Portal de Periódicos da CAPES

Chemokine-Glycosaminoglycan Binding

2005; Elsevier BV; Volume: 280; Issue: 37 Linguagem: Inglês

10.1074/jbc.m505738200

1083-351X

Yonghao Yu, Matthew D. Sweeney, Ola M. Saad, Susan E. Crown, Tracy M. Handel, Julie A. Leary,

Cell Adhesion Molecules Research

Glycosaminoglycans (GAGs) have recently been demonstrated to be required for the in vivo activity of several chemokines. Minimally, the interaction is thought to provide a mechanism for retention at the site of secretion and the formation of chemokine gradients that provide directional cues for receptor bearing cells, particularly in the presence of shear forces. Thus, a key issue will be to determine the sequence and structure of the GAGs that bind to specific chemokines. Herein, we describe a mass spectrometry assay that was developed to detect protein-oligosaccharide noncovalent complexes, in this case chemokine-GAG interactions, and to select for high affinity GAGs. The process is facilitated by the ability of electrospray ionization to transfer the intact noncovalent complexes from solution into the gas phase. The elemental composition as well as the binding stoichiometry can be calculated from the mass of the complex. Ligands of the chemokine receptor, CCR2 (MCP-1/CCL2, MCP-2/CCL8, MCP-3/CCL7, MCP-4/CCL13, and Eotaxin/CCL11), and the CCR10 ligand CTACK/CCL27 were screened against a small, highly sulfated, heparin oligosaccharide library with limited structural variation. The results revealed heparin octasaccharides with 11 and 12 sulfates as binders. Oligomerization of some chemokines was observed upon GAG binding, whereas in other instances only the monomeric noncovalent complex was identified. The results indicate that, in contrast to the apparent redundancy in the chemokine system, where several chemokines bind and activate the same receptor, these chemokines could be differentiated into two groups based on the stoichiometry of their complexes with the heparin oligosaccharides. Glycosaminoglycans (GAGs) have recently been demonstrated to be required for the in vivo activity of several chemokines. Minimally, the interaction is thought to provide a mechanism for retention at the site of secretion and the formation of chemokine gradients that provide directional cues for receptor bearing cells, particularly in the presence of shear forces. Thus, a key issue will be to determine the sequence and structure of the GAGs that bind to specific chemokines. Herein, we describe a mass spectrometry assay that was developed to detect protein-oligosaccharide noncovalent complexes, in this case chemokine-GAG interactions, and to select for high affinity GAGs. The process is facilitated by the ability of electrospray ionization to transfer the intact noncovalent complexes from solution into the gas phase. The elemental composition as well as the binding stoichiometry can be calculated from the mass of the complex. Ligands of the chemokine receptor, CCR2 (MCP-1/CCL2, MCP-2/CCL8, MCP-3/CCL7, MCP-4/CCL13, and Eotaxin/CCL11), and the CCR10 ligand CTACK/CCL27 were screened against a small, highly sulfated, heparin oligosaccharide library with limited structural variation. The results revealed heparin octasaccharides with 11 and 12 sulfates as binders. Oligomerization of some chemokines was observed upon GAG binding, whereas in other instances only the monomeric noncovalent complex was identified. The results indicate that, in contrast to the apparent redundancy in the chemokine system, where several chemokines bind and activate the same receptor, these chemokines could be differentiated into two groups based on the