The MPB83 Antigen from Mycobacterium bovis ContainsO-Linked Mannose and (1 → 3)-Mannobiose Moieties
2003; Elsevier BV; Volume: 278; Issue: 18 Linguagem: Inglês
10.1074/jbc.m207959200
ISSN1083-351X
AutoresStephen Michell, Adam O. Whelan, Paul R. Wheeler, Maria Panico, Richard L. Easton, A. Etienne, Stuart M. Haslam, Anne Dell, Howard R. Morris, Andrew J. Reason, Jean‐Louis Herrmann, Douglas B. Young, R. Glyn Hewinson,
Tópico(s)Tuberculosis Research and Epidemiology
ResumoMycobacterium tuberculosis andMycobacterium bovis, the causative agents of human and bovine tuberculosis, have been reported to express a range of surface and secreted glycoproteins, although only one of these has been subjected to detailed structural analysis. We describe the use of a genetic system, in conjunction with lectin binding, to characterize the points of attachment of carbohydrate moieties to the polypeptide backbone of a second mycobacterial glycoprotein, antigen MPB83 fromM. bovis. Biochemical and structural analysis of the native MPB83 protein and derived peptides demonstrated the presence of 3 mannose units attached to two threonine residues. Mannose residues were joined by a (1 → 3) linkage, in contrast to the (1 → 2) linkage previously observed in antigen MPT32 from M. tuberculosisand the (1 → 2) and (1 → 6) linkages in other mycobacterial glycolipids and polysaccharides. The identification of glycosylated antigens within the M. tuberculosis complex raises the possibility that the carbohydrate moiety of these glycoproteins might be involved in pathogenesis, either by interaction with mannose receptors on host cells, or as targets or modulators of the cell-mediated immune response. Given such a possibility characterization of mycobacterial glycoproteins is a step toward understanding their functional role and elucidating the mechanisms of mycobacterial glycosylation. Mycobacterium tuberculosis andMycobacterium bovis, the causative agents of human and bovine tuberculosis, have been reported to express a range of surface and secreted glycoproteins, although only one of these has been subjected to detailed structural analysis. We describe the use of a genetic system, in conjunction with lectin binding, to characterize the points of attachment of carbohydrate moieties to the polypeptide backbone of a second mycobacterial glycoprotein, antigen MPB83 fromM. bovis. Biochemical and structural analysis of the native MPB83 protein and derived peptides demonstrated the presence of 3 mannose units attached to two threonine residues. Mannose residues were joined by a (1 → 3) linkage, in contrast to the (1 → 2) linkage previously observed in antigen MPT32 from M. tuberculosisand the (1 → 2) and (1 → 6) linkages in other mycobacterial glycolipids and polysaccharides. The identification of glycosylated antigens within the M. tuberculosis complex raises the possibility that the carbohydrate moiety of these glycoproteins might be involved in pathogenesis, either by interaction with mannose receptors on host cells, or as targets or modulators of the cell-mediated immune response. Given such a possibility characterization of mycobacterial glycoproteins is a step toward understanding their functional role and elucidating the mechanisms of mycobacterial glycosylation. concanavalin A phosphate-buffered saline Tris-buffered saline anion exchange chromatography orthogonal acceleration time of flight collisionally activated decomposition tandem mass spectrometry amino acid(s) lipoarabinomannan phosphatidylinositol mannoside Protein glycosylation is ubiquitous in eukaryotes and the diverse array of carbohydrate moieties that can be fashioned to a polypeptide backbone makes glycoproteins ideal molecules to allow specific interactions with other molecules. In contrast, the number of reports of this post-translational modification by prokaryotes is comparatively low. Examples of eubacterial glycoproteins identified to date include cell surface or secreted proteins of pathogens that have antigenic properties and play a role in host pathogen interaction (1Virji M. Saunders J.R. Sims G. Makepeace K. Maskell D. Ferguson D.J. Mol. Microbiol. 