity than the wild-type. The crystal structure of the BAAA mutant (not shown) unfortunately does not reveal any details as to why this mutant shows significantly increased A activity, as next to the R176G mutation site lies a polypeptide loop adjacent to the active site of the enzyme that is observed to be disordered in almost every GTA/GTB native and mutant enzyme structure. The crystal structure of both enzymes containing the G268R mutation, including the P74S/G268R triple mutant responsible for conferring blood type O, immediately reveals the reason for the inactivation of the transferase, as Arg-268 completely obstructs the donor sugar recognition site. In the wild-type and BAAA structures the critical amino acids Leu/Met-266 directly recognize their respective donors by forming complementary interactions between the corresponding acetamido and hydroxyl groups substituted on C-2 of GalNAc and Gal. The second critical residue involved in donor recognition, Gly/Ala-268, is not thought to contact the donor sugar in GTA, whereas it excludes the bulkier A donor in GTB (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar). The G268R mutation renders this delicate recognition mechanism moot by completely blocking donor-sugar access to the active site (Figs. 1a and 2). In contrast to the donor, unambiguous electron density was observed for the H-antigen disaccharide acceptor, showing that the G268R mutation does not exclude this substrate from the binding site (Fig. 1b). These results are consistent with the Km values observed for all G268R mutants, which show KA values that are larger than wild-type GTA enzyme, but comparable with BAAA values, and KB values that are dramatically higher. Fig. 2 shows the severe conflicts between the GalNAc of donor and Arg-268 that limits its access to the active site. The G268R mutant enzyme also showed very low activity with a value of kcat 4.0 × 104-fold lower than that of the wild-type GTA, confirming that the Arg-268 mutation was mainly responsible for the deactivation of the enzyme. G268R displays Km values for the donor that are significantly higher than the wild-type and BAAA enzymes, whereas Km values for acceptor are not as elevated. The KA value was 16-fold higher than GTA and 3.3 times higher than the BAAA, whereas the KB value was 80 times higher than GTA and 20 times higher than BAAA. This indicates that although the mutation at residue 268 mainly affects donor binding, it affects acceptor binding as well. The observation in the crystal structure that the Arg-268 mutation completely blocks the donor sugar site indicates that the physical event corresponding to the observed KB value is the binding of the UDP moiety. This is supported by the fact that the enzyme clearly binds the UDP-hexanolamine affinity column during purification. However, UDP was not observed bound to the enzyme in the crystal structure despite using crystallization conditions similar to those that led to the observation of bound UDP in the structures of the native GTA and GTB enzymes (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar, 19Marcus S.L. Polakowski R. Seto N.O.L. Leinala E. Borisova S. Blancher A. Roubinet F. Evans S.V. Palcic M.M. J. Biol. Chem. 2003; 278: 12403-12405Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar, 25Nguyen H.P. Seto N.O.L. Cai Y. Leinala E. Borisova N. Palcic M.M. Evans S.V. J. Biol. Chem. 2003; 278: 49191-49195Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar). Most surprisingly, the drastic G268R mutation does not significantly affect acceptor orientation, as this substrate is observed to form the same sets of interactions in the mutant as the wild-type enzyme. However, both the Leu-266 and Arg-268 side chains become disordered upon binding of the acceptor, whereas they are ordered in the unliganded mutant enzyme (Fig. 1). Most interestingly, the P74S mutation also shows significant impact on the reaction kinetics. The P74S mutation alone causes a 3-fold reduction in kcat of that in BAAA, bringing it to levels comparable with the wild-type GTA. The same mutation in BAAA G268R causes a 13-fold reduction in kcat in an enzyme that is already 40,000 times less active than the wild type. Although it is clear that the mutation at Arg-268 is the primary contributor to nonfunctionality of the O2 enzyme, the effect of the P74S mutation is notable as this residue lies in the stem region of the enzyme and not in the catalytic domain and is more than 30 Å distant from the active site. An examination of the crystal structure reveals a potential reason for this effect, as the stem regions of adjacent molecules in the crystal lattice are observed to interact to form dimers (Fig. 3). The dimer interface involves the first 15 N-terminal residues of the stem region and is situated such that residues adjacent to Pro-74 in one-half of the dimer are in contact with residues close to the active site in the other half of the dimer. Although the P74S mutation does not induce any significant change in crystal structures of the enzyme or change the topology of the active site and characteristic acceptor contacts (6Patenaude S.I. Seto N.O.L. Borisova S.N. Szpacenko A. Marcus S.L. Palcic M.M. Evans S.V. Nat. Struct. Biol. 2002; 9: 685-690Crossref PubMed Scopus (200) Google Scholar), the mutation from proline to serine would alter its flexibility and causes some rearrangement in the hydrogen bonding patterns. The dimerization observed for the soluble construct of O2 mutant enzyme is also observed in the wild-type GTA and GTB structures. An examination of glycosyltransferases with related folds with known structure shows that this dimerization through the N-terminal residues is not a general feature but is specific to the crystal packing of GTA, GTB, and their mutants. Nevertheless, soluble fragments of GTA and GTB are known to circulate in sera where they would have the opportunity to form dimers. Evidence that the dimerization observed in the crystal lattice is also present in solution comes from electrophoresis experiments. SDS-PAGE of a purified sample shows a single band of appropriate molecular weight; however, the removal of SDS and dithiothreitol from the running buffer results in the appearance of a second band at higher molecular weight corresponding to a GTA dimer. N-terminal sequencing of the band confirmed it was GTA. In summary, the primary effect of the G268R mutation in O2 glycosyltransferase is restriction of the access of the monosaccharide GalNAc of donor in the enzyme active site. Substrate turnover is extremely slow, with kcat values orders of magnitude lower than wild-type glycosyltransferase A. Acceptor binding is also affected as reflected in an increase in Km for H-acceptor; however, its orientation in the active site is unchanged from that of wild-type GTA. Although low levels of GTA activity have been reported in the pooled serum of blood type O individuals, ultra-sensitive enzyme assays on a single O2 serum sample will be required for confirmation of the weak activity observed for our recombinant protein. Assistance with DNA sequencing was provided by the Molecular Biology Service Unit, Department of Biological Sciences, University of Alberta, Alberta, Canada. We thank F. W. Dahlquist for the pCW vector, W. Wakarchuk for the pCWΔlac vector, O. Hindsgaul for the Fucα1-2Galβ-O(CH2)7CH3 acceptor, and N. L. Rose for the gel electrophoresis experiments.