stoichiometry of their complexes with the heparin oligosaccharides. Chemokines are small secreted proteins that are critically involved in many biological processes, including routine immunosurveillance, inflammation, and development. Some specific chemokine receptors provide the portals by which human immunodeficiency virus gets into cells, whereas others have been implicated in inflammatory diseases associated with inappropriate cell migration (rheumatoid arthritis, multiple sclerosis, atherosclerosis, and most recently, tumor metastasis) (1Rossi D. Zlotnik A. Annu. Rev. Immunol. 2000; 18: 217-243Crossref PubMed Scopus (2090) Google Scholar, 2Baggiolini M. Dewald B. Moser B. Annu. Rev. Immunol. 1997; 15: 675-705Crossref PubMed Scopus (1976) Google Scholar). Thus there is considerable interest in understanding how chemokines function, and in identifying strategies that interfere with their function.Based on the pattern of the cysteine residues, chemokines can be divided into CXC, CC, C, and CX3C families. Once secreted, chemokines are thought to form a concentration gradient that controls the direction of the leukocyte cell migration. This process is mediated by the interactions between the chemokines and G protein-coupled seven transmembrane receptors on leukocytes; the binding event subsequently triggers downstream signaling pathways that lead to cell migration and activation (3Proudfoot A.E.I. Nat. Rev. Immunol. 2002; 2: 106-115Crossref PubMed Scopus (600) Google Scholar). Approximately 45 chemokines and 19 chemokine receptors have been identified to date. Characterization of which chemokines interact with which receptors has revealed significant cross-reactivity (e.g. redundancy) in vitro; in other words multiple chemokines bind and activate the same receptor, and a given chemokine can bind multiple receptors. Nevertheless this apparent redundancy may not exist in vivo; for example, significant effects in models of inflammation have been observed using knock-out mice, or antibodies and small molecule antagonists, despite the fact that they target single receptors or chemokines (4Power C.A. J. Immunol. Methods. 2003; 273: 73-82Crossref PubMed Scopus (55) Google Scholar, 5Proudfoot A.E.I. Power C.A. Rommel C. Wells T.N.C. Semin. Immunol. 2003; 15: 57-65Crossref PubMed Scopus (91) Google Scholar). An emerging hypothesis is that glycosaminoglycans (GAGs) 2The abbreviations used are: GAG, glycosaminoglycan; AUC, analytical ultracentrifugation; HPLC, high performance liquid chromatography; FTICR, Fourier transform ion cyclotron resonance; ESI, electrospray ionization; SPE, solid-phase extraction; CCR, CC chemokine receptor; WT, wild type; Ni-NTA, nickel-nitrilotriacetic acid; MCP-1, monocyte chemoattractant protein-1; MS, mass spectrometry.2The abbreviations used are: GAG, glycosaminoglycan; AUC, analytical ultracentrifugation; HPLC, high performance liquid chromatography; FTICR, Fourier transform ion cyclotron resonance; ESI, electrospray ionization; SPE, solid-phase extraction; CCR, CC chemokine receptor; WT, wild type; Ni-NTA, nickel-nitrilotriacetic acid; MCP-1, monocyte chemoattractant protein-1; MS, mass spectrometry. play a role in the in vivo function of chemokines (6Kuschert G.S.V. Coulin F. Power C.A. Proudfoot A.E.I. Hubbard R.E. Hoogewerf A.J. Wells T.N.C. Biochemistry. 1999; 38: 12959-12968Crossref PubMed Scopus (488) Google Scholar, 7Hoogewerf A.J. Kuschert G.S.V. Proudfoot A.E.I. Borlat F. Clark-Lewis I. Power C.A. Wells T.N.C. Biochemistry. 1997; 36: 13570-13578Crossref PubMed Scopus (436) Google Scholar, 8Lever R. Page C.R. Nat. Rev. Drug Discovery. 