1993; 10: 1013-1028Crossref PubMed Scopus (157) Google Scholar, 2Stimson E. Virji M. Makepeace K. Dell A. Morris H.R. Payne G. Saunders J.R. Jennings M.P. Barker S. Panico M. Blench I. Moxon E.R. Mol. Microbiol. 1995; 17: 1201-1214Crossref PubMed Scopus (220) Google Scholar, 3Wu H. Mintz K.P. Ladha M. Fives-Taylor P.M. Mol. Microbiol. 1998; 28: 487-500Crossref PubMed Scopus (117) Google Scholar, 4Popov V.L. Yu X. Walker D.H. Microb. Pathog. 2000; 28: 71-80Crossref PubMed Scopus (81) Google Scholar, 5McBride J.W. Yu X.J. Walker D.H. Infect. Immun. 2000; 68: 13-18Crossref PubMed Scopus (56) Google Scholar, 6Chia J.S. Chang L.Y. Shun C.T. Chang Y.Y. Chen J.Y. Infect. Immun. 2001; 69: 6987-6998Crossref PubMed Scopus (64) Google Scholar). Glycosylation of pilin has been postulated to enhance the adherence ofNeisseria meningitidis to endothelial cells (2Stimson E. Virji M. Makepeace K. Dell A. Morris H.R. Payne G. Saunders J.R. Jennings M.P. Barker S. Panico M. Blench I. Moxon E.R. Mol. Microbiol. 1995; 17: 1201-1214Crossref PubMed Scopus (220) Google Scholar), whereasN-glycosylation of the platelet aggregation-associated protein of Staphylococcus sanguis may promote colonization of the endocardium (7Erickson P.R. Herzberg M.C. J. Biol. Chem. 1993; 268: 23780-23783Abstract Full Text PDF PubMed Google Scholar, 8Erickson P.R. Herzberg M.C. J. Biol. Chem. 1993; 268: 1646-1649Abstract Full Text PDF PubMed Google Scholar). Removal of a sero-specific glycosyl moiety from the flagellin of Campylobacter jejuni has been reported to alter the O antigenicity of the organism (9Doig P. Kinsella N. Guerry P. Trust T.J. Mol. Microbiol. 1996; 19: 379-387Crossref PubMed Scopus (136) Google Scholar, 10Szymanski C.M. Yao R. Ewing C.P. Trust T.J. Guerry P. Mol. Microbiol. 1999; 32: 1022-1030Crossref PubMed Scopus (318) Google Scholar). Similarly, Romainet al. (11Romain F. Horn C. Pescher P. Namane A. Riviere M. Puzo G. Barzu O. Marchal G. Infect. Immun. 1999; 67: 5567-5572Crossref PubMed Google Scholar) demonstrated that removal of covalently bound mannose from the Mycobacterium tuberculosis antigen MPT32 reduced by 10-fold its ability to elicit a delayed-type hypersensitivity reaction in guinea pigs immunized withMycobacterium bovis BCG. M. tuberculosis and M. bovis are closely related members of the "M. tuberculosis complex" (MTb complex) that secrete a series of immunodominant antigens that have been reported to be glycosylated, on the basis of their ability to bind lectins such as concanavalin A (ConA)1 (12Espitia C. Mancilla R. Clin. Exp. Immunol. 1989; 77: 378-383PubMed Google Scholar, 13Fifis T. Costopoulos C. Radford A.J. Bacic A. Wood P.R. Infect. Immun. 1991; 59: 800-807Crossref PubMed Google Scholar, 14Garbe T. Harris D. Vordermeier M. Lathigra R. Ivanyi J. Young D. Infect. Immun. 1993; 61: 260-267Crossref PubMed Google Scholar, 15Dobos K.M. Swiderek K. Khoo K.H. Brennan P.J. Belisle J.T. Infect. Immun. 1995; 63: 2846-2853Crossref PubMed Google Scholar). Structural confirmation of protein glycosylation by mycobacteria was provided by detailed chemical compositional analysis of MPT32, a 45/47-kDa secreted antigen of M. tuberculosis (15Dobos K.M. Swiderek K. Khoo K.H. Brennan P.J. Belisle J.T. Infect. Immun. 1995; 63: 2846-2853Crossref PubMed Google Scholar, 16Dobos K.M. Khoo K.H. Swiderek K.M. Brennan P.J. Belisle J.T. J. Bacteriol. 