2002; 1: 140-148Crossref PubMed Scopus (297) Google Scholar, 9Esko J.D. Selleck S.B. Annu. Rev. Biochem. 2002; 71: 435-471Crossref PubMed Scopus (1226) Google Scholar) and that specificity in these interactions could alter the apparent redundancy (10Handel T.M. Johnson Z. Crown S.E. Lau E.K. Sweeney M. Proudfoot A.E. Annu. Rev. Biochem. 2005; 74: 385-410Crossref PubMed Scopus (422) Google Scholar).GAGs are linear, highly sulfated, and heterogeneous polysaccharides that are often covalently linked to core proteins residing on the membrane of cells or within the extracellular matrix. They have been classified into several major families, primarily heparin/heparan sulfate, chondroitin sulfate/dermatan sulfate, hyaluronan, and keratan sulfate (11Lindahl U. Kusche-Gullberg M. Kjellen L. J. Biol. Chem. 1998; 273: 24979-24982Abstract Full Text Full Text PDF PubMed Scopus (569) Google Scholar, 12Linhardt R.J. Toida T. Acc. Chem. Res. 2004; 37: 431-438Crossref PubMed Scopus (233) Google Scholar, 13Bernfield M. Gotte M. Park P.W. Reizes O. Fitzgerald M.L. Lincecum J. Zako M. Annu. Rev. Biochem. 1999; 68: 729-777Crossref PubMed Scopus (2293) Google Scholar). Although they are ubiquitously expressed on virtually all animal cells, the exact composition of the GAG coat of a given cell/extracellular matrix depends on the type and location of the cell and the pathophysiological state of the tissue/organism (14Kato M. Wang H.M. Bernfield M. Gallagher J.T. Turnbull J.E. J. Biol. Chem. 1994; 269: 18881-18890Abstract Full Text PDF PubMed Google Scholar). GAGs are thought to influence the local cytokine network through interactions with proteins and by modulating protein-protein interactions (15Delehedde M. Lyon M. Gallagher J.T. Rudland P.S. Fernig D.G. Biochem. J. 2002; 366: 235-244Crossref PubMed Google Scholar), and there is accumulating evidence for specificity in these interactions. For example, the sulfation pattern on GAGs is thought to be a major determinant in their interaction with proteins, which are often highly basic (16Bourin M.C. Lindahl U. Biochem. J. 1993; 289: 313-330Crossref PubMed Scopus (392) Google Scholar, 17Tully S.E. Mabon R. Gama C.I. Tsai S.M. Liu X.W. Hsieh-Wilson L.C. J. Am. Chem. Soc. 2004; 126: 7736-7737Crossref PubMed Scopus (138) Google Scholar). Chemokines have been shown to bind GAGs both in vitro and in vivo, and recently it was demonstrated that this interaction is required for their function in vivo (19Proudfoot A.E.I. Handel T.M. Johnson Z. Lau E.K. LiWang P. Clark-Lewis I. Borlat F. Wells T.N.C. Kosco-Vilbois M.H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1885-1890Crossref PubMed Scopus (631) Google Scholar, 20Johnson Z. Kosco-Vilbois M.H. Herren S. Cirillo R. Muzio V. Zaratin P. Carbonatto M. Mack M. Smailbegovic A. Rose M. Lever R. Page C. Wells T.N.C. Proudfoot A.E.I. J. Immunol. 2004; 173: 5776-5785Crossref PubMed Scopus (135) Google Scholar). Immobilization of chemokines on GAGs is thought to enhance their local concentration and facilitate the formation of chemokine gradients to guide the migration of cells, especially under flow conditions (19Proudfoot A.E.I. Handel T.M. Johnson Z. Lau E.K. LiWang P. Clark-Lewis I. Borlat F. Wells T.N.C. Kosco-Vilbois M.H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1885-1890Crossref PubMed Scopus (631) Google Scholar, 20Johnson Z. Kosco-Vilbois M.H. Herren S. Cirillo R. Muzio V. Zaratin P. Carbonatto M. Mack M. Smailbegovic A. Rose M. Lever R. Page C. Wells T.N.C. Proudfoot A.E.I. J. Immunol. 2004; 173: 5776-5785Crossref PubMed Scopus (135) Google Scholar). GAG binding may also help recruit chemokines to specific populations of cells, thereby contributing to the control of cell migration in a receptor independent way.Oligomerization has also been shown to be required for some chemokines as oligomerization-deficient mutants that function like wild type (WT) in vitro, have been shown to be nonfunctional in vivo (19Proudfoot A.E.I. Handel T.M. Johnson Z. Lau E.K. LiWang P. Clark-Lewis I. Borlat F. Wells T.N.C. Kosco-Vilbois M.H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1885-1890Crossref PubMed Scopus (631) Google Scholar). Furthermore, this requirement seems to be coupled to GAG binding, because GAGs can induce some chemokines to oligomerize (19Proudfoot A.E.I. Handel T.M. Johnson Z. Lau E.K. LiWang P. Clark-Lewis I. Borlat F. Wells T.N.C. Kosco-Vilbois M.H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1885-1890Crossref PubMed Scopus (631) Google Scholar, 21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar). Thus, there are several sources where specificity in chemokine-GAG interactions may be derived: (i) the actual GAG-binding epitopes on the chemokines, and indeed, GAG-binding hotspots have been defined for several chemokines and show significant differences in their patterns (21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar), (ii) the oligomerization state that the chemokines adopt, and again differences have been reported (10Handel T.M. Johnson Z. Crown S.E. Lau E.K. Sweeney M. Proudfoot A.E. Annu. Rev. Biochem. 2005; 74: 385-410Crossref PubMed Scopus (422) Google Scholar, 19Proudfoot A.E.I. Handel T.M. Johnson Z. Lau E.K. LiWang P. Clark-Lewis I. Borlat F. Wells T.N.C. Kosco-Vilbois M.H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1885-1890Crossref PubMed Scopus (631) Google Scholar, 21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar), and (iii) the sequence of the GAGs, for which much less is known due to the difficulties in isolating and sequencing GAGs. If the argument for specificity is accurate, then the GAG interaction may significantly contribute to the localization of cells beyond that defined at the level of the chemokine-receptor interaction. The GAG interaction may, in fact, specify the use of one chemokine over another. Targeting chemokine-GAG interactions with tight binding GAG sequences or GAG mimetics that either block the interaction or compete for chemokine-receptor binding in vivo, could also form the basis of novel therapeutics. Along these lines, the objectives of this study are to develop methods for selecting and sequencing GAG ligands of a given chemokine, to determine the extent of their sequence diversity, and to establish if the "redundant" chemokines can be distinguished by their GAG-binding properties.The characterization of chemokine-GAG interactions has been greatly facilitated by the development of reasonably efficient screening and detection techniques (22Powell A.K. Yates E.A. Fernig D.G. Turnbull J.E. Glycobiology. 2004; 14: 17R-30RCrossref PubMed Scopus (223) Google Scholar, 23Blixt O. Head S. Mondala T. Scanlan C. Huflejt M.E. Alvarez R. Bryan M.C. Fazio F. Calarese D. Stevens J. Razi N. Stevens D.J. Skehel J.J. van Die I. Burton D.R. Wilson I.A. Cummings R. Bovin N. Wong C.H. Paulson J.C. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17033-17038Crossref PubMed Scopus (960) Google Scholar, 24Fukui S. Feizi T. Galustian C. Lawson A.M. Chai W.G. Nat. Biotechnol. 2002; 20: 1011-1017Crossref PubMed Scopus (525) Google Scholar, 25Feizi T. Chai W.G. Nat. Rev. Mol. Cell Biol. 2004; 5: 582-588Crossref PubMed Scopus (214) Google Scholar, 26Keiser N. Venkataraman G. Shriver Z. Sasisekharan R. Nat. Med. 2001; 7: 123-128Crossref PubMed Scopus (53) Google Scholar). However, there are still many limitations, and the resulting data have been somewhat nebulous. Affinity chromatography methods are slow and require considerable amounts of material (22Powell A.K. Yates E.A. Fernig D.G. Turnbull J.E. Glycobiology. 