1996; 178: 2498-2506Crossref PubMed Scopus (154) Google Scholar). Mannose, mannobiose, and mannotriose substituents were found O-linked to four threonine residues on the protein backbone. Also, site-directed mutagenesis of the 19-kDa lipoprotein antigen of M. tuberculosis implicated threonine residues in ConA binding, consistent with the observed O-linked glycosylation (17Herrmann J.L. O'Gaora P. Gallagher A. Thole J.E. Young D.B. EMBO J. 1996; 15: 3547-3554Crossref PubMed Scopus (137) Google Scholar). On the basis of monoclonal antibody-binding studies, Fifiset al. (13Fifis T. Costopoulos C. Radford A.J. Bacic A. Wood P.R. Infect. Immun. 1991; 59: 800-807Crossref PubMed Google Scholar) proposed that the 25/23-kDa secreted antigen of M. bovis was a glycosylated form of the major M. bovis antigen MPB70. They suggested that the increases in molecular weight, relative to the 22,000 MPB70, were attributable to glycosylation with mannose, as treatment of these antigens with α-mannosidase resulted in a reduction in their relative molecular weight. Subsequently, we demonstrated the presence of a gene,mpb83, which encodes a protein with 61% identity at the amino acid level to MPB70 (18Hewinson R.G. Michell S.L. Russell W.P. McAdam R.A. Jacobs W.J. Scand. J. Immunol. 1996; 43: 490-499Crossref PubMed Scopus (109) Google Scholar). In the same study, the protein encoded for by this gene, MPB83, was shown to bind a monoclonal antibody specific for the 25/23-kDa M. bovis antigen. Thus, the 25/23-kDa antigen characterized by Fifis et al. (13Fifis T. Costopoulos C. Radford A.J. Bacic A. Wood P.R. Infect. Immun. 1991; 59: 800-807Crossref PubMed Google Scholar) is MPB83, rather than a glycosylated form of MPB70. M. bovis expresses high levels of both MPB70 and MPB83 and these proteins are strongly recognized by the immune system duringM. bovis infection in cattle and badgers (19O'Loan C.J. Pollock J.M. Hanna J. Neill S.D. Clin. Diagn. Lab. Immunol. 1994; 1: 608-611Crossref PubMed Google Scholar, 20Goodger J. Nolan A. Russell W.P. Dalley D.J. Thorns C.J. Croston P. Newell D.G. Vet. Rec. 1994; 135: 82-85Crossref PubMed Scopus (83) Google Scholar). Although the proteins are expressed only at low levels in M. tuberculosis when grown in vitro, they are highly immunogenic during infection with live bacteria in mice (18Hewinson R.G. Michell S.L. Russell W.P. McAdam R.A. Jacobs W.J. Scand. J. Immunol. 1996; 43: 490-499Crossref PubMed Scopus (109) Google Scholar) and man (21Roche P.W. Triccas J.A. Avery D.T. Fifis T. Billman-Jacobe H. Britton W.J. J. Infect. Dis. 1994; 170: 1326-1330Crossref PubMed Scopus (67) Google Scholar). In addition, vaccination of mice with a plasmid encoding MPB83 has been shown to confer significant protection against challenge withM. bovis (22Chambers M.A. Vordermeier H. Whelan A. Commander N. Tascon R. Lowrie D. Hewinson R.G. Clin. Infect. Dis. 2000; 30: S283-S287Crossref PubMed Scopus (38) Google Scholar). Thus, given its strong recognition by the immune system, further characterization of MPB83 is required. Open reading frames with homology to MPB70 and MPB83 can be identified in genome sequences from Streptomyces coelicolorand other microbes, but a physiological function has yet to be ascribed to these proteins. The aims of this study were to explore the glycosylation status of MPB83 and MPB70, to map specific sites for glycosylation within the molecules, and to elucidate the chemical structure of the carbohydrate moiety. Using a combination of genetic and biochemical approaches we have demonstrated that MPB83 but not MPB70 is glycosylated, and that glycosylation occurs via attachment of short (1→3)-linked mannose chains to two threonine residues. Escherichia coli DH5α (Invitrogen) and Mycobacterium smegmatis mc2155 expressing PhoA hybrid proteins were grown in Luria broth, 50 μg/ml hygromycin, or in Middlebrook 7H9 supplemented with albumin, dextrose, catalase (MADC), and 0.05% Tween 80, 200 μg/ml hygromycin, respectively. Alkaline phosphatase activity of E. coli and mycobacterial PhoA recombinants was determined by the presence of blue colonies when plated on L agar (containing appropriate hygromycin concentrations) supplemented with 40 μg/ml 5-bromo-4-chloro-3-indolyl-phosphate-p-toluidine (Sigma). The genes encoding MPB70, MPB83, and those regions of MPB83 used for deletion mapping, were amplified by PCR using the oligonucleotides described in TableI. Oligonucleotides were supplied by Oswel, Southampton, United Kingdom. PCR amplifications were carried out in a DNA thermal cycler (Biometra) using 0.5 units of clonedPfu DNA polymerase (Stratagene). PCR reactions were performed in 50-μl volumes containing: 50 mm-KCl, 10 mm Tris-HCl (pH 8.8), 0.01% (w/v) gelatin, 5 mm MgCl2, 200 μm of each dinucleotide triphosphate (dNTP), and 20 pmol of each oligonucleotide primer overlaid with mineral oil. The parameters for amplification were one denaturation cycle at 96 °C for 2 min followed by 30 cycles of denaturation at 96 °C for 1.5 min, hybridization at 56 °C for 1.5 min, and elongation at 72 °C for 1.5 min. A final extension at 72 °C for 5 min was also performed.Table IOligonucleotides used for fragment amplificationOligonucleotideNo.Sequence70 PF15′-ACGGGTACCATGAAGGTAAAGAACACAAT70 PR25′-TGAGGATCCCGCCGGAGGCATTAGCACGC83 PF35′-ACGGGTACCATGATCAACGTTCAGGCCAA83 PR45′-TGAGGATCCCTGTGCCGGGGGCATCAGCA28 aa-83 PR55′-TGAGGATCCGGTGCTCGAACAACCCGCTA35 aa-83 PR65′-TGAGGATCCGGTGTCTTGCGACACGGGTT56 aa-83 PR75′-TGAGGATCCTGCGGGGTCAGCCATTGCCG65 aa-83 PR85′-TGAGGATCCCCCACGACCAATCAGGTCCGTT(48,49)VV PF95′-GCCCGGCGGCGCCCGTTGTAGTAGCGGCAATGGCTGTT(48,49)VV PR105′-CAGCCATTGCCGCTACTACAACGGGCGCCGCCGGGCThe restriction endonuclease sites are underlined and the site-directed mutations are shown in bold. Open table in a new tab The restriction endonuclease sites are underlined and the site-directed mutations are shown in bold. mpb70 andmpb83 were amplified from the cosmid pA3, described previously (18Hewinson R.G. Michell S.L. Russell W.P. McAdam R.A. Jacobs W.J. Scand. J. Immunol. 1996; 43: 490-499Crossref PubMed Scopus (109) Google Scholar), using the oligonucleotide pairs 1,2 and 3,4, respectively (Table I). Amplified fragments were purified using Wizard PCR Prep columns (Promega) and digested with the restriction endonucleases BamHI and KpnI (Promega). Digested fragments were cloned into BamHI/KpnI-digested p19pro-PhoA (Fig. 1), a modified version of pSMT3(19-PhoA) (17Herrmann J.L. O'Gaora P. Gallagher A. Thole J.E. Young D.B. EMBO J. 1996; 15: 3547-3554Crossref PubMed Scopus (137) Google Scholar) containing a leaderless E. coli alkaline phosphatase (phoA) gene, using standard procedures (Sambrook et al. (23Sambrook J. Fitsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989Google Scholar)). p19proSP-PhoA contains the promoter and signal peptide of the 19 kDa of M. tuberculosis in-frame with the PhoA gene in 19pro-PhoA and is used as a positive control for anti-alkaline phosphatase antibody activity and as a negative control for ConA binding. Truncated versions of mpb83 were amplified using oligonucleotide 83PF with the corresponding reverse complementary oligonucleotide and cloned into p19pro-PhoA as described above. Recombinants with functional alkaline phosphatase activity were selected after transformation into competent E. coli DH5α (Invitrogen), by the presence of blue colonies when grown onl-agar supplemented with 40 μg/ml 5-bromo-4-chloro-3-indolyl-phosphate-p-toluidine. Plasmid DNA was isolated (Qiagen) and subsequently transformed into M. smegmatis mc2155 according to the method of Snapperet al. (24Snapper S.B. Melton R.E. Mustafa S. Kieser T. Jacobs Jr., W.R. Mol. Microbiol. 