2004; 14: 17R-30RCrossref PubMed Scopus (223) Google Scholar), whereas oligosaccharide microarrays generally need pure and structurally defined oligosaccharides that are chemically synthesized or isolated from natural sources (25Feizi T. Chai W.G. Nat. Rev. Mol. Cell Biol. 2004; 5: 582-588Crossref PubMed Scopus (214) Google Scholar). In this report, a mass spectrometry approach was used to identify specific GAGs that bind to the chemokine ligands of CCR2. The assay is a label-free methodology that does not require chemical modification or immobilization of the protein target or the library components. The binding stoichiometry as well as the molecular formula of the oligosaccharide binders can be calculated from the masses of the protein/oligosaccharide noncovalent complex using electrospray ionization-Fourier transform ion cyclotron resonance (ESI-FTICR) mass spectrometry. This is possible given the fact that many structural features, including stoichiometry, ligand binding domain, and even tertiary geometry of solution noncovalent complexes can be preserved during the transition from solution to gas phase (27Sobott F. Robinson C.V. Curr. Opin. Struct. Biol. 2002; 12: 729-734Crossref PubMed Scopus (128) Google Scholar, 28Loo J.A. Mass Spectrom. Rev. 1997; 16: 1-23Crossref PubMed Scopus (1154) Google Scholar, 29Smith R.D. Bruce J.E. Wu Q.Y. Lei Q.P. Chem. Soc. Rev. 1997; 26: 191-202Crossref Google Scholar, 30Yu Y. Kirkup C.E. Pi N. Leary J.A. J. Am. Soc. Mass Spectrom. 2004; 15: 1400-1407Crossref PubMed Scopus (34) Google Scholar, 31Harmer N.J. Ilag L.L. Mulloy B. Pellegrini L. Robinson C.V. Blundell T.I. J. Mol. Biol. 2004; 339: 821-834Crossref PubMed Scopus (92) Google Scholar). Confirmation of the complexed oligosaccharides detected in the noncovalent complex was verified using hydrophobic trapping and elution. Identified oligosaccharides binders were spatially separated from the protein, collected, and analyzed by mass spectrometry. Specific GAG ligands were unambiguously determined by measuring the molecular mass of each ligand from a library of possible binders.EXPERIMENTAL PROCEDURESMaterials—The heparin octasaccharide library was obtained from Dextra Laboratories (Reading, UK) and consists of a partial heparinase I digestion of heparin purified from porcine intestinal mucosa. The catalogue number for this library is ID1008. Sucrose octasulfate was purchased from Toronto Research Chemicals (Toronto, Canada). Thrombin and cyclodextrin sulfate were purchased from Sigma. Enterokinase was expressed and purified as described previously (32Lu D.S. Futterer K. Korolev S. Zheng X.L. Tan K. Waksman G. Sadler J.E. J. Mol. Biol. 1999; 292: 361-373Crossref PubMed Scopus (91) Google Scholar). All polymerases and restriction endonucleases were purchased from New England Biolabs (Beverly, MA).Vector Construction, Expression, and Purification of Chemokines—Genes for the human chemokines MCP-3/CCL7, MCP-2/CCL8, Eotaxin/CCL11, and MCP-4/CCL13 were codon optimized for expression in Escherichia coli as described (33Kane J.F. Curr. Opin. Biotechnol. 