1990; 4: 1911-1919Crossref PubMed Scopus (1004) Google Scholar). Complementary forward and reverse oligonucleotide primers 9 and 10 (Table I) were used in conjunction with the reverse complementary and forward oligonucleotide primers, 4 and 3 (Table I), for mpb83, respectively. Following purification, the two products, having an overlap of 30 bp, were combined and amplified by PCR using oligonucleotide primers 3 and 8 to give a fragment of mpb83 incorporating the desired mutation. All cloned DNA fragments were DNA sequenced using the TaqFS dideoxy terminator cycle sequencing kit in conjunction with a 373A Automated DNA Sequencer (Applied Biosystems Inc.). Each nucleotide was sequenced a minimum of three times from both strands. DNA sequence data analysis and protein sequence comparison were performed using the DNASTAR software package (DNASTAR Inc.). For SDS-PAGE, recombinantM. smegmatis was grown in 40 ml of MADC-Tween 80 at 37 °C, 225 rpm, for 48 h. The mycobacteria were pelleted by centrifugation (4,000 rpm for 10 min, Sigma 3K10, rotor 11133), washed twice with 40 ml of PBS (pH 7.0), and resuspended in 3 ml of PBS (pH 7.0). The resulting suspensions were sonicated on ice 10 times for 60 s with 90-s intervals, using a Soniprep 150 (MSE Ltd) equipped with a 1-cm probe. Sonicated extracts were centrifuged at 14,000 rpm for 30 min at 4 °C, the supernatant was isolated and filtered through a 0.22-μm filter. Total protein concentrations of sonicated extracts were determined using the bicinchoninic acid (BCA) protein assay (Pierce and Warriner). 10 μg of extracts were solubilized by heating at 100 °C for 3 min in an equal volume of sample loading buffer {5 mm Tris-HCl, pH 6.8, 5% (v/v) 2-mercaptoethanol, 2% (w/v) sodium dodecyl sulfate, 10% (v/v) glycerol, 0.002% (w/v) bromphenol blue). Samples were fractionated by electrophoresis through 12.5% acrylamide (w/v) SDS-polyacrylamide gels using a discontinuous Tris-HCl buffer system as described by Laemmli (25Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (206615) Google Scholar) and transferred to nitrocellulose by electroblotting as described previously (26Matsudaira P. J. Biol. Chem. 1987; 262: 10035-10038Abstract Full Text PDF PubMed Google Scholar). Rainbow protein markers (Amersham Biosciences) were run as molecular mass standards. ConA binding was determined by first blocking the nitrocellulose membranes with 20 ml of 5% bovine serum albumin (Sigma) in PBS with 0.1% Tween 20 (Bio-Rad) for 1 h at room temperature. After washing the membranes three times for 10 min with PBS containing 0.5% Tween 20, they were incubated with 20 ml of peroxidase-conjugated ConA at 1 mg/ml in PBS, 0.5% Tween 20 without bovine serum albumin for 1 h at room temperature. The membranes were then washed twice for 10 min with PBS containing 0.5% Tween 20 and finally for 10 min with PBS only. The substrate solution for visualizing peroxidase activity was prepared by dissolving 30 mg of 4-chloronaphthol (Sigma) in 12 ml of methanol followed by addition of 96 ml of PBS and 17 μl of 30% (v/v) hydrogen peroxide. For detecting PhoA protein expression or MPB83, the membranes were blocked with 0.01 m Tris-buffered saline (TBS) containing 3% (w/v) skimmed milk powder (TBSM), and then incubated for 1 h at room temperature. Following three 10-min washes with TBS containing 0.05% (v/v) Tween 20 (TBST) the nitrocellulose membranes were incubated for 2 h at 37 °C with monoclonal antibody raised against bacterial alkaline phosphatase (Sigma) or MBS43 (VLA Weybridge), both at a dilution of 1 in 8000 in 30 ml of TBST containing 3% (w/v) skimmed milk (TBSTM). Membranes were then washed in TBST as described previously and incubated for 2 h at 37 °C with alkaline phosphatase-conjugated goat anti-mouse IgG (Sigma) diluted 1 in 8000 in TBSTM. After three further 10-min washes at room temperature with TBS bound alkaline phosphatase-conjugated antibody was detected with SigmaFast tablets (Sigma) dissolved in 15 ml of water. Native MPB83 was purified from a M. bovis culture supernatant. M. bovis AN5 was grown as a surface pellicle on Bureau of Animal Industry medium (14 g of l-asparagine, 1.5 g of dipotassium hydrogen phosphate, 0.74 g of sodium citrate, 1.5 g of magnesium sulfate, 0.3 g of ferric sulfate, 0.08 g of zinc sulfate, 0.008 g of manganese chloride, 0.00138 g of cobaltous chloride, 10 g of glucose, 100 g of glycerol/liter) for 10 weeks at 37 °C. The culture was filtered through a stainless steel mesh strainer and then passed through a 0.2-μm VacuCap bottle top filter (Gelman Sciences). The culture supernatant was filtered a second time though a 0.2-μm VacuCap bottle top filter to ensure sterility. The culture supernatant was diluted 3-fold in preparative anion exchange chromatography (AEC) loading buffer (20 mm Tris, pH 8, 0.05% (v/v) Igepal CA-630) and applied to a XK50/20 column (Amersham Biosciences) containing 200 ml of Fast Flow DEAE anion exchange medium (Amersham Biosciences) using a fast protein liquid chromatography instrument. Adsorbed protein was eluted from the column by applying a linear increasing gradient of 0–250 mmsodium chloride over a 1-liter volume. Fractions containing MPB83 were identified by separation on SDS-PAGE gels, transfer to nitrocellulose, and probing with the MPB83-specific antibody MBS43 using a Western blotting procedure. The purity of each MPB83 containing fraction was assessed by total protein analysis of SDS-PAGE gels using a silver stain procedure. Fractions containing MPB83 were pooled and concentrated using a 50-ml stirred cell concentrator containing a YM10 membrane (10,000 molecular weight cut-off) (Millipore UK). The enriched MPB83 containing material was further purified by high resolution AEC using a MonoQ HR10/10 column. The sample was diluted 5-fold in AEC loading buffer (20 mm Tris, pH 8) and applied to a MonoQ HR10/10 column using a fast protein liquid chromatography instrument. The adsorbed material was eluted from the column by applying a linear increasing gradient of 0–250 mm sodium chloride over a 100-ml volume. Fractions containing MPB83 were identified as described above. The pooled fractions were concentrated by ultrafiltration using a CentriPrep 10 (10,000 molecular weight cut-off) concentrator unit (Millipore UK) centrifuged at 2500 ×g (MSE Mistrial 2000 centrifuge, Sanyo Gallenkamp). The AEC purified material was diluted 5-fold in hydrophobic interaction chromatography loading buffer (20 mm sodium phosphate, pH 7, 1 m ammonium sulfate) and applied to a HR5/5 phenyl-Superose hydrophobic interaction chromatography column (Amersham Biosciences) using a fast protein liquid chromatography instrument. Adsorbed material was eluted by applying a linear, decreasing gradient, of 1–0 m ammonium sulfate over a volume of 80 ml. Fractions containing MPB83 were identified as described above. The pooled fractions were concentrated by ultrafiltration using Centricon 10 (10,000 M r cut-off) concentrator units (Millipore UK) centrifuged at 4000 × g (MSE Mistral 2000 centrifuge, Sanyo Gallenkamp). The concentrated material was dialyzed against 4× 2 liters of PBS using a 3,500M r cut-off dialysis cassette (Perbio Science). Dialysis was performed at 4 °C and a minimum of 2 h was allowed between each change of dialysis buffer. The dialyzed material was filtered through a 0.2-μm syringe filter and the protein concentration was determined using the BCA protein estimation assay. Purified native MPB83 was subjected to a chemical deglycosylation procedure using a GlycoFree deglycosylation kit (Glyko Inc.). This method employs the use of anhydrous trifluoromethanesulfonic acid, which cleavesN- and O-linked glycans nonselectively from glycoproteins while leaving the primary structure of the protein intact (27Sojar H.T. Bahl O.P. Arch. Biochem. Biophys. 