1995; 6: 494-500Crossref PubMed Scopus (594) Google Scholar). Each gene was then cloned into modified versions of pET21, which include one or all of the following: a leader sequence for increased expression, a His tag for purification, and an enterokinase or thrombin protease site for removal of the tag (TABLE ONE).TABLE ONEExpression conditions of the chemokinesProteinN-terminal sequence addedProteaseRatio (w:w)Cleavage timeCCL7MH8D4KEnterokinase1:25048 h at room temperatureCCL8MK3D4KEnterokinase1:10048 h at room temperatureCCL11MGSSH6ENLYLVPRThrombin1:100072 h at room temperatureCCL13MK3H8Aminopeptidase1:50120 h at room temperature Open table in a new tab MCP-1/CCL2 WT and CCL2 R18A/K19A/R24A were expressed and purified as described before (21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar). Both of these proteins have an additional mutation (M64I) to avoid oxidation during purification but show no effect on receptor-binding function (34Hemmerich S. Paavola C. Bloom A. Bhakta S. Freedman R. Grunberger D. Krstenansky J. Lee S. McCarley D. Mulkins M. Wong B. Pease J. Mizoue L. Mirzadegan T. Polsky I. Thompson K. Handel T.M. Jarnagin K. Biochemistry. 1999; 38: 13013-13025Crossref PubMed Scopus (141) Google Scholar). The CCL7 and CCL8 proteins were expressed as described (21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar).CCL8 was lysed and purified by SP ion exchange and reverse-phase HPLC as described (21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar). After lyophilization, the protein was resuspended in 20 mm Tris, pH 8, 50 mm NaCl, 2 mm CaCl2 at 1 mg/ml and mixed with a 1:100 w/w ratio of enterokinase. After 48 h, the protein was acidified by addition of trifluoroacetic acid to pH 2–4 and purified by reverse-phase HPLC. Eluted CCL8 was lyophilized and stored at –20 °C.CCL7-producing cells were lysed in 10 mm Tris, pH 8, 1 mm MgCl2 and clarified by centrifugation. The supernatant was applied to a Ni-NTA (Qiagen, Valencia, CA) column at room temperature. After washing with 20 mm imidazole, the protein was eluted with 10 mm Tris, pH 8, 300 mm NaCl, 250 mm imidazole. The eluate was then purified by reverse-phase HPLC, cleaved with a 1:250 w/w ratio of enterokinase for 48 h, and repurified as above.CCL11-producing cells were lysed in 20 mm Tris, pH 8, 400 mm NaCl, 20 mm imidazole and purified on a Ni-NTA column as above. Eluate was dialyzed against 20 mm Tris, pH 8, 300 mm NaCl, 2.5 mm CaCl2. Thrombin was added at a ratio of 1:1000 w/w. After 72 h, the protein was purified by reverse-phase HPLC and lyophilized as above.The CCL13 construct was transformed into BL21(DE3) pLysS cells that were then grown in LB at 37 °C to an A600 of 0.600. The cells were induced with 0.5 mm isopropyl 1-thio-β-d-galactopyranoside and harvested after 4 h by centrifugation. The cells were lysed in 10 mm Tris, pH 8, 1 mm MgCl2, and the insoluble protein was isolated by standard methods. The final pellet was resuspended in 6 m guanidine-HCl, 20 mm Tris, pH 8, 10 mm imidazole. The extract was clarified before application to a Ni-NTA column. The protein was eluted with 6 m guanidine-HCl, 10 mm Tris, pH 8, 250 mm imidazole. Refolding was initiated by dropwise dilution into 55 mm Tris, pH 8, 264 mm NaCl, 11 mm KCl, 0.055% polyethylene glycol 3350, 1.1 mm EDTA, 0.1 mm dithiothreitol to a final protein concentration of 0.1 mg/ml. Refolded protein was dialyzed against 20 mm Tris, pH 8, 200 mm NaCl, then against 35 mm Tris, pH 8. Aminopeptidase was added at a 1:50 w/w ratio. After 120 h, cleaved protein was purified by reverse-phase HPLC and lyophilized as above.Exact Mass Measurement of the Heparin Octasaccharide Library Components—The heparin octasaccharide library was analyzed by ESI-FTICR mass spectrometry using negative ion mode detection. Approximately 20 μm of the library was sprayed in a solvent containing 50:50 MeOH:H2O with 10 mm NH4OH. Addition of NH4OH or NH4OAc is known to prevent desulfation (35Saad O.M. Leary J.A. J. Am. Soc. Mass. Spectrom. 