1987; 259: 52-57Crossref PubMed Scopus (79) Google Scholar). Purified native MPB83 was initially desalted by dialysis against 4× 2-liter volumes of water using a 3,500 M rcut-off dialysis cassette (Perbio Science). The desalted sample was then frozen at −80 °C and lyophilized overnight in a model EF03 freeze dryer (Edwards High Vacuum International). The deglycosylation procedure was then performed in accordance with the GlycoFree kit recommendations. Briefly, a reaction vial, containing 0.5 mg of lyophilized material, was placed in a dry ice/ethanol bath and 50 μl of trifluoromethanesulfonic acid was slowly added. The vial was then incubated at −20 °C for 4 h and then returned to a dry ice/ethanol bath, to which 150 μl of pyridine was slowly added. The vial was then transferred to dry ice for a further 5 min and then to wet ice for a further 15 min. The reaction mixture was neutralized by addition of 400 μl of 0.5% (w/v) ammonium bicarbonate. The deglycosylated protein was then isolated from the reaction products and reagents by dialysis against 4× 2 liters of PBS using a 3,500M r cut-off dialysis cassette (Perbio Science). The protein concentration of the dialyzed material was determined using the BCA protein estimation assay. Intact MPB83 and products of tryptic digestion were analyzed by ES-MS using a nanospray ion source on a hybrid quadrupole orthogonal acceleration time of flight (Q-TOF) mass spectrometer (Micromass UK) (28Morris H.R. Paxton T. Dell A. Langhorne J. Berg M. Bordoli R.S. Hoyes J. Bateman R.H. Rapid Commun. Mass Spectrom. 1996; 10: 889-896Crossref PubMed Scopus (390) Google Scholar, 29Morris H.R. Panico M. Dell A. McDowell R.A. Larsen B.S. McEwen C.N. 2nd Ed. Mass Spectrometry of Biological Materials. Marcel Dekker, New York1998: 53-80Google Scholar). Signals attributable to glycopeptides were passed separately for collisionally activated decomposition tandem MS experiments (CAD MS/MS) into a collision cell filled with argon gas, using collision energies from 10 to 40 eV as appropriate (28Morris H.R. Paxton T. Dell A. Langhorne J. Berg M. Bordoli R.S. Hoyes J. Bateman R.H. Rapid Commun. Mass Spectrom. 1996; 10: 889-896Crossref PubMed Scopus (390) Google Scholar, 29Morris H.R. Panico M. Dell A. McDowell R.A. Larsen B.S. McEwen C.N. 2nd Ed. Mass Spectrometry of Biological Materials. Marcel Dekker, New York1998: 53-80Google Scholar). The modifying sugars of purified MPB83 were identified by first hydrolyzing the protein (50 μg) with 4m trifluoroacetic acid at 95 °C for 4 h to release the sugars. The hydrolysis reaction was dried under vacuum at 40 °C in a SCV100H SpeedVac apparatus (Stratech Scientific) and then resuspended in 100 μl of water. A sample of 10 μl was then analyzed on a CarboPak PA1 anion exchange column (Dionex Corp.) using a 625LC high performance liquid chromatography instrument (Waters) equipped with a pulsed amperometric detector (Dionex Corp.). Sugars were eluted from the column with 16 mm sodium hydroxide at a flow rate of 1 ml/min. Sugars were identified by comparing the column retention time with that of sugar standards prepared at known concentrations run under identical conditions (all sugar standards were supplied by Sigma). Reductive elimination, permethylation, preparation of partially methylated alditol acetates, and acquisition of FAB-MS and GC-MS data were performed as described previously (30Dell A. Reason A.J. Khoo K.H. Panico M. McDowell R.A. Morris H.R. Methods Enzymol. 1994; 230: 108-132Crossref PubMed Scopus (236) Google Scholar, 31Khoo K.H. Sarda S. Xu X. Caulfield J.P. McNeil M.R. Homans S.W. Morris H.R. Dell A. J. Biol. Chem. 1995; 270: 17114-17123Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar). A recombinant mycobacterial expression system (17Herrmann J.L. O'Gaora P. Gallagher A. Thole J.E. Young D.B. EMBO J. 1996; 15: 3547-3554Crossref PubMed Scopus (137) Google Scholar, 32Herrmann J.L. Delahay R. Gallagher A. Robertson B. Young D. FEBS Lett. 2000; 473: 358-362Crossref PubMed Scopus (48) Google Scholar) was used to determine whether MPB70 and MPB83 were glycosylated by mycobacteria. The genes encoding these proteins were cloned in-frame with E. coli alkaline phosphatase (PhoA) lacking its own signal sequence in the mycobacterial shuttle vector 19pro-PhoA (Fig. 1) as described under "Experimental Procedures." PhoA was used as a hybrid partner to allow the level of expression of the different constructs to be monitored and to verify that the fusion proteins were being exported. The fusion proteins encoded by these constructs expressed functionally active alkaline phosphatase in both E. coli DH5α andM. smegmatis mc2155, as determined by the formation of blue colonies in the presence of 5-bromo-4-chloro-3-indolyl-phosphate-p-toluidine. Glycosylation of the fusion proteins was monitored by their ability to bind ConA following fractionation of recombinant cell extracts by denaturing SDS-PAGE and electrotransfer to nitrocellulose membranes. In these experiments only MPB83-PhoA expressed by M. smegmatismc2155 was recognized by ConA (Fig.2 A, lane 5). ConA did not bind MPB83-PhoA expressed by E. coli DH5α (Fig.2 A, lane 4), demonstrating that post-translational modification of MPB83 is restricted to the mycobacterial host. Neither the MPB70-PhoA fusion protein nor the PhoA protein alone expressed by M. smegmatis mc2155 bound ConA (Fig. 2 A, lanes 3 and 6, respectively). Probing a duplicate blot with antibody directed against PhoA (Fig. 2 B) demonstrated a similar amount of hybrid protein in each of the extracts, indicating that differences in ConA binding were not a consequence of differences in expression levels of the recombinant fusion proteins. There was evidence of partial proteolysis of fusion proteins expressed in M. smegmatismc2155 as two bands reacting with anti-PhoA antibody were observed. Immunoblotting with antibodies against PhoA also revealed an increase in the apparent molecular weight of MPB83-PhoA expressed by M. smegmatis mc2155 compared with that of the protein expressed by E. coli DH5α (Fig. 2 B, lanes 4 and 5). In contrast, no size difference was observed between MPB70-PhoA expressed by M. smegmatismc2155 or E. coli DH5α (Fig. 2 B,lanes 2 and 3). The difference in apparent molecular weights of fusion proteins expressed by different bacterial hosts, coupled with the ConA recognition of MPB83-PhoA expressed byM. smegmatis mc2155, supports the conclusion that MPB83 but not MPB70 is glycosylated by mycobacteria. To identify regions of MPB83 involved in ConA binding, in-frame hybrid proteins were produced in which increasing amino-terminal regions of MPB83 were fused to PhoA as described under "Experimental Procedures." This approach was used to generate the set of in-frame fusion proteins listed in TableII.Table IIMPB83 amino terminus PhoA fusionsMPB83 amino-terminal PhoA fusionsConstructRecognition by ConA120MINVQAKPAAAASLAAIAIAFLAGCSST—PhoAp28aaPhoA−ve351—20 FLAGCSSTKPVSQDT—PhoAp35aaPhoA−ve561—20 FLAGCSSTKPVSQDTSPKPATSPAAPVTTAAMADPA—PhoAp56aaPhoA+ve631—20 FLAGCSSTKPVSQDTSPKPATSPAAPVTTAAMADPAADLIGRG—PhoAp63aaPhoA+ve Open table in a new tab A
Referência(s)