2004; 15: 1274-1286Crossref PubMed Scopus (97) Google Scholar). Samples were infused into the mass spectrometer at 1 μl/min using a syringe pump (Harvard Apparatus, Holliston, MA). For exact mass determination, the octasaccharide library sample was mixed with a HPLC peptide standard mixture (Sigma), which was used as an internal calibrant. This mixture was comprised of leucine enkephalin (1–, m/z 554.2615), angiotensin II (2–, m/z 521.7594), and Val-Tyr-Val (1–, m/z 378.2029). The identified library components are shown in TABLE TWO.TABLE TWOExact mass measurement and identification of the octasaccharide components in the librarySaccharideMtheoaMtheo, theoretical monoisotopic mass.(6— charge state)/calbThe m/z for the 6— charge state of the octasaccharides is shown: cal, calculated; obs, observed.(6— charge state)/obsΔm/zppmOcta+8SO31988.0581330.3357330.33631.82Octa+8SO3+Ac2030.0687337.3375337.33791.19Octa+9SO32068.0149343.6619343.66210.58Octa+9SO3+Ac2110.0255350.6630350.66361.71Octa+10SO32147.9717356.9880356.98800.09Octa+10SO3+Ac2189.9823363.9898363.99010.82Octa+11SO32227.9285370.3141370.31430.54Octa+12SO32307.8854383.6403383.64040.26a Mtheo, theoretical monoisotopic mass.b The m/z for the 6— charge state of the octasaccharides is shown: cal, calculated; obs, observed. Open table in a new tab Filtration Trapping Screening Assay—For screening experiments, 40 μm of the target protein was mixed with 200 μm octasaccharide library in 100 μl of 100 mm NH4OAc (pH 6.8) at room temperature. To remove the nonspecific/low affinity binders, the protein-oligosaccharide mixture was washed three times with 1 ml of 200 mm NH4OAc (pH 6.8), and subjected each time to centrifugal ultrafiltration (5000 rpm, 40 min, 4 °C) using a molecular mass cutoff filter of 10 kDa (Millipore, Billerica, MA). After the last wash step, the solution was concentrated to ∼50 μl. The retentate was diluted by addition of 1 volume of water and subjected to FTICR mass spectrometric analysis.Hydrophobic Trapping, Elution, and Confirmation of Bound Ligands—Forty micromolar of the target protein was incubated with 200 μm octasaccharide library in 100 μl of a 100 mm NH4OAc solution (pH 6.8). The solution was then applied to an Oasis solid-phase extraction (SPE) cartridge (Waters, Milford, MA) that was previously conditioned according to the manufacturer's instructions. The nonspecific binders were removed by flushing the column three times, each with 1 ml of a 200 mm NH4OAc solution. The high affinity ligands were eluted by using 1 ml of a 760 mm NH4OAc solution. The eluate was desalted by dialyzing against water in a Bio-dialyzer unit with a molecular mass cutoff of 1 kDa (The Nest Group, Southborough, MA) and analyzed by FTICR mass spectrometry using negative ion mode detection.FTICR Mass Spectrometry—Mass spectra were acquired on a Bruker APEX II 7-tesla FTICR mass spectrometer (Billerica, MA) equipped with an Apollo (Bruker, Billerica, MA) electrospray ionization (ESI) source. Generation of the noncovalent complex ions was performed as previously described (30Yu Y. Kirkup C.E. Pi N. Leary J.A. J. Am. Soc. Mass Spectrom. 2004; 15: 1400-1407Crossref PubMed Scopus (34) Google Scholar). Briefly, samples were infused into the mass spectrometer at 1 μl/min using a syringe pump (Harvard Apparatus, Holliston, MA). Ions were desolvated in the nozzle-skimmer region by using a 140-V (positive mode) or 40-V (negative mode) capillary exit voltage. Following desolvation, ions were externally accumulated in a radio frequency-only hexapole for 0.5–1 s and were transferred into the ICR cell for mass analysis. Two to four ion packets were trapped using gated trapping and detected after chirp excitation. Between 8 and 100 broadband time domain transients containing 512 k or 1024 k data points were averaged before zerofill, Gaussian multiplication and fast Fourier transform. The parameters of the ESI source, ion optics, and cell were tuned for the best ion intensity. The FTICR mass spectra were internally calibrated against myoglobin or MCP-1. All the data were acquired and processed using Xmass version 6.0.0 (Bruker).The average masses of the noncovalent complexes of MCP-1 dimer and oligosaccharide ligand were determined by using the method as previously reported by Zubarev et al. (36Zubarev R.A. Demirev P.A. Hakansson P. Sundqvist B.U.R. Anal. Chem. 1995; 67: 3793-3798Crossref Scopus (43) Google Scholar). Briefly, portions of the isotopic distribution above 50% relative intensity of the highest isotopic peak were used to calculate the average mass, Mave(P/L)′, of the protein-ligand complex. In accordance with this methodology, a correction factor (Δm50%) of 0.5 Da was then added to Mave(P/L)′ to derive the true Mave(P/L). The average mass of the protein, Mave(P), was calculated from the known elemental composition. The average mass of the oligosaccharide was then determined from Mave(L) = Mave(P/L)–Mave(P).RESULTSESI-FTICR Analysis of WT MCP-1—The data provided below show that specific GAG-chemokine noncovalent interactions can be accurately determined from gas phase data. Various preliminary experiments were required to prove that these complexes are accurate reflections of the solution complexes. Therefore, sample preparation and parameters affecting the ionization process were critically optimized.The ESI mass spectra of wild type (WT) MCP-1 are shown in Fig. 1. Fig. 1A represents WT MCP-1 sprayed from a solution of 79:20:1 acetonitrile:H2O:formic acid, in which the molecular mass was measured to be 8662.4615 Da. The theoretically most abundant mass of MCP-1, taking into account the fact that it and most other chemokines have two disulfide bonds, is 8662.4685 Da (1Rossi D. Zlotnik A. Annu. Rev. Immunol. 2000; 18: 217-243Crossref PubMed Scopus (2090) Google Scholar). This translates into a mass error of 0.8 ppm of the measured exact mass, clearly indicating the presence of two disulfide bonds.When MCP-1 was sprayed from a solution of 100 mm NH4OAc (pH 6.8), a solvent environment more conducive to the folded state of the protein, a tighter charge state distribution was observed, as shown in Fig. 1B. In accordance with previous ultracentrifugation studies, MCP-1 exists in equilibrium between the monomer and dimer forms (37Paolini J.F. Willard D. Consler T. Luther M. Krangel M.S. J. Immunol. 1994; 153: 2704-2717PubMed Google Scholar). The 4+, 5+, and 6+ charge states of the monomer ion and the 9+ charge state of the dimer ion were clearly observed in the mass spectra. The inset shows the overlap of the 8+ charged dimer with the 4+ charged monomer, indicating that a distinction can be made between the monomeric and dimeric states.Screening of Specific MCP-1 Ligands from Heparin Octasaccharide Library—Previously, it was reported that heparin interacted with MCP-1 (6Kuschert G.S.V. Coulin F. Power C.A. Proudfoot A.E.I. Hubbard R.E. Hoogewerf A.J. Wells T.N.C. Biochemistry. 1999; 38: 12959-12968Crossref PubMed Scopus (488) Google Scholar, 21Lau E.K. Paavola C.D. Johnson Z. Gaudry J.P. Geretti E. Borlat F. Kungl A.J. Proudfoot A.E. Handel T.M. J. Biol. Chem. 2004; 279: 22294-22305Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar). The apparent dissociation constant between unfractionated heparin and WT MCP-1 was determined